Protection of Materials and Structures from Space Environment: ICPMSE-6 (Space Technology Proceedings)

PROTECTION OF MATERIALS AND STRUCTURES FROM SPACE ENVIRONMENT

Space Technology Proceedings VOLUME 5

PROTECTION OF MATERIALS AND STRUCTURES FROM SPACE ENVIRONMENT ICPMSE-6

Edited by JACOB I. KLEIMAN Integrity Testing Laboratory Inc., Toronto, Canada

and ZELINA ISKANDEROVA Integrity Testing Laboratory Inc., Toronto, Canada

KLUWER ACADEMIC PUBLISHERS NEW YORK, BOSTON, DORDRECHT, LONDON, MOSCOW

eBook ISBN: Print ISBN:

1-4020-2595-5 1-4020-1690-5

©2004 Springer Science + Business Media, Inc. Print ©2003 Kluwer Academic Publishers Dordrecht All rights reserved No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher Created in the United States of America

Visit Springer's eBookstore at: and the Springer Global Website Online at:

http://www.ebooks.kluweronline.com http://www.springeronline.com

v

Table of Contents Introduction Acknowledgements Organization Section A New Era for Canada in Space M. Garneau

Space and Atmospheric Environments: From Low Earth Orbits to Deep Space J. L. Barth

xi xiii xv

1 7

Materials Interactions with the Space Environment: International Space Station May 2000 to May 2002 S. L. Koontz, M. Pedley, R. R. Mikatarian, J. Golden, P. Boeder , J. Kern, H. Barsamian, J. I. Minow, R. L. Alstatt, Mary J. Lorenz, B. Mayeaux, J. Alred, C. Soares, E. Christiansen, T. Schneider, D. Edwards

31

Photoconductivity in Transparent Arc-proof Coatings T. Cashman, J. Kaur, L. K. Muhieddine, M. Shanbhag, S. H. Ubaid, B. Welch, J. Vemulapalli, P. D. Hambourger

Effects of Space Environment Factors on Optical Materials for Space Application H. Liu,H. Geng, S. He, S. Yang, D. Yang, V.V. Abraimov, H. Wang

73 81

A Study of Synergistic Radiation Effects of Protons and Electrons on Teflon FEP/Al Degradation D. Yang, C. Li, H. Geng, S. He, S. Yang

Dose Rate Effects in Polymer Materials Irradiated in Vacuum B. A. Briskman, E. R. Klinshpont, V. F. Stepanov

Towards a Database for Assessment of Near-Earth Space Radiation Effects on Optical Glasses A.Gusarov, D. Doyle, M. Fruit

91 99 113

The Role of Proton and Electron “Abnormal” Formations in Radiation Influence on Construction Elements of Spacecrafts 123 Y. A. Grachov, O. R. Grigoryan, L. S. Novikov, I. V. Tchourilo A Study of Methylsilicone Rubber Damage Behavior Induced by Proton Irradiation L. Zhang, S. Yang, H. Geng, S. He, Q. Wei

131

vi Section B

A Unified Space Environment Effects Database for Russian and North American Organic and Inorganic Materials S. H. C. P. McCall, A. A. Clark, A. J. Clark, J. I. Kleiman, Z. Iskanderova, B. Briskman, E. Klinshpont, Y. Shavarin

139

The Effect of Heating on the Degradation of Ground Laboratory and Space Irradiated Teflon FEP K. K. de Groh, M. Martin

On the Thermal Stability of Polyimides for Space Application C.O.A. Semprimoschnig, S. Heltzel, A. Polsak, M. van Eesbeek

Synergistic Degradation of CV-1144-0 Due to Ultraviolet Radiation and Heat J. E. Haffke, J. A. Woollam

Behaviour of Thermal Control Coatings Under Atomic Oxygen and Ultraviolet Radiation S. Remaury, J. C. Guillaumon, P. Nabarra

Ground Testing of SCK5 White Silicone Paint for LEO Applications I. Gouzman, E. Grossman, G. Lempert, Y. Noter, Y. Lifshitz, V. Viel-Inguimbert, M. Dinguirard

Study of Polymer Coatings Resistance After the Long-Term Exposure on Space Station “MIR” E. N. Kablov, V. T. Minakov, I. S. Deev, E. F. Nikishin

Issues and Consequences of Atomic Oxygen Undercutting of Protected Polymers in Low Earth Orbit B. A. Banks, A. Snyder, S. K. Miller, R. Demko

Effect of Space Gaseous Environment on the Thermophysical Properties of Materials and Structures E. Litovsky, J. I. Kleiman, N. Menn

155 171 183 193 203 217 235 245

Cleanliness Support of the Launch Vehicle for Putting Into Orbit the Spacecraft Meteor-3M with the "SAGE-III" Instrument V.G.Sitalo, V.G.Tykhyy, L.P. Potapovych

257

Section C

Irreversible Shrinkage Effects of Carbon Fibers in Polymer Matrix Composites Exposed to the "MIR" Space Environment O.V. Startsev, D.A. Khristoforov, V.V. Issoupov, E.F. Nikishin, A.F. Rumyantsev

263

vii Combined Effect of Thermal and Mechanical Stresses on the Viscoelastic Properties of a Composite Material for Space Structures V. Issoupov, O. V. Startsev C. Lacabanne, P. Demont, V. Viel-Inguimbert, M. Dinguirard, E. F. Nikishin

Hyperthermal Reactions of Oxygen Atoms with Saturated Hydrocarbons T. K. Minton, D. J. Garton, H. Kinoshita

A Review of Lubrication Issues on the Canadarm 2 J. Antoniazzi, D. Milligan

271 283 291

3

Degradation of Polymers by O( P) in Low Earth Orbit A. Gindulyte, L. Massa, B. A. Banks, S. K. R. Miller

Iridium Metal as Potential Substrate for Experiments in Space L. Yan, J. A. Woollam

The Influence of the Atomic Oxygen Plasma on the Surface and on the Photoelectric Properties of Solar Arrays B.G. Atabaev, L.F. Lifanova, F. Rakhimova, A.V. Markov, I.V. Tchourilo

299 307

319

Some Aspects of Simulation of Outgassing Processes in Thermal Vacuum Exposure of Coatings Applied to Space Vehicles R..H. Khassanchine, A.V. Grigoresvskiy, Y.P. Gordeev

327

Section D Vacuum Ultraviolet Radiation Characterization of RF Air Plasma and Effects on Polymer Films J. Dever, C. McCracken, E. Bruckner

Studies of the Surface Oxidation of Silver by Atomic Oxygen M. L. Zheludkevich, A. G. Gusakov, A. G. Voropaev, A. A. Vecher, E. N. Kozyrski, S. A. Raspopov

Determination of the Energy Level of the Atomic Oxygen Flux Generated by the Space Simulation Apparatus using a Thermal Modeling Method X-X. Jiang, L. Lucier, D. Nikanpour, S. Gendron

Design and Testing of a Mini-Spectrometer System for On-Orbit Degradation Studies of Optical Materials M. Dinguirard, M. van Eesbeek, A. P. Tighe

335 351

359

367

Kapton as a Standard for AO Flux Measurement in LEO Ground Simulation Facilities: How Good Is It? E. Grossman, I. Gouzman, G. Lempert, Y. Noter, Y. Lifshitz

379

viii Temperature and Impingement Angle Dependences of Atomic Oxygen-Induced Erosion of Polyimide and Polyethylene Films Measured by Quartz Crystal Microbalance M. Tagawa, K. Yokota, T. Kida, N. Ohmae

391

Integrating Sphere Unit for Precision Measurement of Absolute Reflectance and Transmittance of Spacecraft Materials in a Vacuum Chamber

401

V. V. Eremenko, V. M. Naumenko, V. N. Fenchenko, V. G. Tykhyy

Numerical Simulation of Thermal Stress Induced by Thermocycling in Hot Rolled 1420 Al-Li Alloy H. Geng, S. He, D. Yang

407

Changes in Microstructure and Tensile Properties of Hot Rolled 1420 Al-Li Alloy Subjected to Thermocycling 413 H. Geng, S. He, D. Yang Section E

Photosil¥ Surface Modification Treatment of Polymer-based Space Materials and External Space Components Y. Gudimenko, R. Ng, J. I. Kleiman, Z. Iskanderova, P.C. Hughes, R.C. Tennyson, D. Milligan

Atomic Oxygen Resistant, Low α, Anti-Static Polyimides for Potential Space Applications A. J. Gavrin, S. W. Au-Yeung, R. Mojazza, K. A. Watson Jr., J. G. Smith, J. W. Connell

419 435

Deposition and Characteristics of Atomic Oxygen Protective Coatings Using Plasma Polymerized HMDSO 443 J. Wang, Y. Wang, X. Zhou, Z. Jin, Z. Yu Development of Protective and Passive Thermal Control Coatings on Carbon-Based Composite Materials for Application in Space 451 M. Francke, B. Fritsche, A. Moc, R. B. Heimann, Z. Iskanderova , J. I. Kleiman Structure and Composition of Non-Metallic Solar Array Materials Retrieved after Long-Term Exposure Overboard the “MIR” Orbital Space Station V. A. Letin, L. S. Gatsenko, I. S. Deev, E. A. Bakina, A. V. Malenkov, E. F. Nikishin 461 Micro- and Macrotribological Properties of Solid Lubricants in 5 Electronvolts Atomic Oxygen Exposures M. Tagawa, M. Muromoto, H. Kinoshita, N. Ohmae, K. Matsumoto, M. Suzuki 475 ZnSe Coatings for Spacecraft Electrochromic Thermal Control Surfaces L. Yan, J. A. Woollam, E. Franke

483

ix Perfluorocyclobutane (PFCB) Polyaryl Ethers for Space-Based Applications A. Gavrin, J. Nebo, N. Rice, L. Kagumba, D. W. Smith Jr., J. Jin, C. M. Topping

491

Section F

Fast Three-Dimensional Method of Modeling Atomic Oxygen Undercutting of Protected Polymers A. Snyder, B. A. Banks

503

Development and Verification of a Predictive Model and Engineering Software Guide for Durability Evaluation of Polymer-based Materials in LEO J. I. Kleiman, Z. Iskanderova, D. Talas, M. van Eesbeek, R. C. Tennyson 515 Comparative Study of Low Energy C and O Atoms Impact in a Hydrocarbon Surface M. Medvedeva, B. J. Garrison

A Direct Trajectory Dynamics Investigation of Fast O + Alkane Reactions R. Z. Pascual, G. C. Schatz, D. J. Garton

527 537

Mathematical Simulation Methods to Predict Changes of Integral and Spectral Optical Surface Characteristics of External Spacecraft Materials and Coatings 543 V. N. Vasliew, A. V. Grigorievskiy, Y. P. Gordeev Subject Index

551

Author Index

559

This page intentionally left blank

Introduction This publication presents the proceedings of ICPMSE-6, the sixth international conference on Protection of Materials and Structures from Space Environment, held in Toronto May 1-3, 2002. The ICPMSE series of meetings became an important part of the LEO space community since it was started in 1991. Since then, the meeting has grown steadily, attracting a large number of engineers, researchers, managers, and scientists from industrial companies, scientific institutions and government agencies in Canada, U.S.A., Asia, and Europe, thus becoming a true international event. This year’s meeting is gaining even stronger importance with the resumption of the ISS and other space projects in LEO, GEO and Deep Space. To reflect on these activities, the topics in the program have been extended to include protection of materials in GEO and Deep Space. The combination of a broad selection of technical and scientific topics addressed by internationally known speakers with the charm of Toronto and the hospitality of the organizers brings participants back year after year. The conference was hosted and organized by Integrity Testing Laboratory Inc. (ITL), and held at the University of Toronto’s Institute for Aerospace Studies (UTIAS). The meeting was sponsored by the Materials and Manufacturing Ontario (MMO) and the CRESTech, two Ontario Centres of Excellence; Air Force Office of Scientific Research (AFOSR/NL); MD Robotics; EMS Technologies; The Integrity Testing Laboratory (ITL); and the UTIAS. Summarizing, over 75 people from 15 countries including Canada, USA, France, Holland, Russia, Ukraine, China, Israel, Belarus, Japan, Belgium, Spain, Germany, India and England registered for the conference representing all major space agencies (NASA, Russian Space Agency, Canadian Space Agency, European Space Agency and the French Space Agency) and the major companies, institutions and government organizations involved in space activities, indicating a further increase in international co-operation in this critical area of protection of materials in space. A Plenary Session was held with 3 invited papers. Six Oral Sessions were organized with 34 papers, including three papers by Russian scientists invited by MMO through the “Distinguished Lecturer Program”. In addition, six Poster Sessions with over 38 papers and technical exhibitions by two companies were organized, with all presentations covering a variety of topics. For the ISS and other future space exploration projects, the safety of the crew and the soundness of the structures will be the major concern. Questions about thermal stability, resistance to soft and hard radiation sources, and combined effects of vacuum ultraviolet, atomic oxygen and micrometeoroids will continue to accumulate with the development of new materials and the increased use of polymers, plastics and composite materials. The papers in the proceedings are organized into six major sections as follows: a) Space Environmental Effects: Radiation and Charging Effects b) Space Environmental Effects: Synergism of AO/VUV/TC c) Space Environmental Effects: Synergism of AO/VUV/TC

xi

xii d) Space Environmental Effects: Instrumentation & Calibration e) New Materials and Processes f) Modeling and Computer Simulations

Jacob Kleiman, Chairman/Organizing Committee/ICPMSE-6 Integrity Testing Laboratory Inc., 20 January, 2003

Acknowledgements We would like to acknowledge the following for their generous support of ICPMSE-6, the Sixth International Conference on Protection of Materials and Structures from Space Environment; Air Force Office of Scientific Research (AFOSR/NL) Materials and Manufacturing Ontario (MMO) MD Robotics EMS Technologies The Integrity Testing Laboratory (ITL), CRESTech ASM International, Ontario Chapter; The University of Toronto Institute for Aerospace Studies As well, we would like to acknowledge all the people from ITL and UTIAS that contributed their time and effort and especially Sergei Sivolobtchik and Elissa Schaman two bright University of Toronto co-op students for their help in preparation of the materials for publication. Jacob Kleiman, Integrity Testing Laboratory Inc. Conference Chairman

xiii

This page intentionally left blank

Organization Chairperson: Prof. Jacob Kleiman, ITL Inc./UTIAS, Canada The Organizing Committee: B. A. Banks, NASA, Cleveland, USA D. L. Edwards, NASA, Huntsville, USA D. Nikanpour, Canadian Space Agency, Canada J. Golden, Boeing, Houston, USA E.F. Nikishin, M.V. Khrunichev State Space Scientific Production Center, Russia V.Sitalo, Design Bureau “Yuzhnoe”, Ukraine P. C. Trulove, Air Force, USA M. Van Eesbeek, ESA, Noordwijk, The Netherlands D. Yang, Harbin Institute of Technology, China The Program Committee: M. Dinguirard, ONERA/DESP, France M. Finkenor, NASA, Huntsville, USA S. Koontz, NASA, Houston, USA T. Minton, Montana State University, USA G. Pippin, Boeing, Seattle, USA M. Tagawa, Kobe University, Japan E. Werling, CNES, France The Local Organizing Committee: Z. Iskanderova, ITL Inc./UTIAS, Toronto, Canada R. C. Tennyson, UTIAS, Toronto, Canada R. Worsfold, CRESTech, Toronto, Canada H. Pellegrini, MMO, Toronto, Canada P.Patanik, NRC-IAR, Ottawa, Canada Session Chairs Opening Session: Moderator: J. Kleiman-ITL Inc./UTIAS, Canada Session A: Space Environmental Effects: Radiation and Charging Effects Moderator: M. Dinguirard - ONERA, Toulouse, France Session B: Space Environmental Effects: Synergism of AO/VUV/TC Moderator: T. Minton - Montana State University, Bozeman, U.S.A.

xv

xvi Session C: Space Environmental Effects: Synergism of AO/VUV/TC Moderator: E. Nikishin M.V. Khrunitchev State Space Scientific Production Center, Moscow, Russia Session D: Space Environmental Effects: Instrumentation & Calibration Moderator: B. Banks, NASA Lewis Research Center, Cleveland, U.S.A. Session E: New Materials and Processes Moderator: M. Tagawa, Kobe University, Kobe, Japan Session F: Modeling and Computer Sumulations Moderator: Z. Iskanderova, Integrity Testing Laboratory Inc./UTIAS, Toronto, Canada

A NEW ERA FOR CANADA IN SPACE Speaking Notes DR. MARC GARNEAU Canadian Space Agency 6767 Route de L ’Aeroport, St. Hubert, Quebec J3Y 8Y9, Canada 1 .0 Introduction It has only been 44 years since a small Russian aluminum sphere called Sputnik orbited the Earth for three months and, in doing so, launched the Space Age. Canada's own Space Program dates back to the launch of Alouette-1, which made us the third nation in space in 1962. Incredibly, for all the accomplishments along the way, our journey as a species into space has not yet spanned a human lifetime. Canada first went into space for practical reasons. The rationale behind the launch of our debut satellite, Alouette-1, was simple: Canadian scientists wanted to understand why ionospheric activity, prevalent in the far North, adversely affected radio communications. The best way to find out was to place a spacecraft in orbit for in situ measurements. Ten years later, mission accomplished: Alouette was switched off and Canada emerged as a world expert on ionospheric phenomena. Similarly, Canadians needed to find a way to connect communities scattered over our vast expanse. In 1972, we came up with a national communications satellite system called Anik, the first of its kind in the world. As a result, Telesat Canada, the system’s operator, is now the most experienced satellite control organization anywhere. The Canadian Space Agency was created in 1989 to manage Canada's civil space program, and to ensure that space science and technology benefit all Canadians. 2 .0 Building on a Legacy Building on a legacy of addressing the needs of Canadian society, the Canadian Space Agency has charted a vision for Canada's future in space—to expand and apply knowledge of space for the benefit of Canadians and of humankind and in so doing, inspire through excellence. To accomplish this, we are currently focusing on six strategic areas of importance to Canada: –Earth and Environment –ISS Utilization –Mars Exploration –Small Satellite Program

1

2 –Communicating with the Public; and –Preserving our technical expertise. Please do not assume that we are dropping or putting other agency programs on the backburner. These six strategic areas represent our decision to focus on certain areas in the face of limited budgets. Let’s look at each one briefly. 2.1 EARTH & ENVIRONMENT Our atmosphere is changing. We see evidence of climate change, air pollution, global warming and ozone depletion. Our planet is changing—there is relentless population growth, depletion of vegetable cover, soil erosion, loss of arable land, ground pollution, desertification, depletion of fresh water resources. Our oceans are changing due to pollution, depletion of fish stocks, and changing sea levels. The changes occurring in Canada and world-wide will have important long-term repercussions for our country. Using space as a vantage point to observe and monitor the Earth and our natural heritage will be our most important mission in the foreseeable future. Canada's RADARSAT-1, the world's first commercial Synthetic Aperture Radar, satellite, now has a six-year legacy of monitoring the Earth from space. The heart of the Canadian remote-sensing industry, RADARSAT-1 provides images that guide international rescue teams to disaster sites and dispatch Arctic icebreakers where needed; images that detect oil spills and supports scientists in the fields of cartography, mineral and oil exploration, hydrology, forestry, oceanography and agriculture. RADARSAT-1 currently has a network of stations and a global client base that includes more than 600 commercial and government users from more than 57 countries. It has also responded to over 50 emergencies—both natural and man-made disasters, and is now part of the International Charter of Space and Major Disasters. We are preparing to launch RADARSAT-2 in late 2003, which will build on the success of its predecessor with the most advanced satellite radar technology on the commercial market. 2.2 THE UTILIZATION OF THE INTERNATIONAL SPACE STATION Canada is one of 15 partner nations in the midst of building the most ambitious science and engineering project in the history of humanity—the construction of the International Space Station. The first element of the Canada's contribution to the project, Canadarm2, is already operating on board the Station, and our next element, the Mobile Base System, is scheduled for launch on May 30 of this year. Canada is fulfilling its promised contribution to the International Space Station. However, we must also focus on the main purpose of the International Space Station—to allow us to perform scientific experiments in the unique conditions of microgravity. We would also like to ensure that Canadian astronauts continue to be an integral part of building and working on the Station to perform research on behalf of Canadian scientists. This will be an important ongoing activity for the next 10 to 15 years, both through government and, hopefully,

3 through commercial channels. The success of the Station will depend on it. Canada is on its way to Mars 2.3 CANADA IS ON ITS WAY TO MARS Canada must also look outward into the universe as well as back at Earth. As our closest planetary neighbor, Mars holds a special appeal for the scientific community. Researchers hope that by looking outward into the depths of the universe to planets like Mars, we may find answers to some of the most basic questions about life here on Earth; the formation of our planet and our solar system; and whether life—and the water that sustains it—exists elsewhere in the universe. As we speak, a Canadian scientific instrument, known as the Thermal Plasma Analyzer, is on its way to Mars on board the Japanese satellite Nozomi. When it reaches Mars in 2004, the Thermal Plasma Analyzer, or TPA for short, will provide the scientific community with valuable information on the origin and composition of Mars’s atmosphere. The Canadian public supports our participation in Mars exploration. Mars exploration "inspires" Canadians. In a national public opinion poll conducted in March 2002, 77% of Canadians expressed support for Canada's participation in future missions to Mars. Exploring Mars is the next major international space program after Space Station. Canada can contribute but will only do so if: –It has a visible role –Our involvement is science-driven and technology-enabled –And, of course, if funding permits our involvement. We are now in the planning stages to determine what form our potential involvement may take. We are consulting with the Canadian scientific community and Canadian industry, as well as with our international partners. And we are currently conducting feasibility studies to help us identify opportunities for 2007 and 2009 missions. We are also assessing Canada’s potential contribution beyond 2009. Ideally, we would like to see a “distinctly Canadian” mission in 2011, conducted with our partners but one that would feature Canadian ideas, technologies and expertise. 2.4 SMALL SATELLITE PROGRAM Another area for future development is a small satellite program. There are compelling reasons why the Canadian Space Agency should be providing a Small Satellite capability on a regular basis, e.g. every 2 years. It would allow Canadian industry to space-test new hardware, i.e. technology demonstration. It would provide a more costeffective platform for new science instruments for Canadian researchers, and it would allow remote-sensing pilot projects. It would also enable Canada to compete internationally in the small satellite market, providing it can do so on a cost-competitive basis. 2.5 COMMUNICATING WITH THE PUBLIC The public is very interested in space.

4 We know this from polls and from all the requests we receive on a continuous basis. Canadians believe space is important for Canada. However, they know very little about many of our programs and how they serve Canadians. We need to better inform them— a challenging task. Young Canadians in particular, are fascinated by space. We need to continue to reach out to them in a variety of ways to stimulate their interest in Science and Technology. 2.6 PRESERVING OUR EXPERTISE The CSA is a multi-faceted organization. It manages space programs, but it also: • performs focused in-house R&D in specific areas; • performs satellite operations from St-Hubert; • tests space hardware at the David Florida Laboratory; • trains astronauts and cosmonauts in the use of the Mobile Servicing System; and • has a strong and visible interface with the public. We need to continue to foster the growth of in-house skills to maintain Canada's international leadership in its strategic areas of expertise. 3 .0 The Protection of Spacecraft and Materials from the Space Environment With the ambitious agenda we have laid out for ourselves in the coming years, the protection of spacecraft and materials from the space environment has taken on increased importance. Debris, of course, is a major concern. It has been estimated that more than 4000 space launches have taken place since 1957, leading to more than 8500 trackable objects above 10 cm in size in near-Earth orbit. Of these, between 600 and 700 are operational spacecraft. The remainder is debris, which poses a hazard to human space flight, as astronauts are well aware, as well as the safe operation of unmanned spacecraft. Orbiters have suffered “dings” on their windows. Spacecraft are also exposed to contamination from ultraviolet radiation and atomic oxygen in low Earth orbit, which can degrade sensitive optical surfaces. We are developing increasingly sophisticated hardware, like Canadarm2, which is designed to be maintained and repaired in space for the duration of its lifetime. Protecting the Space Station itself, as well as visiting vehicles, such as the space shuttle, the Soyuz, Progress, the European Space Agency and NASDA ATV (automated transfer vehicle) is a growing concern. A proactive approach to defending space hardware from both human-induced contamination and ever-increasing space debris is essential. Space debris is becoming increasingly concentrated in useful orbits, where activity is at its greatest. This includes geostationary orbit, where telecommunications satellites are located, medium Earth orbit, where numerous Earth Observation spacecraft are in orbit, and in low Earth orbit, where not only the Space Station and shuttle and transfer vehicles are affected, but also low Earth orbit satellites, particularly the micro- and nanosatellites.

5 The recent news of the go-ahead for the Galileo network of satellites by the European Space Agency is indicative of this trend. The only natural debris removal mechanism so far is the decay in the orbit of space debris, and their eventual fall back to Earth. This phenomena is only relevant to very low orbits, as you know, since for the medium Earth orbit, the orbital decay time is of the order of a couple of centuries. For higher orbits, it may take many thousands of years. But even the natural decay of space debris can be problematic. This is an issue of special concern to Canada , the only country to have been impacted by a satellite carrying a nuclear reactor, when the Soviet Union’s Cosmos 954 disintegrated over the Northwest Territories in 1978, spreading contaminated debris over several hundred kilometres. With a projected increase in the of number of launches, the debris population will undoubtedly increase. The Canadian Space Agency has recognized this trend, and has initiated research activity in space debris damage mitigation design. For instance, Canada's RADARSAT-2 spacecraft has introduced a debris damage mitigation design to protect its synthetic aperture radar antenna and reduce debris generation. Through the David Florida Laboratory, we will continue to ensure the rigorous qualification satellites and other forms of space hardware for space flight through its assembly, integration and testing facilities. And we will continue to support the fruitful exchange of ideas in for a such as this Conference as we collectively search for new approaches to address some of these pressing issues. 4 .0 Conclusions And we will continue to support the fruitful exchange of ideas in for a such as this conference as we collectively search for new approaches to address some of these pressing issues. In the coming years, the Canadian Space Agency will continue to build upon a 40-year tradition of Canadian excellence in space. We have charted our course for the future— Canada's future in space. We are committed to ensuring that Canada has a prominent role as we take our first steps off our home planet and venture out into the universe. And as our past achievements in space were guided by our desire to address the needs of Canadians, so, too, will our future be governed by issues Canadians take to heart as we strive to advance scientific knowledge for the benefit of Canadians, and indeed, for the benefit of all of humanity.

This page intentionally left blank

SPACE AND ATMOSPHERIC ENVIRONMENTS: FROM LOW EARTH ORBITS TO DEEP SPACE JANET L. BARTH NASA/Goddard Space Flight Center 301-286-8046 E-mail: [email protected]

Abstract Natural space and atmospheric environments pose a difficult challenge for designers of technological systems in space. The deleterious effects of environment interactions with the systems include degradation of materials, thermal changes, contamination, excitation, spacecraft glow, charging, radiation damage, and induced background interference. Design accommodations must be realistic with minimum impact on performance while maintaining a balance between cost and risk. The goal of applied research in space environments and effects is to limit environmental impacts at low cost relative to spacecraft cost, to infuse enabling and commercial off-the-shelf technologies into space programs. The need to perform applied research to understand the space environment in a practical sense and to develop methods to mitigate these environment effects is frequently underestimated by space agencies and industry. Applied science research in this area is critical because the complexity of spacecraft systems is increasing, and they are exposed simultaneously to a multitude of space environments. 1. 0 Introduction Spacecraft are exposed to a multitude of environments that are not present at the surface of the Earth, including plasmas, high-energy charged particles, neutral gases, x-rays, ultraviolet (UV) irradiation, meteoroids, and orbital debris. The interaction of these environments with spacecraft systems cause degradation of materials, thermal changes, contamination, excitation, spacecraft glow, charging, radiation damage, and induced background interference. The damaging effects of natural space and atmospheric environments pose a difficult challenge for spacecraft designers. Unfortunately, the need to perform applied research to understand and model the space environments, to understand the physics of the interaction of the environment with spacecraft systems, and to develop methods to mitigate environmental effects is frequently underestimated by space agencies and industry. At the same time that the complexity and performance requirements of spacecraft systems are increasing, other system drivers reduce our ability to meet

7

8 requirements. For example, the demand for commercial microelectronics reduces the availability of components suitable for space environments. Also, the need to design lighter and more complex spacecraft structures pushes the development of exotic materials for space use. Uncertainties in space environment and effects models used to predict the performance of new technologies translate into large design margins. Large design margins can preclude the use of new technologies that will meet mission requirements. The goal of applied research in space environments and effects is to reduce design margins used to account for the uncertainty of performance predictions, thereby enabling technology infusion into space programs. Design accommodations must have minimum impact on performance and budgets. The challenge is to achieve a realistic balance between cost of environment accommodations and mission risk. Several organizations have developed concurrent engineering approaches to accommodating environment effects and assessing risk [1,2]. Regardless of the environmental effect or the technology, the approaches follow these steps: Define the environment external to the spacecraft; Evaluate the environment interaction with the spacecraft; Define the requirements and define criticality factors; Evaluate design and performance characteristics of components; “Engineer” with designers and program managers including risk analysis and definition of design margins; Iterate the process with updated knowledge. Note that the first step for every mission is to define the level of the environments in metrics that are applicable the effect on the spacecraft technologies. This is accomplished by using environment and interaction models, and when models are not available, by using in-flight data. The purpose of this paper is to describe the natural space environments that must be taken into account when designing spacecraft. Specific effects on materials and structures can be found in other contributions to these proceedings. Further information on the radiation environment can be found in iii[3] and information on radiation effects can be found in [4].iv i

ii

[, ]

2. 0 Description of the Environments The space and atmospheric environments relevant to spacecraft effects can be roughly categorized into meteoroid and debris, ultraviolet irradiation, neutral thermosphere, cold and hot plasma, and particle radiation. The differences between the atmospheres and magnetospheres of the planets and interplanetary space are dramatic. Even within the atmospheric and magnetospheric systems of the Earth, there are large spatial and temporal variations in the constituency and density of the environments. For the purpose of discussion of environmental definitions, missions can be roughly categorized into low earth orbits (LEOs), middle earth orbits (MEOs), geosynchronous (GEO), geosynchronous transfer orbits (GTOs), interplanetary, and other planets. Table 1 gives a summary of the differences between the space environments around Earth and other planets

9 TABLE 1. Variation of planetary environments from Earth ENVIRONMENT Solar Wind Meteoroids Orbital Debris Galactic Cosmic Rays Solar Particles Solar Radiance Atmospheres Trapped Radiation

COMPARISON TO EARTH ~ Same ~ Same None at this time Small variation Large variation with radial distance from Sun Large variation with radial distance from Sun Large variations Large variations

2.1 THE SOLAR INFLUENCE ON SPACE ENVIRONMENTS The complex environment of Sun-Earth space consists of time varying ultraviolet, xray, plasma, and high-energy particle environments. Variations depend on location in space and on the year in the solar cycle, both somewhat predictable. However, large variations that depend on events on the Sun are not predictable with reasonable certainty and are known only statistically based on past history. Because the Sun is a gas, its solar magnetic field is convoluted and highly variable. Both the long-term variation in the magnetic field that occurs in a 22-year cycle and the short term variations in the form of intense, short lived storms are responsible for observable changes in the interplanetary and near-Earth environments.

Figure 1. The corona extends several solar diameters.

The sun’s outer atmosphere, the corona (see Figure 1), extends several solar diameters into interplanetary space. The corona continuously emits a stream of protons, electrons, doubly charged helium ions, and small amounts of other heavy ions, collectively called the solar wind. It was once thought that the region where the solar wind could no longer be detected, i.e., the boundary of the heliosphere, was not far beyond Jupiter (800 million km). However, the Pioneer 10 spacecraft, presently at > 12

10 billion kilometers from Earth, is still measuring solar wind. Scientists now believe that the boundary could lie as far as 17 billion kilometers from the Earth [5]. [v] The high temperature of the corona inputs sufficient energy to allow electrons to escape the gravitational pull of the sun. The result of the electron ejections is a charge imbalance resulting in the ejection of protons and heavier ions from the corona. The ejected gas is so hot that the particles are homogenized into a dilute plasma. The energy density of the plasma exceeds that of its magnetic field so the solar magnetic field is “frozen” into the plasma. The electrically neutral plasma streams radially outward from the sun at a velocity of approximately 300 to 900 kilometers per second with a temperature on the order of 104 to 106 K. While the solar wind is millions of metric tons of matter moving at a million kilometers per hour, its density is so low that the physics is that of a vacuum. The energies of the particles range from approximately 0.5 to 2.0 keV/n. The average density of the solar wind is 1 to 30 particles/cm3. Figure 2 shows that the solar wind velocity and density can vary greatly over a short time period. Table 2 gives the approximate particle composition of the solar wind.

Figure 2. The solar wind velocity and density are highly variable and are a function of the activity on the sun. SOHO/University of Maryland

TABLE 2. Solar wind particle composition

PARTICLE Proton He++ Other Heavy Ions Electrons

ABUNDANCE 95% of the positively charged particles ~4% of the positively charged particles < 1% of the positively charged particles Number needed to make solar wind neutral

11 It is well known that the level of activity of the sun varies with time defining “solar cycles”. The solar cycle as a recurrent pattern of solar magnetic activity was first identified in 1843 by the German observer, Schwabe, who found an approximately 11year cycle in the number of sunspots1 (see Figure 3). The 11-year cycle of sunspots corresponds to similar 11-year cycles of other features in the sun’s active regions, including the number of faculae, the rate of incidence of solar flares and coronal mass ejections (CMEs), and the intensity of coronal x-ray and radio-frequency emissions.

Figure 3. Sunspots are regions of highly dense magnetic field. after Lund Observatory

From Figure 4, it can be seen that the length of the solar cycle can be highly variable. From 1645 to 1715, the sunspot activity seemed to disappear. Because temperatures on Earth dropped during that time, those 70 years are known as the little ice age. From 1100-1387, there was an increase in the number of sunspots. Studies of recent solar cycles [5, 6], Cycles 19 through 22, have determined that the length of the solar cycle over the past 40 years has ranged from 9 to 13 years, with 11.5 being the average. For modeling purposes and for defining the environment for spacecraft missions, the solar cycle can be divided into a 7-year maximum phase of high levels of activity and a relatively “quiet” 4-year minimum phase. The space environment is dominated by the activity of the Sun, which acts as both a source and a modulator. It is a source of protons and heavier ions via the periodic highenergy solar events that accelerate large numbers of particles. The solar wind is also a source of the particles trapped in outer regions of the Earth’s radiation belts. The galactic cosmic ray heavy ion (GCR) levels follow a cyclic pattern reflecting the activity level of the sun because they originate outside of the solar system and must “fight” against the solar wind to reach interplanetary space. As a result, the GCR levels are highest during solar minimum and lowest during solar maximum. Atmospheric neutrons are secondary products of collisions between GCRs and oxygen or nitrogen atoms in the Earth’s atmosphere; therefore, their levels are also modulated by the solar cycle. Finally, the levels of particles trapped in planetary magnetospheres are modulated by both longterm variations in solar activity and solar storm events. The effect of the cyclic variation vi, vii]

1

cooler areas of the sun seen as dark “spots” through a telescope

12 of the sun’s activity will be discussed in more detail in later sections as it applies to specific environments.

Figure 4. Yearly sunspot numbers

Solar flares and coronal mass ejections (CMEs) are two storm phenomena occurring on the Sun that affect particle levels. Solar flares are seen as sudden brightenings in the photosphere near sunspots (see Figure 5). Flares are intense releases of energy involving tearing and reconnection of strong magnetic field lines. They are the solar systems largest explosive events. Large increases in the solar wind density are measured in interplanetary space after solar flare occurrence because the energy released from the flare accelerates particles in the solar plasma to high energies. CMEs occur in the chromosphere, the layer of the sun outside of the photosphere. The chromosphere can be seen only when filtering out the bright light of the photosphere. In Figure 6, the chromosphere is seen as a bright rim around the sun. CMEs are observed as large bubbles of gas and magnetic field (see Figure 7). A CME can release approximately 1017 grams of plasma into interplanetary space. The

Figure 5. Brightening seen with a solar flare.

Figure 6. Bright rim around the sun is the chromosphere.

13 mechanism for the plasma release is not completely understood. CMEs result in large increases in solar wind velocity. It is the shock wave of the plasma release that is associated with particle acceleration and magnetic storms at the Earth. CMEs are poorly associated with flares but, in very large event CMEs, both CMEs and flares occur together [8]. The particle composition of CMEs and solar flares is discussed in Section 2.9. [viii]

Figure 7. Bubble of gas associated with a coronal mass ejection. NASA/SMM 24 Oct. 1989

2.2 METEOROIDS AND ORBITAL DEBRIS Meteoroids are primarily remnants of comet orbits. Several times a year Earth encounters increased meteoroid exposure as it intersects a comet orbit. Also, sporadic particles are released on a daily basis from the asteroid belt. Orbital debris consists of operational payloads, spent rockets stages, fragments of rockets and satellites, and other hardware and ejecta. The United States Air Force Space Command’s North American Aerospace Defence Command (NORAD) tracks over 7,000 objects in LEO that are greater than 10 cm in size,

Figure 8. Location of objects tracked by NORAD

and there are tens of thousands smaller objects. Figure 8 shows the location of the objects tracked by NORAD. From the figure, it is possible to see the large number of objects in the LEO and GEO regions of space where most space agency, military, and commercial operations take place.

14 Meteoroids and orbital debris are a threat to spacecraft by causing structural damage and decompression, hypervelocity impacts from larger particles, surface erosion from collisions with smaller objects, and surface effects that cause changes in thermal, electrical, and optical properties. Mission risk factors include increased duration, increased vehicle size, vehicle design, solar cycle, orbit altitude, and inclination, and the threat is highly directional. Koontz et al. [9] give examples of micrometeoroid and orbital debris impacts on the International Space Station (ISS). ix]

2.3 ULTRAVIOLET IRRADIATION The sun is the natural source of ultraviolet irradiation, which has wavelengths of about 100 to 400 nanometers. UV irradiance can penetrate the atmosphere to reach the surface of the Earth. UV is an important component of the environment to evaluate due to its degradation effects on spacecraft surface materials. It is known to interact with atoms in the atmosphere, particularly oxygen, and ionizing particles to act synergistically on surface materials of spacecraft. UV radiation diffuses with distance from the sun at a rate of 1/R2 where R is the radial distance from the Sun. Table 3 lists solar UV irradiance at each planet. TABLE 3. Solar UV irradiance as a function distance from the Sun

PLANET Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto

DISTANCE FROM SUN (AU) 0.39 0.72 1.00 1.52 5.20 9.54 19.19 30.06 39.53

IRRADIANCE (W/m2) 9,126.6 2,613.9 1,367.6 595.0 51.0 15.0 3.7 1.5 0.9

2.4 PLASMA ENVIRONMENTS Plasma is ionized gas in which electron and ion densities are approximately equal. Plasma is distinguished from the energetic particle population in that it does not cause radiation effects and has energies < 100 keV. Plasma sources are the ionosphere, geomagnetic sub-storm activity, and the solar wind. The solar wind plasma from the solar corona was discussed in Section 2.1. The ionosphere is the electrically charged portion of the atmosphere and is characterized by low energy (eV) and high density. Plasma from geomagnetic sub-storm activity, on the other hand, has high energy (keV) and low density. Figure 9 shows the plasma around the Earth as seen by the Extreme Ultraviolet (EUV) instrument on NASA’s IMAGE (Imager for Magnetopause-toAuroral Global Exploration) spacecraft. Plasma shows dramatic variation with altitude, latitude, magnetic field strength, and solar activity. The solar wind plasma was

15 discussed in Section 2.1. The other plasma environments will be discussed in Sections 2.5.2 and 0.

Figure 9. The Helium ion plasma around the Earth as seen by the EUV instrument on the IMAGE spacecraft. Note the auroral activity.

2.5 THE EARTH’S ATMOSPHERE The Earth’s atmosphere is composed of complex layers of matter that are loosely defined by their dominant constituents. Starting from the surface of the Earth, the layers are the troposphere, stratosphere, mesosphere, the neutral thermosphere, and the charged thermosphere (ionosphere). The layers overlap and form a connected system. Figure 10 shows the altitude domains of the regions of the atmosphere. Low altitude spacecraft (< 800 km) are exposed to the environments of thermosphere, so those environments will be discussed in more detail. 2.5.1 Neutral Thermosphere The neutral thermosphere is the neutral portion of the Earth’s atmosphere at 90 to 600 km altitude above the surface of the Earth, composed primarily of neutral gases. In the lower thermosphere, the neutral population is dominated by atomic oxygen and by hydrogen and helium in the higher thermosphere. The distribution of the thermosphere neutral gases varies with solar activity because of heating caused by absorption of solar extreme ultraviolet radiation (EUV). A proxy commonly used for EUV is the 10.7-cm radio flux (F10.7). The main effects of the neutrals on spacecraft are drag, degradation of surface materials, and spacecraft glow. Drag results in altitude decay and torques. Drag is a function of the density of the neutral gas, hence is strongly affected by solar activity. The impact of solar storms on the Earth’s atmospheric density often causes sudden changes in the location of tracked objects. Figure 11 is a plot of the number of objects that were lost after a large magnetic storm in March of 1989.

16 Figure 10. The altitude domains of the Earth’s atmosphere, after NASA/MSFC

Figure 11. A plot of the number of tracked objects lost after a large magnetic storm

The degradation of surface materials is also a serious problem in very low earth orbits due to the presence high levels of atomic oxygen at 200 to 400 km. As with other thermosphere constituents, the level of atomic oxygen varies with the solar cycle. The erosion of surface materials causes changes in thermal, mechanical, and optical properties. Micrometeoroid impacts, sputtering, UV exposure, contamination, and ionizing radiation can aggravate these effects. Optical emissions generated by excitation of meta-stable molecules can also cause spacecraft glow. The surface acts as catalyst, therefore, the effect is material dependent.

17 2.5.2 Charged Thermosphere (Ionosphere) The Earth’s ionosphere is the electrically charged portion of the upper atmosphere from 100 to 800 km altitude (see Figure 10). It is a low energy (eV) plasma with high density relative to the magnetospheric plasma and the solar wind. Supersonic spacecraft motion through background ions in the ionospheric plasma has detrimental effects on spacecraft in LEO orbits, including solar array coupling to the plasma causing current drain on solar arrays, generation and emission of plasma waves, and increased surface contamination. There is renewed interest in studying the ionosphere because military and civilian communications are severely degraded during storms in the ionosphere induced by solar activity. The disruptions of communications during storms are far reaching as they affect high frequency radio, backscatter radar, satellite communications, and global positioning system (GPS) location. 2.6 ATMOSPHERES OF OTHER PLANETS Other planets in our solar system have atmospheres which differ dramatically from the Earth’s atmosphere. Therefore, designers of missions to other planets must take into account the differences from the Earth’s atmosphere and those of other planets and adjust the evaluation and mitigation of the effects accordingly. The ability of a planet to have an atmosphere is dependent on the planetary surface pressure and gravity (Table 4). For the four “terrestrial planets” the table shows that Mercury and Mars have very thin atmospheres and that the atmosphere of Venus is 92 times that of the Earth. The Jovian planets have large variations in their atmospheric density as a function of radial distance from the center of the planet because they are primarily composed of gases and clouds.

TABLE 4. Planetary pressure and surface gravity

PLANET Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto

SURFACE PRESSURE (BARS) 1.00x10-15 92 1 8.00x10-3 >>100 >>100 >>100 >>100 3.00x10-6

SURFACE GRAVITY (M/S) 3.70 8.87 9.78 3.69 23.12 8.96 8.69 11.00 0.66

In addition to the density of the atmosphere, it is critical to understand the composition of the atmospheres which is shown in Table 5. Notable is the similarity of the great hydrogen planets, Jupiter, Saturn, Uranus, and Neptune and the carbon dioxide atmospheres of Venus and Mars.

18 TABLE 5. Composition of the atmospheres of each planet

PLANET Mercury

Venus Earth

Mars

Jupiter Saturn Uranus

Neptune

Pluto

ATMOSPHERIC COMPOSITION 42% O2 29% Na 22% H2 6% He 96% CO2 3% N2 78% N2 21% O2 1% Ar 95% CO2 3% N2 2% Ar 90% H2 10% He 96% H2 3% He 83% H2 15% He 2% CH4 80% H2 19% He 1% CH4 ? CH4? N2 Ice?

2.7 THE EARTH’S MAGNETOSPHERE The Earth’s magnetosphere is a cavity formed by the interaction of the Earth’s magnetic field and the solar wind. In the absence of the solar wind, the Earth’s magnetic field would be shaped like the field of a bar magnet; non-varying, nearly symmetric about the magnetic axis, extending outward to long distances, and open at the poles. The bar magnet representation is accurate up to altitude of 4 to 5 Earth radii. The solar wind plasma, with its embedded solar magnetic field, compresses the geomagnetic field until there is balance between the magnetic pressure from the Earth and the momentum pressure from the solar wind forming a “bow shock”. On the dayside, during moderate solar wind conditions, the magnetosphere terminates at the magnetopause at ~10 Earth radii altitude. At the location of this “collision-less” shock, the solar wind plasma cannot penetrate deeply into the geomagnetic field because of its charged particle composition. In fact, 99.9% of the solar wind particles pass around the Earth’s magnetosphere. The flow of the solar wind around the flanks of the magnetopause stretches the geomagnetic field in the anti-solar direction into a long tail of up to ~300 Earth radii altitude. Some tail field lines are not closed and are connected to the solar magnetic field embedded in the solar wind. Figure 12 shows a depiction of the

19 magnetosphere.The magnetosphere is filled with plasma that originates from the ionosphere and the solar wind. The plasmasphere is at low and mid latitudes in the inner magnetosphere. The plasma sheet resides in the magnetotail. Overlapping the plasmasphere and the plasma sheet are the high-energy radiation belts or Van Allen belts (named for their discoverer, James Van Allen). Charged particles become trapped because the Earth’s magnetic field constrains their motion. They spiral around the field lines in a helicoidal path while bouncing back and forth between the magnetic poles. Superimposed on these spiral and bounce motions is a longitudinal drift of the particles because of the gradient of the magnetic field. Figure 13 illustrates the three motions. When the particle makes a complete azimuthal rotation, it has traced a “drift shell” (see Figure 14). The Van Allen belts will be discussed in more detail in Section 0.

Figure 12. The Earth’s magnetosphere, adapted from T. W. Hill by P.H. Reiff

Figure 13. The three motions of the trapped particles form drift shells. after Hess

20 2.7.1 Plasma Storms Geomagnetc substorms in the magnetotail plasma sheet can create “hot plasmas” which are injected into near-Earth regions of the magnetosphere. The effects of the plasma injections include biasing of spacecraft instrument readings, acing which causes upsets to electronics, increased current collection, re-attraction of contaminants, and ion sputtering which in turn leads to acceleration of material erosion. Missions affected by these injections are those in GEOs, GTOs, and MEOs. Conditions for the charging effects are large differentials, large fraction of total flux, darkness, and large spacecraft. Satellites at GEO have also measured strong local time effects on the rates of spacecraft charging with most occurring as the satellite passes into the dawn sector.

Figure 14. Drift shell of a trapped particle. Lamarie et al.

2.7.2 Van Allen Radiation Belts The Van Allen belts consist of protons, electrons and heavier ions that are “trapped” on the Earth’s magnetic field lines. The trapped electrons have energies up to 10s of MeV, and the trapped protons and heavier ions have energies up to 100s of MeV. These particles have complex spatial distributions that vary by several orders of magnitude depending on orbit inclination and altitude. The sun is a driver for long and short-term variations in the locations and levels of trapped particles. A feature of the Van Allen belts is the South Atlantic Anomaly (SAA). The 11° angle between the magnetic and geographic axes and the offset of the geographic and geomagnetic centers of the Earth causes a depression in the magnetic field in the South Atlantic. This magnetic field sink causes charged particles to be trapped at low altitudes (<1000 km) in that region thereby forming the SAA. The trapped particles constitute a major radiation hazard for spacecraft. Radiation effects include total ionizing dose (TID), displacements damage, single events effects (SEEs), and deep dielectric charging. In deep dielectric charging, high-energy trapped electrons can penetrate dielectric materials on a spacecraft and discharge, causing damage to spacecraft circuits and materials. This damage can result in performance degradation or the loss of a mission. Total ionizing dose is a cumulative effect which causes degradation of microelectronics and materials. As TID accumulates, component performance can be

21 driven outside of their design range. TID is caused by exposure to electrons and protons. Figure 15 is a plot of total ionizing dose in krads of silicon as a function of aluminium shield thickness for various orbits around the Earth. The two curves in the lower half of the graph are for LEOs that pass through the SAA. The curves that are higher on the graph are orbits that pass through more intense regions of radiation that are at higher altitudes in the belts. The doses at > 300 mils of shielding are dominated by the highly energetic trapped protons.

Figure 15. Total ionizing dose-depth curves for various orbits around the Earth

Total non-ionizing dose (also known as displacement damage or bulk damage) is another cumulative effect that causes degradation of solar cells, optocouplers, and focal plane arrays. As particles slow down in material and come to rest they knock atoms out of their lattice location creating defects which increase the resistance of the device. Electrons, protons, and neutrons cause displacement damage, and the energy spectra of the particles are used to evaluate the level of the hazard. Modern microelectronic systems are plagued by the effects of single particle strikes, namely SEEs, on sensitive regions of devices. There are several types of SEEs, including single event upsets, single event latchups, and single event transients. The consequences of SEEs in systems range from loss of data to the loss of a mission. SEEs are caused by ions from GCRs, solar particle events, and trapped protons. Figure 16 shows the geographic location of single event upsets on the SEASTAR satellite (98°

22 inclination, 705 km altitude) flight data recorder on a world map of latitude versus longitude. Trapped protons in the SAA cause the concentration of upsets near South America. In fact, the location of the SAA protons is clearly mapped out by the upsets. The upsets that occur in the polar regions are due to galactic cosmic rays and solar particles which, at high latitudes, have access to low altitudes due to the open magnetic field at the poles.

Figure 16. Upsets on the SEASTAR flight data recorder at 705 km altitude clearly show the location of trapped protons in the South Atlantic Anomaly

2.8 MAGNETOSPHERES OF OTHER PLANETS The minimum requirement for the existence of a planetary radiation belt is that the planet’s dipole magnetic moment must be sufficiently great such that the flow of the solar wind is arrested before the particles reach the top of the atmosphere where the particles will lose their energy due to collisions. The magnetic fields of some of the other planets are similar the Earth’s, however, they vary in strength. Figure 17 shows a schematic of the relative size of the planetary magnetospheres. Table 6 gives the dipole moments in nanotesta for each of the planets.

23

Figure 17. Relative size of planetary magnetospheres TABLE 6. Dipole moment for the planets

PLANET Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto

DIPOLE MOMENT (NT) 330 0 30,760 0 428,000 21,000 22,800 14,200 0?

Table 6 shows that Venus, Mars, and possibly Pluto do not have magnetospheres and, therefore, cannot support particle trapping. Mercury has a weak magnetic field so it is expected it has a trapped particle population proportionally lower than that of the Earth. The Probos probe showed that Mars has a radiation environment, however, it is due to the thin atmosphere of Mars, which allows interplanetary GCRs and solar particles to penetrate to the surface. Interaction of these particles with the atmosphere produces neutrons, which penetrate to the planetary surface and then reflect back. Saturn, Uranus, and Neptune have magnetic fields with similar strength to that of the Earth but measurements indicate that the intensities of the trapped radiation

24 environments of Saturn, and Uranus are much lower than the Earth’s and do not pose a threat to spacecraft systems. Jupiter’s enormous magnetic dipole (Table 6) can support an intense particle environment. In fact, its magnetosphere is the largest object in the solar system. Measurements have shown that the radiation environment is considerably more intense than the Earth’s and is more extensive, therefore, mission planning for spacecraft that will spend time in trapping regions of Jupiter must include careful definitions of the radiation environment. For example, the electrons at Jupiter have energies of > 100 MeV whereas those at the Earth are in the 10s of MeV. Accurate dose calculations require a model that can transport high energy electrons through shielding. Figure 15 shows that the expected dose for a Europa mission is at the megarads level for 100 mils of shielding which is higher than a one year obit around the Earth in the intense MEO regions. Single event effects are also a problem at Jupiter. In addition to the protons trapped in Jupiter’s magnetosphere, single event effects calculations must include oxygen and sulfur ions injected by volcanic activity on Io. 2.9 INTERPLANETARY PARTICLES The sun dominates interplanetary space. Its magnetized plasma, the solar wind, distorts the magnetic field of Earth (see Figure 12) and even the outer planets. In addition to the solar wind plasma, interplanetary space contains high-energy charged particles. This radiation environment consists of galactic cosmic ray particles that are present at all times and particles from solar events that occur sporadically (coronal mass ejections and flares). GCRs cause single event effects on microelectronics on interplanetary missions and on Earth orbiting satellites that spend time over the poles (see Figure 16). Solar protons are both a single event hazard and cause degradation of detectors, microelectronics, and solar cells. The heavy ions from solar particle events can increase the rate of single event effects many factors above the background caused by GCRs. 2.9.1 Galactic Cosmic Rays In the early 1900s, scientists found that instruments used for studying x-rays and radioactivity measured a background source of unidentified radiation. Victor Hess, an Austrian physicist, measured gamma rays by designing ionization chambers and flying them on balloons. With his balloon experiments, he discovered an extremely penetrating radiation that increased in density as altitude increased. From his experiments, he concluded that this radiation was from an extraterrestrial source. Later, Jacob Clay was able to show that cosmic rays were the source of the on-ground radiation and that measured by Hess higher in the atmosphere. In 1936, Hess received the Noble Prize for the discovery of galactic cosmic rays. Although we now know that these “rays’ are really particles, they are still referred to as cosmic rays. The GCRs originate outside of the solar system. Although there are plausible models of how they are produced, their origin is still a matter of debate[10]. [x]Scientists believe that they propagate through all space that is unoccupied by dense matter. They are essentially isotropic outside of regions of space that are dominated by particles and fields of the sun. Galactic radiation consists of ions of all elements of the periodic table

25 and is composed of about 83% protons, 13% alphas (4He ions), 3% electrons, and about 1% heavier nuclei. Unlike the charged particles that originate at the Sun, the GCRs do not have a characteristic energy limit. Their energies range from 10s of MeV/n to 100s of GeV/n. Because they must pass through about 7 g/cm2 of interstellar gas, the GCRs of even the heaviest ions are probably fully ionized [11]. [xi] A second source of galactic particles is the so-called “anomalous component”. It is composed of helium and heavier ions with energies greater than 50 MeV/nucleon. It is believed that the anomalous component originates in the neutral interstellar gas that diffuses into the heliosphere, becomes singly ionized by solar radiation or charge exchange, and is then connected by the solar wind to the outer heliosphere. The ions are then accelerated and propagate to Earth. The anomalous component is seen only during solar minimum and the details vary from solar minimum to solar minimum. There is growing evidence that the anomalous component is singly ionized, therefore, the ions have greater ability to penetrate the magnetosphere. Our knowledge of the abundances of galactic cosmic rays comes from spacecraft and balloon experiments that have been conducted over a forty-year period. Figure 18 from Medwaldt [12] gives the abundances of the heavy ions at an energy of 2 GeV/n as a function of particle nuclear charge z. The values are normalized to silicon = 106. Note that the relative flux intensities vary by several orders of magnitude. The relative abundances are roughly proportional to the distribution in solar system material. Significant differences are discussed in Medwaldt who also gives a table of relative abundances. xii

]

Figure 18. Relative abundances of galactic cosmic ray ions in interplanetary space. after Medwaldt

The galactic particles are always present, however, their intensities rise and fall with the solar cycle variations. The sun modulates a set of local interstellar spectra at the outer boundary of the heliosphere.[13][ The modulation can be defined by a single parameter which is a function of distance from the sun, the speed of the radial solar wind, and a radial transport particle diffusion coefficient. GCRs are at their peak level during solar minimum and at their lowest level during solar maximum and we now xiii

]

26 know that the length of the GCR modulation cycle is 22 years and not 11 years as previously thought. The difference between the extremes of the solar minimum and maximum fluence levels is approximately a factor of 2 to 10 depending on the ion energy. Figure 19 shows the slow, long-term

Figure 19. IMP-8 measurements of interplanetary ions from the C-N-O group. Note the solar particle event spikes superimposed on the lower level, slowly varying galactic cosmic rays. after Nakamura

cyclic variation of the cosmic ray (C, N, O) fluences for a 20-year period as measured by the IMP-8 spacecraft. The sharp spikes superimposed on the cosmic ray background are caused by solar events. Measurements from Pioneer and Voyager show that the composition of cosmic rays is weakly dependent on the distance from the Sun. The radial gradient from 0.3 to 40 AU is < 10% per AU. For the anomalous component, the gradient increases to 15% per AU. During solar maximum there is 0% gradient out to 30 AU. Latitude gradients have also been studied and found to be 0.5% per degree and 3-6% per degree for the anomalous component [14].[xiv] The Earth’s magnetic field provides some protection from the galactic particles by deflecting the particles as they impinge upon the magnetosphere. The penetration power of these particles is a function of the particle’s energy and ionization state. The exposure of a spacecraft primarily depends on the inclination and, secondarily, the altitude of the trajectory. Cosmic rays have free access over the polar regions where field lines are open to interplanetary space (see Figure 16). 2.9.2 Solar Particles The sun is always active but it has been observed that there is a definite periodicity to the level of activity. Thus, the solar cycle is divided into minimum and maximum phases (see Section 2.1). During the maximum phase of the solar cycle, the numbers and intensity of coronal mass ejections and solar flares increases. This causes periodic increases in the levels of interplanetary particles up to orders of magnitude over the GCR environment. These particles consist of ions of all elements but protons usually dominate the abundances. As with the GCRs, spacecraft receive some protection from solar particles by the Earth’s magnetosphere depending on their orbit. Analysis of the exposure of spacecraft orbiting the Earth as a function of the geomagnetic disturbances

27 that are often associated with solar events is especially critical. For example, CRRES data showed that solar protons reached L shell values as low as 2 [15].[ Also, unlike galactic heavy ions, which are, for the most part, fully ionized, solar heavy ions are more often singly ionized because they pass through less matter before reaching the Earth. This must be taken into account when calculating the degree of penetration of the solar particles into the magnetosphere. Solar particles diffuse as they move through the interplanetary medium. Therefore, the solar particles abundances are a function of radial distance from the Sun. There are few accurate measurements of solar particles throughout the Solar System so the models of the particle diffusion are not accurate. Table 7 gives estimates of the scaling factors that are commonly used to calculate solar particle levels at distances other than the Earth. xv]

TABLE 7. Scaling factors for solar particles levels at regions ≠ 1 AU

RADIAL DISTANCE FROM THE SUN (AU) < 1 AU > 1 AU

RADIAL SCALING FACTOR RANGE 1/R2 to 1/R3 1/R to 1/R2

Solar Protons. Figure 20 shows the particle counts for E >10 and >30 MeV for some of the larger solar proton events for solar cycles 20, 21, and 22 (measured by the GOES spacecraft). Superimposed on the solar event data is the number of sunspots. Note that although the number of proton events is greatly reduced during solar minimum, they still can and do occur. Also, the figure shows that the peak of proton event activity for each solar cycle usually does not correspond to the peak sunspot number. Solar cycles vary in severity in terms of solar proton events. For example, in cycle 21 there were no proton events as large as the August 1972 event of cycle 20 whereas there were at least six events in cycle 22 where the intensity exceeded 109 protons/cm2 for energies greater than 30 MeV. The energies of the solar event protons may reach a few hundred MeV. The duration of the events is from several hours to a few days.

28

Figure 20. Large solar proton events for solar cycles 20, 21, and 22. The number of sunspots is superimposed on the graph.

Solar Ions – Helium and Higher. Some solar particle events are heavy ion rich with energies ranging from 10s of MeV/n to 100s of GeV/n. For the 26 events observed on CRRES [13] , the peak fluxes for the helium ions with energies E > 40 MeV/n were three times higher than the galactic cosmic ray heavy ion levels. Above energies of a few hundred MeV/n to approximately 1000 MeV/n (depending on the element), the solar helium levels merge with those of the galactic cosmic ray background. Early attempts to characterize the solar heavy ions at higher energies were restricted by a limited dataset. Later more space data became available. Dietrich et al[16] used data from the University of Chicago’s Cosmic Ray Telescope on the IMP-8 and GOES satellites to study the heavy ion events. They analyzed high energy spectra for C, O, and Fe using direct measurements and determined fluences in one or two energy bins for N, Ne, Mg, Si, S, Ar, and Ca. Also, He fluences were studied using carbon indices. This dataset provides the most comprehensive picture of high-energy solar heavy ions to date. [xiii]

[xvi]

3. 0 Summary There are many unknowns in space environments and the interaction mechanisms. The low level of funding in the applied science discipline of spacecraft environments and effects has resulted in model development and validation lagging behind rapid technology changes. Access to space for validating effects models and ground test

29 protocols is also critically low. Ground tests cannot duplicate the space environment, particularly when environments have synergistic effects and effects are complicated by enhanced low dose rates in space. Often these unknowns require that large design margins be applied to performance predications, resulting in overheads that reduce capability or that can preclude use of newer technologies in spacecraft systems. 4. 0 References

1. A. LaBel, A. H. Johnston, J. L. Barth, R. A. Reed, and C. E. Barnes, “Emerging Radiation Hardness Assurance (RHA) issues: A NASA approach for space flight programs,” IEEE Trans. on Nucl. Science, Vol. 45, No. 2, December 1998 2. D. Knison, “Radiation effects in the New Millennium - Old Realities and New Issues, Section V: Achieving Reliable, Affordable Systems,” 1998 IEEE Nuclear and Space Radiation Effects Conference Short Course, July 20, 1998, Newport Beach, CA. 3. J. L. Barth, “Applying Computer Simulation Tools to Radiation Effects Problems, Section II: Modeling Space Radiation Environments,” 1997 IEEE Nuclear and Space Radiation Effects Conference Short Course, July 21, 1997, Snowmass Village, CO. 4. A. Holmes-Siedle and L. Adams, Handbook of Radiation Effects Second Edition, Oxford Press, 2002. 5. J. M. Nash, “Cosmic Storms Coming,”, TIME, pp. 54-55, September 9, 1996. 6. E. G. Stassinopoulos, G. J. Brucker, D. W. Nakamura, C. A. Stauffer, G. B. Gee, and J. L Barth, “Solar Flare Proton Evaluation at Geostationary Orbits for Engineering Applications,” IEEE Trans. on Nucl. Science, Vol. 43, No. 2, pp. 369-382, April 1996. 7. J. Feynman, T. P. Armstrong. L. Dao-Gibner, and S. Silverman, “New Interplanetary Proton Fluence Model,” J. Spacecraft, Vol. 27, No. 24, pp 403-410, July-August 1990. 8. D. V. Reames, “Solar Energetic Particles: A Paradigm Shift,” Revs. Geophys. (Suppl.), 33, 585, 1995. 9. S. L. Koontz, M. Pedley, R. R. Mikatarian, J. Golden, P. Boeder, J. Kern, H. Barsamian, J. I. Minow, R. L. Altstatt, M. J. Lorenz, B. Mayeaux, J. Alred, C. Soares, E. Christiansen, T. Schneider, and D. Edwards,” Materials Interactions with the Space Environment: International Space Station - May 2000 to May 2002,” these proceedings. 10. J. W. Cronin, T. K. Gaisser, and S. P. Swordy, “Cosmic Rays at the Energy Frontier,” Scientific American, January 1997. 11. J. H. Adams, Jr., R. Silberberg, and C. H. Tsao, “Cosmic Ray Effects of Microelectronics, Part I: The Near-Earth Particle Environment,” NRL Memorandum Report 4506, August 25, 1981. 12. R. A. Medwadlt, “Elemental Composition and Energy Spectra of Galactic Cosmic Rays,” Proc. from Conference on Interplanetary Particle Environment, JPL Publication 88-28, pp. 121-132, JPL, Pasadena, CA, April 15, 1988. 13. D. L. Chenette, J. Chen, E. Clayton, T. G. Guzik, J. P. Wefel, M. Garcia-Muñoz, C. Lapote, K. R. Pyle, K. P. Ray, E. G. Mullen, and D. A. Hardy, “The CRRES/SPACERAD Heavy Ion Model of the Environment (CHIME) for Cosmic Ray and Solar Particle Effects on Electronic and Biological Systems in Space,” IEEE Trans. on Nucl. Science, Vol. 41, No. 6, pp. 2332-2339 December 1994. 14. R. B. McKibben, “Gradients of Galactic Cosmic Rays and Anomalous Components,” Proc. from Conference on Interplanetary Particle Environment, JPL Publication 88-28, pp. 135-148, April 15, 1988. 15. M. S. Gussenhoven, E. G. Mullen, and D. H. Brautigam, “Improved Understanding of the Earth’s Radiation Belts from the CRRES Satellite,” IEEE Trans. on Nucl. Science, Vol. 43, No. 2, April 1996. 16. W. F. Dietrich, A. J. Tylka, and P. R. Boberg, “Probability Distributions of High-Energy Solar-HeavyIon Fluxes from IMP-8: 1973-1996,” to be presented at IEEE/NSREC Conference, 1997.

This page intentionally left blank

MATERIALS INTERACTIONS WITH THE SPACE ENVIRONMENT: INTERNATIONAL SPACE STATION - MAY 2000 TO MAY 2002 STEVEN L. KOONTZ(1)*, MICHAEL PEDLEY(1), RONALD R. MIKATARIAN(2), JOHN GOLDEN(2), PAUL BOEDER(2) , JOHN KERN(3), HAGOP BARSAMIAN(2), JOSEPH I. MINOW(6), RICHARD L. ALTSTATT(6), MARY J. LORENZ(2), BRIAN MAYEAUX(1), JOHN ALRED(2), CARLOS SOARES(2), ERIC CHRISTIANSEN(4), TODD SCHNEIDER(5), DAVE EDWARDS(5)

Abstract The set of materials interactions with the space flight environment that have produced the largest impacts on the verification and acceptance of flight hardware and on flight operations of the International Space Station (ISS) Program during the May 2000 to May 2002 time frame are described in this paper. In-flight data, flight crew observations, and the results of ground-based test and analysis directly supporting programmatic and operational decision-making are reported. 1.0 Introduction Orbital inclination (51.6o) and altitude (nominally between 350 km and 460 km) determine the set of natural environment factors affecting the functional life of materials and systems on ISS. ISS operates in the F2 region of Earth’s ionosphere in well-defined fluxes of atomic oxygen, other ionospheric plasma species, solar UV, VUV, and x-ray radiation as well as galactic cosmic rays, trapped radiation, and solar cosmic rays [1,2]. The micrometeoroid and orbital debris environment is an important determinant of spacecraft design and operations in any orbital inclination. The magnitude of several environmental factors varies dramatically with latitude and longitude as ISS orbits the Earth [1,2]. The high latitude orbital environment also exposes ISS to higher fluences of trapped energetic electrons, auroral electrons, solar cosmic rays, and galactic cosmic rays [3-6] than would be the case in lower inclination orbits, largely as a result of the overall shape and magnitude of the geomagnetic field [1-6]. As a result, exposure of ISS to many environmental factors can vary dramatically along a particular orbital ground track, and from one ground track to the next, during any 24-hour period. The induced environment results from ISS interactions with the natural environment as well as environmental factors produced by ISS itself and visiting vehicles. Examples include ram-wake effects, hypergolic thruster plume impingement, materials outgassing, venting and dumping of fluids, and specific photovoltaic (PV) power system interactions with the ionospheric plasma [7-16]. An induced ionizing radiation

31

32 environment is produced by cosmic ray interaction with the relatively thick ISS structure and shielding materials [6]. Vehicle size (L) and velocity (v), combined with the magnitude and direction of the geomagnetic field (B) produce operationally significant magnetic induction voltages (vxB.L) in ISS conducting structure during flight through high latitudes (>+ 45o) during each orbit [15,17,19]. In addition, ISS is a large vehicle and produces a deep wake structure from which both ionospheric plasma and neutrals species are largely excluded [7, 19, 21]. Finally, ISS must fly in a very limited number of approved flight attitudes [22], so that exposure of a particular material or system to environmental factors depends upon: 1) location on ISS, 2) ISS flight configuration, 3) ISS flight attitude, 4) variation of solar exposure (ȕ angle) with time, and 5) levels of solar and geomagnetic activity (space weather). The specific spacecraft-environment interactions that have had the greatest impact on ISS Program activities during the first two years of flight are: 1) spacecraft charging, 2) micrometeoroids and orbital debris impacts, 3) ionizing radiation, 4) hypergolic engine plume impingement, 5) venting/dumping of liquids, and 6) external contamination, atomic oxygen, solar ultraviolet effects. 2.0 Spacecraft Charging Phenomena 2.1.HIGH-VOLTAGE PHOTOVOLTAIC ARRAY DRIVEN CHARGING The relatively high plasma density, low plasma temperature, and high electrical conductivity characteristic of the F2 region ionospheric plasma preclude many of the spacecraft charging processes that are observed in lower density plasma environments [2,7]. Surprisingly, the most important identified spacecraft charging process for ISS requires a high-density, low-temperature plasma environment. An electrical interaction between the F2 plasma and the 160-V US PV arrays can produce an electrical potential difference between the conducting structure of ISS and the ambient plasma (i.e. a floating potential or FP) much greater than that usually observed for spacecraft in lowEarth orbit (LEO), most of which have 28-V PV array power systems [7-12]. Sky Lab, which employed 90-V PV arrays, is an important exception to be discussed below. As is shown below, ISS conducting structure becomes negatively charged with respect to the ambient plasma because the PV arrays and electrical power system utilize a negative-polarity grounding scheme, and the common ground point is ISS conducting structure. The severity of possible charging hazards is largely determined by materials interactions with the F2 plasma environment [7-12]. Spacecraft charging interactions lead to the application of electrostatic fields across the dielectrics that separate conducting structure from the ambient F2 plasma. The magnitude of the field gradient can be large enough to cause dielectric breakdown and arcing [7-12]. Degradation of some thermal control coatings, electrical system noise, and shock hazards to extra-vehicular activity (EVA) crew may result [7-12] if the FP is sufficiently negative. The following simple calculation is aimed at explaining the PV array driven charging process, while highlighting the important role of materialsenvironment interactions in both the charging process and the subsequent analysis of possible hazards. The physical basis of PV array driven spacecraft charging lies in the fact that the ions and electrons in the F2 plasma have nearly the same gas kinetic temperature and,

33 therefore, nearly the same kinetic energy. Because the electrons are much less massive than the ions (mi >> me), the mean gas-kinetic speed for electrons, ve= ¥(8kTe/2ʌme), is much larger than the mean gas kinetic speed for the ions, vi =¥(8kTi/2ʌmi). Therefore, the flux of electrons (electron current, I-) to any surface is much greater than the flux of ions (ion current, I+) until a steady state negative FP is established such that I+ + I- = 0. Each of the 400 photovoltaic cells in one string of the US PV array produces about 0.4 V in sunlight, yielding a total linear ǻV of 160 V from one end of the string to the other. There are 82 strings per PV array wing. In the real PV array, the string is mounted on one side of an insulating flat plate of length L (same as the string length). The plate is flying through the ionosphere at orbital velocity with the PV array string facing forward (ram orientation). The FP can be calculated point by point along the string given ǻV, orbital velocity (vISS), the electron and ion densities (Ne and Ni) and the corresponding gas kinetic temperatures, Te and Ti, in the F2 plasma. Geomagnetic field effects on current collection from the F2 plasma are small enough to neglect for this analysis. Orbit limited current collection, electrostatic focusing effects [1,2], and detailed PV array lay-out are also neglected, for the sake of simplicity, even though the subject effects are large and lead to smaller measured values of ISS FP than were predicted by simple early treatments [9]. The thermal velocity of the plasma ions, vi, is much less than orbital velocity of the spacecraft, vISS, so that only ram ion collection is considered. In contrast, the thermal speed of the plasma electrons is much greater than orbital velocity so that electron collection is by gas-kinetic diffusion to the Debye sheath [1,2] and then to the collecting surface. The faster electrons cannot catch up with the spacecraft from behind because separation from the slower moving ions in the wake region creates an opposing electric field (ambipolar diffusion [1,2]. As a result, simply turning the PV array strings to wake and exposing only the insulating plate to ram completely suppresses PV array driven charging, a prediction confirmed by in-flight ISS floating potential and plasma contactor emission current measurements made during 2001 [15,16]. Finally, the magnitude of the charging depends on the PV array voltage. When ǻV is zero, at night or when the PV strings are shunted, there is no charging. At steady state, plasma ion current to the string must equal plasma electron current to the string. Electron thermal current = I- = I+ = Ion ram current The positively biased end (area Ae, length Le) of the string collects electrons and the negatively biased end (area Ai, length Li) collects ions. Ae + Ai = A, and Le + Li = L, where A is the total exposed conducting area on the PV cell string of length L. I- = 0.25veNeqAe = vISSNiqAi = I+. The ionosphere is a neutral plasma so, Ne = Ni. The mean gas-kinetic speed of the electrons, ve, is multiplied by 0.25 to obtain the correct expression for thermal particle flux to a wall, and q is the value of the elementary charge leading to, 0.25veAe = viAi or Ae/Ai = vISS/0.25ve

34 Assuming a typical daytime F2 region plasma temperature of 0.1 eV (1,160o K) we have vi = 1.3 km/sec, vISS = 7.69km/sec, and ve =163 km/sec so that, Ae/Ai = vISS/0.25ve = (7.69)/(0.25x163) = 0.19. Since PV string voltage is a linear function of distance from the positive end, and ǻV = 160 V, FP can now be calculated as a function of distance from the positive end of the string, where FP(0) corresponds to L=0. FP(0) = ǻV(Ae)/(Ai+Ae) = ǻV(Le)/(Li+Le) = 160 x 0.1597 = +26 volts FP(L) = FP(0) – ǻV = -134 V If the negative end of the string is grounded to a spherical conducting structure that is 10 meters in radius (a reasonable size compared to ISS pressurized elements), the free space capacitance (Cfs = 4ʌİİor = 1112 pF; İ = dielectric constant, İo= free space permittivity) of the structure is charged to –134 volts giving a stored energy of only E = 0.5CV2 = 10 micro Joules. The sphere with a thin dielectric surface coating changes the character of the charging hazard dramatically. On the ram facing side of the sphere, the FP of the external surface of the dielectric film will approach 0 V as a result of positive charge collection from the ionosphere, and –134 V is applied across the dielectric. Now the sphere is best described as a parallel plate capacitor (the conducting structure is one plate and the conducting ionosphere is the other) able to store energy E = 0.5 CV2 = 0.5 İİo(Aram/d)V2 = (0.5)(8.85 x 10-12)İ(2ʌr2/d) V2, where Aram is the area of the hemisphere able to collect positive charge from the ionosphere. If d is 1 micron (1.3 microns is the thickness of the anodic coating on the US Lab and Node 1 meteoroid and debris shields) and İ = 5 for aluminum oxide, the stored energy becomes E = 250 Joules. Now, dielectric breakdown of the thin surface coating can discharge the parallel plate capacitor, releasing enough energy to damage the dielectric coating itself and producing enough voltage and current to present a possibly lethal hazard to any EVA crew in the discharge circuit. The high-density, low-resistance dielectric-breakdown arc plasma provides the conductive path connecting the negatively charged conducting structure to the positively charged dielectric film surface [8-10]. Note that the stored energy is inversely proportional to the dielectric film thickness. Simply increasing the film thickness from 1.0 micron to 100 microns reduces the stored energy from 249.5 Joules to 2.49 Joules while greatly reducing the risk of dielectric breakdown arcing. The thick (>120 microns) dielectric coatings on Sky Lab minimized any charging hazards that might have been generated by the 90-V PV array on that spacecraft. Similarly, the Russian elements of ISS contribute little to the charging hazard because surface dielectric coatings are thick. Stored energy is also directly proportional to V2, and reducing the FP at the negative end of the PV array to –40V reduces the stored energy to 0.9 micro Joules for the uncoated conducting sphere, and

35 22 Joules for the dielectric coated sphere. As discussed below, flight measurement and analysis of US Lab and Node 1 FP, with all FP controls disabled and PV array driven charging enabled, have not exceeded –28 volts during 2001 [16]. Plasma chamber testing (7-12) has shown that the dielectric breakdown voltage for the 1.3-micron thick anodic film on the US Lab and Node 1 meteoroid and debris shields is greater than 60 volts. Therefore, the plasma contactor system has not been in continuous operation since May 2001. The ISS FP not-to-exceed-limit for EVA safety is –40 V, however and two PCUs are operated routinely during EVA. A negative FP of –134 V is remarkably close to the predictions made before the US PV arrays were flown for the first time on ISS [7-10]. Using early charging models, a worst-case FP of –140 volts was predicted. The measured FP from PV array driven charging on ISS have been less negative than –28 V in all cases observed to date. The simple charging calculation presented above as well as the more elaborate pre-flight theoretical models consider only current collection by the PV array string. Ion collection by exposed conducting structure attached to the negative end of the PV array string can offset the effects of electron collection by the string, driving FP(0) toward +160V and FP(L) toward 0 volts, but only if the number of milliamps of electron current collected by the PV array is small. The number of square meters of ramoriented ion-collecting surface needed to hold the PV array FP (0) near +150 V and FP(L) near -10 volts is shown as a function of total electron current collected by the PV arrays in Table 1. TABLE 1. Area of Exposed Conducting Materials Compensating PV Array Electron Collection * Ionospheric Electron Current Collected by 160 V PV Arrays (milliamps) 10 30 60 100 * FP (0) =+150 V and FP (L) = -10 V; Ni = Ne =106/cc

Area of Ram Oriented Ion Collection Surface (square meters) 8 24 48 80

When the electron current collected by the PV arrays on an LEO spacecraft is less than about 60 milliamps, exposed ion conducting area connected to the negative ground plane can offer practical FP control. As the collected electron current grows beyond 100 milliamps, the ion collecting area requirements become unrealistic. The plasma contactor system was selected for ISS FP control precisely because the magnitude of the electron current collected from the ionosphere by the 160-V US PV arrays was estimated to be far too large to allow FP control by passive ion collection surfaces [9, 14]. The Russian segment of ISS provides significant ram-oriented conducting surface area (estimated to be greater than 30 square meters) as a result of Russian Program electrical conductivity/grounding requirements for thermal blanket materials. The plasma contactor system on ISS controls the FP by providing a low impedance return path to the ionosphere for electrons collected by the PV arrays or by other collection mechanisms [17]. The ISS telemetry stream provides measurements of electron emission current from the ISS ground plane to the ionosphere whenever the plasma contactor system is operating. Much of the plasma contactor emission current observed over the past 2 years is attributable to low-voltage non-PV-array-driven

36 charging processes [17]. However, direct measurements of PV array driven electron collection can be made by recording the change in emission current when ISS enters sunlight (eclipse exit) with sun-pointing PV arrays or by shunting and un-shunting the sun-pointing PV arrays while in sunlight. Figure 1 shows measured eclipse-exit plasma contactor emission currents since January 2001. The eclipse exit emission currents show considerable variation both during a given 24-hour day and over the last year. The well-known dynamic structure of the F2 region of the ionosphere [1,2] can account much of the variability. Large variations of Ne and Te with time of day, altitude, ISS latitude and longitude, geomagnetic field, solar activity, and season explain much of the observed variability in the eclipse exit emission currents [1,2,18]. Clearly, the magnitude of PV-array-driven charging will vary in a similar way with variation in natural environment. 0.09

0.08

0.07

Current (A)

0.06

One Array Shunted

0.05

0.04

0.03

0.02

0.01

0 1/1/01 1/29/01 2/26/01 3/26/01 4/23/01 5/21/01 6/18/01 7/16/01 8/13/01 9/10/01 10/8/01 11/5/01 12/3/01 12/31/0 1/28/02 2/25/02 1 Date

Figure 1. ISS plasma contactor emission current increase at eclipse exit; Jan 2001 to Feb 2002, US 160 V PV arrays sun tracking and un-shunted

With the plasma contactors operating, most of the PV array wing area is positively biased (+FP) so that electron collection is maximized and any ion collection by conducting structure is already accounted for in the observed emission currents. If the plasma contactors are turned off, the FP at the negative end of the PV array (and ISS conducting structure) moves toward more negative values. As a result, less of the PV array wing area is positively biased leading to reduced electron collection, while ion collection remains constant or increases slightly. At some negative value of the conducting structure FP, PV array electron collection will equal conducting structure ion collection stopping further movement of FP toward more negative values.

37 In the simple flat-plate charging calculation shown above, the reduction in PV array electron collection is a simple linear function of FP at the negative end of the array. In fact, as FP becomes more negative, the decrease in electron collection by the PV arrays is nonlinear as a result of: 1) dielectric surface charging of PV array cell materials, 2) detailed electrostatic focusing effects in the Debye sheath near the gap between PV array cells, and 3) PV array cell structure and lay-out. Small negative changes in FP cause relatively large reductions in electron collection while ion collection remains nearly constant [9b, 9c, 18]. Steady state FP values with the plasma contactor system off are expected to be closer to –30 V, not –134 volts, as has been observed to date [15,16,18]. As development of more accurate and detailed models of the PV array driven charging process continues, it becomes clear that a materials interaction with the ionospheric environment, specifically surface charging of dielectric materials in the photovoltaic cell structure, limits electron collection by the 160 V US PV arrays on ISS and places natural limits on the FP values that can be achieved [18]. The results of in-flight floating potential probe (FPP) [15,16] measurements of ISS FP characterizing both the PV array driven charging process and the contribution of the vxB.L (v = spacecraft velocity, B = geomagnetic field, L = length of conducting structure) magnetic induction voltages, with the plasma contactor system off, are shown in Figure 2 and Tables 2 and 3. Table 2 compares the worst-case pre-flight predictions of PV array driven charging with the worst-case measurement made to data. Measured electron collection by the two 160-V US PV arrays active during the April 2001 time frame is so low that exposed conducting structure can contribute to limiting the negative FP to the small value observed, as suggested above. FPP measurements of ISS FP were made during several days in 2001, including intervals when the Space Shuttle was docked to ISS. On January 31, FPP data measurements of ISS FP were made with active side (the side with PV cell strings) of the active surface of the PV arrays in shallow wake flight attitude verifying that wake orientation of the arrays prevents PV array driven charging. With the plasma contactor system off and PV arrays sun tracking, FPP data was collected on April 10-12, April 15, and on April 21 (before and after Space Shuttle docking). A total of 46 FP measurements characterizing PV array driven charging were made in 2001, encompassing a wide range of ionospheric conditions. Langmuir probe measurements of electron temperature, Te, at eclipse exit ranged from 0.08 to 0.23 eV while electron density, Ne, ranged from 109 to 1012/m3. To date, the observed range of PV array-driven charging FP values range from –4 to –24 V. It should be noted that the FPP could not provide Te or Ne if Fp exceeded –10V negative as a result of the limited sweep range of the Langmuir probe voltage. The April 11 data is fairly typical, despite the geomagnetic storm starting about 13:30 universal time (UT). Figure 2 shows the ISS FP at the FPP measurement point as a function of universal time on April 11, 2001. In Table 3 the total FP for the April 11, 2001 eclipse-exit charging peaks, shown in Figure 2, are broken down into the

38 magnetic induction and PV array driven components for the locations on ISS defined in Figure 3. Magnetic induction voltage is a significant fraction of the total FP in all cases, and must be considered in any ISS charging assessment. As shown in Figure 2, the agreement between calculated magnetic induction voltage and measurement is excellent in all cases. Figure 3 shows a calculated magnetic induction voltage map of ISS when passing south of Australia on April 11, 2001. Flight south of Australia generates the more magnetic induction voltage on ISS than any other ISS flight path. TABLE 2. PV Array Driven Charging - Pre-Flight Estimates vs. Flight Data Charging Hazard Related Quantity Maximum Negative FP 160 V PV Array Electron Collection Exposed Conducting Surfaces on ISS Duration of Max. Neg. FP

Pre ISS Flight 4A: Worst-Case Estimate -140 200 to 500 milliamps 0 m2 20 to 30 min. of day pass

Post Flight 4A: Worst-Case at US Lab Module -26 V 10 to 80 milliamps 15 to 40m2: PV array mast wires & ISS structure <10 min. of day pass

The data shown in Figure 2 and Table 3 span 6 orbits or 9 hours. During that time, the rotation of the Earth changed the geographic location of ISS eclipse exit from near the west coast of South America to Australia. The magnetic induction voltage peaks twice on each orbit, at + 51.6 degrees latitude. Eclipse-exit PV array driven charging peaks are superimposed on the –51.6 latitude magnetic induction peaks. The + 51.6 magnetic induction voltage peak occurs during eclipse. The measured FP consists only of magnetic induction during voltage when ISS is in eclipse or when the PV arrays are shunted or in wake. When sun tracking, the active surface of the PV arrays move into wake at orbital noon. Figure 2 also shows a comparison of magnetic induction voltages calculated using a first principle model [17,18] with the flight data, demonstrating excellent agreement between the magnetic induction model and the flight data. The ISS magnetic induction voltage map shown in Figure 3 was calculated using the model [17, 18]. TABLE 3. Post Eclipse Exit ISS Charging Peaks (maximum negative FP in volts) from Figure 2, April 11, 2001 at GMT time indicated (PCU system off) 12:38 LAB -6.51 -19.6 -26.1

vxB Chg Total

2B 2.034 -19.6 -17.5

FPP 4B -3.42 -9.56 -19.6 -19.6 -23.0 -29.1

vxB Chg Total

17:15 2B LAB FPP 4B 3.429 -9.7 -6.55 -17.2 -7.0 -7.0 -7.0 -7.0 -3.5 -16.7 -13.5 -24.2

14:10 LAB -7.38 -17.0 -24.4

vxB Chg Total

2B 2.571 -17.0 -14.4

FPP 4B -4.19 -11.5 -17.0 -17.0 -21.2 -28.5

vxB Chg Total

18:46 2B LAB FPP 4B 3.775 -10.2 -7.35 -19.1 -4.7 -4.7 -4.7 -4.7 -0.9 -14.9 -12.0 -23.8

15:41 LAB -8.22 -15.2 -23.5

vxB Chg Total

2B 2.812 -15.2 -12.4

FPP 4B -4.96 -13.5 -15.2 -15.2 -20.2 -28.7

vxB Chg Total

20:18 2B LAB FPP 4B 3.903 -10.1 -7.65 -19.7 -5.2 -5.2 -5.2 -5.2 -1.3 -15.3 -12.9 -25.0

Table 3 definitions: vxB=magnetic induction voltage; Chg=PV array driven charging; Total=vxB + Chg; For ISS locations 2B, 4B, Lab and FPP see figure 3

39 25 FPP Vbody

E W B vxB M odel

Potential (-V)

20

15

10

5

0 1 2 :3 0

1 3 :3 0

1 4 :3 0

1 5 :3 0

1 6 :3 0

1 7 :3 0

1 8 :3 0

1 9 :3 0

2 0 :3 0

2 1 :3 0

T im e (G M T )

Figure 2. ISS FP at the FPP measurement point (FPP Vbody) with the plasma contactor system off. Calculated magnetic induction FP (EWB vxB model) is compared with measured FP. April 11, 2001

FPP

2B

4B

v LAB

Figure 3. Calculated worst-case magnetic induction voltage map of ISS in +XVV flight attitude (velocity vector v), with 10 degrees down pitch. April 11, 2001

The ISS Program has established a process for the evaluation and management of spacecraft charging processes on ISS and has established a requirement for control of ISS FP to values less negative than –40V during EVA operations for all EVA work

40 sites and translation paths. The plasma contactor system [11-13] provides the primary control of ISS floating potential. Shunting the PV arrays or orienting the active surfaces of the PV arrays to wake are additional FP control methods verified for the present ISS flight configuration. As the ISS construction continues, six more 160 V US PV array wings will be added along with the complete truss structure and additional Russian, European, and Japanese modules. Adding six PV array wings increases the risk from PV array driven charging and the completed ISS truss is long enough to develop worst-case tip-to-tip magnetic induction voltages of –50 V. In-flight characterization of ISS floating potentials, and local ionospheric environment are essential for verification of a safe EVA environment as construction of ISS continues. A dedicated floating potential measurement unit (FPMU) will be installed on ISS before the next set of 160 V PV arrays is launched. The FPMU consists of a floating potential probe, two Langmuir probes, and a plasma impedance probe. Langmuir probe measurements will be verified against ground based incoherent scatter radar and ionosonde data. Measurement campaigns are planned to fully characterize the ISS charging and EVA shock hazard environment, verify the effectiveness of hazard controls, and verify the detailed ISS charging models [18]. 2.2 HIGH LATITUDE AURORAL ELECTRON CHARGING The possibility of spacecraft charging by auroral electrons at high latitudes, during geomagnetic storms or other geomagnetic disturbances, is a subject of some concern on the part of the spacecraft charging community [20, 22-24]. Analysis of historical satellite charging and anomaly data for the United States Defense Meteorological Satellite Program (DMSP) satellites and the European Space Agency Freja satellite both suggest that auroral charging may be observed on ISS at high magnetic latitudes [25, 26], especially at night during solar minimum. Charging of the Freja and DMSP vehicles has been correlated with ionospheric plasma densities of 104/cm2, or less, combined with fluxes of energetic auroral electrons (7-10 keV) greater than 108 electrons/(cm2sec sr) [23, 26-28]. The DMSP and Freja satellites both orbit the Earth at or above 800 km, in the topside ionosphere, well above ISS operational altitudes. Nonetheless, the required combinations of ionospheric plasma density and energetic electron flux are expected to occur at ISS altitudes, albeit infrequently, at or near the extreme latitudes of the ISS orbit (+ 51.6o). Inspection of the auroral precipitation maps produced hourly by the US National Oceanics and Atmospherics Administration (NOAA) Polar Orbiting Environmental Satellite (POES) constellation show that ISS passes through the precipitating auroral electrons several times every day, whenever ISS passes south of Australia at night and Kp is greater than 3 [29]. The question of whether or not flight through the same kind of environment that produces charging and the occasional recoverable anomaly on DMSP constitute a risk or hazard for ISS or ISS EVA crew remains open. The absence of severe anomalies on the Freja spacecraft in a similar, if not more severe, charging environment highlight the important effect of spacecraft design on spacecraft charging. More detailed assessments of the frequency of occurrence of the auroral charging environment at 350 to 400 km altitude as well as detailed analysis and modeling of the expected ISS and EVA suit charging in that environment are in work at this time. ISS and the EVA suits used on ISS are not identical to DMSP or Freja, when treated as electrical systems interacting

41 with the auroral charging environment. Materials properties and materials interactions with the auroral charging environment will likely determine the outcome of the assessments. Secondary electron yields, dielectric coating thickness compared to energetic electron range, and total area of exposed conducting surfaces are all important factors. The ISS plasma contactor system also contributes to control of any auroral charging risk by both increasing local plasma density and providing a return path to the ionosphere for any charging of ISS grounded structure produced by auroral electrons. During the first two years of flight (during the current solar maximum), no ISS equipment anomalies have been reported that correlate with geomagnetic storms or flight through either the diffuse or visible auroras. The ISS crews have reported flying through visible Aurora Australis on at least two occasions. The following excerpt from Commander William Shepherd’s deck log of Nov. 10, 2000 is an interesting example. “11:30: Transited through a very unusual aurora field. Started as a faint green cloud on the horizon, which grew stronger as we approached. Aurora filled our view field from SM (Service Module) nadir ports as we flew through it. A faint reddish plasma layer was above the green field and topped out higher than our orbital altitude.” Southern hemisphere auroral precipitation maps produced by NOAA POES Satellites 15 (Nov. 10, 2000 10:56 UT) and 16 (Nov. 10, 2000 12:26 UT) show auroral activity levels of 9, hemispheric powers of 60 to 90 gigawatts, with intense auroral electron precipitation over Tasmania and southern New Zealand, so that ISS was well within the precipitating electron environment during the auroral fly through reported by Cmdr. Shepherd [17]. 3.0 Meteoroids and Orbital Debris ISS has weathered reasonably well through 3.5 years of exposure to the meteoroid/orbital debris (M/OD) environments, with no hardware failures reported due to M/OD impact. In the last 2 years, ISS has successfully completed a total of four propulsive maneuvers to avoid large debris objects tracked by NORAD. Propulsive avoidance maneuvers are executed whenever the likelihood of collision with a large debris object exceeds 1 in 10,000. Substantial evidence of hypervelocity impact damage to ISS external surfaces from small meteoroid/debris strikes have been identified in down-linked video and photographs [30], by examining the surfaces of returned ISS hardware such as the Mini-Pressurized Logistics Module (MPLM) (31), as well as direct observations by the crew. For instance, the crew reported 3 impacts (one “thumbnail-sized”) to a DC-DC Converter Unit (DDCU) heat pipe during an EVA on 20 February 2002, and Figure 5 is a digital image of an impact crater (3mm-5mm diameter) on a nadir Service Module window transmitted to the ground in January 2002 [32]. Other M/OD impacts have been evident on MPLM surfaces after flight during routine refurbishment activities at Kennedy space Center (KSC). MPLM pressurized modules have been used 4 times (through Dec.2001) to carry supplies to ISS, with approximately 6 days each mission exposed to the M/OD environment while docked to ISS (the MPLM is much less exposed to M/OD during the remainder of each mission by virtue of being in the payload bay of the Orbiter vehicle). Over the 4 MPLM missions to ISS to-date, 11 hypervelocity impacts have been observed on MPLM

42 exterior surfaces, with 2 of these completely penetrating the outer 0.8mm thick aluminum “bumper” shield of the module (Figure 6). Predictions of perforation rates have been made of the MPLM, using JSC BUMPER code [33] with the NASA standard meteoroid model and either the ORDEM96 debris model [34] or the latest debris model (ORDEM2000), which has just been released [35]. As Table 4 demonstrates, BUMPER predictions with ORDEM2000 match the MPLM bumper perforation history, whereas ORDEM96 under-predicts the damage. MOD strikes on EVA crew are recognized as an important operational hazard. Comparison of EMU penetration probabilities calculated with ORDEM2000 vs. ODRM96 implies a significant increase in the risk associated with EVA activities as shown in Table 5.

TABLE 4. Predictions of MPLM bumper perforation rate (actual perforation rate is 1 every 2 flights for 4 flight history) PREDICTIONS

Risk of MPLM bumper perforation each flight

Frequency of bumper perforations

ORDEM2000 & std. Meteoroids

55%

1 every 1.8 flights

ORDEM96 & std. Meteoroids

19%

1 every 5.2 flights

TABLE 5. Risk of EMU suit penetration by MM/OD particles during EVA

EMU Meteoroid/Debris Risks For single 6hour EVA

meteoroids/ORDEM96

meteoroids/ORDEM2000

PNP

Odds

PNP

Odds

Risk Change

Any size leak

0.999943

1 in 17,544

0.999715

1 in 3,509

5.0 X

Critical leak (>4mm)

0.999983

1 in 58,824

0.999941

1 in 16,807

3.5 X

For 2700 EVA-hours (10years of EVAs)

meteoroids/ORDEM96 PNP

Odds

meteoroids/ORDEM2000 PNP

Odds

Risk Change

Any size leak

0.9747

1 in 39

0.8796

1 in 8

4.8 X

Critical leak (>4mm)

0.9924

1 in 131

0.9736

1 in 38

3.5 X

43

Figure 5. 3-5 mm MOD strike crater on the Service Module nadir view port

Figure 6. MPLM MOD shield penetration observed after ISS flight.

44 4.0 Ionizing Radiation I: EEE Parts and Avionics The ISS electrical power system and avionics suite (including commercial-off-the-shelf or COTS components) are performing well ahead of expectation with respect to the single event effects (SEE) and total ionizing dose (TID) effects produced by the ionizing radiation environment in LEO. No correlations between any electronic anomaly on ISS and: 1) exposure to solar energetic particle events, 2) flight through the South Atlantic Anomaly (SAA), or 3) flight at high latitude have been identified during the first two years of flight. Similarly, no galactic cosmic ray (GCR) effects, as revealed by a geomagnetic latitude dependence of electronic failures, have been identified, however, events that may be produced by high energy/rigidity GCR particles, which would show no latitude/longitude dependence, are the subject of an ongoing investigation. 4.1 THE RADIATION DESIGN ENVIRONMENT The ISS natural radiation design environment is specified in SSP 30512, Space Station Ionizing Radiation Design Environment [36]. For ISS design, we have specified conservative, single design point, ionizing radiation design environments to address Total Ionizing Dose (TID). The ISS TID radiation environment is specified for 500 km at solar maximum and includes both trapped protons and trapped electrons. Other minor contributions to TID, such as x-rays, as well as uncertainties in the trapped radiation models, are addressed through the application of a recommended design margin of 2x applied to the 500 km design environment. The ISS radiation design environment represents a conservative, low cost solution for ISS hardware design and verification. The selection of 500 km as a design point altitude is in itself a worst-case assumption because ISS will normally fly below 500 km. Both TID and SEE rates increase dramatically (about 3X) between 300 and 500 km largely as a result of the altitude structure of the SAA. Two different natural environments for design have been specified to address SEE on ISS [36]. The nominal SEE environment is based on AP8 solar minimum model for trapped protons and a solar minimum model for cosmic rays. The extreme SEE environment includes solar protons and heavy ions emitted during an extreme or worst case SEP event [36]. The nominal SEE environment is specified for a 500 km altitude at solar minimum so as to define a worst-case environment for both trapped protons and GCR [36]. The extreme SEE environment is also specified for 500 km. Major SEP events are relatively rare and are associated with powerful solar x-ray flares and/or coronal mass ejections [37, 38]. The SSP-30512 extreme SEE environment is based on the October 1989 SEP event, which is generally recognized as a 99th percentile worstcase extreme SEE environment with respect to both energetic proton and energetic heavy ion fluxes [39]. SSP-30512 contains no statements about the radiation-shielding environment produced by ISS structure or avionics cases or enclosures. A simple worst-case shielding environment has generally been assumed for predicting worst-case TID and SEE effects. For purposes of design and analysis, SEE and TID susceptible equipment is assumed to be at the center of a 0.5 cm (1.27 cm) thick aluminum sphere if located

45 outside (inside) the pressurized elements corresponding to shielding surface densities 1.35 g/cm2 (3.43 g/cm2), values substantially lower than the actual shielding mass distribution function inside the pressurized elements of ISS, which ranges from 10g/cm2 to 100 g/cm2 as is shown in the following paragraphs. ISS is exposed to the TID and nominal SEE environments on a continual basis and must meet all performance requirements during this exposure. ISS is protected from SEP events by the Earth’s geomagnetic field over most of its orbital flight path, though some fraction of the SEP heavy ions, which are not fully ionized, may penetrate to latitudes where cosmic rays of comparable rigidity are excluded [39]. The most important ISS exposure to extreme SEE particles fluxes happens when the orbital phasing is such that ISS passes over Canada or south of Australia during the SEP event. Electronic parts are susceptible to SEE induced destructive failure only if powered. The ISS program has recently established an extreme SEE caution and warning procedure to enable power down of any nonessential equipment that has a high risk of destructive failure in the extreme SEE environment [40]. 4.2 ON-ORBIT OBSERVATIONS ISS is performing well within expectations with respect to TID degradation and SEE impacts on EEE parts and avionics performance. Until recently, ISS has been flying at altitudes between 350 and 400 km during solar maximum, well below the 500 km specified for the worst-case radiation design environment in SSP 30512. TID accumulated to date is well below the performance degradation threshold for EEE parts. Ionizing radiation dose measurements, made within the habitable volume with thermoluminescent dosimeters and crew personal dosimeters, range from 5 to 12 µ Gy (0.5 to 1.2 milli rads) per hour, depending on location in the habitable volume, corresponding to an annual dose range of 44 to 105 milli Gy (4.4 to 10.5 rads) [41]. The variation in TID with location in the habitable volume is largely a result of variations in effective shielding mass with location [42-44]. No destructive SEE events of any kind have been observed during the first two years of flight. Only one ISS vehicle equipment item fault that may be uniquely attributed to SEE with a reasonable level of confidence has been observed. An S-Band Antenna System anomaly occurred at 18:16 GMT on 26 December 2001 during a moderate SEP event. The anomaly is best explained as the result of a single event upset (SEU) in the command register of the video signal processor. ISS was near latitude 35o south, longitude 157o east, altitude 390 km at 18:16 GMT on 26 December 20001. ISS was well away from the SAA, and well shielded by Earth’s geomagnetic field from both the SEP event and GCRs with energies below 10 GeV/nucleon, when the anomaly occurred. Using SEE test data for the subject device, the mean-time-between-failure (MTBF) analysis, and 2 gm/cm2 Al shielding as input parameters, calculations of the as-flown average radiation environment using CREAM-96 [45], predict an average upset rate of 0.0001 SEU/day. The S-Band Antenna System has been on orbit for 500 days as of June 26, 2002 for an observed failure rate of 0.002 SEU/day, in reasonable agreement with the CREME-96 calculation given the generic uncertainties in the analysis.

46 A study of those IBM Thinkpad 760 XD (ISS Personal Computer System or PCS) anomalies requiring reboot or power cycling that may be attributed to radiation effects, after excluding all other reasonable causes, shows no correlation with exposure to SEP events, flight through the South Atlantic Anomaly (SAA), or flight through high latitude regions, as shown in Figure 7. The GMT time of PCS anomalies are recorded in the PCS boot file and were reported by ISS crew. No destructive PCS events or anomalies have been observed to date. The observed PCS reboot/power cycling rate (reboots that may be attributed to radiation in the absence of any other identified causes) is in reasonable agreement with rates predicted using the SSP 30512 nominal SEE design environment combined with 0.129 GeV proton beam screening tests [46] and a worst-case mean time between failure (MTBF) analysis with less than 1 g/cm2 shielding, as shown in Table 6. However, the latitude, longitude, and SEP exposure anomaly dependencies predicted are not observed. The IBM 760 XD Thinkpads are commercial-off-the-shelf (COTS) items. One IBM 760 XD from the procurement lot was tested for SEE susceptibility using a proton beam test [46] on the assembled item. Proton testing showed lockups (requiring reboot or power cycling) only for the CPU, video board, and CD-ROM, in order of decreasing susceptibility. 90 No SPE in Progress >100 MeV Proton SPE in Progress

60

latitude

30

0

-30

-60

-90 -180

-150

W

-120

-90

-60

-30

0

30

60

90

120

150

180

E

longitude

Figure 7. PCS anomalies as a function of ISS latitude and longitude

TABLE 6. Predicted Radiation Induced Anomaly* Rate vs. In-Flight Anomaly Rate for Three IBM Thinkpad Laptop 760 XD Computers (PCS) on ISS Laptop

Predicted Reboots/Day (radiation)

Observed Reboots/Day (radiation)

Service Module PCS 0.04 0.02 Lab Robotics Work 0.04 0.01 Station PCS Lab PCS 0.04 0.04 All 0.13 0.08 * Anomalies requiring reboot or power cycling for recovery and not attributable to causes other than SEE causes

47 Figure 7 shows the timing of the laptop anomalies compared to the timing of SEP events observed by the NOAA GOES satellite constellation. Only 4 of the 22 ISS laptop anomalies were observed during periods of enhanced solar particle flux at GOES. However, ISS was not positioned to receive solar proton flux during three of these events (April 10, August 18 and September 26). For the fourth event, which occurred on August 16, the location of ISS at the time of the anomaly cannot be determined (no anomaly time tag). So, no laptop anomalies requiring reboot or power cycling can be uniquely associated with SEP event exposure while ground based proton testing combined with the SSP 30512 extreme SEE design environment predicts several reboots or power cycles per event day with SEP event exposure.

Daily Flux of >100 MeV Solar Protons (p/cm^2-sr-day at Geosynchronous Orbit)

1.E+7 Observed PCS Anomalies

>100 MeV Daily Solar Proton Flux

1.E+6

1.E+5

1.E+4

1.E+3 3/1

4/1

5/2

6/2

7/3

8/3

9/3

10/4

11/4

12/5

Date

Figure 8. IBM Thinkpad (PCS) Anomaly Data vs. Solar Particle Events for 2001 Figure 1 data sources: >100 MeV Daily Solar Proton Flux - Space Physics Interactive Resources (SPIDR) webpage (URL= http://spidr.ngdc.noaa.gov/ ). Observed PCS anomalies – ISS Mission Action Requests (CHITS) database

A review of the US Lab and Node 1 flight equipment nominal SEE susceptibility data indicates that the observed frequency of equipment functional interrupts (i.e. those requiring outside intervention and possibly caused by radiation), is at least an order of magnitude less than the single event rates calculated for the SSP 30512 design environment. As of January 2002, ISS has been exposed to three major SEP events as shown in Figures 9-11. ISS was exposed to SEP event particles only while at the high latitudes during the exposure periods shown in Figures 9-11.

48

Figure 9-11 Source Data: 1) Geosynchronous particle flux - Space Physics Interactive Resources (SPIDR) webpage (URL= http://spidr.ngdc.noaa.gov/ ). 2) ISS SPE exposure calculated with SPERT ( M.J. Golightly, M. J., Weyland M. D., Lin, T.; “Solar Particle Event-Real-Time (SPERT): A Real-Time Solar Particle Event Exposure Analysis System.” ESA Workshop on Space Weather, ESTEC, Noordwijk, The Netherlands, 11-13 Nov 1998.)

49 Each of these events produced a total fluence of >100 MeV protons greater than 1 x 108 protons/cm2 at geosynchronous orbit. For comparison the ISS extreme SEE design environment proton fluence at >100 MeV is 1 x 109 protons/cm2. No ISS equipment anomalies have been tied to any of these major solar particle events. There were not very many SEE susceptible items on ISS during the 2000 events, but a fully outfitted US Lab Module was present during the November 2001 event. Figures 9-11 show >100 MeV particle flux levels during these events, as well as the times ISS was exposed to the enhanced particle environment, when the orbital flight path carries ISS through high latitude (greater than + 45o) regions where geomagnetic shielding is relatively poor (e.g. Canada and Australia/Tasmania). The performance of the static and dynamic random access memories (SRAM and DRAM) in the ISS data system multiplexer-demultiplexer (MDM) units is of special interest in light of the critical nature of MDM function in ISS command and control systems. Several memory chips in the ISS MDM contain bit flip counters for error rate measurements. During 2000 and 2001 only multiple bit error data were downloaded from ISS. A software upgrade is being developed to enable downloading the single bit flip data in the near future. The predicted performance of the MDM SRAM and DRAM devices in the SSP-30512 design environment is compared with the observed performance (multiple bit error data only) in Table 7. TABLE 7. MDM RAM Multiple Bit Error Rates RAM Device

1Mx4 DRAM

Predicted Nominal SEE multiple bit error rate 4.4 x 10-8/day

Observed Nominal SEE multiple bit error rate 0/730 days

4Mx4 DRAM

7.5 x 10-7/day

0/730 days

32Kx8 SRAM

1.1 x 10-6/day

0730 days

Predicted Extreme SEE multiple bit error rate 0.00034 per SEP event 0.013 per SEP event 0.0021 per SEP event

Observed Extreme SEE multiple bit error rate 0/2 SEP events 0/2 SEP events 0/2 SEP events

Application of a worst-case design environment, combined with rigorous EEE parts screening and MTBF analysis at the parts, assembly, and system levels, has produced an ISS electrical power and avionics system that is largely immune to the effects of ionizing radiation. The PCS and S-Band antenna systems are performing well ahead of expectations based on pre-flight test and analysis, verifying the validity of the test and analysis methods used as well as the SSP-30512 design environment as well as the test and acceptance methods described in SSP-30513. Is it possible that observed anomalies are produced by galactic cosmic rays (GCRs) of such high energy (Energy >10 GeV/nucleon, geomagnetic rigidity > 10 GV) that no latitude-longitude dependence of the anomaly rate is expected? Primary cosmic rays having energies greater than 10 GeV/nucleon also have very low flux as well as linear energy transfer (LET) values well below typical SEE thresholds (1 to 10 MeV-cm2/mg) because LET is inversely proportional to the square of particle velocity in the relativistic Bethe-Bloch equation [47, 48]. As a result, no detectable SEE effect is expected from the E > 10 GeV/nucleon GCRs themselves. However, energetic GCRs

50 have high cross sections for inelastic interactions with the Al nuclei in the relatively thick structure and shielding of ISS. Previous studies on GCR reaction and transport in spacecraft have shown that even those GCRs less energetic than 10 GeV/nucleon can produce showers of hadronic secondary particles (pions, protons, neutrons and alpha particles, nuclear fragmentation/spallation products) in the spacecraft shielding mass. The subject hadronic showers ultimately produce a significant flux of 100 to 200 MeV secondary particles inside the functional volume of the spacecraft [48-50] leading to SEE and TID effects [6, 48-51]. It has been previously reported that the CREME 96 SEE prediction software seriously underestimates SEU rates for shielding thickness on the order of 50 g/cm2 precisely because hadronic showers caused by high energy GCRs and other nuclear reactions in spacecraft shielding are neglected [51]. The energy spectrum of the primary galactic cosmic rays can be approximated by the following well-known power law equation [52]. IN(E) = 1.8 E-2.7 nucleons/(cm2 sec. sr GeV) Integrating the expression for IN(E) between 10 GeV/nucleon and 104 GeV/nucleon and multiplying by 2ʌ steradians yields the total flux of 10 to 104 GeV/nucleon particles to a zenith oriented 1 square centimeter target in low Earth orbit (cosmic rays from the nadir hemisphere are blocked by Earth). About 79% of the total nucleon flux consists of protons [52]. With the proton fluence and the relative abundances, the daily fluence of all nuclei with 10 GeV/nucleon < E < 104 GeV/nucleon can be calculated for specific ranges of Z and is shown in Table 8, along with the corresponding inelastic nuclear collision length, Ȝ, for Al (52). TABLE 8. Daily Fluence and Interaction Parameters of GCRs 10 GeV/nucleon < E < 105 GeV/nucleon Z

Elements

1 2 3-5 6-8 9-10 11-12 13-14 15-16 17-18

H He Li-B C-O F-Ne Na-Mg Al-Si P-S Cl-Ar

19-20

K-Ca

Relative Abundance 1 0.054 0.000825 0.00454 0.000619 0.00454 0.000392 0.0000619 0.0000206 0.0000412

Daily Fluence (1)

LET (2) MeV-cm2/mg x 103

Inelastic collision length(3), Ȝ

9,060 486 7.5 41 5.6 41 3.5 0.6 0.2

0.0002 0.0008 0.003 0.009 0.02 0.02 0.03 0.04 0.06

106 39.8 22.5 16.3 12.9 11.3 10.1 9.2 8.1

0.4

0.07

7.8

21-25 Sc-Mn 0.000103 0.9 0.10 6.6 26-28 Fe-Ni 0.000247 2.2 0.14 5.8 (1) Nuclei/cm2 per 24-hour day; (2) MeVcm2/g for 10 GeV/nucleus projectile in Si target (3) grams/cm2 Al:

51 The percentage of GCR particles (10 GeV/nucleon<E<104 GeV/nucleon) lost to inelastic collisions (i.e. nuclear reactions producing hadronic showers) on passing through various thickness of Al is calculated as [(I0 – Ix )/I0]x100 = % primary flux undergoing inelastic interaction, where I0 = the primary GCR flux, Ix = the flux surviving passage through x grams/cm2 Al, Ȝ is the nuclear interaction length for Al and, Ix = I0 e(-x/Ȝ). Table 9 shows how the percentage of GCR nuclei in the energy interval of interest that undergo inelastic nuclear interaction increases with thickness of Al. The median (50% point on the cumulative distribution curve) shielding thickness of the ISS habitable volume is 20 to 40 grams/cm2 Al and ranges from 10 to 100 g/cm2 Al at the 10% and 90% points [53]. Of course, ISS shielding thickness is equal to the path length only for that fraction of the isotropic GCR flux that is normally incident on the Al structure. Most GCR particles will have a longer path length as a result of oblique incidence on structure, and a correspondingly greater likelihood of producing energetic hadronic showers. For a 20-g/cm2 path-length, the E > 10 GeV/nucleon primary GCR flux can produce more than 1800 inelastic nuclear collisions per cm2 per day, each of which can yield an energetic multi-particle hadronic shower. The flux and LET spectrum of the subject hadronic shower inside the thick Al shielding will determine the SEE rates from the environment induced by the E > 10 GeV/nucleon primary GCR flux. TABLE 9. Percent of primary cosmic ray flux with E in the interval 10GeV/nucleon < E < 104 GeV/nucleon lost to inelastic nuclear interaction in Al shielding of thickness 1.0, 5.0, 10, and 20 g/cm2 Z 1 2 3-5 6-8 9-10 11-12 13-14 15-16 17-18 19-20 21-25 26-28

Elements H He Li-B C-O F-Ne Na-Mg Al-Si P-S Cl-Ar K-Ca Sc-Mn Fe-Ni

1.0 g/cm2 Al 0.94 2.5 4.3 5.9 7.4 8.5 9.4 10.3 11.6 12.0 14.0 15.4

5.0 g/cm2 Al 4.6 11.8 19.9 26.3 32.1 35.8 39.0 42.0 46.1 47.3 53.1 56.7

10 g/cm2 Al 9.0 22.2 35.8 45.8 53.9 58.8 62.8 66.3 70.9 72.2 78.0 81.2

20 g/cm2 Al 17.1 39.5 58.8 70.6 78.8 83.0 86.2 88.7 91.5 92.3 95.2 96.5

As is the analysis shown above suggests, inelastic nuclear interactions of high energy GCRs with thick Al structure and shielding can plausibly produce an induced SEE environment with a flux and LET spectrum capable of affecting ISS electronic systems. The GCR energy interval 10 GeV/nucleon < E < 104 GeV/nucleon and the secondary induced environment caused by the subject GCR population has not been explicitly or

52 completely included in previous studies [48-50] and typically isn’t included in SEE assessments at all [6, 51]. The absence of any dependence on SAA, latitude or SEP event exposure for the observed anomalies does not instantly confirm a non-radiation prime cause. Even in the previously reported examples of a strong SAA, latitude, and SEP exposure dependence of SEU rates in various LEO spacecraft [54-56], a substantial number of SEU events are also observed that are randomly distributed with respect to latitude and longitude and are unrelated to SEP events. A detailed study of the SEE environment induced in ISS by the GCR flux in the energy interval 10 GeV/nucleon < E < 104 GeV/nucleon using the Geant4 code (57) is planned and will be combined with continued tracking of the SEE performance of critical ISS avionic systems such as the MDMs. The solid-state memory devices in the MDMs are fully characterized with respect to the LET dependence of SEE processes. 5.0 Ionizing Radiation II: Crew Dose Environment Attempting to reduce ionizing radiation dose-equivalent (dose equivalent = dose x Q, where Q is the quality factor for the radiation field) to ISS expedition crews to As Low As Reasonably Achievable (ALARA) levels has raised a number interesting questions in the area of nuclear and radiation chemistry of materials. The crew radiation doseequivalent environment inside the ISS pressurized elements has a significant induced environment component, produced by inelastic nuclear interactions and nuclear reactions between the primary cosmic ray particle flux (trapped, solar, and GCR) and the relatively thick ISS structure and shielding [41-44, 58]. Design of augmented radiation shielding for ISS and verification of shielding effectiveness is complicated by the limited accuracy of existing theoretical models of cosmic ray reaction and transport in spacecraft materials [59, 60]. In addition, accurate measurement of the doseequivalent environment inside ISS, or any other spacecraft, is controversial and difficult [61]. Of more immediate importance is the fact that most ISS components are either already flying or in the final stages of preparation for flight, so that minimizing the radiation dose-equivalent environment by careful selection of materials and control of topology isn’t always possible. For example, carbon, polyethylene and water have well known advantages over aluminum as radiation shielding in energetic charged particle environments [62-66]. However, most external ISS structure is Al, and practical considerations preclude placing augmented radiation shielding outside the Al structure. As a result, GCR, trapped, and SEP event energetic particles interact with a thick Al target first, producing an induced radiation dose environment that is not effectively mitigated with reasonable amounts of hydrocarbon polymer or water [42, 44, 49, 58], though some flight results collected under relatively thin Al shielding do show that polyethylene can produce a more substantial reduction in dose equivalent than a comparable thickness of Al [67]. Results of the recently completed Japanese Space Agency (NASDA) Bonner Ball Neuron Detector (BBND) experiment on ISS demonstrate this point [68]. The Bonner sphere neutron spectrometer (69) consists of a set of identical thermal neutron sensors each situated at the center of polyethylene moderating spheres of different thickness.

53 Polyethylene is often chosen as a moderating material because it can slow down fast neutrons without significant absorption losses [69]. By design, the NASDA BBND detects neutrons less energetic than 10 MeV and BBND ISS data (68) clearly show significant crew dose-equivalent rates during SAA passage and at high latitude during the SEP events of April 2001. Measurements of quiet time (no SEP event) dose equivalent made on ISS using the German Aerospace Center charged particle detector telescope (DOSTEL) during 2001 yielded an average value of 20.8 micro-Sieverts/hour (2.08 millirem/hour) with a quality factor, Q, ranging from 2.5 to 2.6 [70], corresponding to an average 12-month dose-equivalent of 182 milli-Sieverts (18.2 rem). The total annual dose measured by the DOSTEL team during 2001 (solar maximum) was 0.055 Gray (5.5 rads), significantly less than the 0.091 Gray (9.1 rads) measured during the 1997 Shuttle-Mir missions (solar minimum), as expected. The DOSTEL measurements are comparable to ISS dosimeter measurements made during the same time frame [41], which ranged from an annual dose of 44 -105 milli-Gray (4.4 to 10.5 rads) depending on location in ISS. ISS expedition crews reside on ISS for between 4 and 6 months in most cases, so the practical crew dose environment is one-third to one-half the annual environment. 6.0 Ionizing Radiation III: Surface Dose to Exposed External Materials and Systems Energetic trapped electron dose is the principal threat to the long-term durability of Teflon®, silicone, and other radiation labile materials exposed on the exterior of ISS. Large uncertainties in electron dose predictions made with the AE-8 model [71], combined with uncertainties in the synergistic contributions of mechanical stress, thermal cycling, and atomic oxygen to the degradation of Teflon® by energetic electron radiation [72 a, 73] lead to a corresponding uncertainty in estimated degradation rates. The energy spectrum of trapped energetic electron environment leads to a rapid decrease in radiation dose with depth. Radiation dose and damage is always greatest near the surface of exposed materials. As an example, the dose vs. depth curve for an aluminum sphere (density = 2.7 g/cm3) is shown in Figure 12. Multiplying the thickness axis in Figure 12 by the ratio [density of aluminum]/[density of material] leads to an approximate dose vs. depth curve for any other material [72 b]. When electron range is expressed as g/cm2, the curve for aluminum is approximately correct for any material because density varies only slightly with the atomic number Z of the stopping medium [72c]. Teflon® is an excellent insulator which complicates accelerated ground based testing of energetic electron degradation effects as a result of target electrostatic charging in high dose rate electron beams. On-orbit, dose rates are much lower permitting continuous discharge of exposed Teflon® materials by dielectric relaxation processes, producing a dose-depth profile that is difficult to reproduce in ground based accelerated testing because the lower energy electrons responsible for most near surface dose are most easily decelerated and deflected by target charging. For example, in 10 years the SSP 30512 x 2 dose to surface exposed Teflon® materials is 2 Mrads, corresponding to 1.3 x 10-5 coulombs of charge, distributed over a depth of

54 2 mm if delivered by a 0.1 to 1.0 MeV electron beam (designed to approximate the trapped electron energy spectrum). A typical volume resistivity for Teflon® is on the order of 1016 ohm-m, and the dielectric constant is 2.1, allowing us to calculate the dielectric relaxation time, tr, given the permitivity of free space, İo = 8.85 x 10-12 Farad/m. tr = R İ İo = 51 hours Any charge distribution in Teflon® will dissipate by charge leakage to surrounding conductors with an exponential decay constant tr. Q(t) = Q(0)e(-t/tr) With a relaxation time of 51 hours, it is clear that delivery of 1.3 x 10-5 coulombs during10 years creates no target charging, but attempting to deliver the same charge during 10 hours can result in enough charge build-up in the target to deflect or decelerate the e- beam, producing a much lower dose than anticipated, especially in the near surface region. Consider the ISS Trailing Umbilical Assembly (TUS) cable, which provides power and data services to the Mobile Transporter (MT) platform that moves along rails on the ISS truss. The TUS cable is normally stowed in an uptake reel and is fed out to follow the MT as it moves along the truss. The mechanical design of the TUS cable and uptake reel assembly requires that the insulation and jacketing materials must be capable of surface elongation of not less than 3.5%, without cracking or tearing, for reliable system performance. The TUS cable is jacketed in 1.2 mm (6 layers, 0.2 mm per layer) of expanded polytetrafluoroethylene (PTFE) with a density of 1g/cm3. TUS cable temperature is expected to range between –100o C and +130o C. Polytetrafluoroethylene (PTFE) is especially susceptible to embrittlement by ionizing radiation with an onset of mechanical property loss on exposure to 50 and 150 krads in PTFE [72,73]. The SSP-30512 x 2 dose to each of the 6 layers of TUS cable jacket is shown in Table 10. The radiation dose in Table 10 includes the effects of ISS structural shielding and TUS cable geometry. Continuous exposure without shielding by the uptake reel housing is also assumed. Ionizing radiation breaks chemical bonds and embrittles PTFE causing an increase in stiffness and a decrease in elongation [72,73], suggesting the need for verification testing and analysis to assure TUS cable functional life on ISS. * TABLE 10. SSP-30512 x 2 radiation dose (kilorads PTFE) to the center of each layer of the expanded PTFE TUS cable jacket material as a function of time-on-orbit

Layer 1 2 3 4 5 6 1 year 250 50 22 13 9 6 2 years 600 110 50 29 19 12 4 years 300 120 64 40 29 8 years 1100 320 150 90 60 10 years 520 220 120 85 * Surface erosion rate of 0.03 mm/year from atomic oxygen attack is assumed, leading to the complete removal of the outermost two outermost layers in 10 years.

55 Estimating the useful life of exposed Teflon® materials and components on ISS starts with a worst-case estimate of the expected trapped radiation dose the SSP-30512 design environment (AE8 Max, 500 km) multiplied by 2. Materials are first subjected to the worst-case 10-year dose using energetic electrons (E = 1 MeV to minimize target charging effects) or Co60 gamma rays in vacuum. The test is designed to produce nearly uniform dose through the cross section of the sample because target charging could not be controlled for the reasons given above. Next, the irradiated materials are subjected to a worst-case thermo-mechanical environment to verify that mechanical properties are adequate to perform the required function on orbit. Mechanical testing results (%elongation at maximum load prior to failure) at –100o C, +25o C, and +130o C irradiation are shown in Table 11. TABLE 11. Mechanical test results (% elongation prior to failure) at three different temperatures for TUS cable wrap as a function radiation dose Radiation dose 2.7 Mrad Co60 Ȗ rays 2.7 Mrad 1 MeV electrons 900 krad 1 MeV electrons 300 krad 1 MeV electrons

-100o C

+25o C

+130o C

13.0

13.7

9.3

8.3

22.1

17.3

11.8

28.4

29.1

10.43

36.7

39.3

For the 51.6o inclination ISS orbit, AE8 predicts much higher doses than have been observed on MIR [75, 76] and on the two Russian-Canadian BION satellites [77]. As a result the SSP-30512 design environment, 500 km altitude AE8 Max, overestimates dose significantly, even without the recommended 2 x multiplier. Figure 12 compares dose vs. depth in a solid Al sphere for the 500 km AE8 Max SSP-50512 environment as well as the 390 km environment for both AE8 Max and AE8 Min. Table 12 compares the calculated SSP-30512 x 2-electron dose with corresponding dose measurements from Mir [75, 76] and BION [76]. Clearly, AE8 is seriously overestimating the actual trapped electron dose for the 51.6o, 350-400 km ISS (or Mir) orbit, and the use of the SSP-30512 x 2 design environment to define verification testing dose adds even more conservative margin, which is important in light of the uncertainties in any laboratory test method.

56

Figure 12. A comparison of the AE-8 dose/depth curves for an Al sphere: 1) SSP 30512 Design Environment, 2) 390 km altitude TABLE 12. A comparison of measured annual trapped electron dose with calculated AE8 annual dose and the SSP-30512 x 2 design-environment annual dose. Satellite, Experiment, Measured annual AE8 dose Shielding dose; rads/year rads/year Mir, REM (74), 21 122 0.7 mm Al Mir, TLD (75), 7 x 103 1 x 104 0.0 mm Al Mir, TLD (75), 673 645 0.5 mm Al equiv. Bion 10 & 11 MOSFET/TLD(76), 2 x 105 8 x 104 0.0 mm Al * SSP-30512 x 2 dose = conservative design margin Measured dose

SSP-30512 x 2 dose rads/year

SSP-30512 x 2 dose* Measured dose

1470

21

5 x 105

71

9 x 103

13

5 x 105

2.5

The right-most column of Table 12 shows the ratio of the measured trapped electron dose for a particular flight experiment to the SSP-30512 x 2 design and verification dose and is a direct measurement of the conservative margin built into the SSP-30512 x 2 design environment. The average conservative design margin from the data in Table 12 is 27. Despite the near 30 fold margin in the radiation dose for mechanical testing (Table 11), the TUS cable wrap retained more than adequate mechanical properties over the full temperature range expected on orbit. In addition, removal of the radiation embrittled near-surface layer by ram atomic oxygen further reduced the risk of mechanical failure. The large conservative margin in the radiation dose used for material verification testing assures that the TUS cable and other exposed Teflon® materials will retain useful performance properties for the full life of ISS, despite the many unknown or un-quantifiable factors affecting radiation-induced property loss.

57 7.0 Small Particle Impact and Molecular Contamination from Hypergolic Engine Plume Impingement and Fluid Venting Bipropellant thrusters (monomethyl hydrazine (MMH)/unsymmetrical dimethyl hydrazine (UDMH) and N2O4) are widely used for the purposes of spacecraft reboost, attitude control, and orbit correction. Experimental studies [77, 78] revealed the presence of small particles consisting of unburnt propellant in the engine exhaust plume. The origin of the droplets is commonly attributed to incomplete combustion [77, 78]. Such droplets can range from 1-100 microns in diameter and have velocities up to the exhaust velocity of the thruster exhaust-plume gases (~ 3 km/s). The unburnt propellant may be in the form of super-cooled liquid or ice. Hypergolic engine plume impingement can have important effects on the functional life of exposed materials [79-82]. Past flight experiments, including the Shuttle Plume Impingement Flight Experiment (SPIFEX), have shown the impact of these droplets on witness coupons. SPIFEX (80) was flown on STS-64 in September 1994 and measured in plume parameters during 84 firings of the Primary Reaction Control System (PRCS) and 17 Vernier Reaction Control System (VRCS) engines, all for engine-nozzle to witness-coupon distances of 2.4 to 24 meters, corresponding to more than 10 years of nominal Shuttle–ISS proximity operations. Post-flight analysis of witness coupons on SPIFEX showed mechanical impact pitting on aluminum films and a combination of mechanical and chemical erosion pitting on the Kapton® film samples. Examples of these effects are shown in Figures 13 and 14. It should be noted that particle impact effects were not visible to the unaided eye on the SPIFEX payload. Visual inspection and photographic survey of the SPIFEX payload before and after flight revealed no substantial damage [80]. Control of particle impact degradation of critical materials is based on operational protection of sensitive surfaces during proximity operations (thruster firings) by visiting spacecraft or during ISS venting or purging operations. PV arrays are positioned edgeon to the direction of plume flow. View ports and windows are generally covered, and cameras pointed away from the particle source. Low to medium velocity particles only affect the optical performance of view ports, windows and PV array glass cover slips. Both the Orbiter and US Laboratory module vent liquid water into vacuum as part of their normal operations. A model was developed to describe the two-phase plume (large and small ice particles interspersed with gas) resulting from these operations [83]. A comparison of on-orbit and model predictions shows excellent agreement, as is shown in Figure 15 [83]. The size (0.42 to 1000 microns radius) and velocity (3 to 23 m/sec.) distribution of the ice particles produced in the venting of water from both the Shuttle and ISS are well below the damage threshold for both the PV array cover slips and all ISS optical view ports and instruments (84).

58

Figure 13. SPIFEX Aluminum witness coupon showing craters from thruster plume droplets

Figure 14. SPIFEX Kapton witness coupon showing craters from thruster plume droplets

59

Figure 15. Orbiter Water Dump Plume: Model and Flight Comparison

8.0 External Contamination, Atomic Oxygen and Solar Ultraviolet Radiation 8.1 ISS EXTERNAL CONTAMINATION CONTROL & RISK MITIGATION Vacuum exposed materials contribute to contamination of spacecraft external surfaces as a result of materials outgassing and deposition processes [85]. Materials internal to non-pressurized shells also contribute to contamination of spacecraft external surfaces because outgassing from such materials is vented to the exterior of the spacecraft. SSP30426 [86] defines the external contamination control requirements for ISS. A process of careful materials selection and testing combined with integrated system level contamination transport and deposition analysis assures compliance with contamination performance requirements. The vacuum outgassing and contamination characteristics of ISS materials are determined in long duration ASTM E 1559 (ASTM, 1993b) tests [87]. In testing ISS materials, the sample (outgassing source) is tested over the expected operational temperature range identified by the ISS system level thermal model. Thermally Controlled Quartz Crystal Microbalances (TQCMs), or receivers, are held at temperatures covering the typical range of operational temperatures of ISS contamination sensitive hardware. ISS materials are typically tested for 144 hours according to ASTM E 1559, for each relevant source and receiver temperature, and the test data is curve-fitted to a conservative model describing the decay of outgassing rates, R(t), over the life of ISS using the scaling relation R(t) = R(144hrs)/(t/144[rs), which predicts more outgassing at long times than is expected for most materials [88]. R(t) has the units or mass/(area x time).

60 The NASAN (NASAN is not an acronym) software package [89] is used to analyze external contamination transport and deposition processes on all ISS flight configurations. The principal inputs to NASAN contamination analysis are: 1) the quantity of each material used, 2) materials outgassing data, 3) location on ISS, 4) location and directionality of vents, 5) outgassing source temperature, 6) receiving surface temperature, and 7) the three dimensional configuration of ISS. Using NASAN, contamination processes resulting from visiting vehicle approach, docking, and separations can also be accurately modeled. Direct Simulation Monte Carlo (DSMC) methods [90] have also been employed to model return flux and specific problems encountered during ISS design and development. When a violation of system level requirements is identified in a NASAN analysis, corrective actions are implemented. Materials substitution or vacuum baking of materials and assemblies may be required to remove volatile components and low molecular weight species. For example, the Ku-band antenna dish (coated an S-13 GLO type organic silicone based thermal control paint), Superflex-R silicone insulated electrical cables, and silicone electrical cable clamps were vacuum-baked to ensure compliance with ISS system level requirements. Russian spacecraft design practices require less severe external contamination control processes than those implied by SSP-30426. By working closely with the Russian program during ISS design and development, Russian elements were successfully integrated without volition of the SSP-30426 requirements. Extensive testing of Russian materials, combined with NASAN analysis of materials outgassing, venting, and impingement of thruster plumes has enabled successful and cost effective integration of Functional Cargo Block (FGB), Service Module (SM), Progress, and Soyuz, while also addressing any effects of the NASA, ESA, and NASDA elements on the Russian segment. Taking advantage of FGB and SM time on-orbit as a materials vacuum bake-out prior to the deployment of more contamination sensitive ISS elements also mitigated the impact of FGB and Service Module materials outgassing contamination on sensitive ISS surfaces. The principal FGB contamination sources are non-metallic materials. These materials exhibit high outgassing rates and are exposed to vacuum under multi-layer insulation. The major contamination sources are the impregnated mesh (BF-4) used as support to the solar cells in the solar arrays, AK-573 (acrylic based paint) on the solar array frames, the PArML (Aramid-type fabric) cable wrap which encapsulates the power and data cables, and the KO-5191 (organic silicone based) and EP-140 (epoxy based paint) paints which are widely used on exterior surfaces of the FGB. The source temperature, collector temperature, and collector surface area are vital factors that affect the degree of on-orbit contamination and material degradation rate. Modifications were made to the SM materials selection to minimize contamination risks for any more sensitive ISS surfaces, as there would be less on-orbit time for vacuum baking of materials. Of particular importance were cable insulation and paint changes. However, while changes were made to SM materials selection (cable insulation, surface preparation, paint, and use of anodizing for the pressure shell), evidence of limited self-contamination in the form of surface darkening along seams,

61 around vents, and plume impingement zones has also been identified. The AK-512 paint that replaced the high outgassing KO-5191 and EP-140 paints also shows extensive degradation in the form of flaking and darkening. Controlling surface contamination by hypergolic engine plumes is also an important part of ISS contamination control and risk mitigation. The SPIFEX [80], PIC [79] and DVICON [81] flight experiment projects were all directed toward a quantitative understanding of both short-term and long-term hypergolic engine plume contamination effects needed for verifying compliance with the ISS external contamination requirements [86]. ISS surface contamination resulting from docking and separation of visiting vehicles as well as operation of ISS thrusters for re-boosts and attitude control is estimated using the NASAN software package, verified models of engine plume transport [79,82], and the SPIFEX, PIC, and DVICON deposition and evaporation/sublimation measurements. NASAN simulations of various docking and separation procedures have been used to select procedures that use minimally contaminating thruster firing sequences. In addition, NASAN simulations have also supported development of flight rules that require protecting optically sensitive surfaces such as video camera lenses, ISS view ports, and ISS windows from thruster plume impingement in most cases. It should also be noted that the thrusters on the Progress or Soyuz docked at the aft end of the SM are used for ISS pitch control whenever possible to minimize the contamination impacts of the SM pitch thruster operation over the functional life of ISS. The SM roll jets will be used less in future if roll control can be accomplished using the thrusters on a Progress vehicle docked on nadir. The SM roll control thrusters fire automatically on depressurization of the Docking Compartment Module (DCM) airlock because depressurization is propulsive. As a result, it is possible for EVA crew to contact SM surfaces near the roll control thrusters shortly after thruster firing. Chemical analysis of fresh hypergolic thruster residues on Mir, conducted as a part of the DVICON flight experiment [81], suggests that contamination of EMU suit (space suit) surfaces with toxic substances is possible if insufficient time elapses between roll control thruster firing and entry of EVA crew into the thruster plume impingement area. PIC, SPIFEX, and DVICON data all demonstrate that the fresh contamination deposit, formed when a thruster fires, evaporates or sublimes rapidly if the temperature of the contaminated surface is near 25o C, leaving a persistent residue that is nontoxic if detectable at all [79-81]. However the fresh hypergolic thruster contamination deposit is significant and contains some highly toxic, though highly volatile, compounds [81. The time that must elapse between roll control thruster firings and complete evaporation or sublimation of the toxic species is very sensitive to the surface temperature of the SM and is expected to range from a few minutes to a few hours as surface temperature varies between –40o C to +25oC. EVA flight rules designed to minimize the risk of introducing toxic substances to the habitable volume of ISS have been developed, while laboratory studies aimed at a better definition of the temperature dependence of the toxic residue evaporation time are completed. Cosmonaut and astronaut observations during EVA on the aft end of the SM suggest that there is very little risk of toxic contamination of the EMU suits.

62 8.2 ISS EXTERNAL CONTAMINATION OBSERVATIONS In-flight imaging surveys of the US Lab, Node 1, the Pressurized Mating Adapters (PMAs), and the Z1 and P6 Truss structures, as well as the 160V US PV arrays show little or no evidence of large-scale self-contamination from molecular outgassing and deposition processes, providing some confirmation of the validity of the ISS contamination risk mitigation process. More specifically, high contamination risk items like the SuperFlex-R silicone cable, the silicone cable clamps and the Ku-band antenna dish are not producing visible molecular outgassing and contamination deposits on neighboring hardware after two years of flight. The FGB and SM both show limited self-contamination from molecular outgassing and deposition processes as well as surface discoloration and damage from hypergolic thruster plume impingement. The results of external contamination imaging surveys are summarized in Tables 13-15[91]. TABLE 13. Summary of FGB Contamination Observations Location FGB aft-end and radiator surfaces

First Flight Observation 2A (pre-docking)

FGB TORU antenna opening

2A

FGB MM/OD shields

2A

Gaps/vent paths in the FGB blankets and structure

2A

Description Dark patterns on FGB aft-end and radiators from Proton stage separation thruster induced contamination Darkening of the blanket surrounding the TORU antenna opening Darkening of the area surrounding the MM/OD shield gap and flaking of the MM/OD shield paint Darkening of the area surrounding gaps and vent paths on the FGB blankets and structure

8.3 ATOMIC OXYGEN AND SOLAR ULTRAVIOLET Basic ISS materials performance requirements for the atomic oxygen (AO), solar ultraviolet (UV), and solar vacuum ultraviolet (VUV) environments are defined in SSP30233, Rev. F., “Space Station Requirements for Materials and Processes,” [92] and require that materials be selected to perform in the LEO environment for the intended life cycle exposure. The effects of atomic oxygen (AO) and ultraviolet radiation on materials exposed to LEO were well known and reasonably well understood prior to materials selection for ISS as a result of a important series of space flight experiments [93-95] such as the Long-Duration Exposure Facility (LDEF), the several Space Shuttle flight experiments on Evaluation of Oxygen Interaction with Materials (EOIM-1, EOIM-2, EOIM-3), and the Passive Optical Sample Assembly (POSA) experiment conducted on Mir during ISS Phase 1. As a result, numerous design solutions were implemented to prevent environmental degradation of external materials and AO degradation problems experienced to date on ISS have been minimal.

63 TABLE 14. Summary of Service Module Contamination Observations Location Service Module aft-end and radiator surfaces

First Flight Observation 2A.2b

Service Module handrails

2A.2b

Gaps/vent paths in the Service Module structure

2A.2b

Service Module Instrument Section

2A.2b

Service Module Zenith facing thrusters

5A

Kurs antenna at the end of Service Module solar arrays

5A.1

Description Dark patterns on Service Module (and FGB) aftend and radiators from Proton stage separation thruster induced contamination Darkening patterns and flaking of handrail coating Darkening of the area surrounding gaps and vent paths on the Service Module structure Darkening patterns and surface degradation of painted surfaces Darkening of the area near the Zenith thrusters on the Service Module Darkening of the base of the Kurs antenna at the end of the Service Module solar arrays

The use of oxygen reactive materials in critical ISS applications has been minimized, as has the use of materials that darken or are otherwise degraded by UV/VUV. Where reactive materials must be used, for example in the Kapton® film thermal blanket material upon which the 160-V PV array wings are assembled, multiple layers of oxygen resistant protective coatings are applied to assure full 15-year functional life, as is described elsewhere in these proceedings [96]. Another area in which reactive materials can be used results from the wake environment produced by ISS structure. Because ISS can fly only a limited number of approved attitudes, structural wake shielding from ram atomic oxygen enables the use of some oxygen reactive materials in wake shielded locations with no risk of significant long-term AO induced degradation of performance.

64 TABLE 15. Summary of PMA1, PMA2 and Node 1 Observations Location Node 1 MM/OD anodized aluminum shields

First Flight Observation 2A

Node 1 SVS targets

2A.1

Node 1 stowage bags

2A.1

Node 1 aft-CBM supports

2A.2a

Metalphoto labels on Node 1 and both PMAs ORU Transfer Device (OTD) and APFR

2A.2a

PMA2 washers

2A.2b

PMA1 anodized surfaces

2A.2b

PMA2 Worksite Interface (WIF)

2A.2b

2A.2a

PMA2 APAS glass fabric surfaces and able straps

3A

PMA2 planar and hemispherical retroreflectors

3A

PMA1 zenith sunshade

3A

Description Discoloration patterns observed on several MM/OD aluminum panels “Bubbling” of SVS targets on Node 1 Darkening of the Beta cloth used on the Node 1 stowage bags Darkening of the Node 1 aftCBM supports Darkening of the anodized aluminum labels Darkening of the anodized layer at the base of the OTD and surfaces of the APFR Darkening of metallic washers on PMA2 Darkening of anodized aluminum surfaces on PMA1 Darkening of the anodized base of the WIF Russian glass fabric used on the PMA2 APAS, darkening of cable straps Coating used on PMA 2 planar and hemispherical retroreflectors shows degradation and exhibits flaking PMA1 zenith sunshade exhibits dark spots

8.4 ANODIZED ALUMINUM Aluminum is the most important structural material on ISS and the surface anodization on the aluminum also serves as the external passive thermal control surface for the NASA truss structure as well as on other NASA, Japanese Space Agency (NASDA), and European Space Agency (ESA) elements. In order to avoid the use of paints, major ISS external surfaces are anodized using controlled processes to produce the required thermo-optical properties. The truss structure is essentially all sulfuric acid anodized and the meteoroid/orbital debris shields are usually chromic acid anodized. The anodized surfaces are impervious to atomic oxygen and were shown to be stable to ultraviolet radiation in ground testing. So far, these surfaces have performed well with no visible darkening. Specialized forms of anodized aluminum have been used for some applications. Black anodize has been used in a few applications where a thermally hot environment was required. EVA handrails are required to be yellow; in the absence of a yellow paint that would be durable in the LEO environment, a gold anodize was used. This gold anodize

65 has also proved stable to date. These anodize coatings have also been used for external warning labels where yellow/black stripes were required. Almost all ISS external labels are manufactured using commercially available photosensitive anodized aluminum foil (Metalphoto). Monochrome Metalphoto labels were selected after ground based testing showed them to be stable in the atomic oxygen-ultraviolet radiation environment (many colored labels faded rapidly). However, some UV discoloration has been observed on orbit; although the labels are still readable, the clear-coated (originally silver in appearance) surfaces have turned yellow. Similar yellowing/browning of the anodic coating has been observed on a few components provided by NASA. Current speculation is that the UV sensitivity is caused by a nickel-acetate solution used to water seal anodic coatings (the nickel acetate water sealing process is definitely used for the Metalphoto labels and may have been used for some other anodized aluminum hardware); however, we have been unable to replicate the discoloration with vacuum-UV sources on the ground. Essentially all the other anodized hardware was water sealed with deionized water. 8.5 BLANKETS AND SHROUDS Standard multi-layer insulation (MLI) materials erode rapidly in atomic oxygen. Extensive use is made of Teflon impregnated, tightly woven glass fabrics (Beta cloth) as covers for multi-layer insulation blankets [95,96] where the composite fabric is expected to perform for the full life of ISS. The Beta cloth covers have excellent durability despite AO/UV erosion of the Teflon® coating. The remaining glass fiber fabric is impervious to AO and remains flexible enough to withstand 30 years of thermal cycling. Nearly all ISS MLI blankets have an outer layer of Beta cloth. They are sewn together with AO-compatible ceramic thread or Teflon-coated fiberglass thread. Beta cloth has also been used extensively for wrapping and protecting external wire and cable and for thermal control shrouds. The Beta cloth selected for ISS contains no silicone sizing, because Beta cloth with such sizing tends to darken in the UV environment. In a few locations, thermal requirements have dictated the use of MLI blankets with an outer layer of silver-Teflon®. The Teflon® layer is verified to be thick enough to survive in the environment to which it will be exposed, even with significant erosion by AO/UV. 8.6 PAINTS AND COATINGS Nearly all ISS radiators are coated with Z-93, a white inorganic paint with excellent optical properties and a long history of successful use on spacecraft. Silver-Teflon is again used for a few radiator surfaces, but has adequate durability in the application environment.

66 Very few other external surfaces are painted and the coating system used is almost always a specialty inorganic coating system that has been demonstrated to have good durability in AO/UV. Such coatings have been used for targets and logos. The ubiquitous Aeroglaze A276 polyurethane paint has been used for a few short-term applications. As expected, the A276 coatings on returned hardware are darkened by UV and powdery as a result of AO erosion; however, these coatings are adequate for 12 years exposure. Although nearly all labels are anodized aluminum, as noted above, cable labels are frequently poly(vinylidene fluoride) (Kynar®) overcoated with a thin layer of lowoutgassing silicone as protection against AO damage. No problems with these labels have been noted to date, although shrinking and cracking of the silicone coating is a long-term concern. Silicones have also been used for AO protection on solar array wing batons and silver-plated fasteners. The largest area of nonmetallic material on ISS occurs on the PV array wings. The front sides of these wings are protected by glass cover slips and low-outgassing silicone sealants. The backs of the arrays are Kapton®, protected from AO by a silica overcoating. In a few locations where silica could not be used, the Kapton is protected by a low-blocking, low-outgassing silicone. 8.7 ISSUES Despite the high sensitivity of the ISS Program to AO/UV effects, schedule, cost, and materials availability problems have occasionally forced selection of materials having low AO/UV resistance for exposure to the LEO environment for less critical or shortterm applications. The path followed to control such hardware depends on cost and schedule. Whenever feasible, such items are premeditated before launch. Nylon cable ties have been replaced by durable Teflon-coated fiberglass lacing cord. Nomex lacing cord has been oriented away from atomic oxygen and/or overwrapped with copper wire ties. Silver-plated fasteners have been used extensively, despite efforts to prohibit them because of the high reactivity of silver with both the high kinetic energy ram AO (atomsurface collision energy = 5 eV) and the epithermal/thermalized AO (atom-surface collision energy <0.1 eV) produced by accommodation of ram AO on spacecraft surfaces [97]. Blankets and shrouds cover most of these fasteners, so AO exposure is minimal. Exposed silver-plated fasteners in EVA translational paths are protected from AO by a Beta cloth or aluminum foil cover or by a coating of low-outgassing silicone. The most visible example of AO/UV damage at this time is on the PV array wing blanket box. This box contains polyimide foam as a support and padding for the folded PV array wings during launch. By design, an aluminized Kapton film protected the foam from AO exposure on orbit. The protective film has disintegrated on the two wings deployed to date, leaving the polyimide foam completely exposed. Fortunately, the foam is not needed after launch; testing is in progress and preliminary indications are that the foam will erode harmlessly without releasing particulate contaminants. The exact cause of the aluminized film failure is not known, but it is probably a result

67 of excessive temperatures on the exposed (and very thin) aluminum surface and/or small cracks in the aluminum exposing the Kapton® directly to AO. The design of the remaining PV array blanket boxes is the same and similar failures are expected after they are launched and deployed (the failures are aesthetic in character, essentially harmless, and the cost of a redesign is prohibitive). SUMMARY With the minor exceptions noted above, atomic oxygen performance of ISS systems has been more than adequate to date and no significant atomic-oxygen-induced system or component failures have been observed. The PV array wings and supporting structures, the large area active radiators, and the passive thermal control surfaces and blankets are all performing as required with no detectable sign of atomic oxygen or UV/VUV degradation. Limited UV/VUV discoloration of the Beta cloth fabrics has been traced to contamination during ground processing, however, the discoloration has no effect on performance. 9.0 Summary and Conclusions Detailed consideration of the effects of natural and induced environments on ISS materials and systems during design, development, and flight operations has produced a safe, efficient manned space platform that is largely immune to the deleterious effects of both natural and induced environmental factors. Substantial uncertainties in the effects of many environmental factors on ISS materials and systems were addressed by the use of conservative margins for materials selection, verification testing, and performance analysis. 10.0 References 1. 2.

3. 4. 5.

6. 7.

NASA Technical Memorandum 4527; Natural Orbital Environment Guidelines for Use in Processing, John Wiley and Sons Inc., New York, 1994 Handbook of Geophysics and the Space Environment; Jursa, Adolph S., Editor, Air Force Aerospace Vehicle Development; Anderson, Jeffery B., Editor; Smith, Robert E., Compiler; June 1994; and Lieberman, M. A., Lichtenberg, A. J.; Principles of Plasma Discharges and Materials Geophysics Laboratory, Air Force Systems Command, United States Air Force, 1985 Vette J. I., The AE-8 Trapped Electron Model Environment, NSSDC WDC-A-R&S 91-24, 1991 Abel, R., Thorne, R. M., and Vampola, A. L.; “Solar Cycle Behavior of Trapped Energetic Electrons in Earth’s Inner Radiation Belt,” J. Geophys. Res., 99, 19427, 1994 Ogliore, R. C., Mewaldt, R. A., Leske, R. A., Stone, E. C., von Rosenvinge, T. T.; “A Direct Measurement of the Geomagnetic Cutoff for Cosmic Rays at Space Station Latitudes,” Proceedings of the ICRC 2001: 4112, Copernicus Gesellschaft 2001; and Xinlin, L., Baker, D. N., Kanekal, S. G., Looper, M., Temerin, M.; “Long Term Measurements of Radiaiton Belts by Sampex and their Variations,” Geophys. Res. Lett., accepted for publication Aug 15, 2001, American Geophysical Union Dyer, C. S., Truscott, P. R., Evans. H., Sims, A. J., Hammond, N., Comber, C.; “Secondary Radiation Environment in Heavy Space Vehicles and Instruments,” Adv. Space Res. Vol. 17, No. 2, pp. (2)53 – (2)58, 1996 (1995 COSPAR) Garrett, H. B., Whittlesey, A. C.; “Spacecraft Charging, An Update,” AIAA 96-0143, 34th AIAA Aerospace Sciences Meeting and Exhibit, 15-18, January 1996, Reno Nevada

68 8. 9.

10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

Vaughn, J.A., Carruth, M. R., Katz, I., Mandell, M., Jongeward, G. A.; “Electrical Breakdown Currents on Large Spacecraft in Low Earth Orbit,” J. Spacecraft and Rockets, Vol. 31, No. 1, January-February 1994, pp 54-59 a) Snyder, David B.; “Dynamic Interactions Between Ionospheric Plasma and Spacecraft,” The Radio Science Bulletin, No. 274, Sept, 1995, pp 29-36, b) Ferguson, D. C., Hillard, G. B.; “In Space Measurement of Electron Current Collection by Space Station Solar Arrays,” AIAA 950486, 33rd Aerospace Sciences Meeting and Exhibit, January 9-12, 1995, Reno NV., c) de la Cruz, C. P., Hastings, D. E., Ferguson, D., Hillard, B.;”Data analysis and model comparison for solar Array Module Plasma Interactions Experiment,” J. Spacecraft and Rockets, Vol.. 33, No. 3, pp 438-446, May-June 1996, d) Hastings, D. E., Cho, M., Kuninaka, H., “The Arcing Rate for a High Voltage Solar Array,” Journal of spacecraft and Rockets, 29, No.4, 538-554, 1992, e) Hastings, D. E.; “A Review of Plasma Interactions with spacecraft in Low Earth Orbit,” Journal of Geophysical Research, 100, No. A8, PP. 14457-14484, 1995 Galofaro, J. T., Doreswaamy, C. V., Vayner, B. V., Snyder, D. B., Ferguson, D. C.; “Electrical Breakdown of anodized Structures in a Low Earth Orbit Environment,” NASA/TM – 1999209044, April, 1999 Vayner, B. V., Galofarno, J., Ferguson, D. C., de Groot, W., Thompson, C., Dennison, J. R., Davis, R.; “The Conductor- Dielectric Junctions in a Low Density Plasma,” NASA/TM – 1999209408, Nov. 1999 Murphy, G., Croley, D., Ratliff, M., Leung, P.; “The Role of External Circuit Impedance in Dielectric Breakdown,” AIAA 92-0821, 30th Aerospace Scoiences Meeting and Exhibit, Jan. 6-9, 1992/Reno, NV Patterson, M. J., Verhey, T. R., Soulas, G., Zakany, J.; “ Space Station Cathode Design Performance and Operating Specifications,” IEPC Paper Number 97-170, 25th International Electric Propulsion Conference, Cleveland Ohio, Aug. 1997. Lambert, J. C., Chaky, R. C.; “The ISS Plasma Contactor,” AIAA 96-0627, 34th Aerospace Sciences Meeting, Jan. 15-18, 1996, Reno NV. Ferguson, D. C., Morton, T. L., Hillard, B. G.; “First Results from the Floating Potential Probe (FPP) on International Space Station,” AIAA-2001-0402, 39th Aerospace Sciences Meeting and Exhibit, Jan. 2001, Reno, Nevada Morton, Tl L., Minow, J. I.; “Floating Potential Langmuir Probe Data Reduction Results,” AIAA2002-0936, 40th AIAA Aerospace Sciences Meeting and Exhibit, 14-17, January 2002, Reno Nevada Bering, E. A., Koontz, S., Katz, I., Gardner, B., Evans, D., Ferguson, D.; “The Plasma Environment of the International Space Station in the Austral Summer Auroral Zone Inferred from Plasma Contactor Data,” AIAA 0220-0935, 40th AIAA Aerospace Sciences Meeting and Exhibit, 14-17, January 2002, Reno Nevada Mikatarian, R.R., Kern, J.W., Barsamian, H.R. and Koontz, S.L., “Plasma Charging of the International Space Station”, to be presented at the COSPAR World Space Congress, Houston, Texas, October 2002 Katz, I., Lilley, J. R., Greb, A., McCoy, J. E., Galofaro, J., Ferguson, D. C.; “Plasma turbulence enhanced current collection: Results form the plasma motor generator electrodynamic tether flight,” J. Geohys. Res., Vol. 100, No. A2, pp1687-1690, Feb. 1, 1995 Stone, N. H.; “The Aerodynamics of Bodies in a Rarefied Ionized Gas with Applications to Spacecraft Environmental Dynamics,” NASA Technical Paper 1933, November 1981 Martin, A. R.; “Spacecraft/Plasma Interactions and Electromagnetic Effects in LEO and Polar Orbits,” Final Report for ESA/ESTEC Contract Number 7989/88/NL/PB(SC), Vol. 3, 1991 Personal Communication, William Spetch, Vehicle Integrated Performance Engineering (VIPeR) Office, Mail Code OM, NASA Johnson Space Center, Houston, Texas 77058 Minow, J. I., Neergaard, L. F., Maurits, S., Hwang, K., Suggs, R. M.; “High Latitude Plasma Electrodynamics and Spacecraft Charging in Low-Earth Orbit,” ESA/SCTC, 23-27, April, 2001 Purvis, C. K., Snyder, D. B., Jongeward, G. A.; “Auroral Interactions with ISSA,” NASA Technical Memorandum 106794, December 1994 Cooke, D. L.; “Simulation of an Auroral Charging Anomaly on the DMSP Satellite,” 6th Spacecraft Charging Technology Conference, AFRL-VS-TR-20001578, 1 Sept., 2000 Gussenhoven M. S., Hardy, D. A., Rich, F., Burke, W. J.; Yeh, H-C.; “High-Level Spacecraft Charging in the Low-Altitude Polar Auroral Environment,” Journal of Geophysical Research, Vol. 90, NO. A11, 11,009-11-023, November 1985 Wahlund, J-E., Wedin, L. J., Carrozi, T., Eriksson, A. I., Holback, B. Anderson, L., Laakso, H.; “Analysis of Freja Charging Events: Statistical Occurrence of Charging Events,” WP-130

69

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

40. 41. 42. 43. 44. 45.

46. 47. 48. 49. 50.

Technical Note (SPEE-WP130-TN), Version 2.0, ESA contract 11974/96/NL/JG(SC), 22 February, 1999 Stevens, N. J., Jones, M. R.; “Comparison of Auroral Charging Predictions to DMSP Data,” AIAA 95-0370, 33rd Aerospace Sciences Meeting and Exhibit, January 9-12, 1995, Reno NV. http://www.sec.noaa.gov/pmap/ NASA Johnson Space Center (JSC) Image Science and Analysis Group (ISAG), ISS Surface Damage and Discoloration, August 2001 J. Hyde and E. Christiansen, Meteoroid and Orbital Debris Impact Analysis of Returned International Space Station Hardware, NASA report JSC-29456, May 2001 NASA Johnson Space Center (JSC) Image Science and Analysis Group (ISAG), MMOD strike on Service Module Window 7, http://snisag/stationweb/html/prog_support/sm_win7_strike/sm_win7_strike.shtml D. Lear and E. Christiansen, ISS Debris Risk Assessment Process, Orbital Debris Quarterly Newsletter, Vol.4, Issue 1, Jan.1999. http://hitf.jsc.nasa.gov/hitfpub/main/index.html D.J. Kessler, et al., A Computer Based Orbital Debris Environment Model for Spacecraft Design and Observations in Low Earth Orbit, NASA TM-104825, 1996 J.-C. Liou, et al., The New NASA Orbital Debris Engineering Model ORDEM2000, NASA TP, October 2001 (release pending) Boeder, P.; “Space Station Ionizing Radiation Design Environment,” SSP-30512, Rev C. June 3, 1994. Space Station Program Office, Johnson Space Flight Center, Houston, Texas Baranov D. G., Dergatchov, V. A., Gargarin, Yu. F., Mottl, D. A., Nymmik, R. A.; “About the energy spectra of solar energetic particle events,” Proceedings of the ICRC 2001, 3181 Copernicus Gesellschaft 2001 Tylka, A. J., Cohen, C. M. S., Dietrich, W. F., Maclennan, C. G., McGuire, C. G., Ng, C. K., Reames, D. V.; “energy Spectra of Vety Large Gradual solar Particle Events,” Proceedings of ICRC 2001, 1 Copernicus Gesellschaft 2001 a) J. Feynman, G. Spitale and J. Wang, “Interplanetary Proton Fluence Model: JPL 1991”, I. Geophys Res., 98, No. A8, 13281, 1993, b) Tylka, A. J., Boberg, P. R., Adams, J. H. Jr., Beahm, L. P., Dietrich, W. F., Kleis, T, ; “ The Mean Ionic charge State of soalr energetic Fe Ions Above 200 MeV/nucleon, “ Astrophys. J. Lett. Vol. 444, pp. 109-113, 1995 a) ISS MER Work Instruction – Solar Particle Event Notification, OB-MER-001 Basic, November, 2001, b) Mission Operations Directorate Flight Rule 5191, April 2002 Semones, E., Johnson, S., Weyland, M., Golightly, M.; “Recent Results of Passive Monitoring on International Space Station,” 6th Workshop on Radiation Monitoring of the International Space Station, 12-14 September, 2001, Jesus College, Oxford England Yasuda, H., Komiyama, T., Badhwar, G. D., Fugitaka, K.; “Organ/Tissue Dose Measured with Solid-State Integrating Dosemeters in a Low-Earth-Orbit Space Mission,” paper T-4-2, P-1a-40, 10th Congress of the IRPA, 14-19 May 2000, Hiroshima, Japan Badhwar, G. D., “Shuttle Radiation Dose Measurements in the International Space Station Orbit,” Radiation Research, Vol. 157, No. 1, pp 69-75, January 2002 Vana N., Schoner W., Noll M., Fugger M., Akatov, Y., Shurshakov V.; “Determination of the Absorbed Dose and the Average LET of Space Radiation in Dependence of Shielding Conditions,” Radiat. Prot. Dosim. 85 (1-4) 291-294, 1999 Tylka, A. J., Adams, J. H., Jr., Boberg, P. R., Brownstein, B. Dietrich, W. F., Flueckiger, E. O., Petersen, E. L., Shea, M. A., Smart, D. F., Smith, E. C.; “CREME96: A Revision of the Cosmic Ray Effects on Microelectronics Code,” IEEE Transactions on Nuclear Science, Vol. 44, No. 6, December, 1997 O’Neill, P. M.; Orbiter Avionics Radiation Handbook: Orbiter and GFE Projects, JSC-23160 Rev. A, September 1999, NASA Johnson Space Center Domingo, C., Font, J., Baixeras, C., Font, Ll., Fernandez, F.; “Usage of a corrected Bethe-Bloch formula for charge identification of fast ions with Z > 30 in polycarbonate track detectors,” Nucl. Instr. and Meth. B 146, pp 114-119, 1998 Groom, D. E., The European Physical Journal, C15 (2000) 1, Chapters 6 & 23 Dementyev, A. V., Nymmik, R. A., Sobolevsky, N. M.; Nucleon Spectra Behind 1-100 g/cm2 Aluminum Shielding under Galactic and Solar Cosmic Rays Irradiation, Preprint n 95-28/392. SINP MSU 1995 http://techreports.larc.nasa.gov/ltrs/1995-cit.html , Shinn, J. L., Cucinotta, F. A., Wilson, J. W., Badhwar, G. D., O’Neill, P. M., Badavi, F. F.; “Effects of Target Fragmentation on Evaluation of LET spectra form space Radiation in Low-Earth-Orbit (LEO) Environment: Impact on SEU

70

51. 52. 53. 54.

55. 56. 57.

58.

59. 60. 61.

62. 63.

64. 65.

66.

67. 68. 69.

Predictions,” 32nd Annual International Nuclear and Space Radiation Effects Conference, Madison Wisconsin, 17-21 July 1995 http://www.spenvis.oma.be/spenvis/help/models/letorb/crème.html Groom, D. E., The European Physical Journal, C15 (2000) 1, Chapter 20 Atwell, W.; “ISS Radiation Monitoring: PTE Shielding Distributions, Calculations & Measurements,” 6th Workshop on Radiation Monitoring of the International Space Station, 12-14 September, 2001, Jesus College, Oxford England Seidleck, C. M., LaBel, K. A., Moran, A. K., Gates, M. M., Barth, J. M., Stassinopoulos, Gruner, T. D.; “Singe Event Effect flight Data Analysis of Multilpe NASA Spacecraft and Experiments; Implications to Spacecraft Electrical Design,” http://radhome.gsfc.nasa.gov/radhome/papers/chris.html Personal communication, NASA/JSC/EA44/Truong, A. J.; “STS-95, 88, 96, and 93 Post Flight Analysis Report of Radiation Effectson General Purpose Computer (GPC) Memory Devices.,” EA44-99-022, Oct. 25, 1999 Klausman, A. L.; Effects of Space Flight on Small Personal Computers, Masters Thesis, University of Houston at Clear Lake, December 1995 Truscott, P., Lei, F., Ferguson, C., Gurriaran, R., Neiminen, P., Daly, E., Apostolakis, J., Giani, S., Pia, M. G., Urban, L., Maire, M.; “Development of a Spacecraft Radiation Shielding and Effects Toolkit based on Geant4,” British crown copyright 2000/DERA, and http://wwwinfo.cern.ch/asd/geant4/geant4.html Berger, T., Hajek, M., Schoner, W., Vana, N., Noll, M, Ebner, R., Akatov, Y.,Shurshakov, V., Arkhangelsky, V.; “Measurement of the Depth Distribution of Average LET and Absorber Dose Inside a Water-Filled Phantom on Board Space Station MIR,” Physica Medica, Vol. XVII, Supplement 1, 2000 pp 1-11 http://www.nationalacademies.org/ssb/besrch2.html “Radiation hazards to Crews of Interplanetary Missions; 2 Issues of Concern to NASA” Strategic Program Plan for Space Radiation Health Research, Approved by Arnauld E. Nicogossian, M.D., Associate Administrator, Office of Life and Microgravity Sciences and Applications, NASA Headquarters, October, 23, 1998 Singleterry, R. C., Badavi, F. F., Shinn,, J. L., Cucinotta, F. A., Badhwar, G. D., Clowdsley, M. S., Heinbockel, J. H., Wilson, J. W., Atwell, W., Beaugean, R., Kopp, J., Reitz, G.; “Estimation of neutron and other radiation exposure components in low earth orbit,’ Radiat. Meas. 2001 June; 33(3). Pp 355-360 Kim, M-H. Y., Wilson, J., Thibeault, S., Nealy, J. E., Badavi, F. F., Kiefer, R. L.; “Perforamcne Study of Galactic Cosmic Ray Shield Materials,” NASA Technical Paper 3473, November, 1994 Doll, D., Van Hagen, T., Redler, K., Tooker, J., Baxter, A., Fikaini, M., Schneider, D., Spinos, F., Funk, W.; “Low to High Energy Beam Stops for ATP,” http://accelconf.web.cern.ch/AccelConf/l98/PAPERS/TU4090.PDF , and Maurer, R. H., Roth, D. R., Kinnison, J. D., Jordan, T. M., Hielbronn, L. H., Miller, J., Zeitlin, C. J.; “Neutron Production from Polyethylene and Common Spacecraft Materials,” IEEE Trans. Nucl. Sci., Vol. 48, No. 6, pp 2029-2033, Dec. 22001 Andronenko, M. N., Andronenko, L. N., Neubert, W., Seliverstov, D. M.; “Equilibrium, Isoscaling and Nuclear Isotope Thermometry Related to 1 GeV Proton Induced Nuclear Reactions,” arXiv:nucl-ex/0112014v1 27 dec. 2001 http://www.triumf.ca/safety/rpt/rpt_7/node20.html; http://www.triumf.ca/safety/rpt/rpt_7/node19.html; http://www-esh.fnal.gov/FRCM/Ch08/Ch08.html; Chapter 8 Accelerator Shielding and Radioactivation, Fermi National Laboratory Barschall, H. H., Chadwick, M. B., Jones, D. T. L., Meulders, J. P., Schumacher, H., Young, P. G., Cox, L. J., Hale, G. M., Schrewe, U., Siebers, J. V., Caswell, R. S., DeLuca, P. M., Wambersie, A.; Nuclear Data for Neutron and Proton Radiotherapy and for Radiation Protection, ICRU Report 63, Dec. 1999 Badhwar, G. D., Cucinotta, F. A.; “A Comparison of Depth Dependence of Dose and Linear Energy Transfer Spectra in Aluminum and Polyethylene,” Radiation Research, Vol. 153, pp1-8, 2000 http://spaceresearch.nasa.gov/research_projects/ros/bbndop.html Thomas, D. J., Alvera, A. V.; “Bonner sphere spectrometers-a critical review,” Nucl. Inst. and Meth. .in Phys. Res. A, Vol. 476, pp 12-20, 2002

71 70. Beaugean, R., Burmeister, S., Petersen, F., Reitz, G.; “Dosimetric Measurements on ISS during Quiet and Disturbed Periods,” 6th Workshop on Radiation Monitoring of the International Space Station, 12-14 September, 2001, Jesus College, Oxford England 71. Armstrong, T. W., Colborn, B. L.; “Evaluation of Trapped Radiation Model Uncertainties for Spacecraft Design,” NASA/CR – 2000-210072, March 2000. 72. a) Townsend, J. A., Hansen, P. A.,Dever,, J. A., de Groh, K. K., Banks, B. A., Wang, L., He C.; “ Hubble Space Telescope Metallized TeflonR FEP Thermal Blanket Control Materials: on-orbit degradation and post-retrieval analysis,” High Performance Polymers, Vol. 11, No.1 March 1999, b) Lilley, J.; Nuclear Physics, Principles and Applications, John Wiley and Sons, New York, 2001, pp. 139-137, figure 5.5, c) Katz, l., Penfold, A. S..; Reviews of Modern Physics, Vol. 24, page 28, 1952 73. Dever, J. A., de Groh, K. K., Banks, B. A., Townsend, J. A.; “Effects of Radiation and Thermal Cycling on Teflon FEP,” High Performance Polymers, Vol. 11, pp123-140, 1999 74. Buehler, P., Zehnder, A., Desorger, L., Hajdas, W., Daly, E., Adams, L.; “Measurements of the Radiation Belts from Mir and STRV,” http://www.estec.esa.nl/wmwww/wma/R_and_D/rem/cherbs97www/rem97.pdf 75. Schoner, W., Hajek, M., Noll, M., Ebner, R., Vana, N., Fugger, M., Akatov, Y., Shurshakov, V., Arkhangelsky, V.; “ Measurement of the Dose Depth and LET Distribution at the Surface and Inside of Space Station Mir,” http://www.magnet.oma.be/srew/08.pdf 76. Mackay, G. F., Thomson, I., Ng, A., Sultan, N.; “Applications of MOSFET Dosimeters on MIR and BION Satellites,” IEEE Trans. Nucl. Sci., Vol. 47, pp 2048 - 2051, December 1997. 77. Trinks, H.: Experimental Investigation of the Exhaust Plume Flow Fields of Various Small Bipropellant and Monopropellant Thrusters. AIAA Paper 87-1607, June 1987. 78. Rebrov, S., et al: “Monitoring of Contamination Actions of Flames and Exhausts of Propulsion Systems and Power Plants on Elements of Long-Term Space Stations to Minimize These Actions – Final Report,” Contract No. 251-5208/95, Russian Space Agency – Keldysh Research Center, Moscow, 30 January 1998 79. Soares, C.E.; International Space Station Provisional Plume Contamination Model (Phase I Report), Boeing, SSCN 2208 Phase I Report Deliverable. Houston, Texas. – September 30, 1999 80. Koontz, S.; Melendez, O.; Zolensky, M.; and Soares, C.: SPIFEX Contamination Studies. NASA Johnson Space Center Report JSC-27399, May 1996 81. Rebrov, S., Gerasimov, Y.; “Investigation of the Contamination Properties of Bipropellant Thrusters,” AIAA 2001-2818, 35th AIAA Thermophysics Conference, June 11-14, 2001 82. Soares, C., Mikatarian, R., Barsamian, H., Rauer, S.; “International Space Station Bipropellant Plume Contamination Model,” AIAA 2002-3016, 8th AIAA/ASME Joint Thermophysics and Heat Transfer Conference, June 24-17, 2002, St. Louis Mo. 83. Alred, J.W., Smith, L.N., Wang, K.C., Lumpkin F.E., Fitzgerald, S.M.; “Modeling of Space Shuttle Orbiter Waste Water Dumps,” AIAA Paper 98-2588, 7th AIAA/ASME Joint Thermophysics and Heat Transfer Conference, Albuquerque, NM, June 15-18, 1998. 84. SPHINX Impact Simulation Results, ref: Stellingwerf, Wingate; “Impact Modeling with Smooth Particle Hydrodynamics,” Int. J. Impact Engineering, Vol. 14, pp 707-718, 1993 85. http: history.nasa.gov/SP-404/ch7.htm 86. Thomas, D. and Peterson, C.E., “Space Station External Contamination Control Requirements,” SSP 30426, Revision D, NASA Johnson Space Center, Houston, Texas, 1994 87. Standard Test Method for Contamination Outgassing Characteristics of Spacecraft Materials, E1559, American Society for Testing and Materials, West Conshohocken, Pennsylvania, 2000 88. Alred,J.;” Outgassing Modeling and Preliminary Results,” ISS External Contamination Technical Interchange Meeting, Houston Texas, May 3-7, 1999 89. Hakes, Charles, The NASAN Program for Spacecraft Contamination Analysis – A User’s Guide, NASA Johnson Space Center, Houston, Texas, 1997 90. LeBeau, G. J., Lumpkin, F. E.; "Application Highlights of the DSMC Analysis Code (DAC) Software for Simulating Rarefied Flows", Comput. Methods Appl. Mech Engrg., 191 (2001), 595609 91. Soares, C.E, Mikatarian, R., Scharf, R., Miles, E.; “International Space Station Flights 1A/R – 6A External Contamination Observations and Surface Assessment, International symposium on Optical Science and Technology, SPIE’s 47th Annual Meeting, July 7-11, 2002, Seattle Washington 92. SSP-30233, Rev. F, “Space Station Requirements for Materials and Processes,” NASA Johnson Space Center, Houston, Texas, 1998

72 93. Koontz, S. L., Leger, L. J., Visentine, J. T., Hunton, D. E., Cross, J. B., Hakes, C. L.; “EOIM-III Mass Spectrometry and Polymer Chemistry: STS-46, July-August 1992,” J. Spacecraft and Rockets, Vol. 32, No. 3, pp 483-495, May, June 1995. 94. Linton, R. C., Whitaker, A. F., Finckenor, M. M.; “Space Environment Durability of Beta cloth in LDEF Thermal Blankets,” LDEF Materials Results for Spacecraft Applications, NASA CP-3257, Oct. 1992 95. Koontz, S. L., Jacobs, S., Le, J.; “Beta Cloth Durability Assessment for Space Station Freedom Multi Layer Insulation Blanket Covers,” NASA-TM-104748, March, 1993 96. Banks, B. A., Snyder, A., Miller, S. K., Demko, Rikako; “Issues and Consequences of Atomic Oxygen Undercutting of Protected Polymers in Low Earth Orbit,” ICPMSE-6, 6th International Space Conference; Toronto, Canada, May 1-3, 2002, B-8 97. Koontz, S., King, G., Dunnett, A., Kirkendahl, T., Linton. R., Vaughn, J.; “Intelsat Solar Array Coupon Atomic Oxygen flight Experiment,” J. Spacecraft and Rockets, Vol. 31, No. 3, pp 475481, May, June 1994

* Author to whom correspondence should be addressed. ES4/Dr. Steve Koontz, ISS Environments Manager NASA Johnson Space Center Houston, Texas, 77058 Phone:(281)-483-8860 & Fax: (281)-244-1301 e-mail: [email protected] (1)

Structural Engineering Division Mail code ES4 NASA Johnson Space Center Houston, Texas 77058

(2)

The Boeing Company Mail code HA1-20 13100 Space Center Blvd. Houston, Texas 77059-3599

(3)

Dynacs Engineering Co. Inc. 1110 NASA Rd. 1 Houston, Texas 77058

(4)

Solar System Exploration Division Mail Code SX2 NASA Johnson Space Center Houston, Texas 77058

(5)

NASA Marshall Space Flight Center Engineering Physics Mail Stop ED31 Huntsville, Alabama 35812

(6)

Jacobs Sverdrup, MSFC Group MSFC/ED44, Huntsville, AL 35812

PHOTOCONDUCTIVITY IN TRANSPARENT ARCPROOF COATINGS

T. CASHMAN, J. KAUR, L. K. MUHIEDDINE, M. SHANBHAG, S. H. UBAID, BRYAN WELCH, JYOTHI VEMULAPALLI, AND P. D. HAMBOURGER Cleveland State University Cleveland, OH 44115 USA

Abstract:

Highly transparent thin films with moderately high sheet resistivity ~108 ohms/square (Ω/ ) are needed for protection of photovoltaic arrays and other surfaces from static charging in space. Previous work at NASA Glenn Research Center has shown that codeposited mixtures of indium tin oxide (ITO) with MgF2 or SiO2 are promising for this application. We find that exposure of these films to low-intensity blue light substantially reduces the sheet resistivity, which then takes at least several hours to recover.

1. 0 Introduction Nonconductive spacecraft surfaces need protection from electrostatic charging due to particles in the solar wind. The coating must have a maximum sheet resistivity less than ~108-109 Ω/ and, in the case of photovoltaic arrays, must be as transparent as possible [1]. Indium tin oxide (ITO) has been used [2] but is actually too highly conductive for this application. The large number of free carriers in ITO results in unnecessary absorption and reflection of light, reducing solar array efficiency. In addition, accidental contact between power wiring and a low-resistivity coating could lead to large current flow between the power system and the conductive plasma present in low earth orbit. To produce a better coating for these applications, previous workers investigated mixed films made by co-depositing ITO with MgF2 or SiO2, both highly transparent insulators [1,3]. These films had the desired sheet resistivity, appeared to have enhanced atomic oxygen durability, and were considerably more transparent than ITO films of similar thickness, as shown in Figure 1. However, the sheet resistivity increased considerably in the first year after deposition.

73

Total Solar Transmittance

74 0.9 (a)

0.88 0.86 0.84

Avg. film thickness 650 Å Transmittance of SiO2 substrate 0.93

0.82 0.8 0

10

20

30

Sheet Resistivity (Ω/ )

Calculated Vol. % MgF 2 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+

10 09 08 07 06 05 04 03 02

(b)

Avg. film thickness 650 Å 0

Figure 1.

10 20 Calculated Vol. % MgF 2

30

Solar transmittance (a) and sheet resistivity (b) of ITO-MgF2 films vs composition (from Ref. 3) .

Due to renewed interest in these coatings, we are conducting a further investigation of their electrical properties as a function of deposition parameters and ambient conditions. Most notably, we find that the sheet resistivity of ITO-MgF2 and ITO-SiO2 are strongly affected by exposure to low-intensity light near the blue end of the visible spectrum. This paper focuses mainly on ITO-MgF2, which was previously found to have superior solar transmittance. 2. 0 Experimental Techniques Films ~700 Å thick were deposited on 0.75 mm thick fused silica substrates by simultaneous operation of two RF magnetron sputter guns. One target was 90%/10% indium/tin oxide, the other was MgF2. The guns were driven by separate 13.56 MHz RF generators; film composition was adjusted by changing the relative power outputs of the generators. Sputtering was carried out in Ar gas at 3 mTorr pressure. We did not

75 feed oxygen or air into the chamber because this system produces highly transparent conductive ITO films without it. The background pressure with Ar gas turned off and pump throttle valve set as for deposition was <3x10-5 Torr. (Pressure with Ar off and throttle wide open was typically <1x10-6 Torr). Each substrate was covered with an aluminum mask to produce a sample measuring 0.3x1.9cm2 with electrical contact arms along the edge. Substrate temperature during deposition was ~30 0C. Thus it is likely that the samples are amorphous. To estimate the composition of ITO-MgF2 samples, we measured the deposition rate from each gun at several power levels. Film were determined with a quartz crystal monitor, calibrated against a surface profilometer. Due to the uncertainties of this method, our thickness and composition values are of rather low accuracy. Electrical resistance measurements were made by the four-terminal method to eliminate the effect of contact resistance. Measuring current was 2 pA-100 µA. Guarded cabling and high input resistance electrometers minimized errors due to the high resistance of some samples. Measurements reported here were carried out at room temperature in ambient air. Photoconductivity experiments were performed with the sample in a light-tight box equipped with red, green, and blue light emitting diodes (LED’s) that have relatively narrow optical bandwidth. Before reaching the sample, the light passed through the fused silica substrate, which is highly transmitting in visible wavelengths. Each LED had a molded-in lens that produced a rather well collimated beam. Light intensity at the sample was estimated by measuring the beam’s diameter and total power (using an optical power meter). The results are: Red: 40 W/m2 @ ~660 nm (photon energy ~1.9 eV) Green: 5 W/m2 @ ~570 nm (photon energy ~2.2 eV) Blue: 20 W/m2 @ ~430 nm (photon energy ~2.9 eV)

76 3. 0 Results Sheet resistivity of ITO-MgF2 samples is plotted vs approximate MgF2 content in Figure 2.

Figure 2. Sheet resistivity of ITO-MgF2 samples vs estimated MgF2 content.

There is an abrupt increase in resistivity when MgF2 content goes above ~35 vol. %. These results are qualitatively similar to those of [3] (Figure 1) although we observe a more abrupt increase that sets in at a higher insulator concentration. These differences may simply reflect the different technique used in [3] (ion beam deposition from a two-component target). Note also that our data were taken on samples recently exposed to room light, which may reduce sheet resistivities by as much as one order of magnitude from their “dark” values. Sheet resistivity of several ITO-MgF2 samples is plotted vs time after deposition in Figure 3. As observed by the authors of [1, 3], we find that the resistivity of some samples increases after deposition. However, it is likely that all samples were exposed to strong ultraviolet radiation during deposition. The observed “instability” might then be related to the long-term photoconductive effects to be presented shortly.

77

Figure 3. Sheet resistivity of several ITO-MgF2 samples vs time since deposition.

The effect of visible light is shown in Figure 4, where we plot sheet resistivity vs time while the LED’s were turned on and off one at a time. Note that the influence of blue light is greatest on the sample with the highest dark resistivity. However, the rate of recovery from blue light exposure is greatest in the high-resistivity sample. Also note that the intensity of blue light, ~20 W/m2, is far smaller than the total solar intensity in earth orbit (~1400 W/m2). Similar data for an ITO-SiO2 film are shown in Figure 5. Here too the sheet resistivity is reduced by exposure to blue light, then increases after the light is turned off. It is clear from Figures 4 and 5 that the recovery of resistivity after blue light is turned off happens quite slowly (taking over one day in some cases). This time scale is much longer than, for example, typical electron-hole recombination times in semiconductors.

78 ITO-MgF2 2

Red: 40 W/m @ ~660 nm Green: 5 W/m2 @ ~570 nm Blue: 20 W/m2 @ ~430 nm

Sheet Resistivity (Ω/ )

1.E+11

RED OFF

1.E+10

GREEN ON

GREEN OFF

BLUE ON

BLUE OFF

RED ON

1.E+09

1.E+08 0

100

200 300 Time (minutes)

400

500

Sheet Resistivity (Ω/ )

1.E+09

RED ON

1.E+08

RED OFF

GREEN ON

GREEN OFF

BLUE ON

BLUE OFF

1.E+07

1.E+06 0

100

200

300

400

Time (minutes) 1.E+08

Sheet Resistivity (Ω/ )

RED ON

RED OFF

GREEN ON

GREEN OFF

BLUE ON

BLUE OFF

1.E+07

1.E+06 0

100

200

300

Time (minutes)

Figure 4 . Effect of light exposure on ITO-MgF2 (~700 Å thick).

400

79

Figure 5 . Effect of light exposure on an ITO-SiO2 film.

4. 0 Discussion ITO is a semiconductor with energy gap ~3.5 eV. Since photoconductivity in semiconductors usually involves excitation of electrons from the valence band to the conduction band, it would require a minimum photon energy of ~3.5 eV. Thus it is surprising that we observe photoconductivity with the blue (photon energy ~2.9eV) and green (~2.2 eV) LED’s. It is possible that these sub-bandgap photons could excite electrons to the conduction band from discrete energy levels in the gap (such as the donor level created by oxygen vacancies) or from localized gap states found in disordered semiconductors [4]. However, we would expect that once light is removed, electrons excited by any of the above mechanisms would quickly leave the conduction band, filling holes in the valence band or being captured in donor states or localized states. This is not consistent with the very gradual resistivity increase observed in our films when light is turned off (Figures 4 and 5). Instead, the visible-light photo effects in our samples may be related to a “photoreduction” phenomenon in amorphous and microcrystalline indium oxide films reported by Fritzsche and co-workers [4]. They observed a dramatic long-term reduction in resistivity upon exposure to ultraviolet radiation that could be reversed by exposure to ozone. This effect, which the authors suggest could be observed at photon

80 energies as low as ~3eV, appears to be related to the presence of oxygen vacancies in amorphous indium oxide. 5. 0 Conclusions The sheet resistivity of ITO-MgF2 and ITO-SiO2 films decreases substantially in the presence of low-intensity blue light. The effect is similar to one observed by others [4] in amorphous and microcrystalline indium oxide (which bears many similarities to ITO). Thus, the photoconductivity of our films may be intrinsic to ITO and may require examination of other transparent conducting oxides if it is a potential problem in space. 6. 0 Acknowledgments The authors gratefully acknowledge the support of NASA Glenn Research Center, Cooperative Agreement NCC3-740. We thank Bruce Banks, Joyce Dever, and Sharon Miller for many helpful discussions and Rikako Demko for valuable technical assistance.

7. 0 References 1. Joyce A. Dever et al., NASA Technical Memorandum 1998-208499, August 1998. 2. C. K. Purvis, H. B. Garrett, A. C. Whittlesey, and N. J. Stevens, NASA Technical Paper 2361, September 1984. 3. A. J. Adorjan et al., Mat. Res. Soc. Symp. Proc. 551, 101 (1999). 4. B. Claflin and H. Fritzsche, J. Electron. Mater. 25, 1772 (1996) and references therein.

EFFECTS OF SPACE ENVIRONMENT FACTORS ON OPTICAL MATERIALS FOR SPACE APPLICATION Hai LIU, Hongbin GENG, Shiyu HE, Shiqin YANG and Dezhuang YANG Harbin Institute of Technology, Harbin, 150001, P.R. China E-mail: [email protected] V. V. ABRAIMOV Kharkov National University, Kharkov, 60164, Ukraine Huaiyi WANG The 508 Institute, Space Technology Academy of China, Beijing, 100076, P.R. China Abstract The effects of proton and electron radiations as well as thermal cycling on properties of optical materials for space application were studied using a complex simulator for space environment. The energy of protons and electrons ranged from 60 to 180 keV that is within the energy range of the Earth radiation belt. The temperature interval of thermal cycling, ∆T, was kept between 80 and 400K. The obtained results showed that if the reflectors were not grounded directly, charging-and-discharging could cause severe damage of the reflector surface. Bulging or exfoliating of the surface films would be more severe due to synergistic effect of proton radiation and thermal cycling on the reflectors. For JGS3 quartz glass, the proton and electron radiations result mainly in the ultraviolet absorption at 210-240 nm, while an absorption in visible range also occurs in the case of high radiation fluence. 1. 0 Introduction It is well known that spacecraft placed into geostationary orbits are exposed to the environment of vacuum, thermal sink, thermal cycling and irradiation by electrons and protons from the Earth radiation belt, resulting in degradation of optical components or materials such as main mirrors and optical windows [1-3]. Since the flux of charged particles decreases with increasing their energy in the Earth radiation belt, particles with energy less than 200 keV have large flux up to 108 part/cm2·s., hence The energy absorbed by space materials is mainly concentrated in the surface layers because of the known fact that the lower the particle energy, the shorter the penetration distance of the particles into materials. The effects of low energy particles on optical material properties are, therefore, of high importance and could not be neglected, especially for spacecraft with long service lifespan.

81

82 The reflective property of reflectors is reduced after a certain irradiation fluence by protons and electrons. The proton radiation may cause a specific blistering phenomena [4]. In glass materials, the main influence exerted by the radiation is a decrease in material transmittance in specific spectral ranges with a following coloration/discoloration of the materials. The coloration effects result from specific point defects inside the glass, namely color centers, and their accumulation. In the quartz glass, an important defect is the E′ centers that can lead to some intrinsic absorption bands [5-7]. The present study was conducted to investigate the property changes of beryllium-base reflectors and JGS3 quartz glass under the environment of vacuum, cold black, thermal cycling and irradiation by protons and electrons with energies 60-180 KeV. 2. 0 Experimental The experiment was performed in a specially designed system that can simulate concurrently multi-environments of vacuum, cold blackness (77K), thermal cycling (77-403K), and radiation effects by protons and electrons. The energy of protons and electrons ranged from 60 to 180 keV. Figure 1 presents schematically the configuration of the tested beryllium-based reflector with a diameter of 60mm. The reflector is built on a Be substrate, followed by a quartz glass layer on which an Ag reflecting film and a SiO2 protective film are deposited. To investigate the thickness influence of the SiO2 film, two different thicknesses of 50nm and 100nm, were used in the experiments. The above reflector was irradiated with protons and electrons separately and synergistically by both beams, and then tested under thermal cycling between 80-400K. Another experiment was performed on the 20x2mm specimens of JGS3 quartz glass with the impurity less than 5·10-3 %. The quartz glass was irradiated with electrons and protons with different energies and fluences. During the irradiation tests, two different modes for grounding the specimens were used: in one the specimen was grounded directly through the facility platform; in the other, the samples were grounded through a capacitor with a 20 kV breakdown potential. The former set-up is adopted in general radiation test, and the latter is closer to the real situation for a spacecraft in orbit. SiO2 protective film 100nm(50nm) Ag reflector100nm Quartz glass 0.5mm Be substrate10mm Figure1. Configuration and dimension of reflecting mirrors

83 3. 0 Results and Discussions 3.1 INVESTIGATION ON BERYLLIUM-BASE REFLECTOR Once a certain fluence is reached, either proton or electron irradiation causes degradation of reflectivity and imaging quality. In particular, under proton radiation, surface defects may form on the reflector. The morphology, the amount of the defects and their distribution on the surface, resulting from integrated irradiation by protons and electrons, are related to reflector configuration and its grounding mode. Fig.2 presents surface morphologies of the reflector with a 100nm thick SiO2 protective film after the irradiation by protons and electrons. In the case of the first grounding mode, point defects, 1-10 µm in size were formed and distributed homogeneously over the specimen surface (Fig.2b). A critical radiation fluence Φ = 6·1015/cm2 was found to initiate this type of defects. For the second grounding case, in addition to the above mentioned defects a dense network of distributed defect bands appear on the irradiated surface. The bands seem as streams to flow regularly from the specimen center to its edge (Fig.2c). In some regions close to the reflector edges, the reflective film was found to delaminate from the substrate (Fig.2d).

a

30µ

b

30µ

c

30µ

d

30µ

Figure 2. Surface morphology of the first reflector before and after proton and electron irradiation. a-original sample; b-irradiated sample, Ep=Ee=160keV, Ɏp=Ɏe =5×1016part/cm2, the first grounding mode, c, d-irradiated sample, Ep=Ee=160keV, Ɏp=Ɏe =5×1016part/cm2, the second grounding mode

Point defects that are due to protons appear homogeneously on the irradiated surface of the reflector, but their nature is different for reflectors with different SiO2 film thickness. On the reflector with the 100nm thick protective film, the defects are in the form of small crater-like pits (Fig.2b), while on reflectors with a 50nm thick protective film, there is a small amount of irregularly shaped bubble-like bulges with an average size over 100 µm (Fig.3b).

84 After the proton irradiation experiments were followed by thermal cycling in the range of ∆T=77~403K it was found that the optical surface of the reflector was damaged seriously. In some vulnerable areas, the reflective and protective films peeled completely off with extensive bulging occurred almost over the whole irradiated surface (shown in Fig.4).

a

30µm

b

100µm

Figure 3. Surface morphologies of the second reflecting mirror before and after proton radiation a - original sample; b - radiated sample, Ep=Ee=130keV, Ɏp =1016part/cm2

a

100µ

b

100µ

Figure 4. Surface morphologies of the second reflector after proton irradiation and 8 times thermalcycling between 80 and 403 K: a) peeling of the Ag film, b) partial bulging of the Ag film

From the above results, it is reasonable to imply that under proton irradiation, the formation of pits and bulges on the reflector would originate from the same mechanism, namely hydrogen aggregation and the expansion of formed microbubbles. Generally, the thickness of the protective film is under 0.1 µm that is less than the penetrating depth (about 1-2 micron) of protons into materials. When large amounts of protons are implanted into the reflector, they aggregate to form hydrogen gas and then micro-bubbles under the protective film. The micro-bubbles keep growing up to destroy the films during the irradiation process. Therefore, the different surface defects formed for different reflectors were related to the thickness difference of the SiO2 protective film. Much thicker protective films on the reflector would act favorably to reduce the plastic deformation of the reflecting Ag film, thus producing only small pits under the pressure of hydrogen bubble. For the case of thin protective film, the Ag film could deform freely to form fairly larger bulge features.

85 Using the radiation-induced bubble theory presented by Martynenko [9], the critical radiation fluence for macro-bubble formation could be roughly estimated as:

Φ = (σ

F

H ) • δR 2 ( E ) ,

where σF is the ultimate strength of the material; H - the implantation energy of hydrogen atoms; δR2 - the mean-square deviation of proton penetrating depth. Therefore, from the experimental parameters, the theoretical critical fluence could be estimated as Φ=4×1015/cm2. The calculated value is well in accord with the observed value. In the second grounding mode regime, the severe damage of the reflector results not only from the bubbling mechanism, but mainly from the electrostatic effect. The charged particle implantation could create surface charging of materials, and the implantation of heterogeneous charges into the materials could even cause bulk charging., The charge accumulation, no matter what kind of charging mode is created, would induce discharging effects, and, associated with them, very high instantaneous discharging currents. The discharging phenomena will be accompanied by a large amount of heat and lead to severe damage of reflector surface. Due to the differences in thermal expansion coefficients between the layers in the reflector, the thermal cycling may cause deformation of the various layers on the reflector and introduce interlayer internal stress. Once the interlayer stress is high enough, it would affect the initiation of delamination of the protective film and its eventual destruction. In addition, the hydrogen gas accumulated within the reflector from proton irradiation could diffuse to the surface and boundaries and cause the bulging and peeling of large areas of the surface films. 3.2 INVESTIGATION ON QUARTZ GLASS Spectral transmittance of JGS3 quartz glass before and after irradiation by protons and electrons with different energies and fluence is shown in Fig. 5. It should be mentioned that when proton irradiation fluence reaches 1015 part/cm2, the optical properties of JGS3 glass degraded distinctly. Such changes occur mainly in ultraviolet region. Spectral transmittance tends to decrease with increasing the radiation fluence monotonically with such degrading trend spreading gradually to the visible region. The spectral behavior of JGS3 quartz glass after electron irradiation changes similarly to that after proton irradiation, i.e. the change starts from the ultraviolet region and spreads gradually to the visible region with increasing irradiation fluence. The transmittance also decreases monotonically with the fluence. However, the electron radiation only for the 5·1014 part/cm2 fluence results in an obvious change in the spectra of the quartz glass. Moreover, the electron energy affects stronger the spectral performance of the glass, i.e. at higher electron energies, the reflector transmittance is reduced.

86 100 1

Transmittance, %

Transmittance, %

90

100

2

80

70

3

Ep=180 keV 1- before radiation 2- Ɏ=2×1015part/cm2 3- Ɏ=1016part/cm2 4- Ɏ=2×1016part/cm2

4

60

90

1 2 3

Ep=180 keV 1- before radiation 2- Ɏ=1015part/cm2 3- Ɏ=2×1015part/cm2 4- Ɏ=1016part/cm2 5- Ɏ=2×1016part/cm2

80 4

70

5

60

a) 50 200

300

400

500

600

700

800

b) 50 200

900 1000

300

400

Wavelength, nm

Transmittance, %

Transmittance, %

1 2 3 4 Ee=100 keV 1- before radiation 2- Ɏ=1015part/cm2 3- Ɏ=2×1015part/cm2 4- Ɏ=1016part/cm2

70

60

90

800

900 1000

300

400

500

600

Wavelength, nm

700

800

2 3

80

70

60

c) 50 200

700

1

enter text here

80

600

100

100

90

500

Wavelength, nm

50 200

4 Ee=140 keV 1- before radiation 2- Ɏ=5×1014part/cm2 3- Ɏ=1015part/cm2 4- Ɏ=5×1016part/cm2

300

400

500

d) 600

700

800

Wavelength, nm

Figure 5. Spectral transmittance of JGS3 quartz glass before and after proton and electron irradiation: a) proton irradiation, Ep=80keV; b) proton irradiation, Ep=180keV; c) electron irradiation, Ee=100keV; d) electron radiation, Ee=140keV

The relationships between the optical density and the wavelength for JGS3 quartz glass after proton and electron irradiation were investigated and the results are shown in Fig. 6. As can be seen from Fig. 6, a fairly small irradiation fluence of protons induces just one absorption band near 210nm and the absorption peak at 210nm increases with increasing fluence, with some absorption also occuring at longer wavelengths. The situation is more complicated in the case of electron irradiation of JGS3 glass. Two absorption bands at 204nm and 215nm appear. Furthermore, when radiation fluence reaches 2·1016 part/cm2, three other absorption bands appear at 240nm, 300nm and 550nm. It is generally agreed that, for quartz glass, optical absorption at 210-240nm is caused by E′ center [5-7]. For incident particles with energies < 0.2 MeV, the energy lose is mainly from the interaction with outer shell electrons of the glass atoms and their collision probability with atom’s nucleus is low. Accordingly, few vacancies form in the glass.

87 0.20

0.30

a)

Ep=80keV 15

Optical density change

Optical density change

0.36

2

1- Ɏ=2×10 part/cm

0.24

1- Ɏ=2×1015part/cm2

0.18 2

0.12 0.06 0.00 200

1

300

400

500

600

Wavelength, nm

700

800

b)

Ee=100keV

0.15

15

2

1- Ɏ=2×10 part/cm

1- Ɏ=2×1016part/cm2 0.10

0.05

2 1

0.00 200

300

400

500

600

700

800

Wavelength, nm

Figure 6. Change in optical density of JGS3 quartz glass vs. wavelength after radiation of protons (a) and electrons (b)

In this case, the E′ center in quartz glass could be formed through a process where a silicon atom traps an electron. Meanwhile, a vacancy forms at the site of the oxygen atom. The process can be shown as follows: p ≡ Si − O − Si ≡  → ≡ Si + O − Si ≡

The wavelength of absorption peak changes in a wide range with increasing irradiation fluence of protons that is indicative of formation of defects through a complicated synthesis of defects under higher irradiation fluencies. Figure7 presents the experimented results obtained from an electron-spin resonance investigation of the JGS3 quartz glass. Analyzing the results in Fig.7(a), it can be suggested that the color centers induced by proton radiation are related to hydrogen, namely the E′ centers affected by hydrogen. A few probable reactions for hydrogen inside the quartz glass are shown as follows: p ≡ Si + H  → Si − H p ≡ Si − O + H  → ≡ Si − OH

In the case of electron irradiation, the electron-spin resonance experiments indicate that electron irradiation induces mainly E′ centers in quartz glass. Under higher radiation fluencies, the two additional absorption bands observed at 300nm and 550nm may be related to the defects caused by impurity elements such as Ge, Al, etc.. Since the impurity concentration in JGS3 quartz glasses is very low, the lower fluence of electrons would induce just negligible amount of impurity centers in them. The optical absorption caused by the impurity centers could not be observed unless the electron radiation fluence was increased to a certain level.

88

g=2.00107; Hpp=3·10-4T

a)

g=2.00166; Hpp=3·10-4T

b)

Figure 7. Electro-spin resonance spectra of JGS3 glass after proton and electron radiation: a) 1016part/cm2 proton radiation; b) 5×1015part/cm2 electron radiation

4. 0 Conclusions Irradiation by protons only or by synergistic beams of protons and electrons, when irradiation fluence exceeds Φ=6×1015/cm2 , results in the formation of bulges and pits on Be-base reflectors with both thin and thick SiO2 protective films. In the case of indirect grounding of the reflector, the charging/discharging effects could cause severe damage of the reflector surface. Synergistic application of proton irradiation and thermal cycling to the reflectors causes stronger bulging and/or delamination effects of the surface films as well as a decrease in surface reflectivity and deterioration of imaging quality. For JGS3 quartz glass, the proton and electron irradiation results mainly in increased ultraviolet absorption at 210-240nm. In cases of high radiation fluence, certain absorption in the visible range also occurs. For radiation fluencies lower than Φ=2×1016 part/cm2, the spectral transmittance of glass decreases monotonically with increasing fluence. The critical irradiation fluence to cause a distinct degradation of the spectral property is ΦP=1015 part/cm2 for the protons and Φe=5×1014 part/cm2 for electrons. 5. 0 References 1. 2.

M.D.Blue. Degradation of Optical Components in Space. NASA N94-31029: 217-225 C.A.Nicolettu, A.G.Eubanks. Effect of Simulated Space Radiation on Selected Optical

3.

Materials. NASA TN D-6758. N72-24836, 1-14 Jacob Becher, Walter Fowler. The Simulated Space Proton Environment for Radiation Effects on Space Telescope Imaging Spectrograph. NASA-CR-190618, N92-33745, 1-43

89 4.

V.V. Abraimov, F. Lura, L. Bone, N.I. Velichko., A.M. Markus, N.N.Agashkova, L.A. Mirzova. The Research of Blistering and Flaking in Space Optics Materials Under Space Environment.

5.

Space Science and Technology 1995, 1(2-6), 39-54 J.L.Allen, N.Seifert, Y.Yao, R.G.Albridge, A.V.Barnes, N.H.Tolk. Point

6.

Optical Materials Exposed to the Space Environment. NASA N95-27646, 1131-1132 R.R.Gulamova, E.M.Gasanov, R.Alimov. Proton - induced changes of optical properties and defect

Defect Formation in

formation in quartz glasses. Nuclear Instruments and Methods in Physics Research B. 1997, 127/128, 497-502 7.

8.

Christopher D., Marshall, Joel A. Speth, Stephen A. Payne. Induced optical absorption in gamma, neutron and ultraviolet irradiated fused quartz and silica. Journal of Non-Crystalline Solids. 1997, 212, 59-73 V.V.Abraimov, Yang Shiqin, He Shiyu, Yang Dezhuang, L.K.Kolybaev, E.T.Verhovtseva. Complex Simulation of the Effect of Eight Space Environment Factors on Space Vehicle Materials. The Fifth China-Russian-Ukrainian Symposium on Space Science and Technology Held Jointly With The First International Forum on Astronautics and Aeronautics Symposium Proceedings, Harbin P.R.China, 2000, Harbin Institute of Technology, 706~713

9.

Y.V. Martinenko. Theory of Blistering. Russian Research Center Kurchatov Institute. Moscow, 1979, 40.

This page intentionally left blank

A STUDY OF SYNERGISTIC RADIATION EFFECTS OF PROTONS AND ELECTRONS ON TEFLON FEP/Al DEGRADATION

DEZHUANG YANG, CHUNDONG LI, HONGBIN GENG, SHIYU HE, SHIQIN YANG Space Materials and Environment Engineering Lab. Harbin Institute of Technology, Harbin 150001

Abstract Aluminized Teflon FEP film (fluorinated ethylene-propylene) is commonly used on exterior spacecraft surfaces for thermal control. Synergistic effects of proton and electron radiations on aluminized Teflon FEP degradation were investigated in terms of ground-based simulation testing. The energy of protons and electrons was chosen in the range of a few tens keV. The results showed that both the radiations of protons and electrons led to forming an absorption band in the near ultraviolet to visual regions, and the electrons also resulted in a decrease of spectral reflectance with fluence in the nearinfrared range. The effect of synergistic radiation of protons and electrons did not show additivity on the reflective property of aluminized Teflon FEP film. The effect of simultaneous radiation of protons and electrons was lower, while the additive radiation effect given by the two types of charged particles higher. Under a given fluence of radiation, the overall changes in spectral reflectance were independent from the radiation sequence of protons and electrons. The changes in spectral reflectance with radiation fluence and the XPS spectrum for the simultaneous radiation were similar to those for the sequential electron to proton radiation. 1. 0 Introduction It is known that thermal control coatings are widely used on exterior surfaces of spacecrafts to maintain a given thermal regime [1]. Aluminized Teflon FEP film second surface mirrors are a kind of thermal control coatings, which possess unique thermal emittance and high resistance to space environment [2]. The thermal control surface coatings of spacecrafts would be subjected to a damage caused by the radiation of electrons and protons in the Earth radiation belts, leading to a degradation of optical properties, such as the spectral reflectance ρλ and solar absorption as [3,4]. Since such a

91

92 degradation would destroy the thermal balance of spacecrafts, it is necessary to examine the property evolution of thermal control coatings in geostationary orbit through inflight [5,6] and ground - based simulation [7] experiments. The aim of this paper was to reveal the synergistic radiation effects of protons and electrons on the aluminized Teflon FEP film degradation in reflective property, in terms of ground simulation. The change in spectral reflectance (∆ρλ) was examined. In order to explore the mechanism of degradation, the formation of micro-defects caused by synergistic radiations of protons and electrons were also investigated. 2. 0 Experimental The aluminized Teflon FEP film was used as the experimental material. The specimens were fabricated in a form of second surface mirror consisting of Teflon FEP film with 50µm thickness, an Al reflecting layer with 80~100nm thickness and a thin protection layer of SiOx. The Al layer was deposited in vacuum. The specimens were adhered to the Al alloy substrate of 15mm in diameter. The ground simulation tests were conducted using a facility, which can simulate the radiation of protons and electrons with the energy less than 200keV, independently and simultaneously. During the tests, the temperature of specimens was maintained at 298K using electric heating and water-cooling. The energy of protons and electrons was chosen in the range of a few tens keV. The proton flux is ϕp=5×1011cm-2s-1, and that of electrons ϕe=1×1012cm-2s-1. The vacuum in the test chamber was kept at 10-5Pa. The spectral reflectance ρλ of specimens before and after the radiations in the wavelength region of 0.2~2.5mkm was in-situ measured using an integration sphere. A mass spectrometer of MONOPOL type was used to in-situ analyzes the outgassed products from the specimens during radiations. After the radiations, the specimens were analyzed by means of X-ray photoelectron spectroscopy (XPS). 3. 0 Results and Discussion 3.1 CHANGES IN REFLECTIVE PROPERTY Figure 1(a) and 1(b) show the changes in spectral reflectance of aluminized Teflon FEP film after the sequential radiations from electrons to protons and ones from protons to electrons, respectively. The changes in spectral reflectance after the simultaneous radiations of protons with electrons are shown in Fig.1(c). Both the energy of electrons and protons were chosen as E=30keV, and the corresponding flux was ϕe=1×1012cm-2s-1 and ϕp=5×1011cm-2s-1, respectively. It is indicated that before the radiations, the spectral reflectance in the wavelength region of 280~750nm is approximately 80%, and shows a minimum of 77%. In the wavelength region more than 850nm, the spectral reflectance is always larger then 90% in the near-infrared region. This demonstrates that the aluminized Teflon FEP possesses an excellent reflective property in its original state. However, the reflectivity was degrading with increasing the radiation fluence after the

93 synergistic radiations of electrons and protons. As show in Fig.1, an absorption band appears in the 280~600nm wavelength region, in which the spectral reflectance drops suddenly. The right edge of the band moves to the right with increasing the radiation fluence, and leading to widening of the absorption band. This phenomenon might be related to an increase in the concentration of micro-defects induced by radiation of electrons and protons [7]. Under the sequential radiation of electrons and protons, the influence of

Wavelength, nm Figure1. Changes in reflective spectrum of aluminized Teflon FEP film after synergistic radiation (E=30keV, ϕp=5×1011cm-2s-1, ϕe=1×1012cm-2s-1) (a) e-ĺp+, sequential; (b) p+ĺe-, sequential; (c) p++e-, simultaneous

electrons and protons on the degradation of reflective property is different. The proton exposure mainly causes an obvious decrease of spectral reflectance in the near ultraviolet to visual regions (Fig.1a), while the electrons results in the degradation of

94 reflective property in the whole solar spectrum (Fig.1b). Under the exposure pf electrons, charge accumulation would occur on the Teflon FEP film. In the following in-situ measurement, the accumulated charges could be annihilated by absorbing the incident photons of solar spectrum, and thus leading to the decrease in reflective property within the regions from the near ultraviolet to the near infrared. Under the simultaneous radiation, the degradation of reflective property is similar to that caused by the electron to proton radiation, for the obvious absorption band only forms in the near ultraviolet to visual light regions. The reason for this phenomenon might be related to the annihilation of accumulated negative charges on the surface of Teflon FEP film due to electron radiation by the positive charges of incident protons.

Figure 2. The decrease in spectral reflectance ∆ρ of aluminized Teflon FEP film caused by synergistic radiations at a giver fluence as a function of incident photon energy (Ee=Ep=30keV, Φe=2×1016cm-2, Φp=1×1016cm-2)

Figure 2 shows a comparison for the decrease in spectral reflectance ∆ρ of aluminized Teflon FEP film under different types of radiation with fluence of electrons and protons. Both the energy of electrons and protons is chosen as 30keV, and the fluence of electrons and protons is Φe=2×1016cm-2 and Φp=1×1016cm-2, respectively. It is indicated that under the conditions of synergistic radiations of electrons and protons, the radiation effects due to electrons and protons do not show an additivity. The simultaneous radiation effect of electrons and protons is smaller, and the additive effect of electrons and protons higher. In the case of sequential radiations, the total change in spectral reflectance is independent from the radiation sequence of electrons and protons, if the fluence the same. 3.2 ANALYSES OF XPS AND MASS SPECTROSCOPY The changes in XPS spectrum of aluminized Teflon FEP film after the synergistic radiations of electrons and protons are shown in Fig.3, in which the curve 1 stands for that before radiation, the curve 2 for the sequential radiation from protons to

95 electrons, the curve 3 for the radiation from electrons to protons, and the curve 4 for the simultaneous radiation of protons and electrons. Fig.3 a, b and c indicate the spectra of O1s, F1s and C1s energy bands before and after the radiations, respectively. It is shown that the energy band spectra of O1s, F1s and C1s before the radiations, are typical for the Teflon FEP film, and minor amount of oxygen was detected on the film surface. After the synergistic radiations, the XPS spectra change obviously. In the C1s energy band spectra (Fig.3c), the peak at the binding energy of 292.2eV, which characterizes the basic chain of CF2 for the fluorinated ethylene, dropped remarkably, while the peak at the energy less than 285eV rose noticeably. Fig. 3(b) demonstrates that the fluorine content reduces obviously due to the radiations. Fig. 3(a) shows that the oxygen content increases after the radiations. It was found that the radiations sequence of protons and electrons had an obvious effect on the XPS spectra. The C1s spectrum after the simultaneous radiation of protons and electrons was similar to that obtained by the sequential radiation from electrons to protons (Fig.3c). The C1s spectra after the synergistic radiations can be characterized by eight characteristic peaks, as shown in Table 1. It is believed from the above characteristic binding energy in Table 1 that the peak 1 and 7 originate from the ethylene chain segments and the tetrafluoro-ethylene chain segments in Teflon FEP film, respectively. The binding energy of 285.6 and 285.9eV related to peak 2 originates from the two ethylene units near the tetrafluoroethylene and the ethylene units in the –CFH–CH2– chain segments, respectively, and the peak 3 from the groups of C–O or the ethylene units in the –CF2–CH2– segments. The peak 4 is believed to be given by carbonyl groups (C=O), for its energy (287.9eV) is very near to the latter. The binding energy of 289.9 and 289.4 eV related to the peak 5 might originate from the groups of =CF– in the hexafluoro-propylene and the –CFH– in the –CF2–CFH– chain segments, respectively. The peak 6 originates from the –CF2– CH2– segments in the tetrafluoro-ethylene units, and the peak 8 from the fluorinated methyl groups of CF3 in the molecular chains. The above analyses indicated that under the synergistic radiations of protons and electrons, a complex change in the structure of Teflon FEP film occurred. Since the fluorine atoms and the fluorinated methyl groups are much larger than carbon atoms in the molecular main-chains of Teflon FEP copolymer, the chains of carbon atoms could be closely encased to form a stable structure. During the radiations, the charged particle would preferentially interact with the fluorine atoms and fluorinated methyl groups in the external layer of the molecular chains. When Teflon FEP film was radiated by electrons, the incident electrons were caught preferentially by fluorine atoms, because of their strong electronegativity. The fluorine atoms were bombarded out of the molecular chain segments, and leading to forming the active radicals such as –C*F2, – C*F–, –C*=, –C**– and free carbon atoms. In the case of proton radiation, except for the

Relative abundance

96

Binding energy, (eV) Figure 3. Changes in XPS spectrum of Teflon FEP film caused by synergistic radiation for energy bands of (a) O1s, (b) F1s and (c) C1s (E=30keV, ϕe=1×1012cm-2s-1, ϕp=5×1011cm-2s-1, Φe=2×1016cm-2, Φe=1×1016cm-2) 1.before exposure; 2. p+ĺe-, sequential; 3. e-ĺp+, sequential; 4.p++e-, simultaneous; TABLE 1. The binding energy (eV) and area ratio (%) of eight characteristic peaks in C1s spectra after synergistic radiations

Radiation mode

Binding energy (eV) and/or area ratio (%) 1

2

3

4

5

6

7

8

P+ ĺ e -

284.3 (31)

285.9 (10)

286.9 (8)

287.9 (10)

289.4 (8)

290.9 (9)

292.2 (14)

293.7 (10)

e- ĺ p+

284.3 (40)

285.6 (19)

286.5 (10)

288 (11)

289.9 (6)

291 (2.5)

292.3 (8)

293.7 (3.5)

e- + p+

284.4 (36)

285.8 (19)

286.5 (7.5)

287.9 (14)

289.9 (7)

291 (5)

292.2 (5)

293.6 (6.5)

above defects, the following functional groups could also be formed due to the implanting effect of protons (H*): –CF2H–, –CFH2–, –CFH–, –CH3, –CH2– and =CH–, as well as the free radicals on basis of the above groups. For instance, the peak with C1s binding energy of 287.9eV related with the carbonyl group in Table 1 might result from the product formed by the combination of –C**– with oxygen. Figure 4 shows the mass spectrum of aluminized Teflon FEP film after proton radiation for 60min, indicating that characteristic peaks appear obviously at the mass

97 over charge ratios of 15, 28, 33, 43, 51, 56, 57, 69, 119, 150, 168, 221 and 281. The related segments of molecules and ions might be: CH3*, CH2=CH2, CFH2, CH3CH2=CH2, CHF2*, *CH2CH=CHCH3 or CH3CH=CHCH3, C4H9*, CF3*, CF3–CF2*, C3F6H*, C3F7*, C4F9H, respectively. From the above results, the following mechanisms for proton radiation could be given: (1) the fluorinated methyl served as the protruded nodes in the segments of hexafluoro-propylene was preferentially fractured to form the CF3* ion pieces; (2) under the strong elastic interaction of protons with the large molecules of polymer, the C–C bonds were broken. The fluorine atoms are easy to be bombarded out of the molecules into the ions, and could further fluoridize other free radicals, such as –CF2–, etc., instantaneously. Thus, no peaks appear for large amount of F* in Fig.4. Figure 5 shows the kinetic curves of outgassing for aluminized Teflon FEP film under proton radiation with Ep=40keV and ϕp=5×1011cm-2s-1. It was found that in the early stages, the CF3* ion pieces were formed, the outgassing of which reached to a balanced state within 15 minutes. The relative abundance of CF3* is much higher than the other outgassed products (5~10 times). It is believed that the bonds of C–F and C–C could be broken by protons almost simultaneously, and the active F* would recombine with other ions and activated pieces of small molecules. The existing CF3* pieces of large amount imply that the proton radiation mainly results in the formation of F* and – CF2– pieces from the large molecules in Teflon FEP film. Other pieces with small molecular mass and the molecular segments containing hydrogen appeared after the radiation for longer time (approximately 30min), and their outgassing amount is minor. The above phenomena are in agreement with the rise in C1s peak with 284.4eV and the drop of the peak with 292.2eV in Fig.4c due to the radiation.

Relative abundance

Ep=30keV ϕp=5×1011cm-2s-1 t=60min

Mass/charge ratio

Figure 4. Mass spectrum of gases outgassed from aluminized Teflon FEP film after proton radiation for 60 minutes

Figure 5. Kinetic curves of outgassing for aluminized Teflon FEP film under proton radiation

98 4. 0 Conclusions Both the radiations of protons and electrons led to forming an absorption band in the near ultraviolet to visual regions, and the electrons also resulted in a decrease of spectral reflectance with fluence in the near-infrared region. The effect of synergistic radiation of protons and electrons on the reflective property of aluminized Teflon FEP film does not show an additivity. The simultaneous radiation of protons and electrons shows a lower effect than the addition of independent ones. Under a given fluence of radiation, the change in spectral reflectance is independent from the radiation sequence of protons and electrons. The changes in spectral reflectance with radiation fluence and the XPS spectrum of aluminized Teflon FEP film for the simultaneous radiation of protons and electrons were similar to those for the sequential electron to proton radiation. 5. 0 Acknowledgement The financial support of the National Basis Research Foundation of China under Grant # G19990650 is greatly appreciated. 6. 0 References 1.

Lucas J.W. Heat transfer and spacecraft thermal control. New York, Energy sources, 1970, 5 – 23.

2.

Joyce A., Kim K., Jacqueline A., L. Len Wang. Mechanical Properties Degradation of Teflon FEP Returned From the Hubble Space Telescope. AIAA Paper, 1998, 0895, 1—11.

3.

Durcanin J.T. The definition of the low Earth orbital environment and its effect on thermal control materials. AIAA Paper, 1987, 1599, 1 – 12.

4.

Mikhailov M.M. Possibilities of replacing electromagnetic radiation of the Sun by accelerated electrons in testing space technology materials. Journal of Advanced Materials, 1996, 3(6), 465 – 470.

5.

Leet S.J., Fogdall L.B. and Wilkinson M.C. The Effects of Simulated Space Radiation on Silver Teflon, White Polyurethane Paint, and Fused-Silica Optical Solar Reflectors. AIAA Paper, 1993, 2876, 1—11.

6.

Tribble A.C., Lukins R., Watts E., Naumov S.F. and Sergeev V.K. Low Earth Orbit Thermal Control Coatings Exposure Flight Tests: A Comparison of U.S. and Russian Results. NASA-CP-4647, 1995, 22919, 1—9.

DOSE RATE EFFECTS IN POLYMER MATERIALS IRRADIATED IN VACUUM B.A. BRISKMAN*, E.R. KLINSHPONT, V.F. STEPANOV Obninsk Branch of Karpov Institute of Physical Chemistry, 249020, Obninsk, Russia E-mail: [email protected]

Abstract The most reliable information on the polymeric material radiation resistance and lifetime may be obtained in the tests conducted under operation conditions. An important problem of the environment simulation is the adequacy of the accelerated test results, because the dose rates used in the test and under operation conditions test may differ up to several orders of magnitude. The natural way of acceleration is valid only when the change in a material property does not depend on the dose rate. It is well known that under irradiation in air environment the dose rate effects in polymers are very significant and connected with oxidizing radiation destruction. As a rule, they are taken into account for polymer materials used, e.g., in nuclear energy facilities. We will examine an opposite situation - irradiation of materials in vacuum that is typical for operation in space environment. It is often assumed that the irreversible radiation changes in the polymer properties do not depend on the dose rate. On the basis of available experimental data (including unpublished results obtained in our laboratory) we will show that there are remarkable dose rate effects in physical-chemical and operational (mechanical, electrical, thermal and optical) properties for a wide variety of polymers irradiated in vacuum. These effects reach sometimes an order of value and in a number of cases have a non-monotonous character. 1. 0 Introduction The dose rate effects in polymer properties under irradiation in an oxygen environment are very significant and connected with chain reactions of oxidizing radiation destruction initiated by radiolysis species of radical type [1-5]. It is usually assumed that under irradiation in vacuum such effects are negligible [2,6]. The dose rate effects in polymers under irradiation in air and absence of such effects under irradiation in vacuum or inert environment were pointed out at first in [7]. Similar conclusions were made in [2] mainly for mechanical properties of polymers, but then they were supported by the data for gel formation in the low density polyethylene (LDPE) and ethylenepropylene copolymer irradiated in vacuum in the dose rate range of 0.26 to 2.8 Gy/s [8],

99

100 LDPE, polypropylene (PP) and silicone, fluorine, natural and isobutylene-isoprene rubbers in the dose rate range from 2.6 Gy/s for gamma-radiation to 130 Gy/s for electrons while measuring radiation yields of gas evolution, gel formation, and mechanical properties [9], and also for discoloration of cellulose triacetate, radiation scission of poly-methylmethacrylate PMMA and polycarbonate PC under irradiation in the absence of oxygen at room temperature in the dose rate range from 4.1010 Gy/s for pulse electron radiation to 0.5 Gy/s for gamma-radiation [10]. The observed dose rate dependencies are sometimes based not on the specificity of the radiation-chemical processes, but on the influence of the elevated irradiation temperature at high dose rates. E.g., a significant increase in Kapton optical density was discovered [11] when 1 MeV proton flux density grew up to 2.1013 ɫm-2 s-1. Our calculations displayed very high growth in the material irradiation temperature. As the dose rate dependence of the radiation effects on polymers in the oxygen-free environment is neglected, both international standard IEC [12] and national Russian standard [13] practically limit the acceleration of polymer material radiation tests in such conditions only by control of a specified value of irradiation temperature. But these limits for space materials operated under high vacuum environment are not sufficient even because they relate, as a rule, to the dose rate values I > 0.1 Gy/s. This distinction does not permit to extrapolate the conclusions for the range 10-5
101 accordance with Arrhenius approximation. But the real situation is, as a rule, quite opposite (see the figures below). It means, that the averaged temperature of the material samples does not exceed the specified level during irradiation at indicated mode of cooling. 2.1 GAS EVOLUTION G

G(-COOH)

0,50

2

1 2 0,25

1

0,00

0 0

2

ln(I/I1)

4

6

Figure 1. Radiation-chemical yield of radiolysis species vs. gamma-radiation dose rate in polyethyleneterephtalate PETP: 1 - carboxylic groups - ɋɈɈɇ; 2 - gas species. I1 = 0,01 Gy/s [16,17] G(H2) 16 14 12 10 8 6 4 2 0 -3,0

-2,5

-2,0

-1,5

-1,0

-0,5

0,0

Lg(I/I1)

Figure 2. Hydrogen radiation-chemical yield vs. gamma-radiation dose rate in LDPE. I1 =1 Gy/s [16,17]

102 G

2

1 4

M

1

2 5

3 M

0 0

2

4

8

9

Lg(I/I1)

Figure 3. Radiation-chemical yield of radiolysis species in polydimethylsiloxane (1-3) and neopentane (4-5) vs. gamma-radiation dose rate: 1 - methane; 2 - hydrogen; 3, 5 – ethane, 4 – 2,2,5,5tetramethylhexane. I1 = 1 Gy/s [16,17]. The points, marked by "M" are obtained in [18] for polydimethylsiloxane, Ɇn = 76000. Curves 4, 5 - from [19]. 3

g, cm /g

16

14

12

10

8

6

4

3 2

2

1 0 0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

6

I, 10 Gy/s Figure 4. Concentration of total radiolysis gas species in polypropylene PP vs. 5 MeV electron radiation dose rate at doses 1 (1); 2,5 (2) and 5 ɆGy (3) at room temperature [20].

It can be seen that the radiation-chemical yield of gases changes 4-20 times when the dose rate increases by 2-4 orders. It is well known that the radiation-chemical yields increase with the temperature growth. Hence one can expect only an increase in the gas yield when the dose rate rises without thermosetting, in contrast with the data presented in Ref. [20].

103 2.2 OPTICAL SPECTRUM AND RADIATION CROSS-LINKING S 0,6

5 4

0,5

3 0,4

2

0,3

1

0,2

0,1 0,0 2300

2200

2100

2000

1900

1800

-1

ν,cm

Figure 5. Absorption spectrum of irradiated polystyrene PS (1-4) and polyvinylxylene (5) at dose rate I = 0, 01 (1), 0,17 (2, 5), 2,15 (3), 4.105 Gy/s (4) [16,17]

100

0,6

6

q,%

S

5 0,5

80

0,4

60

4 0,3

40

3 2

20

0,2

1 0,1

0 0

1

2

3

4

6

I, 10 Gy/s Figure 6. Gel fraction q (1, 5, 6) and optical density S at 970 ɫm-1 (2, 3, 4) in LDPE vs. dose rate. Absorbed dose of 5 ɆeV electron radiation, ɆGy: 1—0.02; 2, 5— 1,2; 3 — 2.5; 4, 6 — 5 [21]

104 q, % 80

70

3

60

50

40

2

30

20

10

1

0 0,0

0,5

1,0

1,5

2,0

2,5

6

I, 10 Gy/s Figure 7. Gel fraction q in PP vs. dose rate at dose 1 (1), 2,5 (2) and 5 ɆGy (3) [20]

2.3 RADICAL GENERATION G

4,5

G

4,0

1,0

3,5

4

3,0 2,5

3

2,0

0,5

2

1,5

1

1,0 0,5 0,0 -5 10

0,0 -3

10

-1

1

10

10

3

10

5

10

I, Gy/s

Figure 8. Radiation-chemical yields of paramagnetic centers in vacuum at room temperature in high and low molecular glasses: PS (1), PC (2), sodium-silicone glass (3), PMMA (4, relative values, right vertical axis) vs. dose rate [22]

2.4 OPERATIONAL PROPERTIES Radiation-induced changes in the mechanical, thermal and electrical properties of PTFE and LDPE upon irradiation in vacuum in the dose rate range of 10-3 to 1 Gy/s are presented in table 1 and figs. 9 and 10 [14].

105 TABLE 1. Relative tensile strength σ/σ0 and elongation at rupture ε/ε0 vs. dose rate for some polymers I, mGy/s 0,9 14,5 85 910

PTFE at dose 3 kGy ε/ε0 σ/σ0 0,5 0,3 0,5 0,6 0,7 0,9

PTFE at dose 100 kGy σ/σ0 0,3 0,5 0,7

LDPE at dose 300 kGy σ/σ0 ε/ε0 1,4 0,8 1,3 0,8 1,3 0,9 1,1 1,0

Appropriate data on tensile strength of PP irradiated up to 2.5 MGy are presented in [20]. Different rate of ethylene-vinylacetate copolymer degradation (gel fraction, gas evolution, swelling, strength) under gamma-radiation and electron impact (the dose rates differed about 50 times) was observed in [9]. Small difference of gel fraction values (~10%) at high dose rate observed in high density polyethylene [10] agrees well with the data for LDPE presented in Fig. 6 (curves 1, 5,6) [21].

λ,W/m*K 0,40

0,38

3

0,36

0,34

1

0,32

2 0,30

0,28

0,26

0,24 140 160 180 200 220 240 260 280 300 320 340 360

T, K

Figure 9. LDPE heat conductivity λ (W/m⋅K) vs. measurement temperature, D = 10 kGy. 1- original sample, 2 - 10-2 , 3 - 10-3 Gy/s

106

Figure 10. Tangent of dielectrical losses tan δ (frequency 1 kHz) vs. temperature for PTFE irradiated in vacuum by doses 100 (1 - 3) and 30 kGy (4, 5) at dose rate 0.01 (1,4); 0.1 (2, 5) and 1 Gy/s (3) [23]

The dose rate effects in operational properties are not so significant as for radiolytic species and the products of the chemical stage of radiolysis (e.g. cross-linking and scissions). It is a result of a gradual attenuation of the effect on the way from physical stage of radiolysis via physical-chemical stages to the change of operational properties. The same situation takes place in the problem of LET (linear energy transfer). 3. 0 Dose Rate Effect Models Theoretical prerequisites for the described above dose rate effects are quite clear combination of radiolysis and processes of molecular motion in polymers that results in a change of the material molecular structure. The problem is only in the size of the effect. An indirect proof of the effect inevitability is the temperature dependence of the material radiation degradation [24] when the ratio of thermal activation acts of molecular motion, and radiation acts of excitation and ionization are changing. The physical reasons for dose rate effects in irreversible changes of polymer properties under irradiation in vacuum are the following: a) Radiolysis products interaction; b) Process of radiolysis products mass transfer; c) Relaxation character of molecular mobility. Let us examine each reason separately. 3.1 RADIOLYSIS PRODUCTS INTERACTION Change of the polymer material properties results from a number of chemical and physical processes where particles (structures) with different life or relaxation time take part. In principle, when the result of radiation impact depends on interaction of two

107 radiolysis products (including interaction with secondary radiolysis products) the dose rate effect will take place. Such a possibility was reported still in Ref. [6], and a "threshold dose rate" for spur overlapping was considered in Ref. [10]. It results in growing of intermediate radiolysis products interaction and, therefore, in the sought-for dose rate effect. But up to I = 4.1010 Gy/s such a dependence was not detected. Let us examine as an example the following generalized situation. Let the radiolysis product E be a result of interaction of two radiolysis products P1 and P2. Then the accumulation kinetics will be described by the following set of equations d[E]/dt = k [P1]⋅[P2] d[P1]/dt= G1I - k[P1]⋅[P2]- [P1]/τ1 d[P2]/dt= G2I - k[P1]⋅[P2]-[P2]/τ2, where I - dose rate, G1, G2 - radiation-chemical yields of species P1 and P2, τ1 and τ2 radiolysis species life time that is determined as a result of all other decay processes. This set of non-linear equations, as a rule, can't be solved exactly. For simplicity, we shall examine some extreme cases. Let us examine a stationary case and assume that the radiolysis species stationary concentration is mainly determined by destruction reactions (loss, escape from a reaction cell etc.). Then [P1]= G1Iτ1, [P2]= G2Iτ2 and d[E]/dt = k G1G2I2τ1τ2 It is clear that accumulation rate of the product E depends on the second power of the dose rate. In another limiting case when the radiolysis species vanish only through decay processes (without interaction) we shall observe a linear dependence on the dose rate. Examples of the dose rate dependencies: 1. Interaction of radicals and excited states provides a dose rate dependence of radical yield [17]. 2. Dose rate influence on the magnitude of the irreversible radiation effects may reveal itself when the radiolysis species are formed as a result of, at least, two processes of different kinetic order. Let the products P1 and P2 be formed through the reactions of intermediate species R of first and second order. Then the process kinetics is described by the following set of equations: d[R]/dt = GI - k1[R] - k2[R]2 d[P1]/dt = k1 [R] d[P2]/dt = k2 [R]2, where R, P1, P2 - intermediate active species and reaction products concentration, accordingly, G - radiation yield of the intermediate active particles, k1 and k2 - rate constants of the first and second order reactions. Ratio of the products P2 and P1 accumulation rates is described by the expression W(P2)/W(P1)=[(1+k2GI/k1)0,5 -1] (k2G I)0,5/k1. So the ratio of product accumulation rates depends on the dose rate.

108 As a rule, several reactions involving intermediate active species occur simultaneously - recombination reaction of second order, and reactions with other molecules. E.g., a methyl radical ɋɇ3 is formed at ɋ-ɋ bond scission in branched hydrocarbons. Ethane is formed as a result of methyl radical recombination (second order reaction). Methyl radicals interact with hydrocarbon molecules (first order reaction) and give methane. The ratio of methane and ethane concentration depends on the dose rate. The dose rate range, where this dependence is significant, is determined by the ratio of constants k1 and k2. E.g., composition of gaseous radiolysis products of polysiloxanes and branched hydrocarbons is described by the expression of the type given above in the dose rate range of 0,1-100 Gy/s. 3. Cross-linking is a result of macro-radical recombination. In this case we can write the following expression dc/dt= k [R]2 d[R]/dt= GI - k[R]2, where c is concentration of the cross-links. If the radicals are consumed only for crosslinking then the radiation link yield is directly proportional to the dose rate. But the radicals can participate in other processes, including other radiolysis species. E.g., in the case of interaction with gaseous products dc/dt= k1 [R]2 d[R]/dt= GI - k1[R]2- k2 [R]G dG/dt= GgI - k2[R]G - k3S kD G, where the third term of the last equation takes into account the gas evolution from the polymer volume (S - sample surface, kD - diffusivity). Solution for a stationary mode is given by a nonlinear dose rate dependence of the link yield. 3.2 RADIOLYSIS SPECIES MASS TRANSFER A significant amount of gaseous products in organic materials can be generated as a result of irradiation. They can take part in chemical reactions and diffuse from the polymer volume. The analysis of the kinetic equations describing these processes shows that the following limiting cases, determined by ratio of diffusion time (τ~ l2/6 kD) and exposure time (t= D/I) can be realized (here l is the characteristic size of a sample). If τ << t, then dose rate effects may be neglected, because gases do not influence the radiation-chemical processes - they have time to leave the sample volume. In the opposite case, the dose rate effect will be apparent. The gaseous radiolysis products can influence the radiation effects in different polymer properties and, particularly, crosslinking and scission. E.g., hydrogen and methane accelerate macro-radical destruction, realizing relay race transfer of free valence. It is important that macro-radical destruction may result both in macromolecule cross-linking because of recombination, and in double bond formation at macro-radical disproportionation. In the presence of hydrogen, the disproportionation reaction is more effective. It means that in the above

109 cases we can observe different cross-linking efficiency with the corresponding change in the operational characteristics. The mass transfer processes can also result in dose rate effects, when the gaseous radiolysis products do not take part in the chemical reactions. These products can leave the volume or accumulate as bubbles (clusters) in defect points of polymer. In the last case, the gas accumulation can result in an integrity fracture, generation of cracks, etc. The process direction is defined by competition of micro- and macrodiffusion, rate of gas generation (proportional to the dose rate), nucleation center concentration, polymer strength, relaxation rate, radiolytic gases solubility in the polymer, sample thickness. These effects were observed at investigation of reversible radiation effects in heat capacity of polymers [25,26] and examination of radiation cross-linking of polydimethylsiloxane in vacuum at the dose rate of 0.03 and 10 Gy/s [27]. 3.3 RELAXATION CHARACTER OF MOLECULAR MOBILITY Irradiation of polymer results in formation of non-equilibrium states (excited states of molecules, radicals, ions, etc.), which have certain lifetime (they relax to equilibrium during this time), often depending on temperature. Presence of non-equilibrium states, at first, directly changes some properties of polymeric materials (optical, electrical, chemical) and, second, causes change of the properties through the process of relaxation. The magnitude of degradation depends on the degree of system nonequilibrium. The rate of equilibrium upset is proportional to the dose rate dC*/dt = GI - C*/τ, where ɋ* - non-equilibrium state concentration, τ - relaxation time, G - radiationchemical yield of non-equilibrium states. It is assumed that G is independent of the dose rate. A quasi-equilibrium concentration of non-equilibrium states is established under irradiation ɋ* = GIτ It is defined by competition of generation and destruction. As to irreversible radiation effects, it is necessary to provide an equal concentration of non-equilibrium states at accelerated tests ɋa* = ɋo* It means that as the dose rate rises, it is necessary to increase the relaxation rate, i.e., to decrease the relaxation time τa = τo (Io /Ia), where indices a and o relate to accelerated and operating conditions. The non-equilibrium states generated at polymer irradiation are connected with formation of intermediate active species that have different nature, take part in different

110 chemical processes and have different relaxation time. E.g., the lifetime of electronexcited molecules is about 10-7-10-8 s; the one for charged particles vary in very broad interval and at low temperature (e.g., at 77 Ʉ) can achieve some hours. The lifetime of macro-radicals vary over a larger time range; some of them may be stable even at 300 Ʉ during several years. It means that at accelerated radiation test it is really impossible to reproduce concentration of non-equilibrium states for different particles having different relaxation times and different activation energy of relaxation process. The following approximate approach is yet possible based on comparison of non-equilibrium states concentration and their contribution to the radiation effect. It is necessary to compare the quantity of active species C generated during exposure and stationary concentration ɋ*. If ɋ*<< ɋo, the relaxation processes may be neglected. 4. 0 Conclusions a) Radiation effects in polymer properties under irradiation in vacuum depend in general on the dose rate. The dose rate effect can achieve an order of magnitude for intermediate radiolysis products. During transition from physical-chemical stage of radiolysis to the chemical one, and then to the polymer properties the effect magnitude decreases down to tens or several percent. b) Sometimes this dependence has an extreme character and may be a reason for observed discrepancy between different data of dose rate effect. c) In most cases the dose rate effects are observed at long exposures and dose rates lower than 0,1 Gy/s. d) The problem of dose rate effect in the radiation test may be solved by development of adequate physical-chemical model or approximate simulation. But, as a rule, the development of such model requires extensive investigations. Moreover, such models are valid only within a relatively narrow range of variations in external parameters (e.g., temperature) defined, in particular, by phase or relaxation transition. e) Two ways of approximate simulation are proposed. In the first one the dose rate effect is considered by introducing of reserve factor for absorbed dose value depending on acceleration factor and material property. The second method consists in preliminary test of the effect for a representative material of the set of polymers in the dose rate interval defined by an acceleration factor. The last method is recommended in the draft of ISO standard "Space environment simulation at radiation test of materials. I. Nonmetallic materials" [28].

111 5. 0 References

1. K.T. Gillen, R.L. Clough, J. Polym. Sci. : Polym. Chem. Ed. 23 (1985) 2683. 2. Wilski, Rad. Phys. Chem. 29 (1987) 1 3. K.T. Gillen, R.L. Clough, Polym. Degrad. and Stability 24 (1989) 137. 4. S. G. Burnay, in: Proc. Int. Symp. on Radiation Degradation of Polymers and the Radiation Resistant Materials, Japan, July 24-25, 1989, p. 149. 5. H. Kashiwabara, T. Seguchi, Radiation-induced Oxidation of Plastics, Ch.11. In: A. Singh and J.Silverman (Eds.), Radiation Processing of Polymers, Hanser Publishers (1992). 6. F.A. Makhlis, Radiation Physics and Chemistry of Polymers, Wiley, New York, 1972 7. A. ɋharlesby, Atomic Radiation and Polymers, Pergamon Press, New York, 1960 8. T. Seguchi, S. Hashimoto, K. Arakawa, Radiat. Phys. Chem. 17 (1981) 195 9. Y. Haruyama, Y. Morita, T. Seguchi, R. Tanaka, T. Kanazawa, K. Yotsumoto, K.Yoshida, JAERI-M 88197 10. H. Kudoh, M. Celina, G.M. Malone, R.J. Kaye, K.T. Gillen, R.L. Clough, Radiat. Phys. Chem. 48 (1996) 555 11. D. Fink, M. Muller, L. Chadderton, P. Gannington, R. Elliman, D. McDonald, Nucl. Instr. Meth. in Phys. Res.. B32 (1988) 125 12. IEC 544 "Electrical insulating materials - Determination of the effects of ionizing radiation”. 544.3 (Part 3: Test procedures for permanent tests) 13. GOST 9.706-81 "Polymeric materials. Test methods of radiation ageing" (Russian standard) 14. B.A. Briskman, V.P. Sichkar, L.B. Kras'ko, V.Ʉ. Milinchuk, Chem. High Energy 27 (1993) 8 (in Russian) 15. ESA PSS-01-706. The particle and ultraviolet radiation testing of space materials (1983) 16. N.M. Bol’bit, L.A. Znamenskaya, L.I. Iskakov, A.P. Podsoblyaev, V.B. Taraban, Yu.I. Tolstosheev, E.R. Klinshpont, High Energy Chem. 27 (1993) 13 (in Russian) 17. N.M. Bol’bit, V.B. Taraban, E.R. Klinshpont, V.K. Milinchuk, in: Proc. 6th Int. Symp. on Mater. in Spaɫe Envir., ESTEC. Noordwijk. The Netherlands, 1994, 59 18. A. A. Miller, J. Amer. Chem. Soc. 82 (1960) 3519. 19. R.A. Holroyd, J.Phys.Chem. 65 (1961) 1352. 20. V.I. Serenkov, O.I. Byalinitskaya, V.S. Tikhomirov, High molecular compounds 24B (1982) 873 (in Russian) 21. O.I. Byalinitskaya, V.S. Tikhomirov, Plastics 2 (1978) 42 (in Russian) 22. N.M. Bol’bit, V.B. Taraban, E.R. Klinshpont, I.P. Shelukhov, V.K. Milinchuk, High Energy Chem. 34 (2000) 229 23. V.K. Matveev, N.A. Smirnova, V.K. Milinchuk, Polymer Science 35 (1993) 835. 24. V.K. Milinchuk, V.I. Toupikov (eds.), Organic Radiation Chemistry Handbook, Halsted Press (Wiley), New York, 1989 25. B.A. Briskman, Eng. Phys. J. 46 (1984) 781 (in Russian) 26. B.A. Briskman, S.I. Rosman, S.E. Vaisberg, High molecular compounds A26 (1984) 1047 (in Russian) 27. C.G. Delides, I.W. Shepherd, Radiat. Phys. Chem. 10 (1997) 379 28. B. A. Briskman, A.N. Belyakov, E.R. Klinshpont, Yu.Ya. Shavarin et al, in: Proc. 8th Int. Symp. on Materials in a Space Environment, Arcachon, France, 5-9 June 2000

This page intentionally left blank

TOWARDS A DATABASE FOR ASSESSMENT OF NEAR-EARTH SPACE RADIATION EFFECTS ON OPTICAL GLASSES A.GUSAROV Mulitel a.s.b.l, B-7000 Mons, Belgium D.DOYLE European Space Agency European Space Research Technology Centre, 2200 AG Noordwijk, The Netherlands M.FRUIT Astrium 31402 Toulouse Cedex 4, France

Abstract Establishing a database for assessment of radiation impact on spectral transmission of optical glasses requires addressing several critical theoretical issues and a very significant amount of experimental work. We use BK7 glass as an example to demonstrate our approach to the problem. A sequence of transmission spectra is required to find the parameterization set. Taking into account the number of optical glasses used in space optical instruments, it is proposed that the first step towards a useful radiation effects database consists of the elaboration of a uniformly applicable methodology for testing of optical glasses. We believe we have taken such a significant step in this work. 1. 0 Introduction Effects of radiation on optical instruments operating in the near-Earth space radiation environment remained a concern since the beginning of the space optics. Over the decades, many investigations have been undertaken, but the results lie scattered throughout the literature and are difficult to use due to the specific nature of the tests carried out. The very wide range of materials, radiation types, total doses, dose rates and parameters actually measured makes any extrapolation doubtful and the application of the data to a new situation rarely possible. Further complicating this problem is the fact that the data as presented in the literature are not always accessible or understandable to a non-specialist. For the designer and user of space optical instrumentation, the question

113

114 of qualification of optical materials for use in a space radiation environment remains an issue to be addressed in each case with new tests. Radiation influences various characteristics of optical glasses, the most important are transmission, refractivity, and density. Recent work in this field has shown that there is merit in a parametric approach [1]. In the present work, we analyse a parameterization scheme intended for description of radiation-induced transmission degradation. It is proposed to use this scheme for establishing a database for assessment of radiation impact on spectral transmission of optical glasses. 2. 0 Radiation-Induced Transmission Degradation: Analysis of Transmission Data Any parameterization of radiation-induced transmission degradation must be firmly based on physical causality, i.e. on the color center model, and it must take into account both defect generation and annealing. Transmission degradation is a consequence of the generation of radiation defects. Such defects correspond to electronic states, which are not present in the glass before irradiation. Optical transitions, which are related with those states, give rise to an additional absorption in the visible. A decrease of the density of initial electronic states corresponds to a decrease of absorption in the region of initial absorption bands. However, such induced transparency is usually not of interest because, before exposure to radiation, glasses are transparent in the working spectral range. In general, an electronic state in glassy material can be converted to another one by radiation. The most sensitive to radiation are dangling bonds, oxygen bridges, over-coordinated atoms, i.e. deviations from the “ideal” random network. The concentration of such radiation precursors is usually much lower than the concentration of normal bonds and the saturation effect plays an important role. Radiation-induced defects in glass are not stable. They can be thermally-, optically- or radiation-transformed to other configurations. The concentration of radiation-induced defects changes after the end of irradiation, some of them can be completely annealed, but most defects are meta-stable and are characterized by very long relaxation times. The first order evolution kinetics of the precursors concentration is described by a differential equation, which takes into account, generation, annealing and the saturation effect: (2.1) dn = −k g n + ka (n0 − n) ,

dt

where n0 and n(t) are the initial and instantaneous concentrations of precursors, and where kg and ka are the rates of generation and annealing of the defects, respectively. The first term describes the decrease of concentration due to transformation of precursors to meta-stable defects and the second one takes into account the defect annealing. We will made now an important assumption that the rate of defect generation is proportional to the radiation dose rate R:

115 (2.2)

k g = cR ,

where the constant c depends on material characteristics only. This constant can be analytically computed based on known models of defect generation. However such computations are rather complicated and the accuracy is not very high. It seems more realistic to find c using experimental data. The Eq. (2.2) is valid until interaction of the excited states is not important. The last one is not always true. For example, irradiation of polymers in an oxygencontaining atmosphere shows dose-rate dependence. For glasses such an effect is not important. At high dose-rates it is necessary, however, to take into account radiationinduced heating, which can influence thermally-activated relaxation processes. Solving (2.1) for the defect concentration ni, gives: (2.3) kg ª1 − exp − ª ka + k g º t º . ni (ka ) = n0 ¬ ¼

(

ka + k g ¬

)¼

The index “i” is added in order to distinguish different types of defects. This solution shows that at the initial stage the defect concentration grows linearly, and it saturates with increase of irradiation time. The saturation level depends on both the generation and the annealing rates. If relaxation is very fast (kg « ka) the saturation level is significantly smaller than the level defined by the precursor concentration. In the absence of relaxation (ka = 0) we can rewrite Eq.(2.3) as (2.4) n = n ª1 − exp −cD º , i

0i

(

¬

)¼

where D is the accumulated dose, which makes it evident that c is the inverse of the saturating radiation dose Ds. In disordered materials relaxation is characterized by a broad distribution of the relaxation rates ϕ(k): (2.5) n = dk ϕ (k )n (k ) i

³

a

i

a

i

a

The distribution ϕi(k) is temperature dependent with parameters different for each defect type. The amorphous nature of glass results also in an inhomogeneous broadening of defect-related optical transitions, which are well characterized as Gaussian bands. The induced absorption coefficient ∆ai is proportional to the defect concentration. Therefore (2.6) ∆a (ω ) = A exp(−(ω − ω ) 2 / 2σ 2 ); A = s n , i

i

0i

i

i

i i

where ω0i is the central frequency, σi is the band width and si defines the amplitude of the absorption band. For low doses ni ( k a ) ≈ n0 cD and

Ai ≈ α Di D, α Di = csi n0i = si n0i / Dsi ,

(2.7)

i.e. αDi is the dose coefficient (DC) for the absorption band “i”. Using this DC we can represent Ai as (2.8) Ai (ka ) = α Di Z −1 ª¬1 − exp ( − ZD ) º¼ .

Z = ka / R + α Di / a0i ; a0i = si n0i

116 which is reduced to the previous expression for low doses (αDiD « a0i) and slow relaxation (kat « 1). For a given relaxation kinetics, defect accumulation is characterized by the DC αDi and the parameter a0i. This latter parameter has the meaning of the maximal achievable absorption in the case of a fast irradiation. Finally, it is necessary to take into account that the absorption spectrum is composed of a number of absorption bands N (2.9) ∆a(ω ) = Ai exp(−(ω − ω 0i ) 2 / 2σ i2 ) .

¦ i =1

We have now obtained a set of formulae, which allow the induced absorption under radiation or after the end of irradiation to be described. The practical application of this approach requires the kinetic parameters defining defect generation (αDi and a0i), relaxation rates distribution (ϕi), along with spectroscopic parameters (ω0i, σi) to be known for each type of defect. In the next section we show how these parameters can found from the optical transmission measurements. 3. 0 Experimental Results and Discussion We have selected BK7 (Schott) glass to illustrate the presented approach. The samples were polished discs 0.2 mm thick. The optical transmission spectra were recorded in the spectral range from 200 to 800 nm using a commercial double-beam spectrophotometer. The natural logarithm of the ratio of the transmission before irradiation to that after irradiation, normalized by the sample thickness, gives the radiation-induced absorption coefficient. A Co60 source with a dose-rate of 6 krad/h (water) at the samples location was used for γ-irradiation of the samples (1 rad corresponds to 10-2 Joules of energy absorbed by 1 kg of material, the SI unit of absorbed dose is the Gray = 100 rads). Irradiation was performed in a stepwise dose-accumulation manner. First the samples were irradiated with a dose of 400 krad. Optical measurements were performed with a 2-hours delay. Then the samples were irradiated again, so that the total accumulated dose reached 800 krad. After the end of irradiation transmission spectra were measured several times in a time interval up to 1700 days in order to monitor post-radiation relaxation. In total, we have 12 meaningful spectra. Several transmission spectra measured on a 0.2-mm thick BK7 glass sample are shown in Figure1. In order to construct the kinetic curves it is necessary to resolve the absorption spectra into individual absorption bands. As a result of non-orthogonality of the Gaussians such decompositions are not unique and additional information is required to obtain a physically meaningful result. Based on the data available in the literature we assume that for relevant radiation loads the experimental spectra in the range 230-800 nm can be described with four bands.

117

1

Transmission

0.8 0.6 0.4 0.2 0 200

300

400

500

600

700

800

Wavelength, nm 0krad

800krad

T, +1619d

Figure 1. Transmission spectra measured on a 0.2-mm thick BK7 glass sample. The upper spectrum – before irradiation, the lowest spectrum – after 800 krad dose, and the intermediate curve – 1619 days of annealing at standard laboratory conditions.

The selection of the number of the bands is a very important question and it can not be solved automatically. Moreover, a good fit does not guarantee that the selected bands are physically present. As indirect confirmation that the decomposition is correct is that the physical functions give a good description of the relaxation kinetics. We have used a two-step procedure to find the absorption band parameters form the experimental data. Firstly, we fit each spectrum independently. We assumed that the central energy and the width of those bands are changed neither during the experiment nor after the end of irradiation. This assumption is supported by the results of the fit, which show a small scattering of the values supposed to be constant. On the next step only the amplitudes of the absorption bands were allowed to vary. All other values were taken as the average of the values obtained in an independent fit of each induced absorption spectrum. An example of the decomposition is shown in Figure 2.

118

Induced absorption

15.00 5.00 -5.00 1

2

3

4

5

6

-15.00 -25.00 Energy, eV

exp 3.00eV

sum 4.28eV

2.09eV 5.23eV

Figure.2. Decomposition of the induced absorption spectra for BK7 glass sample into the gaussian bands. D = 800 krad.

In this way we have computed the variation of the absorption bands amplitudes in the course and after irradiation, Figure3. Relaxation in strongly-correlated systems is usually non-exponential. We do have found that the Debye-type (exponential) relaxation does not describe well the experimental results. In contrast, the stretched-exponent (Kohlrausch) law: (3.1) q (t ) = exp(−(t / τ )α ), 0 < α < 1 , 0

where τ0 is the characteristic relaxation time and α is the index of the fractionalexponential function, gave a reasonably good description for the whole interval up to 1700 days or annealing. Simple analytical expressions for the relaxation rates distribution, corresponding to (3.1), are known only for a few values of α. In a general case it computed by numerical inversion of the Laplace transform. The problem of the relaxation function selection will be addressed in an another place. The results are summarized in Table 1.

119

1.2

Relaxation function

1 0.8 0.6 0.4 0.2 0 1

10

100

1000

10000

Time, days 2.089

2.996

4.284

2.09

3.00

4.28

Figure 3. Variation of the amplitudes of the absorption bands. Signs – measured, curves - stretchedexponential relaxation function.

# 1 2 3 4

ω0i, eV 2.089 2.996 4.284 5.232

σi, eV 0.2452 0.6276 0.5869 0.2461

αDi (cm krad)-1 3.53E-03 1.99E-02 4.10E-02 -8.40E-02

a0i cm-1 1.509 11.26 26.35 -74.98

τ0 days 3061 2024 16200 2043

α 0.3227 0.2915 0.2599 0.2292

TABLE 1. Parameters required for description of radiation impact on BK7 glass samples.

The data in Table1 form a set of parameters, which allow describing change of transmission of a BK7 under radiation. Figure4 compares simulation and experimental results for a 5-mm thick sample, which was irradiated up to 800 krad during 7 days.

120

1

Transmission

0.8 0.6 0.4 0.2 0 200

300

400

500

600

700

800

Wavelength, nm

0d 15.0d 800krad

7.0d 1625.0d T, +4d

11.0d 0krad T, +1619d

Figure 4. Comparison of the simulated (signs) and measured results obtained on a 5-mm thick BK7 glass sample. Irradiation during 7 days up to the dose of 800 krad.

A good agreement is observed. Only three first bands were used in the simulation, because the values for the fourth band (5.232 eV) are not reliable. However, that band gives no contribution for wavelengths longer than 300 nm. The error is somewhat higher at 800 nm, which indicates that measurements taken at longer wavelengths are required. 4. 0 Data-base Approach Figure 5 presents schematically essential steps in the design of an optical system intended for use in a space radiation environment. If the end-of-life performance does not satisfy the requirements, the opto-mechanical configuration is changed and the procedure is repeated. In the present work, we have focused our attention on the part labeled “Radiation-induced effects” and more specifically radiation effects on transmission degradation.

121 Orbital parameters

Opto-mechanical configuration

Radiation loads on the optical elements

Radiation-induced effects

Dimensional instability

Transmission degradation

Perturbed optical layout

Index changes

End-of-life performance

Figure 5. Schematics of a radiation-tolerant optical system design.

The absence of easily available data in a standard format is a problem when necessary lifetime predictions must be done urgently and at a minimal cost. A recent example has been qualification assessment of the optical system of the Fluid Science Laboratory (FSL) being constructed for flight on the European laboratory module (Columbus) of the International Space Station. The total absorbed dose expected at FSL optical bench level is predicted to be 700 rad, but a qualification level of 1400 rad was specified for safety. Common engineering sense and experience leads one to expect an insignificant radiation induced performance degradation at such low dose levels accumulated over ten years on-orbit lifetime. Nevertheless, it was decided to proceed with an expensive testing of the full set of glasses, because otherwise reliable data, which would allow quantifying the effect of radiation, are not available. While it is necessary that the radiation testing of the FSL-relevant glasses will be performed under conditions specific for its qualification, the absence of a standard and a generic approach makes it rather probable that the same glasses will have to be tested again for another mission with different radiation conditions. The optical glass industry is passing now through a transition period, when traditional, well-established types of glasses are replaced with new types [2]. It may be expected that new glasses will completely replace, in the near future, the presently used glasses. At the same time, the previously available so-called “radiation-tolerant” (Cerium doped) glasses are being discontinued from routine production. The new glasses are designed to have optical properties in the visible identical to those of the glasses they replace, but with higher chemical stability and lower density. That is

122 achieved by changes in the chemical composition, which may result in a significantly different response to radiation. The radiation hardness of these new glasses has not been addressed up to now. This problem also needs to be addressed as use of these glasses in space optical systems become more widespread. We have developed a new methodology to parameterize the transmission measurements data for it's eventual use for radiation tolerance estimation. The first results obtained with this methodology look very promising. However, a more detailed analysis of several important issues is still required. On of those questions is the influence of the radiation type, i.e. will it be sufficient to introduce a scaling coefficient to describe the effect of proton (electron) radiation based on the gamma-irradiation data or a new set of parameters will be required? Establishing a comprehensive and reliable database for assessment of the radiation impact on spectral transmission of optical glasses still requires a very significant amount of experimental work to be undertaken. Taking into account the number of optical glasses used in space optics, it is proposed that a realistic approach consists of the introduction of a widely accepted and uniformly applicable methodology for testing of optical glasses. It would be particularly advantageous to introduce this methodology as an industrial standard (ISO and ECSS - European Cooperation for Space Standardisation, perhaps), which can be made applicable for radiation qualification testing of all optical glasses performed under ESA contracts. Test data would be deliverable to the agency in a suitable format. In this way the Agency will be able to accumulate experimental data required to build, maintain and update the database on a continuous basis. 5. 0 References 1. 2.

Gusarov, A. I., and Doyle, D. B., Modeling of gamma-radiation impact on transmission characteristics of optical glasses, in Photonics for Space and Radiation Environments II, Vol. 4547, Taylor, E., and Berghmans, F., Eds., SPIE Proc., Toulouse, France (2001). Schott Glass, SCHOTT Catalog 2000 - Optical Glasses, Schott (2000).

THE ROLE OF PROTON AND ELECTRON "ABNORMAL" FORMATIONS IN RADIATION IMPACT ON CONSTRUCTION ELEMENTS OF SPACECRAFTS YE. A.GRACHOV1, O.R.GRIGORYAN1, L.S.NOVIKOV1, I.V.TCHOURILO2 1 Skobeltsyn Institute of Nuclear Physics, Moscow State University, Vorobyevy Gory, Moscow, 119899, Russia 2 Rocket-Space Corporation "Energia", 4a Lenina str., Korolyev 141070, Moscow Region, Russia Abstract Based on measurements conducted onboard the Space Station MIR in 1999, and other original data, the spatial distributions and energy characteristics of “abnormal” areas of electrons with energy of order of 101 – 102 keV at medium and low latitudes (L ~ 1.3– 1.9) and protons with energies of 10 keV – 5 MeV near to geomagnetic equator (L < 1.15) are presented. Computations of the additional absorbed radiation doze values produced by the specified abnormal formations are given. Special attention is paid to estimates of the absorbed radiation doze in the surface layers of materials (with thickness of up to 10-20 micrometers). It is shown that the protons with energies less than 300 keV that are present in near-equatorial abnormal areas are responsible for the increase in the absorbed radiation dozes in such layers that are up to 1.5–2 orders of magnitude higher when compared with the data obtained in terms of the AP-8 model. 1. 0 Introduction For estimation of radiation impact on thin-film protective coatings and covers as well as on surface layers of outer shell materials of spacecrafts operating in Earth radiation belts and in low Earth orbits (LEO), it is important to take correctly into account the contribution to the absorbed radiation doze of electrons and protons with energies less than 100-200 keV. Unfortunately, the AE-8 and AP-8 models [1,2] used for description of space and energy characteristics of electrons and protons in radiation belts provide incorrect information for particles with the specified above energies. Direct use of the AE-8 and AP-8 models for computation of the absorbed doze values in the surface layers of materials provides an underestimated value of the doze, a fact that can result in erroneous estimations of radiation stability of materials and forecasting of their operational life-time duration. Moreover, it was revealed in a number of space experiments, that besides the “standard” structure of radiation belts at heights up to 1000 km there exist rather stable electron and proton formations located at various latitude and longitude intervals [3-5].

123

124

An estimation of the additional contribution from these formations to the total value of the absorbed radiation doze is of interest in this context as well. 2. 0 Instrumentation The results of measurements of electron and proton fluxes conducted at the space station MIR orbit and presented in this paper were obtained using semiconductor spectrometers, included in the SPRUT-6 device. The protons in the energy range 0.155.00 MeV were registered by a 200 µm thick semiconductor detector. For shielding of the electron flux, the magnetic filter allowing “to cut-off” the electrons with energy up to 500 keV was mounted in front of the detector. The pulses from the detector were registered by a 20-channel amplitude analyzer. A 2 mm thick semiconductor detector registered electrons and protons in the energy interval 0.1-1.0 MeV and 3.7-4.9 MeV, respectively. An aluminum foil 100 µm thick was installed in front of this detector that allowed "to cut-off" the low energy protons. The detector pulses were registered by a 20-channel amplitude analyzer. In addition, the energy losses were registered in both detectors that allowed obtaining the data on the absorbed radiation doze rate. 3.0 Experimental Data An example of protons registration in the near equatorial area made by the SPRUT-6 device is shown in Fig. 1. The integral count of protons with energies ȿɪ >450 keV (A) and the energy deposition rate in the detector (B) are shown as a function of time. The L parameter values are given at the bottom of the figure (C). The insert in the upper right corner of the figure shows the region with values L < 1.15. The increases of the proton flux are well visible in the area of geomagnetic equator. Detailed data on the differential energy spectrum of protons in the area of geomagnetic equator obtained earlier in experiments on Azur satellite, S81-1 [6,7] and OHZORA [8] and in an experiment onboard the Space Station MIR are presented in Fig. 2. For particle energies less than 300 keV, the data for disturbed (more intensive fluxes) and quiet geomagnetic conditions are given. The data from the sources [6 and 7] was used with a factor of 0.43 because the average for L < 1.15 differential proton spectrum obtained for the Space Station MIR was used in the computations. To avoid possible "contamination effects" of the measurement results by high energy particles penetrating from the South-Atlantic anomaly, the longitudes 270°-0°-30° were excluded from consideration in the analysis of the SPRUT-6 data. The SPRUT-6 data presenting approximately 100 orbits of MIR station was used for spectrum computation. Authors of the previous works on research of near equatorial proton population revealed, that dependence on geomagnetic activity is shown with protons with energy < 300 keV, at that time the flux of more energetic protons on L < 1.15 from geomagnetic activity practically does not depend [7,8]. Moreover, flux of protons with energy less than 300 keV exist on L < 1.15 even in normal (no activity) conditions.

125

Figure 1. An example of registration of proton fluxes in the SPRUT-6 experiment. A – integral proton count (for 5 sec) with energy Ep= 0.45 – 5.0 MeV, proton spectrum for L< 1.15; Benergy deposition in the detector (relative units); C- the L parameter values, vertical lines mark the boundaries of the L< 1.15 are where the near-equatorial proton population exists.

Figure 2. Differential proton energy spectrum for L < 1.15 1- SPRUT-6 onboard the MIR space station experimental data (350-400 km); 2- AZUR (384 – 3145 km), S81-1 (data extrapolated to 450 km) and OHZORA (350-800 km) satellites experimental data.

126

The results of the SPRUT-6 measurements of differential proton energy spectra for L < 1.15 (1) and average proton spectra in the MIR space station (2) in comparison to proton energy spectra for minimum (3) and maximum (4) solar activities according to model AP-8 are presented in Fig. 3. In the same figure, proton spectra with energy less than 300 keV for L < 1.15 in the case of normal. “quiet” (no activity) (5) and “disturbed” (6) geomagnetic conditions obtained in earlier experiments are presented as well. As can be seen from Fig. 3, the discrepancy between the data for protons on L < 1.15 with data from the modeling spectrum AP-8 for energy E>0.45 MeV is insignificant. However from Fig. 3 it is clear that for energy less than 1 MeV the average intensity of the proton spectra for an orbit of the MIR station exceeds the values based on the AP-8 model spectra. The spectrum of protons at L < 1.15 according to the data from SPRUT-6 agrees well with data received earlier. This allows making an assumption that the whole set of experimental data of the near-equatorial population of protons to be used in the estimation of their contribution to the total absorbed doze. Figure 4 show a compilation of data presenting the registration of electron formations for 1.2 < L < 1.8 done by various spacecrafts as follows: Ⱥ – MIR space station (1991, altitude ~ 400 km, Ee > 75 keV); ȼ – CORONAS satellite (1994, altitude ~ 500 km, Ee > 500 keV) [5]; ɋ - OHZORA satellite (1985-87, altitude ~ 350-800 km, Ee > 190 keV) [9]. In the bottom panel (ɋ) of Fig. 4, the areas of maximum intensity of electron fluxes (solid lines) and so-called «train» in which the less intensive flows are registered are

Figure 3. Differential proton energy 1 – SPRUT-6 data for L < 1.15; 2 - average proton spectrum in the MIR space station orbit based on SPRUT-6 data; 3,4 –AP-8 MIN and AP-8 MAX on model computations accordingly; 5,6 – AZUR satellite data for energy less than 300 keV, L < 1.15, for the quiet and disturbed geomagnetic conditions accordingly..

shown. The following basic features in distribution of electron fluxes are well visible:

127

(i) Despite the fact that electrons in these zones were registered during different years, the obtained precipitation boundaries practically coincide, time and space positions of these areas are stable as a whole, that is indicative at existence of a certain permanently working mechanism for particle precipitation. It is possible, that electron precipitation in the field of L ~ 1.6-1.8 is connected to the work of short-wave transmitters; (ii) Electrons occupy area of latitudes 1.2 < L < 1.8, though are distributed in the area non-uniformly; the maximal precipitation values are located near L ~ 1.3-1.4 and L ~ 1.6-1.7; (iii) The areas of electron fluxes distribution have a well expressed longitude dependence. The results of electron energy spectra measurement done by SPRUT-6 instrument in the “abnormal” formations on L ~ 1.2-1.4 (curve 1), L ~ 1.6-1.8 (2) in comparison to the data of the AE-8 MAX model (3) and the AE-8 MIN model (4) are presented in Fig. 5. It is clearly seen that the electron fluxes in these abnormal formations are small in comparison with the fluxes in radiation belts. Hence, they do not give the large additional contribution to the absorbed radiation doze value. The obtained set of experimental data allows determining the duration of the Space Station MIR stay in various radiation formations, that is necessary for the subsequent estimation of the absorbed radiation dozes caused by these formations. The relative durations of the Space Station MIR stay in various formations are given in Table 1. The criterion of definition of the Earth radiation belt areas was the excess of the integrated proton detector count of the 3 s-1 level. The SouthAtlantic anomaly area was excluded in the Figure 4. Areas of registration of the increased electron estimation of the stay fl mean latitudes: at duration in intervals L < Ⱥ - SPRUT-6 onboard the MIR space station, Ee > 75 keV; 1.15 and L ~ 1.2-1.8. B – CORONAS-I, Ee > 500 keV; C - OɇZORA, Ee > 190 keV accordingly.

128

These data were taken into account for obtaining the average energy spectra and for the absorbed radiation doze computations. TABLE 1. Relative durations of the MIR space station stay in various formations Space area In the Earth radiation belt Outside the Earth radiation belt L < 1.15 L = 1.2-1.8

Relative duration of the MIR space station stay in the area, % 12-15 85-88 18-22 20 %

Figure 5. Differential electron energy spectra: 1,2 – the SPRUT-6 data for L ~ 1.2-1.4 and L ~ 1.6-1.8; 3,4 - AE-8 MAX and AE-8 MIN models data accordingly.

4.0 Computation of the Absorbed Radiation Doze The absorbed radiation doze values were computed for the Space Station MIR’s orbit (height 350 km) using the data shown above. The results of computations of the annual absorbed radiation doze for proton fluxes with different spectra in a structure comprising a Si substrate covered by an infinite flat Al shield of different thickness are presented in Fig. 6. The computations for protons with energies higher than 500 keV were executed with the help of the SHIELDOSE model [10] for infinite flat shield. For energies less than 500 keV, the original model based on the analytical description of the proton energy loss in substances [11] was used. In Fig. 6, curve 1 shows the absorbed radiation doze obtained on the SPRUT-6 instrument for L < 1.15, curve 2 presents data obtained on the Space Station MIR orbit using the SPRUT instrument, curves 3 and 4 show data using AP-8 MIN and AP-8 MAX models and, finally, curves 5 and 6 – show data that takes into account the

129

additional low energy spectra for the quiet and disturbed geomagnetic conditions. Differential energy spectra presented in Fig. 3 were used for calculations. As can be seen from Fig.6, the significant contribution of the low energy protons to the value of the absorbed radiation doze in layers of materials up to 100 µm thick is clearly demonstrated.

Figure 6. Absorbed proton doze for various spectra: 1 - absorbed doze according to the SPRUT -6 data for L < 1.15; 2 - the SPRUT-6 data in the MIR space station orbit; 3 and 4 - AP-8 MIN and AP-8 MAX model computation; 5 and 6 – taking into account low energy supplementary spectrum for quiet and disturbed geomagnetic conditions; 7 – in the case of restriction on the low energy edge of 60 keV value.

It should be noted that the proton energy spectra, according to some data, registered in the internal radiation belt, could sharply break down for energies less than 50-60 keV. To estimate the contribution of protons with energies in this interval to the absorbed radiation doze, we computed the value for the spectrum shown in Fig. 3 (curve 5) restricted in the low energy edge by the value of 60 keV. The computational results for this case are shown in Fig. 6 with the dashed line (7). It is clear, that exclusion of small energy values from the spectrum for structures with thickness 10-20 µm, results in reduction of the absorbed doze value almost by an order of magnitude, whereas for large energies the absorbed doze values coincide with the curve 5. From the data shown in Fig. 3 it can be seen that the real contribution of near equatorial population of protons to the surface dose is insignificant and may constitute several percent for energies in the range 0.45-5.0 MeV: (1) -spectrum of protons near to geomagnetic equator, (2) - the average spectrum of protons for an orbit of Space Station MIR. However the fact that the inclusion of the near equatorial component of lowenergy protons results in essential increase of surface dose calculated by using model

130

AP-8, strongly indicates that without updating the AP-8 model and adding data on low energy protons fluxes it cannot be used correctly in predictive evaluations. 5.0 Conclusions The presented above results allow making the following conclusions. Low orbit spacecrafts with orbits similar to the Space Station MIR orbit are exposed to abnormal formations of protons near the geomagnetic equator for L < 1.15 and abnormal formations of electrons for 1.2 < L < 1.8. The protons with energies of less than 300 keV at equatorial formations are responsible for the increased absorbed doze values (up to 1.5 - 2 orders of magnitude) in the surface layers of materials with thickness of up to 100 micrometers as compared to the computational data obtained in terms of the AP-8 model. For more correct estimates of the relative contribution of abnormal formations of protons and electrons to the absorbed radiation doze values, further development of the AP-8 and AE-8 models for low energy values is necessary. 6.0 References 1. Vette, J.I. (1991) The AE-8 Trapped Electron Environment, NSSDC/WDC-A-R&S 1-24. 2. Sawyer, D.M., and Vette, J.I. (1979) AP-8 Trapped Proton Environment for Solar Maximum and Solar Minimum, NSSDC/WDC-A-R&S 76-06. 3. Biryukov, A., Grigoryan, O., Kuznetsov, S., Ryaboshapka, A., and Ryabucha, S. (1996) Low-energy charged particles at near equatorial latitudes according to MIR orbital station data, Adv. Space Res. 17, 189. 4. Grigoryan, O., Sinyakov, A., and Klimov, S. (1997) Energetic electrons on l < 1.2: connection to lightning activity, Adv. Space Res. 20, 389. 5. Bashkirov, V.F., Denisov, Yu.I., Gotseluk, Yu.V., Kuznetsov, S.N., Myagkova, I.N., and Sinyakov, A.V. (1999) Trapped and quasi-trapped radiation observed by CORONAS-I satellite, Radiation Measurements 30, 537. 6. Miah, M. (1989) Observation of low energy particle precipitation at low altitude in the equatorial zone, J.Atmos.and Terr. Phys. 51, 541. 7. Guzik, T.G., Miah, M.A., Mitchell, J.W., and Wefel, J.P. (1989) Low-altitude trapped protons at the geomagnetic equator, J.Geophys. Res. 94, 145. 8. Gusev, A.A., Kohno, T., Spjeldvik, W.N., Martin, I.M., Pugacheva, G.I., and Turtelli Jr., T. (1996) Dynamics of the low-altitude energetic proton fluxes beneath the main terrestrial radiation belts, J.Geophys. Res. 101, 196. 9. Nagata, K., Kohno, T., Murakami, H., et al. (1988) Electron (0.19-3.2 MeV) and proton (0.58 - 35 MeV) precipitations observed by OHZORA satellite at low zones L = 1.6-1.8, Planet. Space Sci. 36, 591. 10. Seltzer, S.M. (1979) Electron, Electron-Bremsstrahlung and Proton Depth-Dose Data for Space-Shielding Applications, IEEE Trans. Nucl. Sc. NS26, 21-60. 11. Andersen, H.H., and Ziegler, J.F. (1977) Hydrogen Stopping Powers and Ranges in All Elements, Pergamon Press.

A STUDY OF METHYLSILICONE RUBBER DAMAGE BEHAVIOR INDUCED BY PROTON IRRADIATION

LIXIN ZHANG, SHIQIN YANG, HONGBIN GENG, SHIYU HE, QIANG WEI Space Materials and Environment Engineering Lab, Harbin Institute of Technology, Harbin 150001

Abstract In this paper, the damage behavior of methylsilicone rubber induced by irradiation of protons with 150keV energy was studied. The surface morphologies, tensile strength, Shaw hardness, cross-linking density and glass temperature were examined. Positron annihilation lifetime spectrum analysis (PALS) was performed to reveal the damage mechanisms of the rubber. The results showed that the tensile strength and Shaw hardness of the rubber increased firstly and then decreased with increasing the irradiation fluence. The PALS characteristics IJ 3 and I3, as well as the free volume Vf, decreased with increasing the irradiation fluence up to 1015cm-2, and then increased slowly. It was noticed that the proton irradiation caused a decrease in the free volume of the methyl silicone rubber when the fluence was less than 1015 cm-2, while the free volume increased at the fluence greater than 1015cm-2. The results on cross-linking density indicated that the cross-linking induced by proton irradiation was dominant under smaller proton fluencies, increasing the tensile strength and Shaw hardness of the rubber, while the degradation of the rubber dominated under greater fluencies, leading to decreasing the tensile strength and Shaw hardness.

1. 0 Introduction In recent years, silicone rubber materials have been used widely in spacecrafts as binders of solar cells, sealants of air-locked modules, etc., due to their good electric insulation properties and excellent resistance to low temperatures and irradiation. However, long-term irradiation with charged particles (protons and electrons) in space environment

131

132

IJ

•

•

”

133 Japan, and the heating rate was 5 K/min. The process to measure the cross-linking intensity was as follows. The sample with an initial weight of W1 was immersed in 30ml of toluene for some time at 298K. Once the samples were removed from toluene, the solvent on the surface was absorbed quickly with filter paper. The above procedure was repeated, until the sample reached an equilibrium swelling state. The weight of the equilibrium sample was recorded as W2. After that, the solvent absorbed in the equilibrium sample was removed by heating it at 353 – 363 K, and the weight of the sample was recorded as W3. The difference between W3 and W1 was quantified as the dissolved rubber. According to the expression below, the cross-linking density Ve could be calculated: Ve=[ln(1-V2)+ V2+0.465 V22]/106.27(V21/3-0.5 V2) , where V2 is the volume percent of polymer in the swelling-equilibrium sample, which can be calculated from the W1, W2 and the sample density. Surface morphology of the specimens was observed using an Olympus BH2-UMA optical microscope. Hardness was measured in a LX-A type Shaw’ hardness-tester. The Shaw hardness was calculated according to the equation: FA = 549+75.12HA [1] , where FA(mN) is the force applied to the specimens, and HA the Shaw hardness., The tensile tests were performed using an MTS system according to the Chinese Standard GB/T2568. Figure 1 shows the tensile specimen dimensions.

4

25

10

R75

55 120 220

Figure 1. Sketch of the tensile specimen. Units: mm

3. 0 Results and Discussion 3.1 INFLUENCE OF PROTON IRRADIATION ON SURFACE MORPHOLOGY AND PROPERTIES The change in surface morphology of the silicone rubber after proton irradiation is shown in Fig.2. It was noticed that after the radiation mosaic cracks were formed and the crack number increased with increasing the irradiation fluence. Also, the color of the silicone rubber changed gradually with increased fluence, becoming deeper. The data in Table 2 and Fig. 3 demonstrate that the Shaw hardness and tensile strength of the silicone rubber increase firstly and then decrease with increasing of the irradiation fluence.

134

(a) (a)

(b) (b)

0.2mm 0.2mm

0.2mm 0.2mm

(a) F=1015cm-2

(b) F=1016cm-2

Figure 2. Change in surface morphology of the silicone rubber with increasing the proton fluence

TABLE 2. Influence of irradiation fluence on Shaw hardness of the silicone rubber

fluence/cm-2

0

1014

5×1014

1015

5×1015

1016

2×1016

Shaw hardness/A

68

70

72

74

68

69

65

3.2 INFLUENCE OF PROTON IRRADIATION ON CROSS-LINKING DENSITY AND GLASS TRANSITION TEMPERATURE During the process of irradiation by protons there would be two irradiation-induced effects, namely, the cross-linking and degradation, appearing in the silicone rubber, and the influence of them could be related to the irradiation fluence. As can be seen from Fig. 4, with increased irradiating fluence the cross-linking density of the rubber increased and then decreased slowly. Similarly, the glass transition temperature, Tg of the silicone rubber is shifted to higher temperatures with increasing irradiation fluence, and is decreasing gradually after fluence higher than 1015 cm-2, as shown in Fig.5. The above results imply that lower proton fluence irradiation resulted mainly in cross-linking of the macromolecules in the silicone rubber. Increasing the amount of cross-linking may lead to a decrease in the quantities of dissolved rubber, and thus to increasing the cross-linking density. As a result, the molecular chain movement would be restrained, resulting in the increase of glass transition temperature Tg, tensile strength and Shaw hardness. However, after the irradiation of the sample by protons to a fluence higher then 1015cm-2, the decrease of Tg , tensile strength and Shaw hardness was noticed that, in turn, implied that another irradiation-induced effect, namely degradation, starts to play the dominant role.

135

IJ3 IJ3 IJ3

ʌ

IJ3 ǻ

ǻ

ʌ

IJ3 IJ3

136

Figure 5. Glass temperature of the silicone rubber vs proton irradiation fluence

Figure 6.

The IJ3 values in PALS of the silicone rubber vs proton irradiation fluence

The presented PALS analysis results are in accordance with those for the cross-linking density and the glass temperature Tg. Intrinsically, the free volumes in a polymer can be related to the random arrangement of molecular chains due to thermo-fluctuation. In the early stages of irradiation, the protons can cause an increase in the number of cross-linking points and in the degree of polymerization, leading to a reduction of the free volumes. Therefore, the annihilation lifetime, IJ3 becomes shorter, as shown in Fig.6. Meanwhile the intensity I3 and the quantity of free volumes would decrease with increasing the numbers of cross-linking points (Figs.7 and 8). But after increasing the irradiation fluence to a certain degree, the irradiation-induced degradation in the rubber might become dominant and shows a reverse effect on the free volumes. Thus the increase of IJ3 , I3 and Vf was observed with increasing the irradiation fluence after 1015cm-2, and due to the competition between cross-linking and degradation in the silicon rubber, the increasing tendency was very slow.

Figure 7. The intensity I3 in PALS of the silicone rubber vs proton irradiation fluence

Figure 8. The free volume fraction Vf in the silicone rubber vs proton irradiation fluence

137 4. 0 Conclusions The tensile strength, the Shaw hardness, the cross-linking density, and the glass temperature of the silicone rubber increased firstly and then decreased with increasing the irradiation fluence of protons. The PALS characteristics including IJ3 , I3 and Vf for the silicone rubber decreased distinctly with increasing the irradiation fluence up to 1015cm-2, and then increased gradually. The cross-linking induced by proton irradiation was dominant under smaller proton fluencies, increasing the tensile strength and Shaw hardness of the rubber, while the degradation of the rubber dominated under greater fluencies, leading to decreasing the tensile strength and Shaw hardness.

5. 0 References 1. Chinese National Standard Collection, 1986, 14 vol., pp. 867-869. Chinese Standard Press, Beijing. 2. Salnikov V. A.,.Berejnoy V..M., Idkakov L.I., Khatipov S.A. Expert Evaluation of Possible Damage of the Optical Surfaces of T-170 Space Telescope Induced by Spacecraft Materials and Space Environment. 7th International Symposium on “Materials in the Space Environment”, Toulouse, France, 16-20 June 1997, 173-178. 3. Schmitt D-R., Ringel G. F., Kratz F. R., Neubauer R., Swoboda H., Hampe J., “Degradation Effects of Optical Components” in: 7th International Symposium on “Materials in the Space Environment”, Toulouse, France, 16-20 June 1997, 257-263. 4. Jean,Y.C., Positron Annihilation Spectroscopy for Chemical Analysis: A Novel Probe for Microstructural Analysis of Polymers , Microchemical Journal, 1990, 42 -72. 5. Eldrup M., Lightbody D., Sherwood J.N, Positronium Annihilation in Amine-Cured of Polymers, Chem.Phys, 1981, 6351~5751. 6. Jean Y. C., Sandreczki T. C., Ames D. P., Positron Annihilation in Amine-Cured Epoxy Polymers, J. Polym. Sci, 1986, B24, 1247~1253. 7. Bueche A M., Research of the Thermal Stability of Polysiloxanes , J. Polym. Sci, 1955, 15, 105.

This page intentionally left blank

A UNIFIED SPACE ENVIRONMENT EFFECTS DATABASE FOR RUSSIAN AND NORTH AMERICAN ORGANIC AND INORGANIC MATERIALS SUSAN H. C. P. MCCALL, ALAN A. CLARK AND ANTHONY J. CLARK Stellar Optics Research International Corporation (SORIC) 78 Normark Drive, Thornhill, Ontario, Canada L3T 3R1 [email protected], [email protected], [email protected] www.soric.com JACOB KLEIMAN AND ZELINA ISKANDEROVA Integrity Testing Laboratory Inc. 80 Esna Park Drive, Units 7-9, Markham, Ontario, Canada L3R 2R7 [email protected] , [email protected] www.itlinc.com BORIS BRISKMAN, EDWARD KLINSHPONT AND YULI SHAVARIN Karpov Institute of Physical Chemistry Obninsk, 249033, Kaluga region, Russia [email protected], [email protected], [email protected]

1. 0 Introduction An effort is underway to merge two significant materials databases to aid the international space communities. The Russian Karpov Institute Database consists primarily of data for organic materials for space and ground-based applications. The North American SOLEXIS™ database from Stellar Optics Research International Corporation (SORIC) consists primarily of data for inorganic materials, and many black, white, reflective and transmissive organic materials, for space and ground-based applications. The Karpov Institute Database on radiation stability of organic material has incorporated the results of tests conducted at Karpov Institute over the last 40 years on orders from enterprises of the Soviet space and defense industries. Test results for 638 materials were selected for inclusion into the Database, divided into fourteen categories: adhesives; coatings; composite; compounds and compositions; films; foamed materials; hermetics; inorganic materials; paints, lacquers and enamels; plastics; products; resins;

139

140 rubbers; and textile materials. Depending on material application, the data include mechanical, electro-physical, thermo-physical, optical, and other performance properties as well as radiation gas evolution characteristics. The scope of the tests comprised in the database cover various hazardous factors: accelerated electrons and protons, gammaand UV-radiation (including far UV), neutrons and X-rays, reactor radiation and combined effects of these factors. The test results are included that were obtained under varying irradiation conditions (environment, temperature and dose rate). The database presents only the information obtained by the Obninsk Branch of the Karpov Institute for the last 35 years. SOLEXIS™ is a software database used for stray light analysis, material selection and optical design for ground and space based applications. It is primarily used for selecting black, white, reflective, transmissive and other surfaces and materials for ground- and space-based optics applications, typically inorganic. It includes the following data for 500 materials: scatter data (BRDF and BTDF), spectral data (reflectance, transmittance and absorptance), space environment effects data, thermal, chemical, electrical, mechanical, and physical data and vendor information. The direct benefits to a database merger are: There is interest amongst the North American and European scientists and engineers in the Russian aerospace materials, and a database merger will enable users easy access to the data. Scientists and engineers who do research into new materials may be interested to compare the performance of their materials with the Russian materials, to look for ways to improve their research, and even create spin-off products. The data, and subsequent acquisition of materials may result in the reduction of risks on future space programs. More accurate performance predictions for instrumentation over the lifetime of the mission in space. It is critical, especially for long-duration missions, that the degradation in performance be calculated so that compensation can be made in advance. A well-maintained database will reduce the non-recurring design costs future space missions. The indirect benefits to a database merger are: Improved education, training and productivity for space researchers in government and business. Enhanced industrial competitiveness of the users The establishment and/or deepening of strategic international alliances and partnerships within the high technology community with Russia. Enhanced trade with the Russian space community re: materials, consultation and data distribution. A feasibility study is being conducted to explore the merging options for these two comprehensive, and complementing databases. The merging of the two databases will present a major milestone for the materials and optics communities for space and

141 ground based applications. Below, the Karpov Institute Database and the SOLEXIS™ database will be described in sections 2 and 3 respectively. 2. 0 The Karpov Institute Database on Radiation Stability of Organic Materials The Database has incorporated the results of tests conducted at Karpov Institute. Such tests have been carried out at the Institute for the last 40 years on orders from enterprises of the Soviet space and defense industries, such as the Khrunichev State Space Science and Industry Center, the Korolyov Rocket-Space Corporation “Energia”, the Lavochkin Science and Industry Association, the Krasnoyarsk Science and Industry Association of Applied Mechanics, the “Composite” Science and Industry Association, etc. 2.1 STRUCTURE OF THE DATABASE The kernel of the Database consists of five interrelated tables. 1. Table containing the list of materials, giving the brand of each material, its chemical base, type of the material and the production standard for the material. The items of information on chemical structure of a material are useful at selection and comparison of behavior of materials in extreme conditions. In particular, these items of information are useful in connection with a problem of formation own external atmosphere of space vehicles and her influence on change of the optical properties of materials and products in space. 2. Table of the types of materials. 3. Table of the groups of properties. 4. Table of the types of radiation. 5. Table of the test results, containing the name of the material, the type of the material, the group of properties, the irradiation type, the dose rate, the temperature of irradiation, the irradiation environment and the field of the embedded OLE object (tables and graphics). Test results for 638 materials were selected for inclusion into the Database. Selection of test results on the radiation stability of organic materials published in reference books, scientific papers, proceedings of conferences, as well as data contained in reports and records of tests conducted on orders from industrial enterprises was made on the base of maximum reliability. Analysis of the chemical structure, types and functions of the materials, classification of the test results according to groups of properties, types and conditions of the action of ionizing radiation must be certainly conducted in future.

142 2.2 MATERIAL TYPES Based on an analysis of their chemical structure, types and functions, the materials have been divided into 14 followed types: TABLE 1. Fourteen material types in the Karpov Database.

Adhesives

Inorganic materials

Coatings

Paints, lacquers, enamels

Composites

Plastics

Compounds, compositions

Products

Films

Resins

Foamed materials

Rubbers

Hermetics

Textile materials

2.2.1 Adhesives There are adhesives based on phenol-formaldehyde, epoxy, organosilicone and others kinds of resins. Adhesives are compositions capable of forming stable adhesive links with different surfaces. In space vehicles adhesives are used for gluing construction parts together. The Database includes test results for 41 brands of adhesives based on phenol-formaldehyde, epoxy, organosilocone and other resins. Table 2 presents the examples of functions in space vehicles of some brands of adhesives included into the Database. TABLE 2. Examples of adhesives in the Karpov database and their functions. Material

Function of the material

Adhesive 88NP

Gluing foil to the docking plane of the load box on a space station

Adhesive Balzamin-M

Optical adhesive for gluing together components of optical devices

Adhesive BF-4

Adhesive dip of the net fiber of the space vehicle cable system (SVCS), fixing beltlines on space station

Adhesive Cryosil

Gluing parts of the system of cryogenic heat insulation for temperatures 20473 K

Adhesive (optical)

OK-72-FT5

Gluing together components of optical devices

Adhesive UP-4-260-3Ɇ (optical)

Gluing together components of optical devices

Adhesive ED 6-8

Gluing together components of optical devices

Adhesive EKAN-3

Gluing together components of optical devices

Adhesive 88SA

Gluing rubbers, polyurethane, technical fabrics

143 2.2.2 Paints, Lacquers, Enamels. Lacquers, paints and enamels are solutions or suspensions film-forming substances (polymers, oligomers, monomers). They serve protective, decorative and thermal control purposes on the external surfaces of space vehicles. The Database includes test results for 91 brands of lacquers, paints and enamels based on epoxy, phenolformaldehyde and aminoformaldehyde resins, as well as polyacrylate, polyether and polyorganosiloxane. Table 3 presents the examples of functions in space vehicles of some brands of lacquers, paints and enamels included in the Database. TABLE 3. Examples of paints, lacquers, and enamels in the Karpov database and their functions on the external surface of the space vehicle.

Material Paint TNPF-851 Lacquer AK-113F Enamel AK-243 (black) Enamel AK-512 (white) Enamel AK-512 (black) Enamel KO-5191 (white) Enamel KCh-5269 Enamel EP-730

Function of the material on the external surface of the space vehicle Studied as coating for components of optical devices Used in manufacturing components of optical devices Coating for blinds in optical devices Thermal control coating (TCC) TCC TCC Studied as coating for components of optical devices in space vehicles Coating for construction materials on a space station

2.2.3 Compounds, Compositions This group of materials includes compositions based on polymers, oligomers and monomers intended for potting and impregnating current-conducting circuits and other elements of electronic equipment. In space vehicles, materials of this group are used on the external surfaces for protection against environment effects and mechanical impacts. The Database includes test results for 51 brands of compounds and compositions based on lacquers, paints and enamels based on epoxy, phenol-formaldehyde and organosilicone resins. Table 4 presents the functions in space vehicles of some brands of compounds and compositions included into the Database.

144 TABLE 4. Examples of compounds and compositions in the Karpov database and their functions on the external surface of the space vehicle.

Material

Compound EZK-6 (potting) Compound KCz Compound EDL-20 MB Compound 10-80-2M Compound Vixynth K-68 Compound Vixynth PK-68 Compound K-115 NK Compound K-153 (potting) Compound UP-5-220

Function of the material on the external surface of the space vehicle Fixing wires in plug ɫutouts Studied as a material for manufacturing parts of optical devices in space vehicles Fixing wires in plug ɫutouts Used for manufacturing parts of optical devices in space vehicles Fixing wires in plug ɫutouts Potting elements of electronic equipment Used for manufacturing parts of optical devices in space vehicles Used for manufacturing parts of optical devices in space vehicles Potting elements of electronic equipment

2.2.4 Hermetics Hermetics are compositions based on polymers and oligomers intended for applying to bolt, riveted and other joints to achieve their impermeability. The Database includes 28 brands of polyurethane, polyacrylate, organosilicone, epoxy and phenol-formaldehyde hermetics. The purposes served by some of them on the external surface of the space vehicle are presented in Table 5. 2.2.5 Composites Composites are multicomponent heterogeneous materials. They consist of a filler (a natural or synthetic organic and/or inorganic fiber) and a binder (as a rule, an organic resin). The Database includes test results for 87 composites. Table 6 presents several brands of composites.

145 TABLE 5. Examples of hermetics in the Karpov database and their functions on the external surface of the space vehicle.

Material

Function of the material on the external surface of the space vehicle Potting of riveted, welded, bolted, flange and others metal joints Fixing screw joints on a space station Potting of the electrical insulation of contacts and connectors of Mir station equipment

Hermetic VER-1 Hermetic Anaterm-4 Hermetic Vixynth U-1-18 Hermetic VGO-1 underlayer P-11

with

Hermetic Vixynth U-2-28 Hermetic VGO-1 Hermetic 51-G-23 Hermetic UT-34

Potting of the electrical insulation of electronic and radio equipment and movable contacts of connectors External and internal hermetic sealing of construction elements on a space station Potting of the electrical insulation of electronic equipment and movable contacts Fixing and hermetically sealing electronic equipment Filling up junction corner slots in optical devices in space vehicles

TABLE 6. Examples of composites in the Karpov database and their functions in the space vehicle.

Material Glass plastic STP

Composite heat-shielding AFT-2 (AFT-2P) Composite heatinsulating and soundproof ATM-1 Carbon plastic KMU Moulding material AG4S Glass textolite VFT-S Carbon plastic KMU-9

Function of the material in the space vehicle A construction material, thermal shields for the STR panel, thermal bridges, fixing of guardrails. Used for the installation of NHR panels. Thermal bridges, gaskets, fixing of guardrails, used as a protective construction material on a space stations Thermal insulation of the walls of the dry modules on a space station A construction material. Profiles for the installation of the NHO panels on PO-2 on a space station Parts of space vehicle optical devices working outside the pressurized compartments An unstrained construction material A construction material

146 2.2.6 Textile Materials The group of textile materials includes fabrics, tapes, fibers, threads and cords. From the point of view of the chemical structure, the Database includes mainly arimide, capron, phenylone, polyethylene-terephtalate and alumina borosilicate textile materials. The Database includes data on 63 brands of textile materials. Table 7 presents several of them that are used in space vehicles. TABLE 7. Examples of textiles in the Karpov database and their functions in the space vehicle.

Material Fabric NT-7 gummed vulcanized Glass fabric type TSON-SOT (optical function) Glass tape T-13 Tape grade KL-11-5,0

Function of the material in the space vehicle Used for facing the SVCS. Protection against effects of the environment For facing the mats of thermal blankets in space vehicles Faceplate on a space station For attaching detectors to brackets

Tape grade KL-11-5,0-SF Net fiber SS1RU-4-9x9

For attaching detectors to brackets For fixing elements of electronic equipment on the BS

Arimide duck frame fabric article 56420 Arimide fabric article 5355/3-85 (black, carbon filled)

Fabric used on the external surfaces of the space vehicle Fabric used on the external surfaces of the space vehicle

2.2.7 Products The Database includes test data on 15 ready-made products used in space vehicles. TABLE 8. Examples of products in the Karpov database and their functions in the space vehicle.

Material Tape LETSAR-BP-0.2 Adhesive tape LT-19 Polyvinylchloride tape Glass tape LES (electric insulating)

Function of the material in the space vehicle Insulation of cables For shaping cables on a space station For shaping cables on the Mir station For shaping cables, fixing brackets, manufacturing beltlines, fixing the mats of the thermal blankets on a space station

2.2.8 Rubbers The Database presents test results for 44 rubbers. The studied rubbers belong to the following groups: natural, silicone, butadienenitrile, olefine copolymer, fluoroelastomer, styrene-olefine copolymer. Rubbers are used on external surfaces for holding and facing

147 the detectors and cables of the space vehicle cable system. The Database includes test results for 5 inorganic materials (glasses) used in the optical devices of space vehicles. Also included are results of tests for 11 coatings. They are mainly epoxy and polyalkylsiloxane coatings. 2.2.9 Coatings The coatings included in the present Database serve the purposes of protection, electrical insulation and thermal control. The Database contains test results for 14 foam plastics. In space vehicles, these materials are used as heat insulation materials and construction fillers. Pure (non-composite) substances were included into two material type groups: films and plastics. The results of tests for these materials allow predicting their behavior in constructions and products. The Database includes 73 brands of films and 106 brands of plastics. TABLE 9. Description of radiation to which the materials in the Karpov database was exposed.

Radiation Type Gamma

Electrons

Protons Electromag netic radiation

Reactor radiation

Description of the Radiation 60

ɋɨ sources with the quantum energy 1.25 MeV were used for gamma-irradiation. Most data were obtained in the dose rate range 0.01 – 800 Gy/s. The irradiation environments: air and vacuum. The absorbed dose, as a rule, was no more than 3 Mgy. The linear electron accelerator ELU-2-8 was used for electron irradiation. The electron energy equaled 9 MeV. Dose rate - 10 – 1000 Gy/s. Irradiation environments: air or vacuum. Part of the data was obtained on linear electron accelerators with the electron energy 20-400 keV. Irradiation environment: vacuum. Dose rate: 10 – 500 Gy/s. Absorbed doses - up to 500 Mgy. The results were obtained under irradiation with protons with the energy 20 – 200 keV. Irradiation environment: vacuum. Fluence - up to 1019 cm-2. The samples were irradiated in a range of wavelengths simulating the extraterrestrial radiation of the Sun (0.2 – 2.5 µm). DKsShRB - 3000 - 5000 xenon superhigh pressure arclamps were used for this purpose. Part of the data was obtained using PRK-2 mercury lamps. Irradiation environment: vacuum. Exposition - up to 1500 equivalent solar days. Intensity - from 1 to 10 Suns in ultraviolet part of spectrum. The samples were irradiated in the Karpov Institute VVR-C reactor, in its experimental vertical and horizontal channels and thermal column niche. Types of radiation: mixed gammaneutron radiation and filtered neutron-radiation. Specific neutron flux, 1013 cm-2sec-1. Irradiation environment – air, vacuum. Absorbed doses up to 10 MGy.

148 2.3 PROPERTY GROUPS The Database presents the results of tests of the radiation stability of the materials under the action of gamma, electron, proton, pile and electromagnetic radiation.

The Database includes the test results on the mechanical, conductive, dielectric, thermophysical and optical properties of the materials after irradiation, as well as data on the mass loss and gas evolution of the irradiated samples. The data are presented both in graphic and table format. 2.3.1 Mechanical Properties Data on the following mechanical parameters are given: 1. tensile strength σ 2. ultimate deformability ε 3. tensile stress F/d 4. bending strength σb 5. compression strength σɫ 6. impact strength ak 7. shear strength τ 8. modulus of elasticity E 9. exfoliation strength 10. tearing strength 11. Shor hardness 12. Parameters of the creep. 13. The creep rate is determined by the velocity of the relative deformation ε: Vcr = dε/dt. The dose rate, stress and temperature dependencies of the creep rate have been studied. The parameters of the creep are connected with each other by the following empirical dependence: (1) Vcr = AR∆exp(ασ + βT) Here A is the coefficient of the creep rate characterizing its dependence on the dose rate, 0.5 < ∆ < 1 – exponent index, α is the coefficient of the creep rate characterizing its dependence on the stress σ, β is the coefficient of the creep rate characterizing its dependence on the temperature Ɍ. 14. Parameters of the longevity τ: τ = B/Vcr (B is the constant). Longevity is determined by the time of the rupture of the sample with the stress kept constant. The dose rate, temperature and stress dependencies of the longevity have been studied. 2.3.2 Optical Properties The following optical parameters of the irradiated materials are presented in the

149 Database: 1. αS and αλ - integral and spectral absorption coefficient of solar radiation (0.2 – 2.5 µm); ∆αS = αS(irrad) - αS(unirrad) – change of absorption coefficient of solar radiation 2. τS and τλ - integral and spectral transmission coefficient of solar radiation; ρS and ρλ - integral and spectral reflectance coefficient of solar radiation; 3. εH – integral coefficient of heat emission to a hemisphere; 4. D and D/x - optical density and relative optical density reduced to the unit of the sample thickness; 5. Absorption spectra; 6. Changes of appearance and color materials; 7. The groups of relative radiation and light stability. 2.3.3 Dielectric Properties The following dielectric parameters of the irradiated materials are given in the Database: 1. ε - dielectrical permittivity 2. tan δ - tangent of dielectric loss angle 3. EST – electric strength. The Database includes the temperature and frequency dependencies of these parameters for different absorbed doses and irradiation conditions (with varying the environment, temperature and dose rate). Besides, data are presented on the change of these parameters directly during the action of ionizing radiation on the samples for different frequencies, including super-high frequencies. 2.3.4 Conductive Properties The Database presents: 1. The specific volume and surface conductivity of the original and irradiated samples. 2. The effect on them of the absorbed dose and the irradiation conditions (environment, temperature, dose rate, mechanical pressure). 3. Radiation-induced volume and surface conductivity, σɪ =A(R/R0)∆ and σɪ=AS(R/R0)∆ respectively (A and AS are constants characterizing the electrical conductivity for the dose rate equal to one). 4. The effect on them of the dose rate, field intensity and temperature. Data are given on the influence of the physical structure of the materials on their conductive properties (crystallinity, orientation of the polymer chains, annealing and quenching of the samples). 2.3.5 Thermo-physical Properties The thermal properties of original and irradiated materials are characterized by the coefficient of thermal conductivity λ, the specific heat capacity C, the coefficient of thermal diffusivity a, the coefficients of linear β and volume α thermal expansion and the density ρ. The melting temperature Tm and the degree of crystallinity X are introduced for partially crystalline polymers.

150 2.3.6 Gaseous Products The data on the composition of the gaseous products presented in this Database have been for the most part obtained by the gas chromatograph-mass-spectrometer system. 1. The measure of the quantity of evolved gaseous products is given by the radiationchemical yield G, which is the number of molecules produced per 100 eV of radiation energy absorbed. 2. The Database also gives the values of the mass and volume of the evolved gas per unit of the sample mass under normal pressure. The ampoule technique was used for obtaining the data on radiation-induced gas evolution. The procedure is as follows: the samples are inserted into an ampoule (glass, quartz or metallic), which is then evacuated and then filled with the required gas. The ampoule is then sealed and irradiated. After irradiation, the pressure of the gaseous products and their content is measured. 2.3.7 Mass Loss The following technique was used to determine the radiation-induced mass loss: the initial weight of the sample was determined; the sample was then inserted into an ampoule and the ampoule was evacuated for several days. After this, the sample was weighed again under atmospheric pressure. Then the ampoule with the sample was evacuated again and the sample was irradiated in vacuum. After irradiation, the sample was weighed. 3. 0 SOLEXIS™ the Scatter Data and Optical Materials Property Database of Stellar Optics Research International Corporation (SORIC) 3.1 INTRODUCTION TO SOLEXIS™ SOLEXIS™ is a commercial scatter database and materials selection tool for spacebased, airborne and ground-based optical applications. It is used by optical scientists, optical engineers, and system engineers who require access to optical scatter data and other properties for black, white, reflective and transmissive surfaces and materials for stray light control, thermal control, calibration and visual target cues, for space- and ground based systems. As far back as the 1970s, the stray light community began advocating the creation of a database of scatter data to aid in the design and analysis of optical systems (excerpt from McCall, 2001). Government funding was sought, but no government program or company could justify the cost. From 1992 to 2001, Dr. Susan McCall of Stellar Optics Research International Corporation (SORIC) enlisted the cooperation of a number of organizations in the U.S.A. and Canada, and the Canadian Space Agency funded SORIC for the feasibility study and data acquisition phases of the project. About 20 organizations in the U.S.A. donated data. They include Oak Ridge National Laboratories, NASA, Breault Research Organization Inc., Toomay Mathis and Associates, Schmitt Measurement Systems Inc., the Optical Sciences Center at the

151 University of Arizona, and the Jet Propulsion Laboratory and many more. Ultimately, during the decade of development the project cost approximately $1,125,000 CDN cash and $2,625,000 CDN in unpaid in-kind labor. The Canadian Space Agency gave $70,000 total for the feasibility study and data acquisition phases, and other government sources gave a total of $70,000. SORIC raised the $1,125,000 cash through sales of its other products and services. A team of about 15 international expert advisors worked on the project, advising on data sources and content, organization and presentation. The Chief expert advisor for scatter data and black surfaces is Dr. R. Breault, Chairman of Breault Research Organization, Inc., and the Chief expert advisor for spectral data and white surfaces is Dr. Art Springsteen of Avian Technologies, (he was formerly Director of Reflectance of Labsphere, Inc.). Although SOLEXIS™ emphasizes scatter data, it is also a materials database that includes a host of other parameters. SOLEXIS™ is currently being used by clients in ten countries. The relevance of SOLEXIS™ to optical modeling will be described in section 3.2, followed by a description of SOLEXIS™ in section 3.3. 3.2 SCATTER DATA, STRAY LIGHT ANALYSIS AND OPTICAL MODELING SOFTWARE In general scatter data is used to: • Select surfaces and materials for the above applications. • Perform a computational stray light analysis to see how the material scatters light within the optical system. Considerable attention to stray light is paid by those interested in detecting weak signals from bright surroundings. One example is the use of space infrared sensors for astronomy or defense; another is the observation of the solar corona whose brightness in the visible spectral region is a million times less than that of the solar disc. • Aid in the manufacturing specifications of the surfaces and materials. Stray light in optical systems can come from scatter, diffraction and selfemission. Optical scatter arises when radiant flux is directed over a range of angles by interaction with a sample. Sources of scatter from a surface include surface topography, surface contamination, and subsurface effects. Stray light analysis is the analysis of all unwanted light paths in an optical system that reduce image quality or contrast. A stray light analysis is performed on optical systems to reduce scatter and thereby improve the signal-to-noise ratio, ghosting, contrast, sharpness, and image quality, and to produce the numerical results that indicate that performance specifications will be met. The process involves identifying the stray light paths and then finding ways to minimize the propagation of stray light. This can best be accomplished with stray light analysis software. Light scatter measurement, analysis, and stray light analysis are well established fields with wide-ranging applications. A detailed study is given by J. Stover's book (1995), and a summary is found in articles articles, such as S. McCall (2001), and Goodman (1992). Scatter data is essential to the creation of accurate stray light models and surface selection in optical systems. Conducting a stray light analysis early in the system design process saves vast amounts of time and money in design and manufacturing costs. Over the years, the field has evolved such that now the optical

152 modeling software that performs stray light analysis is now reliable and user friendly, can be used with confidence, and eliminates the need to rely on rules of thumb. A full analysis can be made in a timely, cost-effective way. The input to the computer program for the stray light analyses includes the optical design, size and shape of system objects, the scattering characteristics of each surface for all input and output angles, and the considerations including wavelength characteristics and source characteristics (e.g. spectral, spatial distribution and polarization). Finally, with the SOLEXIS™ scatter database, scatter data is now readily available, whereas before the commercially available optical design software packages typically had either no scatter data or only about a dozen curves – SOLEXIS™ brings them thousands of scatter curves. A convenient and well-accepted means of expressing optical scatter levels is with bi-directional scatter functions, where bi-directional refers to the angles of the incident and scattered light. The terms most commonly used to define or measure scatter are collectively called Bi-directional Scatter Distribution Function (BSDF). The most common are the Bi-directional Reflectance Distribution Function (BRDF) for reflected scatter and the Bi-directional Transmittance Distribution Function (BTDF) for transmitted scatter. The internationally accepted BRDF standard, which describes the measurement of the amount and angular distribution of optical scatter from an opaque surface, is the American Society for Testing Materials E1392-90 (ASTM, 1990). The standard includes 91 data fields that should accompany BRDF measurements. The SOLEXIS™ database was written to accommodate these 91 fields for each BRDF and BTDF (i.e. BSDF) measurement. The BSDF is a differential function dependent on the wavelength, incident direction, scatter direction, and polarization states of the incident and scattered flux. In practice, it is calculated as the average radiance (Wcm-2 steradian1 ) divided by the average irradiance (Wcm-2) of the sample. Where the units are steradian-1, the BRDF is defined as: BRDF = (Ps/ΩA cosϑs) / (Pi/A),

(2)

where Ps and Pi are the power of the scattered and incident light respectively, Ω is the solid angle subtended by the detector/receiver, A is the illuminated solid area, and ϑs is the angle of the scattered light. 3.3 PROPERTIES OF SOLEXIS™ The following is a brief summary of the characteristics of the SOLEXIS™ database: 3.3.1 Materials Included: • Paints and primers; lacquers and varnishes • Bulk materials: metals and alloys, glasses, polymers, ceramics, glass composites and more • Films, foils, tapes, meshes fibers, papers, textiles, fabrics dyes, inks, marking materials • Surfaces and films created by processes such as: anodization, electrodeposition, plasma spray, chemical vapor deposition, vacuum, proprietary processes and many more

153 3.3.2 Properties Included: • Optical • Scatter: BRDF and BTDF (Bi-directional Reflectance and Transmittance Distribution Function data) • Spectral reflectance, transmittance, and absorptance • Solar absorptance and emittance • Thermal, Electrical, Chemical, Mechanical, Physical properties • Space environment effects on the materials 3.3.3 Subject Areas Covered: • Surfaces and materials properties • Production issues • Manufacturing processes • Vendor information 3.3.4 Software Capability: • PC-Based, Windows95/98/NT/2000/ME/XP compatible, user friendly software • Multiple ranked search/sort materials based on the criteria: (1) name, (2) construction type or process, and (3) specification • View/print data/reports, images (i.e. photographs), and graphs/plots • Data formats: textual descriptions, graphical, documents, tabular data. 3.3.5 Main Analysis Capabilities for BRDF and BTDF Data: • Convert and plot to ß-ßo, cosine corrected, linear degrees, and linear BRDF and BTDF • Calculate Total Integrated Scatter (TIS) • Compute Power Spectral Density (PSD), including average surface wavelength and Root Mean Square (RMS) Roughness • Compare up to four data curves, choose scale, plot, print 4. 0 Conclusion The merging of the Karpov database and SORIC’s SOLEXIS™ database would be a significant achievement, and a major materials selection and research tool for the international space community. The subject areas complement each other, and as stated in the Introduction there are numerous direct and indirect benefits that would result from such a merger. A feasibility study is underway and funding for the project is being sought. A serious treatment of the materials selection process is one of the best investments that a program manager of a high performance space system can make, from budgetary, schedule and systems engineering points of view. With a database merger, the users will have even more comprehensive, accurate, up to date data and information for surfaces and materials which will:

154 • •

•

Substantially improves the ultimate system performance and minimize the risk of system failure and thus the cost of damage control; Substantially reduces the time, cost, and scope of a materials pre-selection process and/or measurement program simply because the optical, mechanical and thermal designers for high performance optical systems have not had ready access to the data and information in a single source before; Substantially increases the depth and competitiveness of proposals so that proper attention is given to the selection of surfaces and materials, which is usually critical to the system performance.

Both the Karpov Institute and SORIC have large additional quantities of materials not yet in either database. The Karpov Institute, for instance, has accumulated more than 1000 materials, presented in the form of 4000 electronic entries and has began to develop a list of foreign analogous materials to be able to compare Russian and foreign data on radiation stability. SORIC, has approximately 100,000 optical scatter curves, and data for 6000 surfaces and materials commonly employed in optical systems, that is not yet in the SOLEXIS™ database and efforts are underway to include this data. 5. 0 References 1. ASTM E1392-90, “Standard Practice for Angle Resolved Optical Scatter Measurements on Specular or Diffuse Surfaces,” Copyright American Society for Testing Materials (ASTM), 100 Barr Harbor Drive, West Consholtocken, PA 19428, USA, 1990. 2. Goodman Douglas, “Stray Light”, Optics and Photonics News, 3, (1) 52, 1992. 3 McCall Susan, The Importance of Scatter in Stray Light Analysis, Optics and Photonics News, pgs. 40-47, November, 2001. 4. Stover John C., “Optical Scattering Measurement and Analysis”, 2nd edition”, McGraw Hill, 1995.

THE EFFECT OF HEATING ON THE DEGRADATION OF GROUND LABORATORY AND SPACE IRRADIATED TEFLON FEP KIM K. de GROH NASA Glenn Research Center Cleveland, OH 44135 MORGANA MARTIN Ohio Aerospace Institute Brook Park, OH 44142

Abstract The outer most layer of the multilayer insulation (MLI) blankets on the Hubble Space Telescope (HST) is back surface aluminized Teflon FEP (fluorinated ethylene propylene). As seen by data collected after each of the three servicing missions and as observed during the second servicing mission (SM2), the FEP has become embrittled in the space environment, leading to degradation of the mechanical properties and severe on-orbit cracking of the FEP. During SM2, a sample of aluminized-FEP was retrieved from HST that had cracked and curled, exposing its aluminum backside to space. Because of the difference in optical properties between FEP and aluminum, this insulation piece reached 200°C on-orbit, which is significantly higher than the nominal MLI temperature extreme of 50°C. This piece was more brittle than other retrieved material from the first and third servicing missions (SM1 and SM3A, respectively). Due to this observation and the fact that Teflon thermal shields on the solar array bistems were heated on-orbit to 130 °C, experiments have been conducted to determine the effect of heating on the degradation of FEP that has been irradiated in a ground laboratory facility or in space on HST. Teflon FEP samples were x-ray irradiated in a high vacuum facility in order to simulate the damage caused by radiation in the space environment. Samples of pristine FEP, x-ray irradiated FEP and FEP retrieved from the HST during SM3A were heat treated from 50 to 200°C at 25° intervals in a high vacuum facility and then tensile tested. In addition, samples were tested in a density gradient column to determine the effect of the radiation and heating on the density of FEP. Results indicate that although heating does not degrade the tensile properties of non-irradiated Teflon, there is a significant dependence of the percent elongation at failure of irradiated Teflon as a function of heating temperature. Irradiated Teflon was found to undergo increasing degradation in the elongation at failure as temperature was increased from room temperature to 200 °C. Rate of degradation changes, which were consistent with the glass I transition temperatures for FEP, appeared to be present in both tensile and density data. The results indicate the significance of the on-orbit

155

156 temperature of Teflon FEP with respect to its degradation in the low Earth orbital space environment. 1.0 Introduction The HST was launched on April 25, 1990 into low Earth orbit as the first mission of NASA’s Great Observatories program. It is a telescope capable of performing observations in the near-ultraviolet, visible and near-infrared wavelengths (0.115-2.5 µm). The HST was designed to be serviced on-orbit to upgrade scientific capabilities. SM1 occurred in December 1993, after 3.6 years in space. SM2 was in February 1997, after 6.8 years in space. SM3A was in December 1999, after 9.7 years in space. The fourth servicing mission, designated as SM3B, occurred in March 2002. A future servicing mission is currently planned for mid 2003. The HST is covered with two primary types of thermal control materials, radiators and multi-layer insulation blankets, which passively control temperatures onorbit [1]. Both of these thermal control materials utilize metallized Teflon FEP as the exterior (space-facing) layer. Metallized Teflon FEP is a common thermal control material used on spacecraft, such as the Long Duration Exposure Facility, the Solar Max Mission spacecraft and HST, but it has been found to degrade in the low Earth orbital (LEO) space environment. Teflon FEP is used as the outer layer of thermal control insulation because of its excellent optical properties (low solar absorptance (αs) and high thermal emittance (ε)), in addition to its flexibility and low molecular weight. A metallized layer (Al or Ag) is applied to the backside of the FEP to reflect incident solar energy. The αs and ε of 5 mil (127 µm) thick FEP with an aluminized backing are 0.13 and 0.81, respectively [2]. Solar radiation (ultraviolet (UV) radiation and x-rays from solar flares), electron and proton radiation (omni-directional particles trapped in the Van Allen belts), thermal exposure and thermal cycling, and atomic oxygen exposure are all possible LEO environmental factors which could possibly contribute to the degradation of FEP. Analyses of aluminized-FEP (Al-FEP) and silvered-FEP (Ag-FEP) MLI blankets retrieved during SM1 revealed that the 5 mil (127 µm) thick FEP exterior layer was embrittled on high solar exposure surfaces [3, 4]. Surfaces, which received the highest solar exposures had microscopic through-thickness cracks in the FEP at stress locations [3,4]. Bonded solar facing 2 mil (51 µm) Al-FEP on the solar array drive arm (SADA) power harness, which was also retrieved during SM1, had many cracks and a total loss of mechanical integrity in heavily stressed areas [5]. The maximum temperature during thermal cycling of the power harness FEP was higher (>130 °C)[5] than that of the MLI FEP (maximum temperature of 50 °C)[6]. During SM2, severe cracking of the 5 mil Al-FEP MLI outer layer was observed on the light shield (LS), forward shell and equipment bays of the telescope. Astronaut observations combined with photographic documentation revealed extensive cracking of the MLI in many locations, with solar facing surfaces being heavily damaged [2]. Figure 1 shows two large cracked areas on the LS. A very large vertical crack can be seen near the center of the photograph, and a smaller cracked area, in which free standing Al-FEP had curled-up tightly (with the FEP surface in

157 compression), is located above the vertical crack. The worst of the MLI outer layer cracks were patched during SM2. Prior to patching the upper LS crack, the tightly curled Al-FEP outer layer was cut off and retrieved for post-mission analyses. Patches of 5 mil thick (127 µm) Al-FEP were placed over the two LS cracks, and patches of 2 mil thick (51 µm) Al-FEP were placed over large cracks in MLI on Equipment Bays 8 and 10. Figure 2 shows one of these cracked areas from Bay 10. As determined through a HST MLI Failure Review Board, embrittlement of FEP on HST is caused by radiation exposure (electron and proton radiation with contributions from solar flare xrays and UV radiation) combined with thermal cycling [6].

SM2 sample location

Figure 1. Two cracked areas in the MLI outer layer on the HST LS as witnessed during SM2. The astronauts have cut off the curled upper LS material and are preparing to place a patch over the area.

Figure 2. Cracks present in 5 mil thick Al-FEP Bay 10 MLI, photographed after retrieval during SM3A.

During SM3A, original MLI from Bay 10, which experienced 9.7 years of space exposure, as well as 2 mil thick Al-FEP patch material, which experienced 2.8 years of exposure, were retrieved and available for degradation analyses. Surprisingly, FEP retrieved during SM2 after 6.8 years exposure was found to be more embrittled than FEP retrieved 2.8 years later during SM3A, after 9.7 years exposure [7,8]. Because

158 the retrieved SM2 material curled with the FEP surface in compression, exposing the lower emittance Al surface to space, it experienced a higher temperature extreme during thermal cycling (≈200 °C) than the nominal solar facing MLI experiences (≈50 °C). As this was the only sample retrieved during SM2, it was important to determine the difference between the measured properties of this excessively heated sample and nominally heated MLI FEP. Another back surface metallized Teflon insulation, 2 mil thick Al-FEP thermal shields that covered the solar array bi-stems of the second pair of HST solar arrays (SA-2), thermal cycled to a maximum temperature of 130 °C on-orbit. Because of the various temperatures experienced by HST FEP materials, it is necessary to understand the effect of temperature on FEP degradation on HST. Understanding temperature effects is important for determining degradation mechanisms, and for facilitating the prediction of FEP degradation in LEO. Investigations have been conducted by de Groh et al on the effects of heating pristine FEP and FEP irradiated in ground facilities or in the LEO space environment, as reported in [7,8]. For this study, samples of pristine FEP, x-ray irradiated FEP and SM3A-retrieved FEP were heated from 50 °C to 200 °C in 25°C intervals in a high vacuum furnace and evaluated for changes in tensile properties and density in order to improve the understanding of the degradation of this insulation material in the LEO space environment. X-rays were used for the source of irradiation because x-rays from solar flares are believed to contribute to the embrittlement of FEP on HST [6], and because previous ground tests have shown that solar flare x-ray energies are energetic enough to cause bulk embrittlement in 127 µm FEP [10]. Also, the mechanism of embrittlement of polymers is believed to be the same for all forms of ionizing radiation, therefore x-ray exposure is a very useful technique for understanding radiation damage effects in Teflon. 2.0 Materials 2.1. PRISTINE FEP AND FEP FOR X-RAY IRRADIATION Teflon FEP is a perfluorinated copolymer of tetrafluoroethylene (TFE) and hexafluoropropylene (HFP). The FEP material used for x-ray irradiation followed by vacuum heat treatment and for non-irradiated heat treatment tests was non-aluminized 5-mil thick (127 µm), and was purchased from Sheldahl (lot #96-16). 2.2. HST SM3A Al-FEP As previously mentioned, MLI blankets originally installed on HST on Bay 10 (exposed to the space environment for 9.7 years) and 2 mil Al-FEP patches installed on Bay 10 during SM2 (exposed for 2.8 years) were retrieved by astronauts during servicing mission SM3A. Figure 3a shows a close-up of the MLI and patches on Bay 10 prior to being removed from HST. Four different materials were retrieved: original MLI from the top section of Bay 10 (Top MLI (TM)), original MLI from the bottom section of Bay 10 (Bottom MLI (BM)), and patches installed during SM2 over portions of TM and

159 BM, designated as Top Patch (TP) and Bottom Patch (BP), respectively. The outer most layer of the TM and BM materials was 5 mil thick Al-FEP. Retrieved patch material was 2 mil thick Al-FEP. The Al was ≈1000 Å. Samples were sectioned from various regions of the BM, TP, and BP surfaces for post-flight analyses and the results are reported in [7,8,11]. For the BM surface, regions designated as R1 and R2 refer to the areas without a patch and covered by a patch, respectively. A large section of the original MLI material, which was not covered by a patch and thus had been exposed to the space environment for 9.7 years (section BM-R1), was cut from the MLI blanket and provided for these heating studies. This sample section can be seen in Figure 3b. Because this large sample was sectioned from the blanket in 2001, while most samples for post-flight analyses testing were sectioned shortly after the December 1999 flight in 2000, this particular sample is referred to as the SM3A 2001 BM-R1 sample. The 2001 BM-R1 sample was carefully examined and the location of all impact sites and cracks were documented in order to avoid these areas when punching out tensile samples or cutting density samples.

TP

TM

BM BP

(a)

(b)

Figure 3. HST Equipment Bay 10 material: (a) Bay 10 MLI and patches during SM3A prior to removal from HST, and (b) A section of MLI blanket showing the location and size of the SM3A 2001 BM-R1 sample.

3.0 The HST Environment Table 1 provides the space environmental exposure conditions for the retrieved SM2 and SM3A BM-R1 materials. For the samples, the table lists the direction the surface faces with respect to the coordinate system for HST, indicated by V2 and V3 axes (described below). The sample retrieved during SM2 faced the +V3 direction, which is

160 the solar facing surface of Hubble. Because this area had cracked and curled with the FEP surface on the inside of the curl at some point during the mission, all environmental exposure conditions (except thermal cycling) are indicated as being some amount less than the values calculated for the entire mission duration. Bay 10 faces the –V2 direction (SADA direction). The environmental exposure conditions of solar exposure hours, solar event x-ray fluence, electron and proton fluence, and atomic oxygen fluence for FEP surfaces on HST vary depending on the direction the surface was facing and position of nearby obstructing surfaces, whereas the number of thermal cycles is independent of direction. TABLE 1. Exposure Conditions for Retrieved HST FEP Materials. Exposure Thermal Cycles/Temperature Range (cycles/°C) Equivalent Solar Hours (ESH) X-ray Fluence (J/m2)

SM2 FEP (LS, +V3)12

SM3A BM-R1 MLI (Bay 10, -V2)

37,100/-100 to +50° C -100 to +200°C when curled

52,550/ -100 to +50°C

< 33,638

13,598

1-8 Å*: < 209

0.5-4 Å*: < 13

1-8 Å : 62

0.5-4 Å : 3.9

Electron Fluence (#/cm2), >40 keV

< 1.95 x 1013

2.74 x 1013

Proton Fluence (#/cm2), >40 keV

< 1.95 x 1010

2.77 x 1010

* Values reported in Ref. 12 incorrectly assumed that +V3 surfaces always face direct sun.

Because the Bay 10 MLI surfaces, which approximately faced the –V2 direction, were at an oblique angle to the sun, Bay 10 MLI retrieved at SM3A actually received less equivalent solar exposure than the SM2 sample retrieved 2.8 years earlier. However, the SM3A BM-R1 MLI experienced many more thermal cycles, and higher electron and proton radiation, and atomic oxygen fluence than the SM2 sample. The exposure levels for various retrieved materials are affected by the solar activity, and it should be noted that there was a solar minimum between the SM2 to SM3A time period. More details of environmental exposures are provided in [8, 12]. 4.0 Experimental Procedures Exposures: 4.1 X-RAY EXPOSURE (PRISTINE FEP) A modified X-ray photoelectron spectroscopy (XPS) facility was used to irradiate the pristine FEP tensile samples. A copper target was irradiated with a 15.3 kV, 30 mA electron beam producing Cu x-rays (Cu Kα at 8048 eV, Cu L at 930 eV). The tensile samples were located 30.5 mm from the target, and the Cu x-rays were filtered through a 2 µm Al window (part of the x-ray tube). A 25 mil (635 µm) thick beryllium filter was placed over the FEP samples to absorb the low energy Cu L components, which would contribute significantly to damaging only the surface [13]. The x-ray flux was 13.28 W/m2[14]. The choice of target material, electron beam energy, and filter was chosen to produce a high flux, uniform distribution of energy absorbed, versus depth in the film. The energy deposition rate, or dose rate, versus depth below the surface for 127 µm FEP film at the specified exposure conditions are provided by de Groh and

161 Gummow in [15]. The technique used to characterize the x-ray source and energy deposition within the FEP film is described by Pepper and Wheeler in [13]. Pepper et al provide quantitative characterization of the Cu x-ray source and the absorbed energy deposition rate within a 75 µm film in [14]. X-ray irradiated samples were stored under vacuum until they were tensile tested or vacuum heat-treated. 4.2 VACUUM HEAT TREATMENT Pristine FEP, x-ray exposed FEP and HST SM3A Al-FEP were vacuum heat treated from 50 °C to 200 °C in 25° C intervals in a high vacuum facility adapted with a tube furnace. A Teflon lined Cu pipe was placed inside the tube furnace to promote uniform heating. The exposure temperature was monitored with a thermocouple attached to a Teflon witness sample, held in contact with the test samples. The pressure was 10-6 to 10-7 torr during heating. Samples were heated at the desired temperature for a target of 72 hours. 4.3 TENSILE PROPERTIES Samples for tensile testing were ‘dog bone’ shaped and die-cut using a tensile specimen die manufactured according to ASTM D638-95, type V. The tensile samples were 3.18 mm wide in the narrow section (neck), with a 9.5 mm gauge length. Samples were tested using a bench-top tensile tester with a 4.54 kg load cell and a test speed of 1.26 cm/min. Ultimate tensile strength (UTS) and elongation at failure were determined from the load displacement data. 4.4 DENSITY MEASUREMENTS Density measurements were obtained using density gradient columns calibrated using glass float standards of known densities (± 0.0001 g/cm3). The density solvents used were carbon tetrachloride (CCl4, ρ = 1.594 g/cm3) and bromoform (CHBr3, ρ = 2.899 g/cm3). The presence or absence of the thin (1000 Å) aluminized coating (as removed by NaOH solution) was found to have no effect on the density of the Al-FEP samples. 5.0 Results and Discussion 5.1 ROOM TEMPERATURE TENSILE PROPERTIES 5.1.1 Pristine FEP The room temperature (23 °C) tensile data for pristine FEP are listed in Table 2. The UTS and percent elongation at failure for the pristine FEP, for an average of 13 samples, was 24.1 ± 1.5 MPa and 271.2 ± 16.9%, respectively. 5.1.2 HST SM3A Al-FEP The retrieved SM3A Teflon from HST, after 9.7 years in the space environment, is substantially degraded. The tensile results for the as-retrieved SM3A FEP are listed in

162 Table 2. If compared to the pristine FEP tested in this study, the UTS of the retrieved HST FEP has decreased from 24.1 to 13.9 MPa, and the elongation at failure has decreased from 271.2% to 55.3%. These correspond to decreases of 42.3% and 79.6%, respectively. These tensile properties provide insight into the damage mechanism of Teflon in space. Because the UTS decreased, with the decrease in elongation at failure of the space-exposed FEP, chain scission is identified as the primary degradation mechanism on HST. TABLE 2. Tensile Properties of Vacuum Heat-Treated Pristine, X-ray Irradiated and HST Retrieved Teflon FEP. Vacuum Heat Treatment Temperature Room Temperature 23 °C 50 °C (+/- 1 °C) 75 °C (+/- 2 °C) 100 °C (+/- 1 °C) 125 °C (+/- 1°C) 150 °C (+/- 3 °C) 175 °C (+/- 12 °C) 200 °C (+/- 4 °C)

Material

Number of Samples

UTS (Mpa)

% Elongation at Failure

Pristine FEP

13

24.1 +/- 1.5

271.2 +/- 16.9

X-ray FEP

10

17.1 +/- 1.5

212.7 +/- 31

HST Al-FEP

4

13.9 +/- 0.4

55.3 +/- 9.3

Pristine FEP

4

23.4 +/- 0.7

264.1 +/- 12.6 162.6 +/- 35.5

X-ray FEP

4

15.1 +/- 1.1

HST Al-FEP

4

13.9 +/- 0.3

46.5 +/- 4.1

Pristine FEP

6

22.5 +/- 1.5

259.6 +/- 14.2 134.9 +/- 46.3

X-ray FEP

8

15.3 +/- 0.2

HST Al-FEP

4

14.4 +/- 0.1

25.7 +/- 5.5

Pristine FEP

4

22.1 +/- 0.7

250.2 +/- 5.7 43.1 +/- 6.6

X-ray FEP

4

15.5 +/- 0.3

HST Al-FEP

4

14.5 +/- 0.2

10.4 +/- 1.1

Pristine FEP

4

22.5 +/- 0.8

254.7 +/- 8.3 23.8 +/- 4.9

X-ray FEP

4

15.8 +/- 0.2

HST Al-FEP

3

14.2 +/- 0.7

9.4 +/- 3.3

Pristine FEP

4

22.8 +/- 0.6

271.0 +/- 6.4 22.8 +/- 5.6

X-ray FEP

4

15.4 +/- 0.3

HST Al-FEP

4

14.2 +/- 0.3

8.7 +/- 4.2

Pristine FEP

3

22.1 +/- 0.8

287.4 +/- 11.1 15.2 +/- 3.5

X-ray FEP

4

16.4 +/- 0.8

HST Al-FEP

4

14.9 +/- 0.8

7.9 +/- 2.0

Pristine FEP

4

23.7 +/- 0.8

306.8 +/- 9.6

X-ray FEP

2

16.1 +/- 0.4

9.7 +/- 5.2

HST Al-FEP

3

14.5 +/- 0.5

4.5 +/- 0.9

5.1.3 X-Ray Irradiated FEP The x-ray exposure was not intended to simulate the full extent of damage occurring on Hubble, but to cause irradiation induced polymer damage, and still have enough elongation at failure remaining to see the effects due to heating. Based on a series of prior tests, it was determined that a 2-hour exposure would provide the desired reduction in tensile properties [15]. Prior tests also indicated that the maximum number of samples that could be uniformly exposed at a time was two. The samples were centered in a holder that provided a 2.0 x 2.0 cm exposure area (the tensile sample

163 gauge length is ≈1 cm). The total energy absorbed per unit area integrated through the full thickness (the areal dose, D) of the 127 µm film for the 2-hour exposure was 33.8 kJ/m2.[16]. Density tests were conducted on pieces sectioned from a 200 °C heated xray irradiated tensile sample, which indicated that the irradiation exposure was uniform across the length of the exposed area. After x-ray exposure, the UTS and percent elongation at failure, for an average of 10 samples, was 17.1 ± 1.5 MPa and 212.7 ± 31%, respectively. This is a 29.0% reduction in the UTS and a 21.6% reduction in percent elongation at failure due to irradiation embrittlement. Although it was not the goal of this study to try to simulate the extent of damage on HST with the x-ray exposure, it was decided to compare the areal dose for xray irradiated FEP with that experienced by the FEP on HST. The areal dose for the 5 mil thick HST SM3A FEP is provided in Table 3 [17]. The total areal dose for the SM3A BM-R1 FEP was 427.2 J/m2. It should be noted that the HST FEP has an elongation at failure of 55% with an areal dose of only 427 J/m2, while the x-ray exposed FEP was much less embrittled (213% elongation), after orders of magnitude higher areal dose (33,800 J/m2). The factors that could contribute to these differences include the extreme differences in the dose rates (i.e. time factor), the variation in ionizing species and energies, temperature differences during irradiation exposure and the contribution from thermal cycling on HST (52,550 cycles from –100 to +50 °C), and possibly, surface effects from atomic oxygen and UV exposure in space. This stresses the difficultly in conducting simulated space environment durability tests, and emphasizes the potential complication when conducting durability testing based strictly on expected mission fluence or dose values. TABLE 3. Areal Dose for HST SM3A FEP. SM3A BM-R1 MLI (Bay 10, -V2)

Areal Dose (J/m2)

X-rays, 1-8 Å

29.80

X-rays, 0.5-4 Å

0.72

Electrons, >40 keV

389.6

Protons, >40 keV

7.11

5.2 VACUUM HEAT TREATMENT 5.2.1 Tensile Properties The results of tensile tests for the pristine, ground-laboratory irradiated and HST FEP after vacuum heat treatment are listed in Table 2 along with the room temperature data. The data are graphed in Figure 4. There was no degradation in the tensile properties of vacuum heat-treated non-irradiated FEP, in fact, with this batch of FEP an increase in the percent elongation at failure was observed for the higher temperatures (175 & 200 °C). Although heat treatment did not cause much change in the UTS of x-ray irradiated FEP with vacuum heat treatment, there was a dramatic decrease in the percent elongation at failure, as can be seen in Figure 4. The elongation decreased from 212.7% at 23 °C to only 9.7% after 200 °C exposure. This corresponds to a 95% decrease. And as can be seen in the graph, there is a rapid decrease in the elongation from 23 °C to 100 °C, with near complete losses of elongation from 125 °C to 200 °C. Only two of the

164 original four x-ray samples heated to 200 °C could be tensile tested because two stuck together slightly together during heating and then broke during separation. Although the FEP retrieved from HST was significantly embrittled in its asretrieved condition, it became even more embrittled with vacuum heat treatment (even after vacuum heat treatment at 50 °C, the maximum on-orbit temperature). The spaceexposed HST FEP followed a similar trend as the ground-laboratory x-ray irradiated FE, showing little changes in UTS and decreases in elongation from 23 °C to 100 °C, with near complete loss of elongation with heating to 100 °C and higher. 350

% Elongation at Failure

300 250 200

Pristine 150

X-Ray

100

HST

50 0 0

25

50

75

100

125

150

175

200

Vacuum Heat Treatment Temperature (C) Figure 4. Percent elongation at failure of pristine, ground laboratory x-ray exposed and retrieved HST Teflon FEP as a function of vacuum heat treatment temperature.

5.2.2 Density The density data for the pristine, ground-laboratory irradiated and HST space irradiated FEP, at room temperature and after vacuum heat treatment, are listed in Table 4. The data are graphed in Figure 5. The standard deviation is given when more than one sample was measured and averaged. As can be seen in the graph, the density of the retrieved HST FEP is essentially the same as pristine FEP, and the room temperature xray irradiated FEP is just slightly more dense than pristine FEP, even though these irradiated samples are significantly embrittled. This indicates that although irradiation induces scission in the polymer chains, resulting in embrittlement, the actual packing of the chains is not affected by irradiation exposure. There were very gradual increases in the density with heating up to 75 °C for all samples. Significant increases started at 100 °C, with larger increases corresponding to higher temperatures. Although the density increased with temperature for all samples, larger increases occurred for the samples that had been irradiated either in

165 space or in the ground facility than for pristine FEP. These results are consistent with de Groh's previous studies that show pristine FEP increases in density with heating, but FEP from HST has greater increases in density for the same heat treatment (200 °C exposure, references [7] and [9]). This is attributed to irradiation-induced scission of bonds in space, which allows for greater mobility and crystallization upon heating than that which occurs with non-irradiated FEP. Previous x-ray diffraction studies verify that the increases in density correlate to increases in polymer crystallinity.7,9 The density results further support chain scission as the primary mechanism of degradation of FEP in the space environment. TABLE 4. Density Data of Vacuum Heat-Treated Pristine, X-Ray Irradiated and HST Retrieved Teflon FEP. HST FEP Pristine FEP X-Ray FEP (SM3A 2001 BM-R1) Temperature (°C) Std. Density Std. Density Std. Density (g/cm3) Dev. (g/cm3) Dev. (g/cm3) Dev. 23

2.1373

0.0011

2.1407

-

2.1376

0.0005

50

2.1379

-

2.1414

-

2.1376

0.0005

75

2.1379

-

2.1428

-

2.1389

0.0005

100

2.1393

-

2.1477

-

2.1407

-

125

2.1414

-

2.1585

-

2.1456

-

150

2.1473

-

2.174

-

2.1577

-

175

2.1507

-

2.1775

-

2.1647

0.0016

200

2.1631

-

2.1856

-

2.1696

0.0031

2.19 Pristine

2.18

X-Ray

3

Density (g/cm )

HST

2.17

2.16

2.15

2.14

2.13 0

25

50

75

100

125

150

175

200

Vacuum Heat Treatment Temperature (C) Figure 5. Density of pristine, ground laboratory x-ray exposed and retrieved HST FEP as a function of vacuum heat treatment temperature.

166 When comparing the curves for the elongation and density data, it was observed that in each set of data there appeared to be a noticeable change in the slope of the data around 100 °C. The data was therefore graphed with linear fits for two sections of the data. The lines chosen were based on the best fit for each individual section of data. The resulting curves for the elongation at failure and density data are shown in Figures 6 and 7, respectively. The “change-of-slope” temperature has been highlighted in these graphs at the intersection of the two linear fits. The change-of-slope temperature of the pristine FEP (115 °C for the elongation data and 126 °C for the density data) correlates well with the glass I transition temperature (α relaxation), which is listed from ≈83 °C to 150 °C in the literature, dependent on hexafluoropropylene (HFP) content [9]. Eby and Wilson report transition temperatures for FEP with densities (2.136-2.135 g/cm3) similar to the pristine FEP examined in this report at ≈150 °C and ≈127 °C for 10.7 and 17.7 mol % HFP, respectively [18]. Commercially available FEP is reported to be 20 mol % HFP[19], which would indicate that the transition temperature for pristine FEP would be close to 125 °C based on the Eby study, which is consistent with the change-of-slope temperatures for the pristine FEP. Another interesting observation is that the irradiated samples have lower change-of-slope temperatures than pristine FEP. For example, the temperature in which the density of the ground-laboratory irradiated FEP starts to increase quickly is 82 °C, while it is 100 °C for the HST retrieved FEP and 126 °C for pristine FEP. These results indicate that irradiation causes changes in the polymer structure allowing increases in crystallization to occur at a lower temperature than which it occurs in pristine FEP. 6.0 Summary and Conclusions The objective of this research was to determine the effects of heating on ground laboratory irradiated FEP and FEP retrieved from the Hubble Space Telescope, in order to better understand the effect of temperature on the rate of degradation, and on the mechanism of degradation, of this insulation material in the LEO environment. Samples of pristine FEP, x-ray irradiated FEP and HST SM3A-retrieved FEP were heated from 50 °C to 200 °C in 25°C intervals in a high vacuum furnace and evaluated for changes in tensile properties and density. Results indicate that although heating does not degrade the tensile properties of non-irradiated Teflon, there is a significant dependence on the degradation of the percent elongation at failure of irradiated Teflon as a function of heating temperature, with dramatic degradation occurring at 100 °C and higher exposures. The density of non-heated irradiated FEP (ground or space irradiated) was essentially the same as pristine FEP, although these samples are significantly embrittled. This indicates that irradiation induces scission in the polymer chains, resulting in embrittlement, but chain packing is not affected. Gradual increases in the density occurred with heating from 23 °C to 75 °C for all samples, with significant increases occurring at 100 °C and higher exposures. Larger increases occurred for the irradiated samples than for the pristine FEP. These results were consistent with previous studies that show pristine FEP increases in density with

167 heating, but irradiated FEP experiences greater increases for the same heat treatment. This is attributed to irradiation-induced scission of bonds, which allows for greater mobility and crystallization upon heating than that which occurs with non-irradiated FEP. Changes in the rate of degradation were present in both elongation and density data. The change-of-slope temperatures of the pristine FEP (115 °C and 126 °C, for elongation and density, respectively) correlate with the glass I transition temperature of FEP. The change-of-slope temperature of irradiated FEP was lower than for pristine FEP, further indicating that scission damage has occurred. The tensile results and heated density data support chain scission as the primary mechanism of degradation of FEP in the space environment. The results show the significance of the on-orbit service temperature of FEP with respect to its degradation in the LEO space environment.

Figure 6. Change in the slope of the percent elongation at failure data of pristine, ground laboratory x-ray exposed and retrieved HST FEP as function of vacuum heat treatment temperature.

168

Figure 7. Change in the slope of the density data of pristine, ground laboratory x-ray exposed and retrieved HST FEP as function of vacuum heat treatment temperature.

7.0 Acknowledgments We would like to thank Dr. Stephen Pepper and Dr. Donald Wheeler of GRC for the use of, and characterization of, their x-ray facility. We thank Ed Sechkar of QSS, Inc. for build-up of the vacuum furnace facility. We appreciate technical contributions from Bruce Banks of GRC, and we thank Joyce Dever of GRC for providing areal dose values for the HST FEP. We would like to acknowledge John Blackwood and Jackie Townsend of NASA GSFC, and Ben Reed of Swales Aerospace, and the HST Project Office for providing the retrieved HST material for this study. 8.0 References 1. 2. 3. 4. 5. 6. 7. 8.

P. A. Hansen, J. A. Townsend, Y. Yoshikawa, D. J. Castro, J. J. Triolo, and W. C. Peters, (1998), SAMPE International Symposium, 43, 570. J. H. Henninger, (1984), NASA RP 1121. K. K. de Groh and D. C. Smith, (1997), NASA TM 113153. T. M. Zuby, K. K. de Groh, and D. C. Smith, ESA WPP-77, 385 (1995); NASA TM 104627, Dec. 1995. M. Van Eesbeek, F. Levadou, and A. Milintchouk, (1995), ESA WPP-77, 403. J. A. Townsend, P. A. Hansen, J. A. Dever, K. K. de Groh, B. A. Banks, L. Wang and C. He, High Perform. Polym. 11, 81-99 (1999). K. K. de Groh, J. A. Dever, J. K. Sutter, J. R. Gaier, J. D. Gummow, D. A. Scheiman and C. He, High Perform. Polym. 13, S401-S420 (2001). J. A. Dever, K. K. de Groh, R. K. Messer, M. W. McClendon, M. Viens, L. L. Wang and J. D. Gummow, High Perform. Polym. 13, S373-S390 (2001).

169 9.

K. K. de Groh, J. R. Gaier, R. L. Hall, M. P. Espe, D. R. Cato, J. K. Sutter and D. A. Scheiman, High Perform. Polym. 12, 83-104 (2000). 10. B. A. Banks, K. K. de Groh, T. J. Stueber, E. A. Sechkar, and R. L. Hall, SAMPE International Symposium, 43, 1523 (1998); also NASA TM-1998-207914/REV1. 11. J. R. Blackwood, J. A. Townsend, P. A. Hansen, M. W. McClendon, J. A. Dever, K. K. de Groh, B. B. Reed, C. C. He and W. C. Peters, SAMPE 2001 Conference Proceedings, May 6-10, 2001, Long Beach, CA, pp. 1797-1810. 12. J. A. Dever, K. K. de Groh, B. A. Banks, J. A. Townsend, J. L. Barth, S. Thomson, T. Gregory and W. Savage, High Perform. Polym. 12, 125-139 (2000). 13. S. V. Pepper and D. R. Wheeler, Review of Scientific Instruments, Vol.71, No. 3, March 2000, 15091515. 14. S. V. Pepper, D. R. Wheeler and K. K. de Groh, Proceedings of the 8th ISMSE & 5th ICPMSE Conference, June 5-9, 2000, Arcachon, France. 15. K. K. de Groh and J. D. Gummow, High Perform. Polym. 13, S421-S431 (2001). 16. S. V. Pepper, NASA Glenn Research Center, personal communication (1999). 17. Joyce Dever, NASA Glenn Research Center, personal communication (2002). 18. R. K. Eby and F. C. Wilson, J. of Applied Physics, 33, 2951-55 (1962). 19. Don Farrelly, DuPont, personal communication (1999).

This page intentionally left blank

ON THE THERMAL APPLICATION

STABILITY

OF

POLYIMIDES

FOR

SPACE

C.O.A. SEMPRIMOSCHNIG, S. HELTZEL, A. POLSAK, M. V. EESBEEK Materials Physics and Chemistry Section European Space Research and Technology Centre (ESTEC), European Space Agency (ESA), Keplerlaan 1, PO Box 299, NL-2200 AG Noordwijk, The Netherlands Abstract Currently planned missions of ESA (European Space Agency) to the inner part of the solar system will require the use of materials at an extreme radiation/temperature environment. This paper deals with the investigation of the thermal stability of two types of polyimides at a temperature of 350°C. Both materials were assessed by TGA (Thermo Gravimetric Analysis). Further tests were conducted in a high vacuum facility at 350°C. Test data were gathered up to a duration of 2500 hrs. The thermal stability was assessed by mass loss measurements and by UV/VIS/NIR spectrophotometric transmission measurements. Furthermore the degradation behaviour of the thermooptical properties versus time at this temperature was characterised. 1. 0 Introduction ESA (European Space Agency) is currently planning missions to the inner solar system. The first of such missions named Bepi-Colombo was approved at the end of 2000 as the fifth cornerstone mission of the Horizon 2000 science programme. This mission aims at a comprehensive exploration of the innermost terrestrial planet and aims to find answers about the understanding of planetary formation as well as the evolution in the hottest parts of our solar system [1]. As the environment closer to the sun will be harsher in terms of the impinging solar radiation (particle, UV etc.) such missions set a major challenge to materials and processes (M&P) and will require design solutions that are somewhat different to solutions used for spacecraft (s/c) in Earth orbit applications. The solar irradiance is inverse proportional to the square of the distance from the sun. At the Mercury perihelion for instance at about 0.3 AU it is more than a factor of ten above the averaged earth solar constant of about 1.3 kW/m2. Such high solar irradiances will naturally increase the temperature of any s/c in that environment. For the materials engineer it is therefore important to predict and understand the behaviour of materials at such an environment/temperature. This requires understanding the limits of the available materials, what drives their degradation

171

172 mechanisms and whether a material could fulfil its application in a certain environment within the s/c design life. This triggered us to adopt a certain testing & analysis philosophy within the Materials & Processes division. This philosophy is sketched in fig. 1 and has been presented in a previous paper [2]. In this paper results about the thermal stability of two different types of polyimides are presented. Polyimides have been widely characterised in the past and are well known for their thermal stability [3,4,5,6] and their intriguing mechanical properties [7]. Such materials are commonly used on the outside of s/c for thermal control purposes and are good candidates for sunshields [8,9]. The temperatures that such an external layer will reach depend on its solar absorptance (α) versus thermal emittance (ε) ratio, i.e. the ratio between the amount of the absorbed solar energy and the amount of the thermally emitted radiation. This socalled α/ε ratio can be varied in a wide area with various coatings. For inner solar system missions we want to assess the behaviour of such polymers at extreme temperatures. Figure 2 shows equilibrium temperatures as a function of the α/ε ratio for three different distances. The first would be at 1 Astronomical Unit (AU), the second APPROACH OF MATERIAL TESTING AND ANALYSIS

Materials

Degradation Tests

Property Analysis

High Temperature Resistance

Decomposition resistance

Adhesives

Structural adhesion

Mechanical Properties

Optical coupling

Optical stability (transmission)

Thermal Control Materials

Paints

Optical Solar Reflectors

Foils

Irradiation test

Thermo-Optical

in combination with high temp resistance

(α/ε)

Accelerated UV radiation Accelerated X-ray, γ and particle radiation Mechanical Properties (tensile, stress/strain)

Figure. 1. Approach to materials testing & analysis for inner solar system missions

173

Surface Temperature (sun facing) [°C]

500 0.31 AU

400

300

0.47 AU

200 1 AU

100

0

-100 0,0

0,2

0,4 0,6 Alpha/Epsilon ratio

0,8

1,0

Figure 2. Surface equilibrium temperature as a function of α/ε ration for 1, 0.47 and 0.31 AU.

around 0.47 AU and the third around 0.31 AU. The latter two correspond to the aphelion and the perihelion of Mercury. It is worth noting that this graph does not include any contribution of infra-red radiation or planetary albedo that will increase surface temperatures in low planetary orbits. 2.0 Experimental Approach on Assessing the Thermal Stability Various experimental thermal analysis techniques exist that can be used for this purpose. The thermal stability can be most easily measured by TGA that measures mass loss as a function of time and/or temperature. The change in mass is an indication of the thermal stability of the material. Experimentally this can be assessed in two different ways. By measuring in an isothermal mode precise data about the decomposition can be gathered. Results from such tests can be used to derive mathematical models in order to describe decomposition kinetics. The disadvantage of this method is that it is time intensive and test duration is naturally limited. The second method relies on heating the sample with a (mostly) constant heating rate. Testing a material with various heating rates offers the possibility to determine activation energies and pre-exponential factors. This enables the development of life-time models such as the one proposed in ASTM E 1641 [10]. Such models show a good correlation when single decomposition reactions occur and when higher heating rates do not activate normally covered reactions.

174 We have screened both materials by tests applying a constant temperature ramp as well as by isothermal ageing tests. The latter technique has however two limitations for being applicable to space missions. First it is generally not done under vacuum conditions. This reduces the outgassing potential. Second long term testing is naturally limited on one instrument. To overcome this we have built a so-called high temperature exposure system (HITES) that allows exposures of up to 50 samples at elevated temperatures under high vacuum conditions. To study the thermal endurance we have aged the polyimides thermally for more than 2500 hrs. Materials investigations focussed on determination of the mass loss and the optical (UV/VIS/NIR) and thermo-optical properties versus time. To cover a wide range of inner solar system missions at an extreme temperature on the one hand but to stay within a reasonably long duration for a defined service temperature on the other hand we have selected to perform thermal endurance tests at 350°C. 3.0 Experimental Results 3.1. MATERIALS UNDER INVESTIGATION The materials under investigation were supplied by Du Pont and by UBE Industries. They are commercially available under the name Kapton (K) and Upilex  (U). We have used various film thicknesses (with and without VDA (Vacuum Deposited Aluminium) coating) between 7.5 µm and 50 µm for our investigations. An overview is given in the table below. TABLE 1. Overview about materials used and investigations performed.

Material K-film 7.5µm K-film 25 µm K-film 7.5 µm/VDA K-film 50 µm/VDA U-film 7.5 µm U-film 25 µm U-film 7.5 µm/VDA U-film 25 µm/VDA

Investigation TGA, HITES, UV/VIS/NIR Transmission, TGA, HITES, UV/VIS/NIR Transmission TGA, HITES, Thermo-optical measurements TGA, HITES, Thermo-optical measurements TGA, HITES UV/VIS/NIR Transmission TGA, HITES, UV/VIS/NIR Transmission TGA, HITES, Thermo-optical measurements TGA, HITES, Thermo-optical measurements

3.2 Assessment of Thermal Stability by Thermal Analysis For the mass loss experiments we used uncoated materials. In fig 3 a comparison of two TGA test results for both materials are shown. Testing was done according to ISO 11358:1997 [11]. A heating rate of 10K/min was applied and 5.0 dry nitrogen was used

175 as a purge gas with a flow rate of 55 ml/min. As can be seen the two materials show no apparent mass loss up to 300°C. The 5 % mass loss figure, the T5, was determined to be around 597°C for the K film and 634°C for the U-film. Furthermore the first derivative of the mass loss is also shown in that graph. It uses the right axis of the graph. We found the maximum rate of decomposition for the K-film around 616°C and for the U-film at 655°C. Figure 4 shows an enlargement of the TGA recordings between 200°C and 600°C. As can be seen above 300°C both films can be discriminated. The higher the temperature the bigger the difference gets. 0,2 Mass %

100

0,1

Derv. Mass %/T

MASS [%]

-0,1 -0,2

80

-0,3 70 U-film

-0,4

K-film

-0,5

60

DERV. MASS [%/°C]

0 90

-0,6 50 0

200

400

[°C]

600

800

-0,7 1000

Figure 3. Comparison of thermal stability between K-film and U-film by TGA. 100

Mass [%]

99,5

99 U-film

K-film

98,5

98 200

300

400

T [°C]

500

Figure 4. Comparison of thermal stability between K-film and U-film by TGA.

600

176 Figure 5 shows a compilation of two TGA isotherm runs of 25µm thick films at 350°C. Tests were done again according to ISO 11358:1997 and 5.0 dry nitrogen was used as purge gas with a flow rate of 55 ml/min. Again the films can be discriminated. Even at the longest isotherms of the U-film that lasted nearly 140 hrs only a marginal mass loss was detected. We determined the mass loss to be around 0.1% for the U-film and on the other hand we found a noticeable mass loss of about 0.3 % for the K-film material. Thus we observed in the same test duration about three times higher mass losses for the K-film material compared to the U-film material. 100,0%

mass%

99,9%

U-film (10,1 mg) K-film (8,8 mg)

99,8%

99,7% 0

50

t [hours]

100

150

Figure 5. Isothermal TGA tests at 350 °C, comparing thermal stability between K-film and Ufilm.

3.3 ASSESSMENT OF THERMAL STABILITY IN THE HITES The HITES is a custom built facility that enables to expose up to 50 samples at elevated temperatures under high vacuum conditions. Samples are sliding into a metallic samples holder and temperature is measured at several locations within the facility. We have found that we were able to control the set temperature of 350°C within ± 3°C over the duration of the test program. Vacuum levels within the facility were maintained below 2x10-5 mbar and reached bottom levels of high 10-8 mbar values after long exposure durations. The experiments performed within the HITES used the same type of uncoated materials as before for the mass loss and UV/VIS/NIR transmission analysis. For the thermo-optical analysis one-sided VDA coated materials were used. The thermo-optical properties were determined on the polyimide side as the front layer and were performed according to ESA-PSS-01-709 [12]. To determine the α we used a UV/VIS/NIR spectrophotometer (Cary 5) and measured in the wavelength range of 250 nm and 2500 nm. To measure the thermal emittance we used a Gier Dunkle DB100. Measurements were performed under

177 atmospheric condition outside the test facility. This required cooling the samples down to below 25°C and flushing it with dry nitrogen before opening and recovering the test samples. The total test duration of this test was 2540 hrs. After the initial determination of the beginning of life (BOL) values and properties the test was stopped three times for intermediate inspections/measurements. The first was done after approximately 172 hrs, the second after about 872 hrs and the third after 2540 hrs. Mass loss was determined with an ultra-microbalance that is able to resolve 100 ng. This balance is placed within a continuously dry nitrogen purged cabinet. Temperature and humidity are recorded before and after measurements. As measuring mass loss at such a resolution is a very delicate operation we had special reference samples that were weighed together with our test samples at each individual break. This was done to see whether variations within the reference material could be responsible for arbitrary mass losses or gains. The latter is commonly attributed to the adsorption and absorption of water. 3.3.1. Results on mass loss measurements The results gathered with the dedicated mass loss samples were conclusive in so far that we could discriminate both film materials. We observed a noticeable mass loss on the K-film samples (above 1%) and no significant change on the U-film material after the first inspection point at 172 hrs. Comparing these results with the test results of the TGA isothermal tests at 350°C shows that for approximately same durations (140 hrs versus 172 hrs) we find an increased outgassing/decomposition contribution on the HITES samples. The amount of mass loss on the K-film did however only marginally increase with time. The U-film material seemed to be stable and the highest variation was sometimes observed on the reference samples itself. We even noted sometimes increases on the mass of the aged test samples. One explanation could be that the aged material shows increased water adsorption that cannot be avoided during transfer of the sample. This could be facilitated for instance by a change in the surface morphology. To clarify this point we intend to perform further AFM images of the surfaces of the aged materials. 3.3.2. Results on UV/VIS/NIR transmission stability during isothermal ageing The transmission of dedicated transmission samples was measured BOL and at each of the three inspection points. A compilation of recorded values for 7.5 µm U- and K-film samples are shown in Figure 6. It shall be noted that each individual graph represents averaged values of three different samples.

178 100

Transmittance [%]

80 1

2

60

3

1

K7 - BOL

U7 - BOL

2

K7 - 172 h

U7 - 172 h

3

K7 - 872 h

U7 - 872 h

4

K7 - 2540 h

U7 - 2540 h

4

40

20

0 250

750

1250

λ [nm]

1750

2250

Figure 6. Compilation of UV/VIS/NIR transmission data of 7.5 µm K and U films vs. isothermal ageing duration at 350 °C

As can be seen the K-film shows a higher transmission than the U-film at the BOL. This was visually verified by the colour difference between the two films. The Ufilm appears darker at the BOL. At the first inspection break after 172 hrs a difference in the visible range can be noted already. The K-film increases its absorptance and shows already lower transmission behaviour in the visible range. This trend continues with time as can be deduced by the measurements performed at the next two breaks. Additionally it is worth noting that the cut-off wavelength of the K-film material shifts with increasing ageing duration whereas the cut off wavelength of the U-film material is shifting less. 3.3.3. Results on thermo-optical stability during isothermal ageing The results of thermo-optical properties of various samples measured between the BOL, after 172 hrs, 872 hrs and 2540 hrs of isothermal ageing time is shown in Figures 7 and 8. The first one shows the stability of the α vs ageing time of four test samples. Again the values show averaged data of three different samples. If we look at the two directly comparable 7.5 µm thick films one can see that the initial α at the BOL were nearly identical. At the next two inspection points the α’s of both films are noticeably different. This trend seems to increases and after 2540 hrs a

179 clear distinction can be made. The other two graphs give values for a 50 µm K-film and a 25 µm U-film. Even though the other two curves are not directly comparable due to a different thickness they show a similar trend. It is worth noting that the indicated trend lines show similar slopes for the two different films and the U-films shows a lower slope. This confirms the previously made measurements. The results of the thermal emittance measurements versus time are shown in Table 2. As was expected only a marginal change in the thermal emittance can be noted. TABLE 2. Thermal emittance data versus isothermal ageing duration at 350 °C

Material

U-film 7.5 µm

U-film 25 µm

0 hrs (BOL) 172 hrs 872 hrs 2540 hrs

0.444 0.446 0.442 0.449

0.639 0.638 0.614 0.620

K-film µm 0.476 0.464 0.454 0.465

7.5

K-film 50 µm 0.778 0.743 0.719 0.732

0,8 K-film - 50 um

0,7

Solar absorptance

U-film - 25 um

K-film - 7 um

0,6

U-film - 7 um

0,5

0,4

0,3 0

500

1000 1500 ageing time [hours]

2000

2500

Figure 7. Compilation of solar absorptance of 7.5 µm K- and U- films vs. isothermal ageing duration at 350 °C

A plot of the α/ε ratio versus time is shown in Figure 8. It basically shows a similar picture as Figure 7, the U-film material tends to be more stable and degrades at a slower rate.

180 1,6

K-film - 7 um

1,4

Alpha / Epsilon

U-film - 7 um 1,2

U-film - 25 um 1,0

K-film - 50 um

0,8

0,6

0,4 0

500

1000 1500 ageing time [hours]

2000

2500

Figure 8. Compilation of α/ε ratio of K- and U- films vs. isothermal ageing duration at 350 °C

4.0 Conclusions In this paper results about the thermal stability at 350 °C of two different polyimides were presented. The thermal mass loss was assessed by various TGA measurements and it was found that the U-film material shows a higher thermal stability than the K-film. These results were confirmed by thermal endurance tests by TGA measurements. Further tests were conducted in a high vacuum facility at 350°C. Test data were gathered up to a duration of 2500 hrs. The thermal stability (mass loss) was again found to be lower for the K-film material. We also noted a significant vacuum effect on the Kfilm. Further UV/VIS/NIR transmission and thermo-optical measurements revealed the degradation behaviour of both films versus time at 350 °C. The test results gathered indicate that the U-film material tends to be more stable and that the space relevant thermo-optical properties degrade at a slower rate. 5.0 Acknowledgement The authors would like to acknowledge J. Sorenson for providing environmental data for the Mercury mission and A. Santovincenzo and H. Ritter for discussing thermal control aspects of the Mercury mission.

181 6.0. References 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

“BepiColombo, An Interdisciplinary Cornerstone Mission to the Planet Mercury”, System and Technology Study Report, ESA-SCI (2000)1, April 2000, European Space Agency, ESTEC, Noordwijk, The Netherlands Semprimoschnig C.O.A., S. Heltzel, A. Polsak, “Materials Behaviour at Mercury – Challenges and First Experimental Results”, 33rd SAMPE Conf., Seattle, USA, 2001 Yokota R., “Recent trends and space applications of polyimides”, J. of Photopolymer Sc & Techn., Vol 12, 2, 209, 1999 Traeger et al., “Thermal Aging of polyimide films”, Polym Preprints 12, 292 (1971) Ortelli et al, “Pyrolysis of Kapton in Air: An in situ DRIFT Study, Appl. Spectr. 4, Vol 55, 412, (2001) Matsumoto T., “Nonaromatic polyimides derived from cycloaliphatic monomers, Macromolecules, 32, 4933, 1999 Hasegawa et al., “Structure and properties of novel asymmetric biphenyl type polyimides. Homo and copolymers and blends, Macromolecules, 32, 387, 1999 Russell et al, ‘Simulated space environmental testing on thin films”, NASA/CR-2000-210101, April 2000 Wooldrige et al, “Effects of manufacturing and deployment on thin films for the NGTS sunshade, AIAA, 2001-1349 ASTM E 1641-99, American Society for Testing and Materials, USA, 1999 ISO 11358:1997, International Organisation for Standardisation, PO 56, CH-1211 Geneva 20, Switzerland ESA PSS-01-709, ESA/ESTEC, Noordwijk, The Netherlands, July 1984

This page intentionally left blank

SYNERGISTIC DEGRADATION OF CV-1144-O DUE TO ULTRAVIOLET RADIATION AND HEAT JOSEPH E. HAFFKE JOHN A. WOOLLAM Center for Microelectronic and Optical Materials Research, Department of Electrical Engineering, University of Nebraska-Lincoln, Lincoln, NE 68588-0511

Abstract Nu-Sil’s CV-1144-O is a silicone-based co-polymer used on spacecraft in LEO. CV1144 has been spin-coated onto platinum sputter coated silicon wafers, and simulated LEO experiments made to study degradation. First, samples were exposed to ultraviolet radiation from a deuterium lamp with no additional heating. Next, samples were heated to 100°C during which there was no radiation exposure. Finally, samples were exposed simultaneously to ultraviolet radiation and heating. Variable angle spectroscopic ellipsometry (VASE£) was used to determine optical constants from vacuum ultraviolet (VUV) to mid-infrared (IR). VUV SE data were analyzed to determine thickness and optical constant changes, while IR data provided bonding-chemistry information. Multi-sample analysis of optical data eliminated correlation when determining polymer optical constants and thickness. Heating alone did not alter optical properties. Small changes were seen when exposed to ultraviolet radiation alone. Significantly larger changes occurred when the sample was exposed to ultraviolet light while being heated. Results show that ultraviolet radiation breaks apart benzene rings in the co-polymer, a process enhanced by heating. 1.0 Introduction Low earth orbit (LEO) exposure of space systems to atomic species, charged particles, temperature fluctuations, electromagnetic radiation, and debris are considered when designing spacecraft materials capable of long-term LEO exposure [1-8]. Protective coatings can protect space system surfaces from atomic oxygen destructive effects, and have been studied for years [9,10]. Recent attention has focused on polymeric protective layers, due to ease of application. Inorganic siloxane polymers, which convert to protective silica-like coatings, yield erosion rates one to two orders of magnitude lower than carbon based polymers when exposed to atomic oxygen [2, 9, 11]. CV-1144-O is a silicone dispersion developed by NuSil Technology [12] proposed

183

184 for use as a protective barrier in LEO [10]. It is based on a dimethyl diphenyl silicone co-polymer with chemical structure shown in Figure 1.

CH 3

95.2%

Si

+

O

CH

3

4.8%

Si

O n

m Figure 1

Chemical structure of CV-1144-O silicone-based copolymer.

Qualification of materials for use in space is difficult. Although in-flight testing for candidate materials is possible, it is usually time-and cost-prohibitive therefore laboratory simulation is important to development and testing of coatings [2,13,14]. In previous work we studied the effects of atomic oxygen on CV-1144-O [2,15]. In this study, the affects of vacuum ultraviolet radiation and thermal cycling on CV1144-O co-polymer were investigated, both independently and simultaneously. The optical properties of the copolymer were examined using variable angle spectroscopic ellipsometry (VASE) from vacuum ultraviolet (V-UV) to mid-infrared (IR). 2.0 Experimental 2.1 MATERIAL PREPARATION First optically thick platinum films were prepared on polished silicon wafers by magnetron sputtering. Thin films of CV-1144-O co-polymer were then spin-coated onto the smooth platinum substrates. CV-1144-O was thinned with Naphtha in different dilutions to lower the polymer viscosity for spin-coating. By changing the revolutions per minute during spin-coating and the ratio of polymer to solvent, uniformly distributed samples of unique thickness were obtained. After spin coating, and during drying the Naphtha evaporated [10]. Five nominally identical samples were prepared, labeled A, B, C, D, and E. 2.2 EXPOSURE EXPERIMENTS Thin films of CV-1144-O were subjected to heating and/or ultraviolet (UV) radiation in a simulation chamber capable of a base pressure near 1x10-6 torr. Samples were heated by placing them on a copper plate which was on top of a Kapton heating element with thermocouple-based temperature control. UV exposure was from an RF powered deuterium lamp, mounted in the vacuum chamber directly above the substrate. Three experiments were performed. First, the polymer was heated to 100°C with no ultraviolet radiation. Such heating might come from spacecraft equipment, the sun,

185 or other sources. In the second experiment a sample was exposed to vacuum ultraviolet radiation with no additional heating, as UV light is a strong component of sun radiation in LEO. The third experiment used a combination of heat and UV, as this better represents the actual space system in LEO environment. These experiments allowed study of cooperative effects of the two sources, UV light and heating. 2.3 ELLIPSOMETRY Variable angle spectroscopic ellipsometry data were acquired prior to and after each exposure condition. In ellipsometry the change of the polarization state of light on reflection from the surface is measured, and results displayed as psi (ψ) and delta (∆), which are related to the Fresnel reflection coefficients [16]. Optical constants and film thicknesses were determined from sets of measured ψ and ∆ values using regressionanalysis. The mean square error of the fit between experimental Ψand ∆ values compared to those calculated from Fresnel equations is minimized using the LevenbergMarquardt algorithm and varying the model parameters [17]. The optical parameters were then checked for correlation to make certain solutions were indeed unique and representative of each sample. The optical constants are expressed as ε1 and ε2, the real and imaginary components of the complex optical dielectric function of the material [16]. Spectroscopic data were taken from 0.75 to 9 eV, and 300 to 6000 cm-1, using vacuum-ultraviolet (VUV) and infrared (IR) ellipsometers, respectively. Ellipsometric data were taken every twenty hours, for each experiment, to establish trends in the optical constants or thickness. Data collected from the VUV and visible ranges helped determine thickness and optical constant changes, while IR data provided information on chemistry, and chemistry-related modifications of exposed samples. Multi-sample analysis was performed to minimize parameter correlation. In this procedure, five samples of varying thickness had variable optical model parameters coupled together assuming a Cauchy dispersion model in the regression. The Cauchy parametric model was used in the non-absorbing (300—700 nm) region of the spectrum This method forced the regression-determined optical parameters of all samples to be the same, and permitted an independent fit for film thickness of each sample along with the Cauchy parametric model parameters for the optical constants and dispersion representing all samples. Parameter coupling for the unique samples during optical model fit successfully eliminated correlation. Results are found in Table 1. 3.0 Results and Discussion 3.1 CV-1144-O—PRISTINE REFERENCE OPTICAL PROPERTIES Multi-oscillator models were developed to determine optical constants of the pristine polymer in both the VUV and IR regions where the polymer is not totally transparent. Both models contained mixtures of Gaussian and Lorentzian oscillators, whose dispersion parameters were used in regression fits of experimental data. In the VUV, a total of 4 oscillators provided exceptionally good fits to data. In the IR, 12 oscillators were used, and yielded excellent data fits. Oscillator parameters are given in Table 1.

186 TABLE 1. IR oscillators: Pristine, and after UV exposure with simultaneous heating to 100°C CENTER ENERGY

AMPLITUDE

BROADENING

Pristine

After 60hrs

Pristine

After 60hrs

Pristine

699.74

0

0.35

0

33.09

799.05

801.44

3.2

1.2

19.13

817.08

0

0.05

0

15.31

844.49

852.9

0.11

0.23

9.85

867.17

0

0.22

0

26.45

1019.4

1025.5

1.52

0.93

24.18

1047.2

0

0.85

0

70.45

1093.6

1083.8

1.09

0.98

1125.2

1153.9

0.17

1260.2

1261.9

1415.8 2962.3

Chemical ID

After 60hrs

35.82

C6H5 Methyl rocking & Si-C stretching Unidentified

27.77

Unidentified Unidentified

36.22

Si-O Stretch

42.1

97.88

0.06

13.2

63

1.31

0.58

9.9

15.49

1423.7

0.05

0.02

55.52

112.95

2961.7

0.11

0.05

36.52

63.93

Si-O Stretch Si-C6H5 Planar ring vibration Si-CH3 Symetric deformation Asymmetric CH3 deformation CH3

Si-O-Si

IR spectra oscillators were analyzed to determine chemical bonding and functionality [18,19]. Results are shown in Figure 2. Comparing oscillator locations in photon energy and relative strengths for all samples conclusions are made about chemical changes resulting from experiments performed.

4.0 CH3 rocking & Si—C stretching

3.0 ε2

Si—O stretch

2.0 1.0

Si—CH3 symmetric deformation

C6H5

CH3 asymmetric deformation CH3

0.0 0

1000 2000 3000 -1 Wave Number (cm ) Figure 2 Chemical identity of IR dielectric function peaks.

4000

187 3.2 CV-1144-O—AFTER HEATING TO 100°C Polymer sample A was heated to 100°C for 60 hours to determine changes in optical properties under thermal influence. Allowing the oscillator parameters from the pristine sample to vary in a regression fit with the new experimental data taken after 40 hours of heating, resulted in a new set of optical constants, ε1 and ε2. The absence of a wavelength shift of experimental data in the transparent spectral region indicates thicknesses were unchanged, and only small changes in the optical constants. In the same manner, the IR optical constants of the pristine sample were compared to those found after 60 hours of heating. ε1 and ε2 IR spectra before and after heating to 100°C are nearly identical. This evidence suggests that sample A is not chemically altered as result of heating alone. 3.3 CV-1144-O—IRRADIATING THE POLYMER WITH ULTRAVIOLET LIGHT Polymer sample B was exposed to ultraviolet radiation for 60 hours to observe possible optical property changes. New VUV optical constants were found by regression fit to experimental data taken after 20 hours of exposure. Significant changes in the optical constants were found, especially at photon energies higher than 5 eV. IR optical constants from the pristine reference sample were compared to the IR optical constants after 60 hours exposure. Close examination of Table 2 reveals a general decrease in optical constant amplitudes after exposure. Thus CV-1144-O is significantly affected by 20 hours exposure to ultraviolet radiation. TABLE 2. IR oscillators after UV exposure only CENTER ENERGY

AMPLITUDE

BROADENING

Pristine

After 60hrs

Pristine

After 60hrs

Pristine

699.74

695.57

0.35

0.23

33.09

After 60hrs 64.82

799.05

799.62

3.2

2.51

19.13

23.46

817.08

817.78

0.05

0.27

15.31

12.73

844.49

845.7

0.11

0.13

9.85

13.39

867.17

864.69

0.22

0.21

26.45

32.61

1019.4

1018.8

1.52

1.43

24.18

27.95

1047.2

1042.2

0.85

0.75

70.45

57.64

1093.6

1091.2

1.09

0.78

42.1

43

1125.2

1097.7

0.17

0.39

13.2

76.34

1260.2

1260.5

1.31

1.12

9.9

10.94

1415.8

1476.5

0.05

0.03

55.52

272.15

2962.3

2961.8

0.11

0.13

36.52

28.48

188 3.4 CV-1144-O—HEATING THE POLYMER TO 100°C WHILE EXPOSING IT TO ULTRAVIOLET RADIATION Polymer sample C was heated to 100°C during simultaneous exposed to UV radiation for 60 hours, to best simulate LEO conditions. Visible region experimental VUV data shifts results from a significant thickness reduction. Thickness after 20 hours exposure was determined by fitting experimental data in the non-absorbing spectral region (300— 700 nm) using a Cauchy dispersion model, similar to the method used to find thicknesses of the pristine references. Once the new thickness was determined, it was

3.2

(a)

3.0

After 20 hours of UV radiation and heat Pristine reference

ε1

2.8 2.6 2.4 2.2 2.0

0

1.2 1.0

2 (b)

6 4 Photon Energy (eV)

8

10

After 20 hours of UV radiation and heat Pristine reference

ε2

08 0.6 0.4 0.2 0.0 0

2

4 6 Photon Energy (eV)

8

10

Figure 3 Comparison of optical constants of sample C in the VUV-UV-Visible-NIR range before and after 20hr of UV radiation and heat. Dielectric function (a) ε1. (b) ε2.

189 used in a regression fit of the data. The optical constants, ε1 and ε2, obtained from this fit are compared to data from the pristine reference, and the results shown in figure 3. It is interesting that two additional oscillators were needed between 1 and 5 eV, to fit the experimental data in the VUV region. The presence of oscillator-like spectra indicate photon-induced transitions between new energy level sets generated by combined heating and UV exposure. The origins f these new spectra is unknown to us. The IR data were analyzed using a regression fit to the data taken after 60 hours

4.0

(a)

ε1

3.0 2.0 1.0 0.0

After 60 hours of UV radiation and heat Pristine reference

0

4.0

ε2

3.0

1000 2000 3000 -1 Wave Number (cm ) (b)

4000

After 60 hours of UV radiation and heat Pristine reference

2.0 1.0 0.0 0

1000 2000 3000 -1 Wave Number (cm )

4000

Figure 4 Comparison of optical constants of sample C in the middle IR range before and after 60hr of UV radiation and heat. Dielectric function (a) ε1. (b) ε2.

190 exposure. The resulting ε1 and ε2 optical constants compared to those from the pristine reference, and results are shown in figure 4. Examination of IR data in table 1 reveals significant changes in peak amplitude and breadth, demonstrating major chemical changes. Most notable is disappearance of the mono-substituted benzene ring resonant signature seen in the pristine IR data at 699cm-1. A general trend of decreasing amplitudes is also observed, with the methyl associated peaks and Si—O stretching resonance peaks being the most affected. 4.0 Conclusions Heating the polymer to 100°C for 40 hours produces only very small changes in the VUV optical constants. Exposure of sample B to ultraviolet light without external heating produced significant change in chemical and optical properties of CV-1144-O. Although the visible region data exhibited no optical constant changes, substantial changes in the “B” material optical properties in the high photon energy range were seen after exposure. Decreasing peak heights seen in the IR ε2 spectra confirm that CV1144-O was chemically changed as result to 60 hours of exposure to ultraviolet radiation. The specific chemical bonds that were changed are listed in table 2. Most interesting, was extensive alteration of optical and chemical properties resulting from combined heating and UV irradiation of Sample C, where both optical properties and thickness were changed. In the VUV, a new model was needed to describe the material after 20 hours exposure due to large differences in experimental data. When compared to pristine film data, the new model required two additional oscillators between 2 and 5 eV to allow a good fit to data, indicating that two new sets of energy levels were created. Oscillator features in the UV are much more difficult to associate with specific chemistries as compared with IR resonance, thus the chemical origins of the new UV peaks are presently unknown to us. From the IR data summarized in table 2, the disappearance of several oscillators in the data shows that benzene rings in the CV-1144-O co-polymer are broken by UV radiation. It’s clear that exposing the sample to UV light and heating results in synergistic effects on material thickness and optical constants, not seen with heating or UV exposure alone. Chemical bond identifications are given in table 2 and show that benzene ring resonant spectra go away completely and others are drastically reduced in strength. Our explanation then is that UV radiation breaks apart benzene rings of the co-polymer, and heating enhances this process. 5.0 Acknowledgements This work was supported by the NASA Glenn Research Center, Grant NAG3-2219.

191 6.0 References 1. 2.

Tennyson, R. C., Can. J. Phys. (1991). 69, 1190 Yan, L., Bungay, C. and Woollam, J. A. “Surface chemistry changes and erosion rates for CV-1144-0 silicone under oxygen plasma and ultraviolet light exposure”, (2000) 5th international conference on “Protection of Materials and Structures from the LEO Space Environment”, European Space Agency. 3. Srinivasan, V., Banks, B.A., (1990) M,aterials Degradation in Low Earth Orbit (LEO), The minerals, Metals & Materials Society, Pennsylvania. 4. Kleiman, J. I., Tennyson, R. C., (1999) Protection of Materials and Structures from the Low Earth Orbit Space Environment, Space Technology Proceedings, Kluwer Academic Publishers, the Netherlands. 5. Zimcik, D. G. and Maag, C. R., (1998) J. Spacecr. Rockets 25, 162-168. 6. Banks, B. A., de Groh, K. L., Baney-Barton, E., Sechkar, E.. A., Hunt, P.K., Willoughby, A., Bemer, M., Hope, S., Koo, J., Kaminski, C. and Youngstrom, E. (1999) NASA Technical Memorandum 209180. 7. Dever, J. A., (1991), NASA Technical Memorandum 103711. 8. Banks, B. A., de Groh, K. K., Rutledge, S. K., and Difilippo, F. J.(1996) NASA Techinical Memorandum 107209. 9. Zimcik, D. G., Wertheimer, M. R., Balmain, K. B., and Tennyson, R. C. (1991) J. Spacecr. Rockets 28, 652-657. 10. Bungay, C. L., Tiwald, T. E., Thompson, D. W., DeVries, M. J., Woollam, J. A., Elman, J. F. (1998), “IR ellipsometry studies of polymers and oxygen plasma-treated polymers”, Thin Solid Films 313-314, 714. 11. Gilman, J. W., Schlitzer, D. S., and Lichtenhan, J. D. (1996), .J. Appl. Polym. Sci. 60, 591-596 12. NuSil Technology, 1050 Cindy Lane, Carpinteria, CA 93013, 805/684-8780. 13. Bungay, C. L., Synowicki, R. Spady, B., Hale, J. S., Woollam, J. A. (1999) “Laboratory Simulation of low earth orbit”, Protection of Materials and Structures from the Low Earth Orbit Space Environment, edited by J. I. Kleiman and R. C. Tennyson, Kluwer Academic Publishers, Norwell, MA. 14. Rutledge, S. K. and Banks, B. A., (1996), “A technique for synergistic atomic oxygen and vacuum ultraviolet radiation durability evaluation of materials for use in LEO”, NASA Technical Memorandum 107230. 15. Yan, L., Gao, X., Bungay, C., and Woollam J. A. (2001) “Study of surface chemical changes and erosion rates for CV-1144-O silicone under electron cyclotron resonance oxygen plasma exposure”, J. Vac. Sci. Technology A 19(2), 447. 16. Azzam, R. M. A. and Bashara, N. M., (1977) Ellipsometry and Polarized Light (North Holland, New York. 17. Jellison, Jr., G. E.,(1991) Appl. Opt. 30, 3354. 18. Colthup, N. B., Daly, L. H., and Wiberley, S. E., (1990) Introduction to Infrared and Raman Spectroscopy, Academic Press, San Diego. 19. The Sadtler Handbook of Infrared Spectra, (1978), edited by W. W. Simons, (Sadtler Research Laboratories, Inc, Philadelphia.

This page intentionally left blank

BEHAVIOUR OF THERMAL CONTROL COATINGS UNDER ATOMIC OXYGEN AND ULTRAVIOLET RADIATION S. REMAURY, J.C. GUILLAUMON and P. NABARRA Centre National d'Etudes Spatiales (CNES) 18 avenue Edouard Belin 31401 Toulouse Cedex 04 FRANCE

Abstract The Thermal Control Department of CNES (French Space Agency) Toulouse Space Centre has developed many materials and thermal control coatings for satellites and launchers. Some of them have been specially developed to be used in low Earth orbit (LEO). These coatings are resistant to the conditions encountered at this orbit: vacuum, atomic oxygen, thermal cycling, and ultraviolet radiation. These coatings are: electrically conductive or non - conductive black and white paints; protective coating (MAPATOX K) for organic materials which are sensitive to atomic oxygen; new high performance low cost cold coatings named flexible solar reflectors (FSR) which could replace the second surface mirrors (SSM) in LEO. In this paper we present the results of the behaviour of these thermal control coatings under simulated LEO environment. Firstly, we describe the behaviour of black and white silicone and polyurethane paints under atomic oxygen (AO), and secondly, the behaviour of the new white silicone paints, of the MAPATOX K and of the FSR under AO and ultraviolet (UV) combined exposure. In the first part, it has been shown that the polyurethane paints are not stable towards AO but the silicone paints are stable. The new white silicone paints, SG121 FD (electrically non-conductive) and PCBE (electrically conductive) developed to replace SG120 FD and PCBZ, have a good stability towards AO. The new polysiloxane coating named MAPATOX K is a good solution to protect Kapton against AO, The new polysiloxane coating named FSR could replace the SSM for LEO applications. In the second part, the AO and UV combined effect is studied on the new CNES developed thermal control coatings: SG121 FD, PCBE, FSR and MAPATOXK. It has been shown that SG121 FD, PCBE and FSR are more sensitive to AO than to UV, MAPATOX K is not sensitive to AO and UV. The maximal degradation of the solar absorption factor is associated with the AO bombardment prior to UV irradiation. 1. 0 Introduction The Thermal Control Department of the CNES Toulouse Space Centre has been working for several years on the development of thermal control coatings for satellites and launchers. 193

194

For the thermal control of the low Earth orbit (LEO) satellites, we have developed different coatings particularly resistant to the conditions encountered at this orbit (vacuum, atomic oxygen, temperature, ultraviolet, etc.). These coatings are: - electrically conductive or non-conductive black and white paints - protective coatings on materials which are sensitive to atomic oxygen (MAPATOX K) - new low cost cold coatings named flexible solar reflectors (FSR) which could replace the second surface mirrors (SSM) in LEO. In this paper, we present the behaviour of thermal control coatings under simulated LEO environment: - firstly, we describe the behaviour of black and white silicone and polyurethane paints under atomic oxygen (AO), - secondly, the behaviour of the new white silicone paints, of the MAPATOX K and of the FSR under AO and ultraviolet (UV) combined exposure is described. 2. 0 Atomic Oxygen Effects In LEO, the surface of satellites materials are exposed to the oxygen atoms impact with a velocity of 8 km/s. The materials can be more or less oxidised during the mission and for some of them the resulting erosion can lead to the changes of the thermo-optical properties and the performances. The stability of thermal control coatings is thus a significant problem, particularly the stability of paints. The tests have been performed in the CASOAR chamber (Figure1) belonging to the Department of Space Environment (DESP) of ONERA in France. The fluence of the AO has been 2.1020 atoms/cm2 simulating the AO dose received by a satellite at 800 kilometres during 4 years. The tested materials have been black and white silicon paints and black polyurethane paints. 2.1. ATOMIC OXYGEN SOURCE The AO source is a pulsed source that is using a laser detonation technology. This laser detonation source was originally developed by Physical Science Inc. (PSI), and sources based on this design are now used in several laboratories. The source is capable of producing beams containing AO with average velocity that could be adjusted between 5 and 13 km/s, in practice it is maintained at 8 km/s. The amount of fluence received by exposed samples is estimated by exposing samples of a material with known reactivity (Kapton) and measuring the resulting erosion. 2.2. MATERIALS The tested thermal control paints and coatings are described in Table 1. The SG121 FD and the PCBE are the new developed paints which respectively replace the SG120 FD and the PCBZ paints. This replacement has been necessary because the commercial silicone resins are no longer available and because the CNES laboratory perfected a pigment treatment allowing a best stabilisation from UV and particles irradiation.

195

The FSR is a new low cost cold coating named Flexible Solar Reflector to replace the Second Surface Mirrors (SSM). It is a thin coating of polysiloxane (50 µm) with a low density (around 1), which is inexpensive, easy to use and able to be applied on metal deposits (such as polished aluminium or silver). The MAPATOX K is a new atomic oxygen resistant coating. It is a very thin film of polysiloxane (< 10 µm) which is applied on Kapton to avoid the debris production. The initial thermo-optical properties of the samples are given in Table 2. Figure 1. The CASOAR chamber

2.3. TESTS The paints were applied on 19×19 mm aluminium plates. The FSR was applied on the aluminium size of an aluminised Kapton (50 µm) and the MAPATOX K on Kapton (25 µm) and both were stuck on aluminium plates with transfer adhesive. The plates were placed on an aluminium sample holder and put on primary vacuum at 110°C during 24 hours, then hold at room temperature and put at 55 cm (for S2, PGN 7991, PGN AS, PU1, PUC, SG11 FD, SG120 FD and PCBZ) or at 40 cm (for SG 121 FD, PCBE, FSR, SSM and MAPATOX K) from the source. The fluence received by all exposed samples was 2.1020 atoms/cm2. 2.4. MEASUREMENTS The following measurements were carried out : - the mass of each sample (measurement accuracy ± 20 µg) - the spectral reflectance between 250 and 2500 nm using an integrating sphere with central sample put on a Varian Cary 2300 spectrometer. From this spectral value the solar reflectance Rs is calculated (measurement accuracy ± 1%). From Rs it can be deduced that the hemispherical solar absorption factor is: αs = 1-Rs.

196

- the total normal infra-red emissivity factor with a portable reflectometer (Gier Dunkle DB100). TABLE 1. Definition of the tested materials

Coatings S2 PGN 7991 PGN AS PU1

PUC SG11 FD SG120 FD PCBZ

SG121 FD PCBE FSR

SSM MAPATOX K

Characteristics Black electrically conductive Black electrically nonconductive Black anti-static Black electrically conductive Black electrically conductive White electrically conductive White electrically nonconductive White electrically conductive White electrically conductive White electrically conductive Transparent electrically non-conductive Transparent electrically non-conductive Transparent electrically non-conductive

Pigment Carbon black Carbon black Carbon black Carbon black Carbon black Zn2TiO4

Binder Commercial silicone resin Commercial silicone resin Commercial silicone resin Commercial polyurethane resin

Commercial polyurethane resin Commercial silicone resin

ZnO

Commercial silicone resin

Zn2SnO4

Commercial silicone resin

Treated ZnO

CNES synthesised silicone resin

Treated ZnO

CNES synthesised silicone resin

None

CNES synthesised silicone resin

None

Sheldahl G401900 FEP 125 µm silver-inconel and adhesive CNES synthesised silicone resin

None

TABLE 2. Initial thermo-optical properties of the tested materials

Coatings

αs ± 0.01

S2 PGN 7991 PGN AS PU1 PUC SG11 FD SG120 FD PCBZ SG121 FD PCBE FSR SSM MAPATOX K

0.96 0.96 0.96 0.94 0.93 0.15 0.19 0.15 0.21 0.20 0.12 0.08 0.36

εIR ± 0.03 0.89 0.88 0.88 0.89 0.77 0.85 0.85 0.90 0.89 0.87 0.84 0.81 0.80

197

2.5. RESULTS The erosion (deduced from mass measurement and density), the AO reaction efficiency, the variation of the solar absorption factor and the variation of the infrared emissivity factor are given in Table 3. For S2, PGN 7991, PGN AS, PU1, PUC, SG11 FD, SG120 FD and PCBZ, the results are given in details in [1]. For SG121 FD, PCBE and MAPATOX K, the results are described in [2], and for FSR and SSM in [3]. TABLE 3. Results after the AO test

Paints

Erosion (µm)

S2 PGN 7991 PGN AS PU1 PUC SG11 FD SG120 FD PCBZ SG121 FD PCBE FSR SSM MAPATOX K

-0.09 +0.10 +0.08 -1.22 -1.97 -0.14 -0.24 +0.01 +0.07 +0.08 -0.14 -2.39 -0.20

AO reaction coefficient, 10-24 cm3/atom 0.05 -0.05 -0.04 0.58 0.94 0.07 0.11 -0.01 -0.04 -0.04 0.08 1.40 0.12

∆αs

∆ε

-0.01 -0.01 -0.01 0.00 +0.02 +0.04 +0.01 -0.01 +0.01 +0.01 +0.04 +0.02 0.00

-0.01 0.00 0.00 +0.02 +0.08 0.00 0.00 +0.01 0.00 -0.01 0.00 0.00 -0.01

2.6. DISCUSSION Many comments can be deduced from Table 3. First of all, the polyurethane binder (C-C bond energy = 306.7 kJ/mol) is degraded much more by AO than the silicone one (Si-O bond energy = 797.5 kJ/mol). The attack of polyurethane bonds leads to the scission of the polymer chains, and then to the erosion of paints. Therefore surface changes degrade the emissivity factor. The commercial silicone resin used in the SG11 FD and SG120 FD paints is the same, and it appears that it is more sensitive to AO erosion than the CNES synthesised silicone resin used in the SG121 FD. The ZnO pigment (SG120 FD) and the treated ZnO pigment (SG121 FD and PCBE) are much more stable than the Zn2TiO4 (SG11 FD). The AO reaction coefficients and the variations of the solar absorption factor of PCBZ and PCBE are practically the same, so these two paints will be considered as equivalent. The erosion is practically the same for the FSR and the MAPATOX K. The solar absorption factor depends on the coating thickness then the more the thickness is important, the more the degradation is visible. Moreover, vacuum UV (< 200 nm) are produced by the AO source. The coatings are degraded in depth by these VUV, therefore the degradation is much more important for the FSR than for the MAPATOX K. The thermo-optical properties of the FSR are more degraded than the SSM ones. But the SSM is much more sensitive to erosion than the FSR.

198

3.0 Atomic Oxygen and Ultraviolet Combined Effects In LEO, the thermal control materials on satellites surface are exposed not only to AO impact but also to the solar flux, except for heliosynchronous orbits where particles (electrons and protons) are also present. A satellite or a space station attitude change leads to: - Direct and non-direct UV exposure of materials - AO exposure of materials, most critical when their direction is perpendicular to the AO velocity vector - UV and AO exposure Then materials degradation can be caused by expose to both those environmental factors, and also to cause successive degradation. Therefore it is interesting to study the combined effects of AO and UV on thermal control coatings. 3.1. ATOMIC OXYGEN AND ULTRAVIOLET SOURCES The AO testing was carried out in the CASOAR chamber (see 2.1). Irradiation tests by UV were conducted in the SEMIRAMIS chamber belonging to the DESP of ONERA (Figure 2). The test consisted of irradiating the coating samples with a lamp reproducing as faithfully as possible the solar spectrum without atmosphere in the UV range (200 to 300 nm band and 300 to 400 nm band). The UV generation was used with an acceleration factor between 3.1 and 3.4. The samples were positioned on a sample holder, held at 40°C throughout the test and placed in a vacuum chamber with a vacuum level of around 5.10-7 Torr. A new development of a transport and transfer system under vacuum (VESTA) between SEMIRAMIS and CASOAR allows us to perform the combined tests. Figure 2. The SEMIRAMIS chamber

199

3.2. MATERIALS The tested thermal control coatings are described in Table 4 and their initial thermooptical properties are given in Table 5. TABLE 4. Definition of the tested materials Coatings SG121 FD PCBE FSR MAPATOX K

Characteristics White non electric conductive White electric conductive Transparent non electric conductive Transparent non electric conductive

Pigment Treated ZnO Treated ZnO None None

Binder CNES synthesised silicon resin CNES synthesised silicon resin CNES synthesised silicon resin CNES synthesised silicon resin

TABLE 5. Initial thermo-optical properties of the tested materials Coatings SG121 FD PCBE FSR MAPATOX K

αs ± 0.01 εIR ± 0.03 0.21 0.20 0.12 0.36

0.89 0.87 0.84 0.80

3.3. TESTS The paints were applied on 19×19 mm aluminium plates. The FSR was applied on aluminised Kapton (50 µm) and the MAPATOX K on Kapton (25 µm) and both were stuck on aluminium plates with transfer adhesive. Ten samples were divided into two sets of samples. Three steps were carried out: step one, when the first set of samples was irradiated by UV during 500 equivalent solar hours (ESH) in SEMIRAMIS, and the second set was screened from irradiation; step 2, when the two sets were transferred with the VESTA system under vacuum to CASOAR and exposed to 2.1020 at/cm2, and step 3, when the two sets were transferred with the VESTA system to SEMIRAMIS and exposed to UV irradiation during 500 ESH. 3.4. MEASUREMENTS Before and after the step 1 and after the step 3 the spectral reflectance was measured in SEMIRAMIS under vacuum between 250 and 2500 nm using a Perkin Elmer Lambda 19 spectrometer with a lateral sample integrating sphere. The value of the solar reflectance Rs was calculated from the reflectance values in 250-2500 nm area of each reflection spectrum. The solar absorption factor is obtained by the formula: αs = 1-Rs with a relative error of ∆αs⁄αs = ± 1%. Before the step 1 and after the step 3 (beginning and end of the test) the total normal infra-red emissivity factor was measured with a portable reflectometer (DB100 Gier Dunkle).

200

3.5. RESULTS The results are described in [4]. The variation of the solar absorption factor is given in Table 6 and the variation of the infra-red emissivity before and after the test in air is given in Table 7. 3.6. DISCUSSION The thermal control coatings used in this study are more sensitive to AO than to UV irradiation (except the MAPATOX K). The maximal degradation of the solar absorption factor seems to be associated with AO bombardment prior to UV irradiation. One hypothesis could be imagine: the surface of the silicone resin is previously transformed to a defective SiOx form by AO; this structure could lead to the formation of colour centres under UV irradiation and therefore to the degradation of the thermo-optical properties. TABLE 6. ∆αs results after the AO and UV combined test Coatings

SG121 FD PCBE FSR MAPATOX K

∆αs

∆αs

after 500 esh

after 500 esh + AO 2.1020 at/cm2

+0.01 +0.01 0.00 +0.01

+0.01 +0.01 +0.03 0.00

∆αs

after 500 esh + AO 2.1020 at/cm2 + UV 500 esh +0.02 +0.02 +0.05 +0.02

∆αs

after AO 2.1020 at/cm2 + UV 500 esh +0.03 +0.02 +0.05 +0.01

TABLE 7. ∆ε results after the AO and UV combined test Coatings

SG121 FD PCBE FSR MAPATOX K

∆ε

after 500 esh + AO 2.1020 at/cm2 + UV 500 esh 0.00 0.00 -0.01 0.00

∆ε

after AO 2.1020 at/cm2 + UV 500 esh 0.00 0.00 -0.01 0.00

4.0 Conclusions In this paper, the AO effect and the AO and UV combined effect have been studied on thermal control coatings. The AO effect is tested on black and white silicone paints, black polyurethane paints and silicone varnishes. It has been shown that: - The polyurethane paints are not stable towards AO, on the contrary the silicone coatings have a good behaviour. - The new white silicone paints SG121 FD (electrically nonconductive) and PCBE (electrically conductive), developed to replace SG120 FD and PCBZ, have a good resistance towards AO. - A new polysiloxane coating named MAPATOX K is a good solution to protect Kapton against AO.

201

-

A new polysiloxane coating named FSR could replace the SSM for LEO applications. The AO and UV combined effect is studied on the new CNES developed thermal control coatings, SG121 FD, PCBE, FSR and MAPATOX K. It has been shown, that : - SG121 FD, PCBE and FSR are more sensitive to AO than to UV irradiation; the MAPATOX K is not sensitive to AO and UV. - The maximal degradation of the solar absorption factor is associated to an AO bombardment prior to UV irradiation. This test was important to perform in finding out which environmental component is responsible for the degradation of the coatings. It is the AO bombardment to consider first, but the AO source produces VUV which could also lead to the degradation of the coatings. A modification of the source for separating the AO and the VUV radiation could allow to find and distinguish which element is responsible for the degradation. 5.0 Acknowledgements: I thank Virginie Viel, Joseph Marco, and C. Pons from ONERA/DESP for their participation in this work.

6.0 References: 1. Paillous A., Oscar H., Riboulet M., and Siffre J. (1992) Bombardement par l'oxygène atomique de matériaux de régulation thermique, CERT/ONERA 437812 Report. 2. Viel V. and Siffre J. (1999) Tenue à l'oxygène atomique de revêtements de contrôle thermique, ONERA/DESP RF/472100 Report. 3. Viel V. and Chardon J.P. (2001) Essai de bombardement par l'oxygène atomique, ONERA/DESP RTS 2/06281 Report. 4. Marco J., Viel V., and Pons C. (2000) Tenue sous oxygène atomique et UV de revêtements de contrôle thermique, ONERA/DESPRF/CS0418801 Report.

This page intentionally left blank

GROUND TESTING OF SCK5 WHITE SILICONE PAINT FOR LEO APPLICATIONS I. GOUZMAN, E. GROSSMAN, G. LEMPERT, Y. NOTER AND Y. LIFSHITZ Space Environment Division, Soreq NRC Yavne 81800, Israel V. VIEL-INGUIMBERT AND M. DINGUIRARD DESP, CERT ONERA Toulouse, France F-31055

Abstract SCK5 is a white antistatic silicone paint developed by CNES and manufactured by MAP (France). The present work summarizes durability tests of this paint in a simulated low earth orbit (LEO) atomic oxygen (ATOX) environment. The paint was applied on various substrates including Kapton film, Duroid 5880 (a glass/Teflon PTFE composite) and TMM3 (a ceramic/thermoset polymer). Two types of ATOX simulation systems were used: an RF oxygen plasma, and a laser detonation source (manufactured by PSI) producing a 5 eV ATOX beam. Both types of simulation facilities generate VUV radiation in addition to oxygen species. Dedicated experiments were performed to distinguish between VUV and ATOX effects. The SCK5 coated samples were also exposed to RF argon plasma, in order to separate between chemical effects of atomic oxygen and physical effects introduced by the RF plasma. The effects of ATOX exposure were studied by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS). A comparative study of the erosion yield, the surface morphology and the chemical composition resulting from exposure to equivalent ATOX fluences in both types of simulation systems was performed. The SCK5 exposed to RF plasma showed significant cracking, partial delamination and enhanced embrittlement even for low ATOX fluence, equivalent to 2x1019 atoms/cm2. Similar exposures to the 5 eV ATOX (PSI source) exhibited no cracking. In both cases the exposed samples showed a decrease of the carbon atomic concentration and an increase of the oxygen concentration in the upper surface layer, indicating the formation of a silicon oxide skin, which was more significant for the samples exposed to the RF plasma asher. It may be concluded that the erosion of SCK5 by the RF oxygen plasma is considerably more severe than by the 5 eV ATOX, at least for the specific case of porous coating of siliconic material, tested in the present work.

203

204 This is most probably associated with a combination of factors, including the nature of the reactive species in the plasma asher, their omnidirectional flux and the high porosity of SCK5 coating, leading to a strong compressive stresses and consequently cracking of the brittle silicon oxide skin. 1.0 Introduction The SCK5 white antistatic silicone paint was developed by CNES and is manufactured by MAP (France) as a thermal control coating for high frequency circuit materials [1]. SCK5 has a very short shelf life (24h), implying that the application should be done at the paint manufacturing site. Technically this may be achieved either by painting the surfaces at the plant or by purchasing a ready-made SCK5 painted Kapton film and adhering it to the surface of interest in the customer’s facility. The present work involved the ground testing of atomic oxygen (ATOX) durability of different SCK5 painted surfaces, including a Kapton film, Duroid 5880 (a glass/Teflon PTFE composite) and TMM3 (a ceramic thermoset polymer composite). The effects of ATOX, the most prominent hazard for polymeric materials in LEO, have been widely investigated in recent years by in-flight experiments as well as by laboratory simulation techniques [2-7]. The difficulty of the ground simulation experiments to forecast the long-term LEO durability of external spacecraft materials stems from basic differences between the actual environment and the simulated one. These include: (i) differences in energy distribution and directionality of reactive species; (ii) low ATOX flux in space as compared to accelerated tests, (iii) neutral ground state ATOX in space as compared to ionic and excited species in some of the experiments; (iv) synergistic effects of ATOX, ultraviolet (UV) and ionizing radiation in space. Two complementary types of ground simulation techniques were used for simulation of the ATOX LEO environment in the present work. The first type included two RF plasma systems located at Soreq NRC. The RF plasma asher is a widely accepted simulation facility for screening tests in many laboratories [2,3]. It contains a mixture of reactive species (excited, neutral and ionized oxygen atoms and molecules) that impinge upon the exposed surface omni-directionally, as well as vacuum UV (VUV) radiation. The source is capable of simulating high LEO equivalent ATOX fluences within reasonable experimental times. The second system was a laser detonation source (CASOAR) located in CERT ONERA, Toulouse. This source generates a highly directional, pulsed 5 eV ATOX beam accompanied by high doses of VUV radiation. The instantaneous ATOX flux during the pulse is 4 orders of magnitude larger than in the LEO environment, but the average value is similar to the LEO ATOX flux (~1x1015 atoms/cm2sec) [4,5]. A significant discrepancy between the results of the RF plasma and the laser detonation systems was observed in the present work, initiating detailed studies and reevaluation of both techniques.

205 2.0 Experimental Details The following samples were studied in the present work: (i) Kapton coated with SCK5, further referred to as Kapton/SCK5; (ii) Kapton coated with SCK5 and bonded by a pressure sensitive adhesive to a Duroid 5880 (a glass/Teflon PTFE composite manufactured by Rogers Corp.) substrate, further referred to as Duroid/Kapton/SCK5; (iii) Duroid 5880 coated directly with SCK5, further referred to as Duroid/SCK5, and (iv) TMM3 (a ceramic/thermoset polymer composite manufactured by Rogers Corp.) coated directly with SCK5, further referred to as TMM3/SCK5. A conventional RF plasma asher (Harrick, Model PDC-3XG) system (15W, 13.56 MHz), operating at 100 mTorr was used for most of RF plasma tests. The LEO equivalent ATOX flux in the central part of the RF plasma reactor was ~5x1015 atoms/cm2sec. The SCK5 coated samples were exposed to different ATOX fluences in the range from 2x1019 atoms/cm2 to 1.7x1021 atoms/cm2. Argon plasma was used to distinguish between reactive and non-reactive plasma interactions. A high power RF plasma reactor (1200 W, Litmas company) was used to evaluate the relative effects of different plasma components, such as in-glow exposure (all plasma components) and after-glow exposure, 100 mm away from the reactor (reduced amount of excited species, an increased amount of neutral atomic oxygen). A specially designed target holder assembly was used to eliminate the direct VUV irradiation allowing all other after-glow plasma components to react with the sample. The system was operated at 300 and 1200 W RF power at an oxygen pressure of 120 mTorr. The VUV flux was assessed by a Phototube sensor (Hamamatsu Model R1187) positioned at the location of the exposed sample. The measurements were performed using a 1% transmission filter, in order to reduce the intensity to the working range of the sensor. The VUV flux was 1.6x1016 photons/cm2sec at 300W, 100 mm away from the reactor. At CERT ONERA, samples were tested using a laser detonation source (CASOAR, manufactured by PSI). Under the operational conditions used in the present study a mean LEO equivalent ATOX flux of ∼1x1015 atoms/cm2sec was obtained. The VUV flux of 1.85x1016 photons/cm2sec was measured by a phototube detector. Modification of the CASOAR system allowed the separation between ATOX, VUV and ATOX/VUV synergistic effects [8]. Due to limitations of the system, only relatively low LEO equivalent ATOX fluences, amounting to 4x1019 atoms/cm2 for ATOX without VUV and 2x1020 atoms/cm2 for ATOX/VUV were achieved. The samples were also exposed to VUV alone at a fluence of 2x1021 photons/cm2. In all ATOX simulation experiments the LEO equivalent atomic oxygen fluence was evaluated by measuring the mass loss of a Kapton HN polyimide reference sample. The erosion rate of Kapton was assumed to be equal to 3x10-24 cm3/atom and independent of the ATOX fluence [9]. The effects of ATOX exposure were studied by several complementary techniques including scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS). SEM micrographs were obtained using a FEI Quanta 200 microscope operating in the low vacuum mode. Surface elemental composition was estimated by means of energy dispersive X-ray

206 spectroscopy (EDS) using JSM 5410LV JEOL instrument with a primary electron energy of 25 keV. To minimize surface charging effects, all samples were coated with a thin (~25 nm) gold layer. The elemental composition and chemical bonding states in the near surface region were assessed by XPS measurements. The measurements were carried out using a VG system with Mg KĮ (1253.5 eV) 200 W X-ray source and a triple channeltron 150 mm hemispherical analyzer, CLAM2. Mass loss measurements were carried out for all the samples exposed in the laser detonation system and for some samples exposed in the RF plasma asher system. The results were used to determine the erosion yields of the SCK5 samples. 3.0 Results 3.1 MORPHOLOGICAL CHANGES AND EROSION YIELDS 3.1.1. RF plasma simulation All types of SCK5 coated samples were exposed in a low power RF plasma asher to different LEO equivalent ATOX fluences, ranging from ~2x1019 to 1.7x1021 atoms/cm2. Duroid/Kapton/SCK5. Duroid 5880 substrates (3cm×5cm) were prepared, onto which SCK5 coated Kapton was adhered by a pressure sensitive adhesive layer. The SCK5 coated Kapton applied on both sides of the substrate in order to prevent its direct exposure to RF oxygen plasma. The edges were protected with CV1144-0 clear silicone coating, manufactured by NuSil, to eliminate the possibility of oxygen undercutting. It was found that RF oxygen plasma exposure to 2x1020 atoms/cm2 and up resulted in well-distinguished changes in surface morphology, detected by the naked eye. The changes included the formation of numerous cracks on the initially smooth paint surface, resembling cracks in dry earth (see Figure 1(a)). Cracks were observed also at lower fluences but they were hardly detected by the unaided eye. An average crack width of about 5 µm was estimated from SEM micrographs (see Figure 1(b)). The width, shape and areal density of cracks did not change significantly with the ATOX fluence above 2x1020 atoms/cm2. The distance between cracks differed from about 100 µm to 1mm. The morphology of the regions between the cracks was similar to that observed for unexposed samples. It should be noted that the unexposed SCK5 coating was highly porous (see Figure 2). Pores with dimensions similar to the width of a crack were found in the unexposed sample. The size and areal density of these pores did not change after ATOX exposure to different fluences. Kapton/SCK5. A free-standing Kapton film (~40 µm thick, 1.5cm×2.0cm) painted with SCK5 was exposed in a low power RF plasma asher to an ATOX fluence of about ~5x1019 atoms/cm2. The sample was supported in a glass sample holder in order to reduce Kapton etching from the backside. After exposure, the sample was found to be strongly bent. No cracks were found on the exposed surface by visual inspection. However, an attempt to flatten the bent film resulted in cracks, as shown in Figure 3.

207 Note that due to the omni-directional nature of the ATOX in the RF plasma system, it is difficult to expose only one side of the sample to ATOX and completely avoid the exposure of the backside. As Kapton itself undergoes erosion under ATOX exposure (~1.5 µm at 5x1019 atoms/cm2 of ATOX), the observed bending effect could be partially associated with a thinning of the peripheral part of the Kapton substrate, in addition to bending caused by ATOX-induced stresses in SCK5. To prevent Kapton erosion from the backside of the sample, in the separate experiment Kapton/SCK5 sample was exposed in a glass frame, which maintained fixed position of the film. In this case welldefined cracks were observed on the SCK5 surface after ATOX exposure, similar to those shown in Figure 1.

(a)

(b)

Figure 1 SEM micrographs of Duroid/Kapton/SCK5 samples after exposure to RF oxygen plasma (1.0x10 21 atoms/cm2 LEO equivalent ATOX fluence): a general view (a) and a magnified view of cracks (b).

Figure 2. SEM micrograph of an unexposed SCK5 coating.

208 The erosion yield of the SCK5 was calculated using an ATOX fluence estimated from the mass loss of a Kapton witness sample exposed at the same experiment. The erosion yield was ~5x10-25 cm3/atom and ~1x10-24 cm3/atom after exposure to a LEO equivalent ATOX fluence of 3.6x1019 and 8x1019 atoms/cm2, respectively. Duroid/SCK5. After exposure to the LEO equivalent ATOX fluence of about 2.5x1019 atoms/cm2, the initially smooth paint surface was cracked into well-distinguished domains. The size of an individual domain was approximately 1mm×1mm (see Figure 4). The average width of the cracks was 5-7 µm. However, due to the poor adhesion of the coating to the Teflon-based Duroid substrate, many delaminated regions were observed, leading to the formation of the paint flakes.

Figure 3 A free standing SCK5 coated Kapton film exposed to RF oxygen plasma (LEO equivalent ATOX fluence of 19 2 ~5x10 atoms/cm ). The apparent cracks while trying to flatten the bent exposed sample.

TMM3/SCK5. No cracks or other changes were observed by the naked eye after an ATOX exposure to a maximum fluence of 1.7x1021 atoms/cm2. However, SEM observations revealed cracks, similar to those shown in Figure 1(a). In this case neither delamination nor separation between substrate and coating were observed. Argon RF plasma exposure. In order to distinguish between chemical effects of ATOX and physical effects of RF plasma, three samples were subjected to argon RF plasma instead of to oxygen plasma. The exposed samples were: (i) a free-standing Kapton/SCK5 film, (ii) a Kapton/SCK5 film in a glass frame, and (iii) a Duroid/Kapton/SCK5. Cracks or bending of the Kapton/SCK5 film were not observed after Ar plasma exposure of all samples for 34 hr (a 34 hr exposure of RF oxygen

209 plasma is equivalent to ATOX fluence of 6x1020 atoms/cm2), and the only effect was a very light coloration of the samples. High power RF plasma reactor. Dedicated experiments were carried out to distinguish between VUV and ATOX effects. This was achieved by using a high power RF plasma apparatus and a specially designed target holder assembly located in the afterglow region. Kapton/SCK5 samples were exposed to 300W and 1200W RF oxygen plasma afterglow flow including and excluding direct VUV radiation. In all cases similar cracking of the exposed area was observed after exposure to a LEO equivalent ATOX fluence of about 2x1019 atoms/cm2.

Figure 4 SEM micrograph of the Duroid/SCK5 sample after exposure to 1.0x1021 atoms/cm2 of equivalent ATOX fluence (RF oxygen plasma simulation).

3.1.2 Laser detonation ATOX source Several samples of each type (Kapton/SCK5, Duroid/SCK5, TMM3/SCK5 and Duroid/Kapton/SCK5) were exposed simultaneously in the laser detonation 5 eV ATOX source (CASOAR). The samples were exposed to a LEO equivalent ATOX fluence of 4x1019 atoms/cm2, to synergistic ATOX/VUV fluences of 2x1020 atoms/cm2 and 2x1021 photons/cm2, respectively, and to a VUV fluence of 2x1021 photons/cm2. Surface morphology of the exposed samples was studied by SEM (data not shown). No cracks or other changes in surface morphology were found after the ATOX, ATOX/VUV or VUV exposures. The erosion yields for Kapton/SCK5 exposed to various environments are shown in Table 1. The erosion yield was calculated using the LEO equivalent ATOX fluence obtained from the mass loss of Kapton witness samples exposed at the same

210 time. The erosion yield of samples exposed to VUV was calculated using the VUV flux of 1.85x1016 photons/cm2sec, as measured by a VUV detector. TABLE 1. Erosion yields of Kapton/SCK5 samples exposed in the laser detonation source. Environment

Fluence

Erosion yield

ATOX/VUV

2x1020 atoms/cm2 and 2x1021 photos/cm2

9.1x10-26 cm3/atom

VUV

2x1021 photos/cm2

1.3x10-26 cm3/photon

19

ATOX

2

3.3x10-25 cm3/atom

4x10 atoms/cm

3.2. CHEMICAL COMPOSITION CHANGES 3.2.1. EDS results The elemental composition of SCK5 coated samples exposed in both simulation systems was determined by EDS and the results are summarized in Table 2. The elemental composition of the reference sample (unexposed SCK5) was measured at several points to assess the uniformity of the coating. For all samples the most prominent effect is a decrease of carbon atomic concentration from about 20 at.% for the pristine coating to about 11-15 at. % after exposure to different fluences of ATOX. This was accompanied by an increase in oxygen atomic concentration in the analyzed layer. The modification of elemental composition in the analyzed region does not show a fluence dependence. No changes in Si, Ti, Zn, and Sn atomic concentrations were observed. TABLE 2. Elemental composition of SCK 5 coated samples determined by EDS before and after exposure to the RF oxygen plasma or the laser detonation source. Element Treatment Unexposed, area I area II 2x1019 atoms/cm2 9x1019 atoms/cm2 1x1021 atoms/cm2 ATOX, 4x1019 atoms/cm2 ATOX/VUV, 2x1020 atoms/cm2 and 2x1021 photons/cm2

C 21.3 19.6

O

Si

46.1 6.5 48.1 6.7 RF oxygen plasma 13.0 52.1 6.8 10.8 57.4 6.4 13.8 54.5 6.5 Laser detonation ATOX source 15.0 50.7 7.4 15.7 50.5 7.2

Ti

Zn

Sn

9.8 9.7

7.1 6.9

9.2 9.0

10.6 9.7 9.6

7.7 6.7 6.6

9.8 9.0 9.0

10.3 10.2

7.0 7.0

9.6 9.4

3.2.2. XPS Results Chemical changes in the irradiated surface layer were detected by XPS. The following Duroid/Kapton/SCK5 samples were analyzed: an unexposed SCK5 coating (reference), two samples exposed using the CASOAR system to ATOX/VUV and ATOX alone,

211 respectively, as well as two samples exposed in a low power RF plasma asher to the LEO equivalent ATOX fluence of 2x1019 and 1x1020 O-atoms/cm2. A typical XPS survey spectrum obtained from SCK5 surface included the Si 2p, C 1s, O 1s, Zn 2p, Sn 3d and Ti 2p core level lines. High-resolution XPS spectra (not shown) obtained from the surfaces before and after exposure to various environments were used for chemical composition analysis. The near surface composition analysis was done by assuming that this region is homogeneous and using published atomic sensitivity factors [10]. The results are presented in Table 3. TABLE 3. Surface composition (at.%) determined from XPS data for Duroid/Kapton/SCK5 samples before and after exposure to ATOX alone and to ATOX/VUV using the CASOAR system, as well as after exposure to an RF plasma system.

Peak Origin

Silicone matrix, Contamination Silicone matrix, Silicon oxide Si-O-Si, metal oxides, adsorbed O, H2O Metal oxides

Bonding Unexposed CASOAR, CASOAR, state ATOX, 4x1019 ATOX/VUV, 2 atoms/cm 2x1020 atoms/cm2 C 1s 36.2 10.2 9.8

RF plasma, 2x1019 atoms/cm2 4.3

RF plasma, 1x1020 atoms/cm2 2.2

Si 2p

24.1

30.5

30.4

23.7

23.9

O 1s

37.0

58.8

59.1

67.3

69.2

Sn 3d, Zn 2p, Ti 2p

2.7

0.7

0.7

4.7

4.7

4.0 Discussion The white paint SCK5 is a thermal control coating, applied on external satellite surfaces that may suffer from electrostatic discharge (ESD) problems. The paint is composed of a purified silicone binder, “doped” metallic oxides that are responsible for its antistatic properties and white color appearance, aromatic solvents and probably some other additives. Its outgassing properties are well within the ASTM E-595 limits (TML=0.187%, CVCM=0.036%, according to Soreq NRC outgassing tests). One issue of primary importance for this coating is its atomic oxygen durability. The SCK5 coatings (applied on Kapton, Duroid 5880 and TMM3) were tested using two types of ATOX simulation systems, an RF plasma asher and a laser detonation source. RF oxygen plasma sources are commonly used for material screening with respect to ATOX degradation in LEO. The advantage of such a facility is its relatively low cost and simplicity. The source generates omni-directional flux of oxygen atoms at thermal energies (~0.04 eV) [3]. However, other species are also present in the RF plasma environment, including molecular oxygen, atomic and molecular oxygen ions

212 and electrons at energies of tens of eV, excited neutral and ionic species, as well as ~130 nm VUV radiation with an approximate flux of 1013 -1016photons/cm2.sec [11,12]. The absolute and relative concentrations of these species depend crucially on the plasma operating conditions, such as plasma power, gas flow, pressure, chamber geometry and the sample’s position inside the reactor. Therefore the plasma - surface interactions are extremely complex and may introduce various artifacts as compared to the real ATOX environment in LEO. The laser detonation atomic oxygen source provides a highly directional beam of approximately 5 eV oxygen atoms. The mean flux provided by this source is about 1x1015 atoms/cm2sec. However, the atomic oxygen is generated in a pulsed mode and the fluence in a single pulse may be as high as 1x1014 atoms/cm2 in a period of time of about 100µsec, which is equivalent to a flux of 1x1019 atoms/cm2sec. This flux, higher by 4 orders of magnitude than in the LEO environment, may react differently with various materials, although, as yet this is not supported by experimental evidence. The formation of atomic oxygen in a laser detonation source is also accompanied by a high flux of VUV (1.85x1016 photons/cm2sec). Therefore this source may provide a strong synergistic effect between ATOX and VUV. In the present study, a modification of the sample holder enabled separation of the ATOX beam from the VUV radiation. However, due to technical limitations, the maximum ATOX fluence that could be simulated in this system within reasonable exposure time was far below the expected values in LEO applications. The interactions of the isotropic, omni-directional flux of thermal RF oxygen plasma vs. the highly directional ATOX beam in LEO with uncoated and coated polymer surfaces were discussed by Banks et al. [3]. It was demonstrated that atomic oxygen undercutting at defect sites (e.g. cracks) progresses much more rapidly in thermal energy plasma systems, due to its omni-directional impingement as compared to this effect in space. Note that from this point of view, the ATOX interaction in the CASOAR simulation system resembles much more closely the LEO ram direction interaction. However, as was already noted, the very high instantaneous acceleration factor may introduce artifacts in the mechanism of chemical interaction and relaxation processes in ATOX - surface interactions. Let us now discuss observed results with particular attention to the basic differences between the two simulation systems. The main visual effect of SCK5 exposure in the RF plasma simulation systems was its cracking. The cracking was observed after low equivalent ATOX fluence of about 2x1019 atoms/cm2. Increasing the ATOX fluence beyond this value up to a maximum ATOX fluence of about 1.7x1021 atoms/cm2 did not affect significantly the surface morphology. However, at some regions and for some materials delaminations of SCK5 were observed in the vicinity of the cracks at high ATOX fluences. This was most prominent for SCK5 coated Duroid substrates and Duroid/Kapton/SCK5 samples, where after exposure to RF oxygen plasma the SCK5 coating became brittle and was affected by vibrations and handling, causing formation of flakes. Such separated/delaminated fragments and flakes possess a potential source of particulate contamination in space, which should be avoided. Micro-cracks in SCK5 could allow the penetration of ATOX and erosion of the underlying substrate by undercutting. It is noted that the cracks on the

213 Kapton/SCK5, Duroid/SCK5 and Duroid/Kapton/SCK5 samples were readily observable by the naked eye, though their average width was only 5-7 µm. Such narrow cracks are visible only in the case of a significant undercutting of the underlying substrate. In the case of TMM3 substrates coated with SCK5 and exposed to a maximum equivalent ATOX fluence of 1.7x1021 atoms/cm2 no cracks were detected by the unaided eye. This may be explained by the better adhesion of SCK5 to TMM3, reducing SCK5 undercutting. The results described in Chapter 3.1.1 indicate a strong surface contraction of the SCK5 during interaction with RF oxygen plasma. Surface contraction leads to an increase of compressive stresses that finally result in fracture of the coating. In the case of a freestanding Kapton/SCK5 film, surface contraction caused the bending of the flexible film and its cracking while attempting to flatten it (see Figure 3), whereas fixed position of the film resulted in a stress relaxation via cracking during the ATOX exposure. To separate between the chemical effects of atomic oxygen and possible artifacts of RF plasma caused by the electromagnetic field, UV radiation, electrons and energetic ions, the SCK5 coated samples were exposed to an RF argon plasma. No cracks or other changes in surface morphology were found. Only some coloration was observed, probably associated with VUV exposure characteristic of RF plasma systems. Thus, the fracture and cracking of the SCK5 coating is clearly associated with oxygen reactivity. The exposure of the various samples in the CASOAR system demonstrated milder changes, as compared to RF plasma systems. The only detected morphological effect was some slight bending of Kapton/SCK5 film. The same trend, namely a milder deterioration of the SCK5 coating in the CASOAR system, was also reflected in the erosion yield and chemical composition results. The observed erosion yields for the same type of tested samples differed for the RF plasma and the laser detonation sources. The erosion yield measured after exposure to various environments in a laser detonation source (CASOAR) was found to be dependent on the irradiation type (see Table 2). It was maximal in the case of ATOX exposure alone, less in the case of synergistic ATOX/VUV exposure and minimal during exposure to VUV irradiation alone. The inhibiting effect of VUV is clearly observed from these results. Based on our previous studies [13] it is suggested that VUV radiation induces cross-linking in the siliconic matrix, reducing the erosion rate due to ATOX impingement and leading to a negative synergistic effect. For Kapton/SCK5 samples the erosion yield was measured both after RF plasma and after laser detonation source (CASOAR) treatments. The erosion yield after the RF plasma exposure was higher (~5x10-25 and ~1x10-24 cm3/atom after exposure to a LEO equivalent ATOX fluence of 3.6x1019 and 8x1019 atoms/cm2, respectively), compared to an ATOX or an ATOX/VUV irradiation in the CASOAR system (~3.3x10-25 and ~9.1x10-26 cm3/atom after exposure to 4x1019 ATOX and 2x1020 atoms/cm2 ATOX/VUV, respectively). It should be noted that the precise mass measurements of the exposed Kapton/SCK5 were problematic due to the unstable weight of the sample, probably due to humidity absorption by the exposed surfaces and/or charging problems.

214 Chemical composition of the SCK5 coating after exposure to RF plasma and laser detonation sources was studied by EDS and XPS. The sampling depths of EDS are matrix dependent and lie within the range of 0.1 µm for light elements, to about 3-5 µm for some metals [14]. XPS is a surface sensitive technique and probes the material to a depth of about 10 nm [15]. In addition, the sensitivity and accuracy of XPS measurements are higher as compared to EDS. The EDS and XPS analyses detected the following elements: C, O, Si, Ti, Zn and Sn. Comparison of the EDS and XPS results indicated that the main changes induced by ATOX are localized in a thin layer at the surface. Besides, it was found that this layer is composed predominantly of organic matrix. This is due to the fact that XPS analysis showed only small amounts of metals on the surface (~2.5%), whereas EDS revealed about 25% of Ti, Zn and Sn. EDS studies did not reveal any significant changes in the metals and silicon concentrations after exposure to both RF plasma system and laser detonation source, indicating that only a thin near-surface region is involved in the ATOX-material interaction. Accurate quantitative analysis of chemical composition using XPS was problematic due to possible surface contamination, as well as high roughness and porosity of the SCK5. Particularly, the concentration of small metal oxide particles embedded in the organic matrix can be affected by the surface morphology. In spite of these limitations significant differences in the concentrations of different components were observed after exposure to 5 eV ATOX and RF oxygen plasma. The results presented in Table 3 clearly show that exposure to the RF oxygen environment caused a drastic decrease of carbon atomic concentration (erosion of organic matrix), while exposing the metal oxide particles and increasing the oxygen content, which partially can be explained by an adsorption of water vapor and oxygen from air. The data clearly indicate the higher deterioration of the SCK5 surface under RF exposure, supporting the surface morphology and erosion yields evidence. Based on the above observations, the following phenomenological model is suggested. Cracking of the silicone-based SCK5 coating as a result of RF oxygen plasma is attributed to strong compressive stresses generated in the exposed coating due to a very effective erosion of the organic component in the silicone binder and formation of a brittle silicon oxide layer. Due to the omni-directional character of atomic oxygen impingement and the high porosity of SCK5 coating, this erosion process includes inner defects and pores. Consequently, a fast degradation of the silicone matrix takes place, leading to the exposure of the metal oxide particles to the RF plasma. This model cannot be applied to the laser detonation source since it interacts in a highly directional manner, creating a surface layer of silicon oxide, which might be non-continuous and/or thinner in this case and consequently less brittle. It may be speculated that the interaction of the SCK5 coating with the RF plasma environment may also be affected by (i) electromagnetic field interaction with exposed metal oxide particles, (ii) accompanying VUV irradiation, (iii) electrons and (iv) energetic ions. To gain a deeper insight into the RF plasma - surface interactions, an experimental study has been initiated at Soreq NRC, in order to understand in more detail the contributions, both individually and synergistically, of the different RF plasma components. Techniques have been developed to separate the plasma components in the plasma afterglow, where the effect of RF electromagnetic field is eliminated. However

215 this location of the sample did not prevent cracking of the SCK5 coating. In addition, a specially designed target holder assembly enabled the sample to be irradiated in the RF plasma afterglow, with and without direct VUV irradiation. No visible effect of the direct VUV flux on the surface morphology was observed. In the future we will expose the sample in a Faraday cup assembly in order to control the amount of electrons and energetic ions arriving at the sample surface, achieving a full separation of the RF oxygen plasma components. 5.0 Summary and Conclusions Atomic oxygen durability of SCK5 white antistatic silicone paint was studied using two types of simulation systems: a conventional RF oxygen plasma system and a laser detonation oxygen source (CASOAR, PSI system). The effects of equivalent ATOX exposure on the surface morphology and surface composition of SCK5 coating applied on different substrates were studied by several complementary techniques, including SEM, EDS and XPS. The tested materials were exposed to different equivalent atomic 21 2 oxygen fluences, ranging from 2x1019 up to 1.7×10 atoms/cm . The SCK5 exposed to RF plasma showed significant cracking, partial delamination and enhanced embrittelment at a relatively low ATOX exposure. Similar samples exposed to the laser detonation source (5 eV ATOX) exhibited no cracking. Thus, the RF plasma simulation demonstrated a more severe degradation of SCK5 paint, evidenced by the morphological changes, as well as by erosion yields and chemical composition changes as compared to the laser detonation system. These results are most probably associated with a combination of omnidirectional flux of reactive species and high porosity of SCK5 coating, which result in strong compressive stresses and consequently cracking of a brittle silicon oxide layer. It is suggested that RF oxygen plasma overestimates the ATOX interactions in LEO, at least for the specific case of a porosive coating of siliconic material, tested in the present work. 6.0 References 1. Guérard, F. and Guillaumont, J.C. (1997), “Thermal control paints and various materials for space use”, Proceedings of the 7th International Symposium on “Materials in a Space Environment”, Toulose, France, 457-458. 2. Golub, M.A., Wydeven, T., and Cormia, R.D. (1988), “ESCA study of Kapton exposed to low Earth orbit or downstream from a radio-frequency oxygen plasma”, Polymer Commun. 29, 285-288. 3. Banks, B.A., Rutledge, S.K., de Groh, K.K., Stidham, C.R., Gebauer, L., and LaMoreaux, C.M. (1995), “Atomic oxygen durability evaluation of protected polymers using thermal energy plasma systems”, NASA Technical Memorandum 106855, 1-15. 4. Caledonia, G.E., Krech, R.H., and Green, B.D. (1987), “A high flux source of energetic oxygen atoms for material degradation studies”, AIAA J. 25, 59-63. 5. Caledonia, G.E., Krech, R.H., Oakes, D.B., Lipson, S.J., and Blumberg, W.A.M. (2000), “Products of the reaction of 8 km/s N(4S) and O2”, J. Geophys. Res. 105(A6), 12,833-12,837. 6. Minton, T.K., Garton, D.J. (2001), “Dynamics of atomic-oxygen-induced polymer degradation in low earth orbit”, in “Chemical Dynamics in Extreme Environments: Advanced Series in Physical Chemistry”, ed. Dressler, R.A., World Scientific, Singapore.

216 7. Koontz, S.L., Albyn, K., and Leger, L.J. (1991), “Atomic oxygen testing with thermal atom systems: a critical evaluation”, J. Spacecraft, 28, 315-323. 8. Grossman, E., Gouzman, I., Viel, V., and Dinguirard, M., “Modification of the Atomic Oxygen Laser Detonation Source: separation of atomic oxygen and UV radiation”, to be published. 9. Minton, T.K. (1995), “Protocol for atomic oxygen testing of materials in ground based facilities, ver. No. 2”, JPL publication 95-17. 10. Chastain, J. and King, R.C., Jr. (eds.) (1995) Handbook of XPS, Physical Electronics, Inc. USA. 11. Townsend, J.A. (1996) A comparison of atomic oxygen degradation in low earth orbit and in a plasma etcher, Proc. 19th Space Simulation Conf., Baltimore MD, 249-258. 12. Kearns, D.M., Gillen, D.R., Voulot, D., McCullough, R.W., Thompson, W.R., Cosimini, G.J., Nelson, E., Chow, P.P., and Klaassen, J. (2001) Study of the emission characteristics of RF plasma source of atomic oxygen: measurements of atom, ion, and electron fluxes, J.Vac. Sci. Technol. A , 19, 993-997. 13. Grossman, E., Noter, Y., and Lifshitz Y. (1997) Oxygen and VUV irradiation of polymers: atomic force microscopy (AFM) and complementary studies, Proceedings of the 7th International Symposium on Materials in a Space Environment, 16-20 June, Toulouse, France, ESA-SP-399, 217. 14. Goldstein, J.I., Newbury, D.E., Echlin, P., Joy, D.C., Roming, A.D., Jr., Lyman, C.E., Fiori, C., and Lifshin, E. (1992) Scanning Electron Microscopy and X-ray Microanalysis, Plenum Press, New York. 15. Walls, J.M. (1990) Methods of Surface Analysis, Cambridge University Press.

STUDY OF POLYMER COATINGS RESISTANCE AFTER THE LONGTERM EXPOSURE ON SPACE STATION “MIR” E. N. KABLOV, V. T. MINAKOV, I. S. DEEV All-Russian Institute of Aviation Materials 17, Radio Str., Moscow, 107005, Russia Phone: +7 095 261-86-77, Fax: +7 095 267-86-09, E. F. NIKISHIN M.V. Khrunitchev State Space Scientific Production Center 18, Novosavodskaya Str., Moscow, 121087, Russia Phone: +7 095 142-50-36, Fax: +7 095 956-24-41 Abstract The study of the resistance of various thermo-regulating (thermal control) protective coatings to space environment was conducted on carbon fiber-reinforced polymer (CFRP) epoxy composites and three-layered honeycomb specimens after the longterm exposure of “Komplast” removable cassettes outside of “Kvant-2” module of Space Station “MIR”. Scanning electron microscopy (SEM), X-ray spectral analysis and visual inspection were used as the major tools in the study. Surfaces of white thermal control coatings, based on acrylic and epoxy polymers, in pristine condition and after 839 and 1218 days of the exposure were comparatively evaluated. In addition, similar analysis was conducted for a number of coatings based on organosilicone polymers, for inorganic coatings based on liquid glass, and aluminum foil, after 1024 days of space exposure. During the space exposures, some specimens with coatings of similar types were exposed in holders allowing to keep them in a stack, one under another with positioning of the upper specimen with the coating on outside and of the lower specimen’s coating facing the wall of the module that allowed to expose coatings of the same structure to different space conditions. Different types of macro- and micro-defects that appeared in the coatings were revealed, in particular, long cracks with a partial delamination of the coating from the substrate, a network of thin web-like cracks, scaly swelling-up and ring-like delamination. The change of the coatings surface composition, including the substances deposited on the surface during the long-term exposure in space, was evaluated. The possible coatings degradation mechanisms under the complicated space environment factors were discussed, with the account for the coating’s chemical nature. 1. 0 Introduction It was shown earlier that the thin surface layers of the matrix and the reinforcing fibers of polymer composite materials without protective coatings after long-term exposures on “Salute-5”, “Salute-6” orbital stations and Space Station “MIR” were subjected to various degradation effects that lead to changes of their structural 217

218

properties [4-7]. In order to improve the resistance of polymer-based composites to the long-term space environment factors and to prevent the deterioration of the thermal optical properties of the structural components it is necessary to use the thermal control protective coatings that should possess both the high environmental resistance and the ability to reduce the material’s surface temperature that is of great importance for its long-term thermocycling. The long-term space exposure effects on optical characteristics of protective coatings have been discussed already [7-8]. The use of thermal control coatings for the protection of polymer composites during the long-term flight in the outer space requires the confirmation of their resistance to the combination of space factors. Therefore, the evaluation of the long-term resistance of various thermal control protective coatings on CFRP specimens and skins in three-layered honeycomb structures to outer space factors was included into the program of space experiments with the use of “Komplast” removable cassettes. Those experiments are important for both research contribution and practical application. Results of a comparative evaluation of different types of thermal control protective coatings deposited onto polymer composite specimens after a long-term exposure on Space Station “MIR” are presented in this paper. 2. 0 Materials and Methods Various white thermal control protective coatings such as polymeric (acrylic, epoxy, organosilicone) and non-organic (based on liquid glass), and also aluminum foil of 30 µm thick were the subjects of this study. The KMU-4* epoxy CFRP composites in the form of plates (60x60x2 mm) and KMU-3l* in the form of skins for three-layered honeycomb specimens (60x60x15 mm) served as substrates for the coatings. Prior to the coating deposition process KMU-3l CFRP three-layered specimens with skins underwent additional thermal stabilization to eliminate any volatile substances (air, sorbed water vapor, organic solvent residuals, etc.) that may evolve. The specimens of above-mentioned CFRP’s with the protective coatings were exposed in “Komplast” removable cassettes on the “Kvant-2” module outside of Space Station “MIR” (Figure 1) for 839, 1024 and 1218 days, and returned back to Earth. A comparative study of the surfaces of thermal control protective coatings prior to and after the natural exposure in space was conducted using optical and scanning electron microscopy, X-ray microanalysis and visual inspection. The micro-structural analysis was performed using the JSM-35C SEM (“JEOL”) microscope and MBS-10 optical microscope with magnifications from x4 to x5000. A thin layer of gold (15-20 nm) was deposited using a sputter deposition vacuum system, FC-1100 FINE COAT (“JEOL”), before the structural examination by SEM to prevent charging. The surface chemistry of the pristine coatings and of the specimens, subjected to full-scale test in the space was determined by X-ray microanalysis using the JXA-840 micro-analyzer (“JEOL”) and energy-dispersive detector (“LINK”). The results are presented below.

*

traditional Russian types

219

Figure 1. “Komplast” removable cassette with protective coating on composite specimens on “Kvant-2” module exterior surface of Space Station “MIR”

3.0 Results and Discussion Epoxy coating. The homogeneous surface structure, as seen in Figs. 2a and 2b, is typical for the initial white epoxy coating both on KMU-4 CFRP and KMU-3l CFRP skin.

Figure 2. Microstructure of the initial (a, b) and shielded by another specimen surfaces of the white epoxy coating on KMU-4 CFRP after 1218 days (c, d) of exposure on Space Station “MIR”: a, c – x2000; b, d – x5000.

220

Powder-like, rounded filler particles (bright-contrast spots) are distributed comparatively uniformly in the film-forming polymer, with their sizes not exceeding 1 µm. No defects like microcracks, de-lamination, etc. are present in the initial coating. According to X-ray elemental microanalysis (Fig. 3a), the epoxy coating surface layer contained titanium (most probably in the form of titanium dioxide pigments) and silicon, magnesium and aluminum in smaller concentrations.

Figure 3. X-ray energy-dispersive microanalysis spectra of the white epoxy coating surface on KMU-4 CFRP prior to (a) and after 1218 days (b) of exposure on “MIR” orbital complex.

221

Visual examination of KMU-4 CFRP and KMU-3l exposed specimens (as well as shielded specimens) for 1218 days on “MIR” orbital complex showed that the morphology of the coating on the thermo-stabilized specimen surfaces remained very close to the initial, without the formation of visible macro-defects (Figure 4a). However, KMU-4 CFRP specimens that were not thermo-stabilized exhibited local macro-defects in the coating in the form of extended cracks with partial delamination and lifting, scale-like formations and ring-like de-laminations from the CFRP surface (Figure 4b). It should be mentioned that the observed local swelling and de-lamination also occur on shielded KMU-4 CFRP specimens, but to a considerably lesser extent as compared to unshielded specimens (Figure 4b).

Figure 4. General view of the surface of epoxy (a, b, c) and acrylic (d) protective coatings of KMU-4 CFRP specimens prior to (a) and after 1218 days (b, c, d) of exposure on “MIR” ; x1.

It was established from electron microscopy analysis that the shielding of KMU-4 CFRP specimen with epoxy coating by another similar specimen for 1218 days completely protected the coating surface from the degradation and erosion, with the structure of the shielded coating only negligibly changed (Figure 2b, d). The surface of the open specimen coating shielded within the period of 1218 days of exposure by a cassette frame (Figure 5a) was covered by a network of randomly distributed microcracks, with the structure of the matrix with dispersed in it filler particles being retained (Figure 5b). The matrix film-forming epoxy polymer on the

222

open coating surface, opposite to the surface covered by the frame, on the other hand, was subjected to degradation and erosion. As a result, the filler particles became loose and were partially removed from the surface. (Figure 5c, d). Also, as a consequence of the degradation processes and the matrix polymer coating film erosion, the surface became enriched in filler particles, giving rise to stronger signals of titanium, silicon and magnesium in the EDS spectra collected from the surface (Figure 3b).

Figure 5. Microstructure of the white epoxy coating surface on KMU-4 CFRP after 1218 days exposure on “MIR” orbital complex: a – surface, covered by a cassette frame, x200; b – surface, covered by a cassette frame, x2000; c – open surface, x2000; d - open surface, x5000.

A comparative investigation of pristine (Figure 6a) and exposed (Figure 6b, c, d) surfaces of the epoxy coating on the CFRP (KMU-3l) skin in the three-layered honeycomb specimen showed, that their microstructure and elemental composition changed in the same way as the KMU-4 CFRP after 839 (Figure 6b, c) and 1218 (Figure 6d) days of exposure. Acrylic coating. The surface structure of acrylic coating on CFRP (KMU-3l) skin in the three-layered honeycomb specimen prior to the space exposure (Figure 7a) doesn’t contain macro- and micro-defects, the filler particles are distributed uniformly. X-ray analysis of this coating showed (Figure 8a), that it’s characteristic feature is the content of titanium and aluminum, which is related to the filler composition. The visual inspection of acrylic coating specimens on KMU-4 CFRP surface and KMU-3l CFRP skin in three-layered honeycomb specimens, exposed for 839 and 1218 days showed, that the coating was retained on the majority of specimens without visible macro-defects.

223

Figure 6. Microstructure of the white epoxy coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 839 (b, c) and 1218 (d) days of exposure on “MIR” orbital station; x2000.

At the same time macro-defects in the form of scale-like swelling-up and cracks with local de-lamination (Figure 4d) were observed on KMU-4 CFRP specimens with acrylic coating, having the lesser thermo-stabilization degree as well as with epoxy coating though in this case it occurs more rarely. The electron microscopic study of acrylic coating specimens on KMU-3l CFRP after 839 days of space exposure (Figure 7b) has shown the fine coating structure was changed insignificantly (only a weak polymer film erosion and negligible filler particle stripping took place). Due to the acrylic coating structural stability, its elemental composition practically did not change after 839 days of exposure in space, as compared to the pristine. An increase in the exposure time to 1218 days (Figure 7c, d) leads to the appearance of separate microcracks on the coating surface The electron microscopic study of acrylic coating specimens on KMU-3l CFRP after 839 days of space exposure (Figure 7b) has shown the fine coating structure has changed insignificantly (only a weak polymer film erosion and negligible filler particle stripping took place). Due to the acrylic coating structural stability, its elemental composition practically did not change after 839 days of

224

exposure in space, as compared to the pristine. An increase in the exposure time to 1218 days (Figure 7c, d) leads to the appearance of separate microcracks on the coating surface and some indication of erosion of the polymer film, however, the elemental surface layer composition is slightly changed (only small amount of compounds with silicon was revealed, Figure 8b).

Figure 7. Microstructure of the white acrylic coating surface on (KMU-3l) CFRP skin in the threelayered honeycomb specimen prior to (a) and after 839 (b) and 1218 (c, d) days of exposure on “MIR” orbital station; x2000

225

Figure 8. X-ray energy-dispersion microanalysis spectra of the white acrylic coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 1218 days (b) of exposure on “MIR” orbital complex.

Figure 9. Microstructure of the open (a) and shielded (b) surfaces of the white acrylic coating surface on KMU-4 CFRP after 839 days of exposure on “MIR” orbital complex; x2000.

226

Figure 10. X-ray energy-dispersion microanalysis spectra of open (a) and shielded by another specimen (b) surfaces of the white acrylic coating surface on KMU-4 CFRP after 1218 days of exposure on “MIR” orbital complex.

A similar behavior of structural transformations of acrylic coating at the macro- and micro-levels is observed after 839 days of exposure in space on KMU-4 CFRP specimens both on the open surface (Figure 9a) and the one shielded by another specimen (Figure 9b). The presence of small amount of zinc (Figure 10b) is observed in the elemental composition of protective acrylic coating, containing titanium, aluminum, and silicon compounds on the shielded surfaces and the content of which is somewhat larger on the open coating surface (Figure 10a). Organo-silicone coating. The micro-heterogeneous structure (Figure 11a), in which the continuous dispersion medium (silicon-organic polymer) contains multiple inclusions of dispersed phase (powder-like filler particles) is the characteristic feature for the thermal control organo-silicone coating surface prior to the space exposure.

227

The coating surface has the roughness due to salient particles and their aggregates of dispersed filler. The size of such particles may be up to ~ 1 µm and their aggregates are up to 10-15 µm. The micro-cracks network (Figure 11b) appeared on the surface of organo-silicone coatings after 1024 days of exposure of honeycomb specimens in the space, however, it’s fine structure also was noticeably changed and become more friable and heterogeneous (Figure 11b). It should be mentioned, that any new structural elements were not been revealed on the coating’s surface after the long-term exposure in space. The X-ray energydispersion microanalysis of thermo-regulating organo-silicone coating, subjected to the long-term exposure in space showed that it included silicon and zirconium compounds (silicon is the polymer matrix base and zirconium is included into the powder-like filler composition (Figure 12a)).

Figure 11. Microstructure of the white thermo-regulating organo-silicone coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 1024 days (b, c) of exposure on “MIR” orbital complex: a, c – x2000; b - x30.

228

The inorganic coating structure was negligibly changed after 1024 days of space exposure (Figure 13b). The fine-dispersed filler particle shape and sizes are retained on the coating surface without changes: microcracks are absent and the micrometeoroid particle effect traces or the condensed deposit were not revealed. According to X-ray spectrum microanalysis (Figure 12b) the inorganic coating includes the large amount of compounds, consisting of silicon, zinc and potassium atoms as well as chlorine- and titanium-containing compounds but in lesser amount into its composition. It’s known, that silicon and potassium are the base of liquid glass, entering into coating, and the zinc and titanium availability points to the content of fine-dispersed fillers in it (zinc and titanium oxides), ensuring the white color to coating. Inorganic coating. The thermo-regulating inorganic coating on the base of liquid glass on KMU-3l CFRP surface before the space exposure (Figure 13a) has the homogeneous structure consisting of friable packed fine-dispersed (less than 1 µm) particles of a filler with the asimmetric shape. The filler particles of larger size (to 10 µm) along with some separate microcracks also can be though not so often. Aluminum foil. The aluminum foil macrostructure on the back KMU-4 and KMU3l CFRP specimen surfaces was considerably changed after 1218 days of space exposure. Rounded de-lamination (Figure 14a) in the form of swelling-up and folds, repeating the honeycomb cell configuration were formed in the foil (Figure 14). The skin surface composition with the aluminum foil wasn’t practically changed after the space exposure and it was determined only by the aluminum content and small silicon impurity.

229

Figure 12. X-ray energy-dispersion microanalysis spectra of the thermo-regulating inorganic coating surface (a) and organo-silicone coating (b) on (KMU-3l) CFRP skin in the three-layered honeycomb specimen after 1024 days of exposure on “MIR” orbital station.

230

Figure 13. Microstructure of the thermo-regulating inorganic coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 1024 days (b) of exposure on “MIR” orbital complex; x2000.

231

Figure 14. Microstructure of the aluminum foil surface on KMU-4 CFRP after 1218 days of exposure on “MIR” orbital complex: a – x4; b – x20.

The observed changes of coating surface macrostructure (epoxy, acrylic and aluminum foil) are obviously associated with the liberation of gaseous lowmolecular compounds out of CFRP under the space conditions (air, absorbed water vapors, organic solvent residuals and other volatile substances), which can’t diffuse

232

through the coating film or foil because of their high tightness (low gas permeability). The liberated gaseous substances are accumulated under the coating during the long-term space exposure, break its adhesion to CFRP surfaces due to swelling-ups, so that local de-laminations and cracks are formed. At the micro-level the coating degradation under the space factor effect begins with the micro-crack net formations in the film-forming polymer, its erosion, which causes the stripping of powder-like filler particles, being placed on the surfaces. Most likely, the coating cracking is associated with the high inner stresses, appearing in the material during the long-term space exposure with the large thermo-cycle quantity (18 thermocycles per one day of flight) because of the difference of thermal linear expansion coefficients (TLEC), the elasticity of coating and CFRP on the surfaces of which it was deposited. 4.0 Conclusions The study of resistance for different types (epoxy, acrylic, silicon-organic, nonorganic) of protective thermo-regulating (thermal control) coatings on epoxy CFRPs composites and three-layered honeycomb specimens were conducted by SEM, X-ray spectrum microanalysis and visual inspection methods after the longterm exposure in the “Komplast” removable cassette composition on the exterior “Kvant-2” module surface of Space Station “MIR”. The comparative evaluation of white thermo-regulating coating surfaces based on acrylic and epoxy polymers is given prior to and after 839 and 1218 days of exposure as well as on the base of organo-silicone polymer, inorganic coating based on liquid glass and aluminum foil after 1024 days of space exposure. Different types of macro- and micro-defects appeared in coatings were revealed, in particular, long-length cracks with the partial coating de-lamination from a substrate, thin microcracks, scaly swelling-up and ring-like de-lamination. The elemental composition changes of coating surfaces and also substances, deposited on them during the long-term exposure in the space were studied. The feasible coating degradation mechanism was considered with the account of the complex space factor effect, chemical nature of coatings and their substrates. 5.0 References 1. Deev I. S., Nikishin E.F.) Effect of outer space factors on the polymer composite structure under longterm staying conditions in the nearearth orbit. Space Forum, OPA. 1 (1996), pp. 297-302 2. Deev I.S., Nikishin E.F.) Effect of long-term exposure in the space environment on the microstructure of fibre-reinforced polymers. Composites Science and Technology. 57 (1997), pp. 1391-1401 3. Shalin R.E., Minakov V.T., Deev I. S., Nikishin E.F. Study of polymer composite specimens surface changes after the long-term exposure in space. Proceedings of the 7th International Symposium on "Materials in a Space Environment", Toulouse, France, 16-20 June 1997 (SP-399, August 1997), pp. 375-383 4. Barbashev E.A., Bogatov V.A., Deev I.S., Dorofeev Yu.I, Konkin N.I., Milinchuk A.V., Naumov S.F., Nikishin E.F., Perov B.V., Skurat V.E. The contribution of different factors of a space environment in the changes of properties polymer materials. In: "Proc. of the Sixth International Symposium on “Materials in a Space Environment", ESTEC, Noordwijk, The Netherlands, 19-23 September, 1994, pp. 235-238 5. Startsev O.V. & Nikishin E.F. Structure and properties of polymeric composite materials during 1501 days outer space exposure at «Salut-7» orbital station, in: Proc. of the Third LDEF Symposium, Williamsburg, Wirginia, November 1993, pp. 8-12.

233 6. Startsev O.V. & Nikishin E.F. Properties of adhesive compounds of polymeric composite materials and thermoplastic polymers during 1501 days of outer space exposure, in: Proc. of the Sixth International Symposium on “Materials in a Space Environment”, ESTEC, Noordwijk, The Netherlands, 19-23 September 1994, pp. 223-235. 7. Naumov S.F., Gorodetsky A.A., Sokolova S.P., Demidov S.A., Kurilenik A.O., Gerasimova T.L. Study on materials and outer surface coatings aboard space station “MIR”. Proceedings of the 8-th International Symposium on "Materials in a space environment”/ Proceedings of the 5th International Conference on "Protection of Materials and Structures from the LEO Space Environment". Arcachon, France, 5-9 June 2000. 8. Kleiman J. and Iskanderova Z., Technological aspects of protection of polymers and carbon-based materials in space. Proceedings of the 8th International Symposium on "Materials in a space environment"/Proceedings of the 5th International Conference on "Protection of Materials and Structures from the LEO Space Environment". Arcachon, France, 5-9 June 2000.

This page intentionally left blank

ISSUES AND CONSEQUENCES OF ATOMIC OXYGEN UNDERCUTTING OF PROTECTED POLYMERS IN LOW EARTH ORBIT BRUCE A. BANKS, AARON SNYDER, SHARON K. MILLER NASA Glenn Research Center Cleveland, OH 44135, USA RIKAKO DEMKO Cleveland State University Cleveland, OH 44115, USA

Abstract Hydrocarbon polymers that are exposed to atomic oxygen in low Earth orbit are slowly oxidized which results in recession of their surface. Atomic oxygen protective coatings have been developed which are both durable to atomic oxygen and effective in protecting underlying polymers. However, scratches, pin window defects, polymer surface roughness and protective coating layer configuration can result in erosion and potential failure of protected thin polymer films even though the coatings are themselves atomic oxygen durable. This paper will present issues that cause protective coatings to become ineffective in some cases yet effective in others due to the details of their specific application. Observed in-space examples of failed and successfully protected materials using identical protective thin films will be discussed and analyzed. Proposed approaches to prevent the failures that have been observed will also be presented. 1. 0 Introduction The use of atomic oxygen protective coatings applied over conventional polymers that have traditionally been used in space has been the primary approach to date to achieve atomic oxygen durability in space. Metal atoms or metal oxide molecules have been used extensively for the protective coating materials. Typically silicon dioxide, fluoropolymer-filled silicon dioxide, aluminum oxide or germanium have been sputter deposited on polymers to provide atomic oxygen protection. For example, the large solar array blankets on International Space Station have been coated with 1300 angstroms of SiO2 for atomic oxygen protection [1]. Although protective coatings can provide excellent atomic oxygen protection of hydrocarbon or halocarbon polymers, the details of how the coatings are used and/or applied can result in widely varying protection consequences.

235

236 2. In-Space Protective Coatings Experiences 2.1 EUROPEAN RETRIEVABLE CARRIER (EURECA) The EURECA spacecraft, which was deployed into low Earth orbit on August 2, 1992 and retrieved after 11 months on June 24, 1993, was exposed to an atomic oxygen fluence of approximately 2.3x1020 atoms/cm2 [2]. To assist in its retrieval, the spacecraft used two thin adhesively mounted acrylic optical retro-reflectors for laser range finding. Prevention of atomic oxygen attack of the retroreflector surfaces, which would have degraded the specularity of the reflectance, was accomplished by coating the retroreflector surface with a ~1000 Angstrom thick film of sputter deposited SiO2 filled with 8% fluoropolymer (by volume). The LEO exposed and retrieved retroreflector was inspected and optically characterized. The results indicated that the protective coating provided excellent protection and the retroreflector performed as planned except in a small 3 cm patch where the protective coating was accidentally abraded prior to flight as a result of handling during preflight ground integration [3]. Figure 1 shows a close up picture of the retro-reflectors as well as their appearance during illumination after retrieval.

Figure 1. EURECA retro-reflectors after retrieval close up and during illumination.

2.2

INTERNATIONAL SPACE STATION (ISS) RETROREFLECTORS

ISS retro-reflectors, which serve in a similar role as the EURCA retro-reflectors, have been used which employ a glass corner cube retroreflector that is housed in a 10 cm diameter Delrin 100 polyoxymethylene mount. Polyoxymethylene is an oxygen rich polymer that is readily attacked by atomic oxygen. To prevent atomic oxygen attack of the Delrin, the machined polymer surfaces were coated by the same processes, in the same facility and with the same ~1000 Angstrom thin film of sputter deposited 8% fluoropolymer-filled SiO2 that was used for the EURECA retroreflector. Several of these retro-reflectors have been mounted on the external surfaces of the ISS structures at various locations that are exposed to LEO atomic oxygen. Figure 2 shows a close up of one of the coated retro-reflectors prior to use on ISS in space as well as a photograph from space of a retroreflector after attack by atomic oxygen. It is clear from the in-

237 space photograph that the coating was only partially attached allowing direct atomic oxygen attack of the unprotected areas.

Figure 2. ISS retro-reflectors prior to launch and during use in space on ISS after atomic oxygen attack.

2.3 ISS PHOTOVOLTAIC ARRAY BLANKET BOX COVERS Prior to deployment, the ISS photovoltaic arrays were folded into a box that allows the array to be compressed in a controlled manner against a cushion of open pore polyimide foam that was covered with a 0.0254 mm thick aluminized Kapton blanket. The Kapton was coated on both surfaces with 1000 Angstroms of vacuum deposited aluminum. The array was exposed to the LEO atomic oxygen environment from December 2000 through December 2001. Photographs of the array, taken in orbit, indicated that the Kapton blanket had been almost completely oxidized leaving only the thin largely torn aluminization in place as shown in Figure 3.

a. Distant photo b. Close up photo Figure 3. ISS photovoltaic array showing effects of atomic oxygen erosion of the double aluminized Kapton blanket cover for the ISS photovoltaic arrays box cushions.

238 3.0 Analyses and Discussion 3.1 SURFACE ROUGHNESS AND DEFECT DENSITY The drastic differences in atomic oxygen protection provided by the same SiO2 coating filled with 8% fluoropolymer on the EURECA retro-reflectors and the ISS retroreflectors is thought to be due to drastic differences in the protective coating defect densities. The acrylic EURECA retro-reflectors surfaces were extremely smooth as required to produce high fidelity specular reflections. Such smooth surfaces result in low-defect-density protective coatings that have also been demonstrated, in ground laboratory testing, to perform acceptably. For example smooth surface (air-cured side) Kapton when coated with 1300 Angstrom thick SiO2 resulted in ~ 400 pin window defects/cm2. However, the same coating on the rougher surface (drum-cured side) has been found to result in 3500 pin window defects/cm2 [1]. Similar experiences with graphite epoxy composite surfaces formed by casting against another smooth surface produce defect densities of ~262,300 defects/ cm2 [3]. Surface leveling polymers applied over such surfaces have been found to reduce the defect densities by an order of magnitude to ~22,000 defects/cm2 [3]. The machining of the Delrin 100 (polyoxymethylene) retroreflector mount surfaces produces machine marks or rills in the surface resulting in a highly defected atomic oxygen protective coating. Such rills allow atomic oxygen to oxidize and undercut the high erosion yield Delrin, causing the coating to gradually be left as an unattached gossamer film over the retroreflector mount which could be easily torn and removed by intrinsic stresses and thruster plume loads. The use of smoother surfaces, surface-leveling coatings over the machined Delrin or use of alternative atomic oxygen durable materials could potentially eliminate the observed problem. 3.2 TRAPPING OF ATOMIC OXYGEN BETWEEN DEFECTED PROTECTIVE SURFACES The lack of atomic oxygen protection provided by the aluminized Kapton blanket cover for the ISS photovoltaic arrays box cushion is thought to be due to the trapping of atomic oxygen between the two aluminized surfaces on the 0.0254 mm thick Kapton blanket. Defects in the space exposed aluminized surface allow atomic oxygen to erode undercut cavities. If the undercut cavity extends downward to the bottom aluminized surface, then the atomic oxygen becomes somewhat trapped and has multiple opportunities for reaction until it either recombines, reacts, or escapes out one of the defects in the aluminization. This eventually results in a complete loss of the Kapton with only the aluminized thin film remaining. The vacuum deposited aluminum has a slight tensile stress that causes stress wrinkling of the unsupported aluminum films. Figure 4 is a photograph of a vacuum deposited aluminized Kapton sample that was placed in a radio frequency plasma environment to completely oxidize the Kapton over a portion of the sample.

239

Figure 4. Photograph of a vacuum deposited aluminized Kapton sample bonded to a metal frame after ground laboratory oxidation of the Kapton.

As can be seen in Figure 4, where the ~1000 Angstrom aluminum film in the lower portion of the sample is free standing, stress wrinkles and tears develop similar to those seen in the ISS photograph of Figure 3. A two dimensional Monte Carlo computational model has been developed which is capable of simulating LEO atomic oxygen attack and undercutting at crack defects in protective coatings over hydrocarbon polymers [4]. Optimal values of the atomic oxygen interaction parameters were identified by forcing the Monte Carlo computational predictions to match results of protected samples retrieved from the Long Duration Exposure Facility [4]. These interaction parameters and values were used to predict the consequences of atomic oxygen entering a 2-dimensional crack or scratch defect in the top aluminized surface. This was accomplished using 100,000 Monte Carlo atoms entering a defect which was 20 Monte Carlo cells wide (representing a 13.4 micrometer wide defect) over a 38 cell thick (representing a 0.0254 mm thick) Kapton blanket. Figure 5 compares the Monte Carlo model computational erosion results for a 45-degree angle of attack (relative to the surface normal) of the atomic oxygen for both double surface-coated Kapton (which was the case for ISS) and single top surfacecoated Kapton.

240

a.

b.

Aluminized on both sides

Aluminized on exposed side only

Figure 5. Monte Carlo computational atomic oxygen erosion predictions for a 45 degree from perpendicular angle of attack of atomic oxygen at a crack or scratch defect in the aluminized Kapton surface.

As can be seen from Figure 5, even though the atomic oxygen gradually becomes less energetic with the number of interactions and has approximately a 13% chance of recombination, the trapped atoms undercut far more in the actual ISS case of a double aluminization as would have occurred if the Kapton was simply aluminized on one side. Thus, contrary to intuition, the use of two atomic oxygen protective coatings rather than a single coating appears to cause more rather than less undercutting attack. The extent of undercutting of trapped atomic oxygen is also dependent on the opportunity for the atoms to loose energy, recombine, or escape back out the defect opening. Figure 6 compares the results of 2-dimensional Monte Carlo modeling and 3dimensional pin-window computational predictions [5] for a 45-degree angle of attack atomic oxygen of a 13.4 micrometer wide crack or scratch for the 2-dimensional case and a 5.1 micrometer diameter circular aperture for the 3-dimensional case for both single side and double side aluminized Kapton.

241

Fraction of atoms reacted

0.25

Double-coated

0.20

0.15

0.10

0.05

Single-coated 0.00 0.0E+00

4.0E+20

8.0E+20

1.2E+21

1.6E+21

2.0E+21

Fluence, atoms/cm2

a.

2-Dimentional model of crack or scratch defect

0.20 Fraction of atoms reacted

0.18 0.16

Double-coated

0.14 0.12 0.10 0.08

Single-coated

0.06 0.04 0.02 0.00 0.E+00

2.E+21

4.E+21

6.E+21

8.E+21

2

Fluence, atoms/cm

b.

3-Dimentional model of circular pin window defect

Figure 6. Computational atomic oxygen erosion predictions for 45-degree incident atomic oxygen attack at defect sites protected Kapton.

As can be seen in Figure 6, for both 2-dimensional modeling of a crack or scratch defect and 3-dimensional modeling of a circular defect the growth characteristics of the under cut cavity have similar trends with fluence. Initially, as the undercutting starts the existence or absence of the back surface coating plays no role

242 and as the cavity grows the probability of atoms reacting increases due to trapping of the incoming atom. However, as the bottom surface is reached, atoms begin either to escape, or in the case of no back-surface coating, they recombine after collision with the SiO2 on the back surface. The double surface aluminized Kapton consistently reacts more atomic oxygen atoms than the single surface aluminized Kapton except at very low fluences where the erosion in either case does not reach the bottom of the polymer. For both cases, as the fluence increases, the atomic oxygen can escape out the bottom (only in the case of the single surface aluminized Kapton), recombine, or thermally accommodate and thus becomes less probable to react with the Kapton. Thus it appears that a single surface aluminized Kapton would have been much more durable because the unreacted atoms passing through the bottom of the polymer would simply enter into the open pore foam and gradually react with it, without causing much damage to the aluminized Kapton. The double-SiO2 coated ISS solar array blankets may show similar detachment of the outer surface SiO2 layer with time. However, the defect density appears to be much lower than for vacuum deposited aluminum coatings as shown in Figure 7 which compares the experimental results of RF plasma oxidation of double aluminized Kapton with double SiO2 coated Kapton.

0

Mass change / area, mg/cm2

Double SiO2 Coated Kapton

-0.05

-0.1

Double Aluminized Kapton

-0.15

-0.2

-0.25 0

1E+20

2E+20

3E+20 4E+20 Fluence, atoms/cm2

5E+20

6E+20

Figure 7. Comparison of RF plasma oxidation of aluminized and SiO2 coated Kapton.

4. 0 Conclusions Atomic oxygen protective coatings have been developed and used in space that perform acceptably. However, rough surface substrates cause defects in the protective coatings that allow atomic oxygen to react and gradually undercut the protective coating. In the case of machined Delrin ISS retroreflector mounts, such roughness has lead to detachment of portions of the protective film covering the retroreflector mount.

243 Atomic oxygen undercutting of the double aluminized Kapton blanket covers for the ISS photovoltaic array box cushions has occurred resulting in a torn and partially detached aluminum film. Based on computational modeling, atomic oxygen atoms that become trapped between the two aluminized films on each side of the Kapton blanket appear to cause accelerated undercutting damage in comparison to the use of a single top-surface coating. 5. 0 References 1.

Rutledge S., Olle R., “Space Station Freedom Solar Array Blanket Coverlay Atomic Oxygen Durability Testing,” 38th SAMPE Symposium, May 10-13, 1993.

2.

Banks B. A., Rutledge S., and Cales M., “Performance Characterization of EURECA Retroreflectors with Fluoropolymer-Filled SiOx Protected Coatings”, Long Duration Exposure Facility (LDEF) Conference, Williamsburg, Virginia, November 8-12, 1993.

3. De Groh K., Dever J., and Quinn W., “The Effect of Leveling Coatings on the Durability of Solar Array Concentrator Surfaces,” 8th International Conference on Thin Films and 17th International Conference on Metallurgical Coating,” San Diego, California, April 2-6, 1990. .

4. Banks B., Stueber T., and . Norris, "Monte Carlo Computational Modeling of the Energy Dependence of Atomic Oxygen Undercutting of Protected Polymers," NASA TM 1998207423, Fourth International Space Conference, ICPMSE-4, Toronto, Canada, April 23-24, 1998. 5. Snyder A. and Banks B., “Fast Three-Dimensional Method of Modeling Atomic Oxygen Undercutting of Protected Polymers,” Sixth International Conference on “Protection of Materials and Structures from Space Environment”, Toronto, Canada, May 1-3, 2002.

This page intentionally left blank

EFFECT OF SPACE GASEOUS ENVIRONMENT ON THE THERMOPHYSICAL PROPERTIES OF MATERIALS AND STRUCTURES E. LITOVSKY, J. I. KLEIMAN Integrity Testing Laboratory Inc., 80 Esna Park Drive, Units 7-9, Markham, Ontario, L3R 2R7, Canada; Tel. 905 415-2207, Fax. 905 415-3633, N. MENN LUMINOS Computerized Systems Ltd., Carmel Business Park, Tavor Bldg., Tnufa Str. 3, P.O. Box 104,Tirat Carmel, 39101, Israel 1. 0 Introduction Design and weight of aerospace and space apparatus, energy consumption, temperature regimes and reliability are connected with thermal physical properties of applied materials. Thermal physical properties such as thermal conductivity and thermal diffusivity depend strongly on the environment the materials are used in and on the microstructure of materials. The thermophysical properties of materials under the influence of environment may change 2-10 and even more times. The investigation of these properties in conditions of space where high vacuum, drastic temperature changes, outgassing problems, irradiation, etc, remains an acute problem to which not enough attention is paid. To shed more light onto the behavior of thermal conductivity and diffusivity of materials in conditions mentioned above, heat barrier resistance behavior of interfaces containing two major types of materials, i.e. densified and with microcracks, and highly porous insulation materials were investigated at normal pressures and in vacuum condition. The contact heat barrier resistances between mechanical contacting bodies we consider as partial case of the micro cracks analyses. Experimental and theoretical work was conducted and mathematical models were developed allowing estimating the influence of environment and the structure of the material on thermal physical properties of the above materials and structures [1-13]. This paper presents a short review of our current work devoted to solutions of these problems. 2. 0 Basic Theoretical Models In order to understand the conducting experiments and the results described below in section 4, a very brief introduction of the developed basic theoretical models that described the thermal behavior of materials and pores in various environments is given first.

245

246 2.1 APPARENT THERMAL CONDUCTIVITY OF MATERIALS AND PORES Equation (1) below provides the basis for estimations of relative influence of heat transfer mechanisms on apparent thermal conductivity λapp that is governed by all mechanisms of heat transfer [1, 4]:

λapp = λs (1 – Π)3/2 M + λΠΠ1/4 + λrad + λconv

(1)

λs is thermal conduction of the solid phase, λrad, λconv – are radiative and gas convection components respectively, M is a function taking into account the contact heat barrier resistance (HBR) of cracks between grains. λconv =0 in our case. The apparent thermal conductivity of pores in equation (1) can be written as

λΠ = λgas + λem + λs-d

(2)

λgas in Eq.(2) accounts for heat transfer through the gas filling the pores either in a continuum or a free molecular regime; λem denotes the contribution of gas emission due to chemical reactions, evaporation or sublimation, occurring within pores and λs-d denotes the contribution of segregation and surface diffusion of impurities. The quantity λΠ,̓defined in Eq. (1) affects also the HBR coefficient M appearing in Eq.(1). Each of the additive components in Eqs.(1) and (2) is considered below. At atmospheric and higher pressures, the pore size normally greatly exceeds the molecular mean free path and characteristic Knudsen number Kn<<1 and λgas is equal to the molecular gas thermal conductivity λgas, atm. If Kn ≥1, λgas will dependent on the pore size and can be defined as follows [6]: (3) where B is given by

Here γ is the specific heat ratio, Pr is the gas Prandtl number and A is the gas-solid accommodation coefficient [2]. An example of air thermal conductivity dependence on pressure is presented in Figure.1. When the thermal conductivity (diffusivity) of the material is measured in unsteady heating regimes, several physical-chemical processes, in particular, chemical reactions, may occur on the surface of the pore. These reactions involve various impurities, normally present on the pore surfaces of industrial materials. Emission and absorption of the impurities into and from the gas phase usually accompany some of these reactions. Since the opposite sides of the pores (along the direction of the temperature gradient) have different temperatures, the saturation pressures of the active gases in these points are different. This leads to intra-pore transport of the impurities.

247 This process is accompanied by a concomitant heat flux across the pore, which contributes to the total heat transferred through the material. The heat flux arising from the gas emission in pores may be expressed via the corresponding component of the pore phase conductivity, λem valid for arbitrary gas pressures (Kn numbers) [10]: (4)

(5) and r is the coefficient of reflection of gaseous molecules, pA is the saturation pressure of the "active" (emitted) gas, ∆Hem is the molar enthalpy of the considered physicalchemical processes, R is the gas constant, µ ҏis the molecular mass of the active gas, δ is the thickness of pores (or cracks), p is the total gas pressure in the cracks, D is the diffusion coefficient of the active gas through the "passive" gas (which fills the vacuum chamber and pores, and does not participate in the reactions). Heat may be transferred across the pores by diffusion in solid of various kinds of impurities (or crystal lattice defects). These impurities tend to segregate at the pore surfaces. The equilibrium surface concentration of the segregated substances is mainly determined by the local surface temperature. Therefore, a temperature gradient in porous material causes a mass flow of the impurities along the pore surface. Heat is released (on the hotter side of the pore) during bulk-to surface segregation of impurities and consumed (on the cooler side) during their surface-to bulk desegregation. This results in a net heat transport through the pore. The effect of this mechanism may be expressed via the corresponding component λΠ,s of the pore phase conductivity [2, 6]: (6) where

Here C0 is the equilibrium concentration of impurities in the bulk phase, Ds, * D are the surface and the grain boundary diffusivities, respectively, δs is the thickness of the segregated layer on the pore surface, ρ is the density of the solid phase, a is the radius of a spherical pore, ∆Hs is the molar enthalpy of the segregation process and R is the universal gas constant.

248 2.2 HEAT BARRIER RESISTANCES (HBR) Heat barrier resistance (HBR) phenomena are present normally in regions of mechanical contacts of different components. In space vacuum conditions their role is especially important. HBR in porous materials arise from different technological and physicalchemical factors in the course of material fabrication and usage. They are normally associated with discontinuities of different sizes, i.e., grain boundaries, cracks resulting from pressing or kilning processes, material application, etc. The basic structure of the discontinuities in porous materials can repeat itself on different length scales. Smaller scales describe structural elements of higher orders (or generations), elements of which are effectively included within the larger scale elements (of lower generations). Accordingly, a generalized geometric fractal model of the material microstructure had been developed [1, 6]:

Μ = Fout!,

(7)

where Mj is the HBR parameter of j-th order structure and can be expressed as follows.

Mj = Fout!, -2 Φ× a, = 1 – Fout![I1(nπr/L) K1(nπb/L) –K1(nπr/L) I1(nπb/L)]

(8)

In the above

RΠ = Fout!= Fout!and

Rb = Fout!= Fout!

are the non-dimensional thermal resistances of the crack and the contact layer of intergrain material, respectively, where δ is the micro-cracks thickness, L is the distance between the micro-cracks, b and r are the effective radii of the contact area and of the micro-crack (or grain boundary), a, =b/r, λb is the thermal conductivity of the intergrain material, and I1 and K1 are the modified Bessel functions of the first and the second kind, respectively. In our model the thermal resistance of the gas in Rg is determined as the non dimensional thermal resistance of pores, gaps, microcracks Rg

249 249 and RȆ = Rg

250 Fourier’s equation that does not take into account that materials could be transparent to heat radiation. Application of a number of modern ASTM standards (ASTM E1225 for plate steady state method of measurement of apparent thermal conductivity, ASTM E1461 for Flash Method) to semitransparent materials is therefore limited. Similar limitations apply in application of ASTM C-1113 (hot wire method) and a number of other ASTM and ISO standards. Most of the instruments measure only the apparent thermal conductivity or diffusivity. Considerable errors in calculation of temperature fields using the apparent properties and the Fourier’s equation instead of RCHT theory were shown to exist in a number of investigations [12-13]. Therefore, development of new experimental methods and instruments for measurement of true thermal physical properties of semitransparent materials remains an acute problem in thermal physics. An expansion of measurement possibilities should be based on radiative conductive heat transfer (RCHT) theory rather than differential Fourier’s equations. The required basic information in the RCHT approach includes the apparent thermal conductivity/diffusivity and certain optical characteristics of the material. Based on the discussed above principles, a system for determination of the apparent, the radiative, and the conductive thermal conductivity and diffusivity and optical characteristics of radiation attenuation and scattering in semitransparent materials was developed in our group. The details on the theory and the developed instruments one can find in [12-13].

4.0 Examples of Experimental Data and Their Analysis Thermal conductivities of the majority of ceramic crystalline oxide materials follow the Eucken law at atmospheric pressure, however in vacuum (pressures below 0.1 kPa) this law is not observed as a rule. Instead, λ either grows or changes nonmonotonically with temperature. Typical results are shown in Figs. 1-3. To summarize these results below we formulate the most important effects which were analyzed and discussed. 1. Strong gas pressure dependence of λ at low temperatures of most ceramic materials. 2. Independence of λ on p at low temperatures for dense materials without cracks. 3. Pressure dependence of λ of materials, collected at high temperaturesҏ differ markedly from the comparable data collected for low temperatures. 4. Temperature dependence of λ of many ceramic materials at low gas pressure (space conditions) is qualitatively different from that observed at atmospheric pressure.

Relative air thermal conductivity

251 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.01

0.1

1micron 10micron 100micron 1000micron

1

10

100

1000 10000 1E+0 5

P, Gas pressure, Pa

Figure 1. Gas pressure dependence of relative air thermal conductivity λp / λatm in pores/gaps of different thickness

Figure 2. The geometric model of heat barrier resistance. The HBR parameter M vs relative gas resistance

252 The main mechanism that explains the above effects is a decreasing λΠ,g with decreasing p, especially in the transition and free-molecular regimes (where Kn may be close to 1). Estimates show that for thickness δ of about 1µm, λΠ,g is dominated by the pressure term in (14) and decreases markedly with p decreasing below 105 Pa. For 100 µm pores λΠ,g begins to significantly decrease with gas pressure below about 103

Thermal conductivity, λ

Pa. Fine-grained pure oxide materials are normally characterized by very small nanometer size pores, in which Kn reaches unity already at atmospheric pressure. As can be seen from Figures 1-3 an increase in contact heat barrier resistance in vacuum causes a decrease of in the value of the apparent thermal conductivity of materials and an increase of contact heat resistance between contacting surfaces. At high temperatures this effect is negligible. More data and a more detailed explanation for the phenomenon one can find in [1-11].

W mK 5 4 3 2 1 0 200

400

600

800

1000

Temperature, C Figure 3. Thermal conductivity of alumina ceramics with porosity 24%. In Hydrogen, atmospheric pressure (upper open circles), in air, atmospheric pressure (triangles), and in vacuum 10-2 Pa (solid circles).

In summary two main groups of mechanisms are considered to explain these effects: 1. Heterogeneous heat and mass transfer processes occurring in pores existing at grain boundaries and in cracks. In particular, surface segregation and diffusion of impurities on pore surfaces, transport of gases produced from chemical reactions, evaporation, and sublimation.

253 2. Micro-structural changes due to non-uniform thermal expansion of particles and grains. These changes are caused by mismatch between the thermal expansion coefficients of different material phases, and anisotropic thermal expansion of crystals. For highly vacuum and low temperatures characteristic for space conditions, the main heat transfer is determined by HBR values. Therefore composition and structure of contact area can influence strongly on apparent thermal physical properties. Changes of HBR in ceramics due to space irradiation is the subject of our current investigations. The introduced structural parameters of our HBR models when combine with and modern techniques for surface structure analyses open new ways for mathematical modeling of the influence of heat radiation on thermal physical properties. Figure 4 demonstrates relationships between the heat radiation of the quartz glass fiber refractory materials and their porous structure. As can be seen from Figure 4 figure for T>1200 K, λ of materials with fiber diameter of 15 µm rapidly increases with T, whereas for materials containing 3 µm fibers, this effect is less pronounced. In space vacuum conditions the relative influence of heat radiation on apparent properties of insulation materials increases strongly.

W mK 0.4

Thermal conductivity, λ

1 0.3 2

0.2

3 0.1

0 500

1000 Temperature, K

1500

Figure 4. Thermal conductivity of quartz glass fiber refractories. Experimental data [13] and curves calculated from Eqs. (10), (26), (27), (30). , 1 - d=15 µm, Π=79.5%; , 2 - d=15 µm, Π=94.2%; , 3 - d=3 µm, Π=94.2%.

254 The latest development in our group and world wide of using other methods based on the combine radiative-conductive heat transfer theory (RCHT) can be found in [11, 13]. 5. 0 Conclusions The influence of the mechanisms on the apparent thermal conductivity in space conditions can be estimated on the basis of developed models of apparent thermal conductivity and the combination of classical and novel heat transfer mechanisms. The contact HBR’s are the main structural elements determining changes of thermal conductivity and diffusivity of materials in space conditions. The described models and their analysis can have important technological and metrological applications in: selection and development of materials with predominant properties; understanding of scattering in thermophysical properties in experimental and reference data; development and selection of space materials and measurement methods. 6. 0 Acknowledgements We wish to thank the Canada Israel Industrial Research and Development Fund (CIRDF) for supporting part of the work described here under a contract ITL-CIRDF9801. 7. 0 References 1. E.Ya. Litovsky, M. Shapiro. Gas Pressure and Temperature Dependences of Thermal Conductivity Porous Ceramic Materials. Part I. Ceramics with porosity below 30%, J. American Ceram Soc. 12, pp. 3425-3439, (1992). 2. T. Gambaryan, E. Litovsky, M. Shapiro. Influence of Segregation - Diffusions Processes on the Effective Thermal Conductivity of Porous Ceramics, International Journal of Heat and Mass Transfer, 1993, v.36, 17, pp 4123-4131. 3. T. Litovsky, R. Cytermann, E. Litovsky, M. Shapiro, A. Shavit., Thermal conductivity of high temperature ceramic thermal insulations, Advances in Porous Materials, Material Research Society Proceedings, v.371, Editors S. Komarneni, pp.309-314, 1995, USA. 4. Litovsky E., M. Shapiro, A. Shavit, “Gas pressure and temperature dependences of thermal conductivity porous ceramic materials. Part II. Ceramics with porosity above 30%”, J. American Ceram. Soc., 79/5, 1996 1366-1376. 5. E. Litovsky, T. Litovsky, M. Shapiro, A. Shavit., "Thermal conductivity dependence of MgO heat insulations on porosity in temperature range 500-2000K" Third Symposium on Insulation Materials: Testing and Applications, ASTM, 1997, Quebec, Canada, 292-306. 6. Litovsky E., Gambaryan-Roisman T., Shapiro M. and Shavit A, Heat Transfer Mechanisms Governing Thermal Conductivity of Porous Ceramics, Trends in Heat, Mass & Momentum Transfer, Research Trends, vol 3, pp 147-167, 1997.

255 7. Litovsky, E. “Effect of Gas Pressure on Heat Transfer in Isulating Materials”, Report of corporation, Sweden, 1997

the ABB

8. Litovsky E., T. Gambaryan-Roisman, M. Shapiro, A. Shavit, “Effect of Grain Thermal Expansion Mismatch on Thermal Conductivity of Porous Ceramics”, J. American Ceram. Soc. , vol 82, No 4, pp 994-1000, 1999. 9. Litovsky E., Litovsky T., Shapiro M., Shavit A., "Thermal conductivity dependence of MgO heat insulation on porosity in the range 500-2000K" Third Symposium on Insulation Materials: Testing and Applications, ASTM STP 1320, R. S. Graves and R.R. Zarr, eds. Quebec, Canada, 1997, pp. 292-306. 10. Litovsky E., Gambaryan-Roisman T., Shapiro M., Shavit A., “Novel heat transfer mechanisms in porous ceramic materials”, High Temperatures-High Pressures, 1(33): 2001, pp. 27-34. 11. Litovsky, E., J. Kleiman, and N. Menn, Measurement of Thermal Physical and Optical Properties of th Semitransparent Materials in the Temperature Range 20 -1800C, 26 International Thermal Conductivity Conference, MA, USA, 2001. 12.. Litovsky, E., J. Kleiman, Relative influence of heat transfer mechanisms on apparent thermal th conductivity of porous materials, 26 International Thermal Conductivity Conference, MA, USA, 2001. 13. Litovsky, E., Kleiman J., and Menn N., “Thermal Physical and Optical Properties of the Fiber o Insulation Materials in the Temperature Range 200 – 1800 C,” Insulation Materials: Testing and th Applications, 4 Volume ASTM STP 1426, A. O. Desjarlais, Ed., ASTM, PA, 2002 (in print).

This page intentionally left blank

CLEANLINESS SUPPORT OF THE LAUNCH VEHICLE FOR PUTTING INTO ORBIT THE SPACECRAFT METEOR-3M WITH THE "SAGE-III" INSTRUMENT V.G.SITALO, V.G.TYKHYY, L.P. POTAPOVYCH Yuzhnoye State Design Office, Dniepropetrovsk, Ukraine Abstract Care should be taken in order to avoid any contamination sources in the design and construction of the fairings. Such sources as the outgassing of organic materials used in the design including fiberglass, heat-shielded fiberglass, etc. as well as the contamination with particulates always remains a problem when dealing with cleanliness issues. A special production process was designed to maintain the required cleanliness levels on the fabrication stage of the fairing. The developed process provided for the cleaning of separate parts and units and preservation of the reached level of cleanliness during the transportation and in-process storage. The requirements to the cleanliness and the developed methodology are described in this paper. 1.0 Introduction On 10 December 2001 the Ukrainian rocket Zenit-2 put into orbit the Russian spacecraft Meteor-3M with the American instrument SAGE-III, placed on it. A fairing of the rocket was made up of the aluminum shell and the internal thermal insulation involving glass fiber plastics, coated by an aluminum foil. The cleanliness requirements for modern spacecrafts specify not only the spacecraft itself but also the fairing used during the design, manufacturing, and transportation and launching The fairing shall meet the following cleanliness requirements: • the clean room and the fairing interior are maintained as a class M6.5 clean environment; • molecular contamination shall be lower than 20 mg/m2; • particulate contamination shall not exceed the level 500. The need for weight reduction of spacecraft has led structural engineers to the extensive use of composites. The outgassing of such materials is well known [1]. Volatile products of outgassing create a contamination atmosphere around a spacecraft. The volatiles may condense on elements of the spacecraft - particularly on elements of solar arrays - and degrade their performance. The process of condensation may apply also to the outgassing of fairing materials during spacecraft storage on the launching pad, during

257

258 launch into orbit, and during operational orbit time. It is essential to control this process and to be able to undertake appropriate steps leading to a minimum interference with the cleanliness of the whole system. Work was conducted in past to examine the outgassing of non-metallic materials of fairing and spacecraft to a condition that makes them adequate for operation and launch. A number of fairing materials were investigated including fiberglass and heatshielded fiberglass, rubber, glues, lacquers and cables, as well as carbon plastics and glues [1]. Analysis of the accumulated data had shown that the majority of materials used in design and manufacturing of fairing meets the requirements concerning the outgassing, with the exception of some glues, greases and insulating tapes [2]. Analyzing the fairing production process it was possible to confirm, in general, that they conform to the modern cleanliness requirements that provide the necessary level of quality for launching services [2]. To guarantee the given cleanliness requirements on the fabrication stage, a special production process was designed. It allowed for cleaning of separate parts and units and for preserving of the reached levels of cleanliness during intershop transportation and in-process storage. 2.0 Experimental All units containing nonmetallic materials were exposed to thermal vacuum treatment. The assembled fairing was submitted to vibration, washing with high pressure water jet and vacuum treatment. The percentage of the particulate contamination collected by various cleaning methods is distributed as following: • first vacuum cleaning - 31%; • vibration cleaning - 16%; • second vacuum cleaning - 23%; • water jet washing - 30%. To control the quality of the water jet washing process, the dry residue on the filter left after the wash process was weighed and kept below the required 10 mg level. At transportation the fairing was packed up into a dust-proof bag made of an antistatic film. On the preparation stage to launching, the preservation of the reached level of cleanliness was provided through the installation of technological caps on harness access windows during pneumatic and electric tests, by purging with clean air, and conducting all activities in the clean room when mounting the METEOR - 3Ɇ with the SAGE-III to the launcher and mating the fairing with the launcher. The actual level of air cleanliness in the clean room during the operation did not exceed M 5.5. A detailed measurement program was implemented during all activities for an effective control of the cleanliness. At the final stage of the work at the plant and the launch site, Ukrainian and American specialists conducted a joint verification control of the cleanliness of the fairing.

259 3.0 Results and Discussion Figure 1 shows the general view of the internal surface of the fairing.

Figure 1. General view of the internal surface of the fairing.

Figure 2. Demonstration of a wipe test on the internal surface of the fairing to measure the molecular contamination

260 The process of application of a wipe test for molecular contamination analysis as performed on the launch site at Baykonur is shown in Fig. 2. If the results from the test didn't conform to the required level of cleanliness, an additional round of cleaning was performed. The results of the particulate and molecular contamination control are shown in tables 1 and 2 at different fabrication and exploitation stages of a rocket. The quantity of the residual mechanical contamination collected from the surfaces was verified by a “Tape Lift” method. The effective control was carried out with the help of membranes “Vladipor” and was based on the property of particles to stick to a surface of a membrane. Using a microscope, the numbers of particles and fibers were calculated and their size and the material established. The weight of the mechanical residual contaminations was determined with the help of the MCAD program. The control of residual molecular contamination was conducted by wiping the inside surface of the fairing. To qualitatively estimate the composition of the molecular contamination the Fourier Transform Infrared FTIR spectrometry analysis was conducted. The results of FTIR analysis conducted at the Langley Research Center confirmed presence of silicones on the internal surfaces of the thermal insulation. It is well known that silicones form very stable compounds not easily eroded by the atomic oxygen in LEO orbits [3,4] and, as a result, may cause unwanted deposits on surfaces of sensors and optical windows of spacecrafts and instruments during the flight. Their presence, therefore, even in minimum quantities on surfaces of the fairing is not permitted. Additional cleaning of the fairing surface and the repeated control of the quantity and qualitative composition of the molecular contamination was conducted. As a result of such an approach, the silicones were eliminated completely from the surfaces. In an attempt to identify the source of silicones found among the contaminants the initial materials used in fabrications were analyzed. The analysis of different materials used in the production process as well as different lubricants indicated that one of the lubricants, namely the KO-21 that was used in moulds to produce rigid thermal insulation from foil, contained silicones. Figure 3 shows a typical FTRI spectra obtained from one of KO-21 samples. As can be seen from Fig. 3, the spectrum of the lubricant contains vibrational bands in the range 1250-750 ɫm-1, where vibrational bands from silicon-based molecules are located.

Absorbance units

1,0 0,8 0,6 0,4 0,2 0,0 4000

3500

3000

2500

2000

1500

1000

500

Wave number

Figure 3. FTIR Spectrum of the Ʉ-21 lubricant, obtained at the Institute of Physics of the National Academy of Sciences of Ukraine.

adapter

after

Internal surface of the fairing thermal insulation after transportation to launch site at Baykonur Surface of the adapter after transportation to launch site at Baykonur Surface of butt end of the launch vehicle after transportation to launch site at Baykonur

Surface of the manufacturing

Specification Internal surface of the fairing before the installation of a thermal insulation Internal surface of the fairing thermal insulation after the installation Internal surface of the fairing thermal insulation after the vacuum treatment of the fairing

Control zone

2 1 3 8 9 5

6 2 4 8 2 2

624 208 416 832 208 208

123 18 14 17 11 5 1 1 1 1

208 104 312 832 936 520

12800 1874 1458 1771 1144 520 104 104 104 104

≥ 50 µm On On 0.1 membrane m2

104 104 104 312 104

104 104 104 416 104

1190 208 208 208 208 208 104 104 104 -

1 1 1 3 1

1 1 1 4 1

12 2 2 2 2 2 1 1 1 -

-

-

28.1 -

-

-

0.27 -

Particles quantity according MIL-STD-1246 ≥ 100 µm ≥ 250 µm On 0.1 On On 0.1 On m2 membrane m2 membrane

-

-

0.01 -

1.08 -

≥ 500 µm On 0.1 On m2 membrane

4.2 1.6 2.5 3.4 1.0 1.2

1.5 1.5 2.5 5.1 6.5 2.9

Weight of particulate contamination (with allowance for shape and density of particles), mg/m2 10 4.0 1.1 1.7 9.2 6.5 4.5 4.1 4.1 5.1

TABLE 1. Control data residual quantity of particulate contamination on fairing space surface by filter paper “Vladipor”

261

262 TABLE 2. Control data of residual molecular contamination of the fairing space surfaces by wiping Control zone Molecular contamination quantity, mg/m2 20 Specification Internal surface of the fairing thermal insulation after 9.58 the installation and cleaning 6.98 1.33 Internal surface of the fairing thermal insulation after transportation to launch site at Baykonur 12.25 Surface of the adapter after transportation to launch site at Baykonur 16.70

4.0 Conclusions Based on the optical analysis and FTIR analysis of the samples collected during the cleaning and conducted at different stages of the cleaning process during the manufacturing and handling of the fairing and the launcher It can be suggested that the developed and used cleaning technology complies fully with the stringent requirements of cleanliness as applied to the fairing. 5.0 Acknowledgments The authors want to acknowledge the help of scientists from the NASA Langley Research Center and the Institute of Physics of the National Academy of Sciences of Ukraine in conducting the FTIR analysis 6.0 References 1.

V.G. Tikhii, “About some aspects of changing optical properties of glass in solar arrays and other space materials on exposure to LEO space environment”, eds. J.I. Kleiman and R.C. Tennyson, in: Proceedings of the 3rd International Space Conference, ICPMSE-3, Toronto, Canada, April 25-26, 1996, pp. 203-206.

2.

V.G. Sitalo, V. G. Tikhiy, “Cleanliness support of vehicle fairing for spacecrafts launch” in: Protection of Materials from the Space Environment, eds. J.I. Kleiman and R.C. Tennyson, Proceedings of the 4th International Conference, ICPMSE-4, Toronto, Canada, April 23-24, 1998, pp. 95-102.

3.

J. Kleiman, “Surface Modification of Polymers Used in the Low Earth Orbit Space Environment”, Metallized Plastics 5/6, Fundamental and Applied Science, ed. K. Mittal, (1998) 331-351.

4.

J.I. Kleiman, Z.A. Iskanderova , Y.I. Gudimenko, W.D. Morison, and R.C. Tennyson, “Polymers and Composites in the Low Earth Orbit Space Environment: Interaction and Protection”, Canadian Aeronautic and Space Journal, v.45, No.2, (June 1999), 148-160.

IRREVERSIBLE SHRINKAGE EFFECTS OF CARBON FIBERS IN POLYMER MATRIX COMPOSITES EXPOSED TO THE "MIR" SPACE ENVIRONMENT O.V. STARTSEV AND D..A. KHRISTOFOROV Altai State University, 656099 Barnaul, Russia V.V. ISSOUPOV ONERA/DESP, 31055 Toulouse, France E.F. NIKISHIN M.V. Khrunichev State Research and Production Space Center, 121087 Moscow, Russia A.F. RUMYANTSEV All-Russian Institute of Aviation Materials, 107005 Moscow, Russia

1. 0 Introduction More than 20 years ago, M.V. Khrunichev State Research and Production Space Center (Moscow) initiated a comprehensive study of carbon/epoxy, glass/epoxy composites and nonmetallic materials after long-term exposure onboard SALYUT and MIR space stations [1]. A complex of employed physical experimental methods has convinced that properties of the materials during long-term staying in LEO environment are affected by physical, chemical and morphological transformations in the epoxy matrix. Extensive examples and experimental results are reported in [1-3]. Carbon fibers is the most promising reinforcement component of thermoplastic and thermoset matrices to attain high mechanical performance required from composite materials of spacecraft structures [4]. The fibers are assumed to have stable properties and outstanding resistance under the most severe working conditions (mechanical stresses,

263

264 humid and corrosive environments, radiation, raised temperature, thermal cycling, etc.) [4-6]. This is the reason why carbon fibers properties were believed to be unchanged after long-term exposure to outer space and were not specially examined. Nevertheless, instability is in the very nature of the carbon fibers. Modern carbon fibers are produced by thermal treatment of oriented polymeric fibers [4,6], extra stresses are created to stretch the fibers on purpose to achieve a higher elasticity modulus. Carbon fiber structure and properties are formed and frozen in a mechanical stress-strain field. Therefore, not expansion but shrinkage effects are usually observed when heating carbon fibers and carbon/epoxy composite materials [7-10]. It is exactly this structural non-equilibrium caused by the manufacturing procedure which is the reason for negative thermal expansion coefficient (CTE) of carbon fibers. As suggested in [5,7], thermal deformations on carbon fibers are believed to be reversible due to the ability of the fibers to keep the intrinsic non-equilibrium at heating and cooling. However, this is not proved experimentally because of the insufficient sensitivity of conventional methods for measuring thermal expansion of fibers with low CTE values [4-7]. Therefore, reliable experimental data on thermal deformation of carbon fibers undergoing repeated heating represent a particular interest. 2. 0 Experimental An experiment aimed at precise measurement of the thermal expansion of carbon fibers and carbon/epoxy composites was performed with the help of a developed linear dilatometer [11]. A distinctive feature of this instrument is the possibility of continuous specimen length monitoring when heating or cooling at a rate of 1°C/min. The realized principal of absolute thermal expansion control permits to measure the CTE of carbon fibers specimens prepared in the form of a bunch, a tape or a bar with an experimental incertitude of 0.5⋅10-7 °C-1. The absolute expansion was controlled in each of 5 successive scans performed on the same specimen including heating and cooling between 20° and 350-400°C. An assembly drawing of the optical and thermal parts of the dilatometer instrument is shown in Fig. 1, the technical parameters are given in Table 1. A detailed description of the experimental procedure and data treatment is considered in [11].

265

Figure 1. Assembly drawing of the optical and thermal parts of the linear dilatometer 1 – thermal chamber, 2 – nichrome heating coil, 3 – asbestos, 4 – quartz tubes, 5 – specimen, 6 – thermocouple, 7 – inertial part, 8 – optical half-plane, 9 – frame, 10 – liquid container, 11 – incandescent lamp, 12 – collimator, 13 – shadow stop, 14 – lens, 15 – multiple-unit photo-detector, 16 – metallic stoppers TABLE 1. Basic technical parameters of the linear dilatometer Resolving power Interval of registered displacements Linearity of the optical registration system Temperature regulation under transient conditions Maximal overheating in the chamber (rate of 10ɨɋ/min.) Effective temperature range Time before temperature stabilization (rate of 10ɨɋ/min.)

0.5 µm 1600 µm ≤ 1.5 % ≤ 0.2ɨɋ ≤ 0.6ɨɋ –120° to +400ɨɋ ≤ 90 s

3. 0 Results and Discussion Typical thermo-grams for T-300JB, LJU-35, LU-24P carbon fibers, unidirectional composites reinforced with these fibers and VS-2526 epoxy matrix [4] are shown in Fig. 2, 3.

266

∆L/L0⋅10

-4

0.0

1

-1.0

2

-2.0 3

-3.0

-4.0 0

50

100

150

200

250

300

350

400

o

T, C

Figure 2. Temperature variation of the relative thermal expansion of T-300JB (1), LJU-35 (2) and LU-24P (3) carbon fibers

As seen from these figures, the CTE of carbon fibers varies in a wide temperature range as a function of the fiber trademark, manufacturing procedure, achieved elasticity modulus, etc. Note that the common feature of the measured carbon fibers is always a negative CTE. Thermal deformations of each of the represented fibers (Fig. 2) were measured by making 5 repeated scans between 20° and 350°-400°C. Negative thermal deformations decreased in every case when the number of thermal cycles was increasing, the effect was certainly above the experimental error. Fig. 4 is a typical example of temperature variations of the relative thermal expansion and CTE of LU-24P carbon fibers during the first and the fifth scan.

267

∆L/L0⋅10

-4

0.0 2

-1.0 -2.0

1 3

-3.0 -4.0 -5.0 50

100

150

200

o

T, C

Figure 3. Temperature variation of the relative thermal expansion of carbon/epoxy composites based on VS-2526 epoxy matrix and T-300JB (1), LJU-35 (2) and LU-24P (3) carbon fibers

Some quantitative results on influence of the repeated heating upon the thermal expansion of the fibers are reported in Table 2. TABLE 2. Linear thermal expansion coefficients of carbon fibers during repeated heating measured at 50° and at 250°C Fiber

CTE at 50°ɋ as a function of the heating serial number (10-7 °ɋ-1) 1 2 3 4 5

CTE at 250° as a function of the heating serial number (10-7 °ɋ-1) 1 2 3 4 5

T-300JB

-6.1

-4.3

-4.2

-3.6

-3.2

1.0

2.5

2.9

4.3

4.6

LJU-35

-7.5

-6.8

-6.6

-6.5

-6.0

-3.5

-3.0

-2.5

-2.2

-2.0

LU-24P

-18

-15

-14

-12

-11

-7.8

-6.6

-6.5

-5.6

-5.6

268

∆L/L0⋅10

-4

(ɚ)

0.0 -1.0

2

-2.0

1

-3.0 -4.0 0 -6

α⋅10 , K

100

200

300

o

400 T, C

-1

-0.5

(ɛ) 2

-1.0

-1.5 1

-2.0 0

100

200

300

o

400 T, C

Figure 4. Temperature variation of the relative thermal expansion of (a) and CTE (b) of LU-24P carbon fibers during the first and the fifth scan

Analysis of the whole experimental data set has shown that decrease in the thermal deformations is different in the temperature interval explored. For instance (Fig. 4), the CTE value of LU-24P carbon fibers goes up by 40 % at 50°C but only by 20 % at 250°C. The same effect is detected for the other fibers. In some cases, like for T-300JB fibers at elevated temperatures, the CTE can change the sign as the number of thermal cycles increases but keeps the upward trend. Therefore, the experimental results are not related to thermal oxidation of the fibers but confirm the assumption about thermal relaxation of the carbon fibers structural non-equilibrium. Because the number of performed measurements is still not big enough, the discovered effect has not been characterized quantitatively. The strict confirmation needs additional measurements to be conducted when the amplitude and the number of thermal cycles would be varied. However, the importance of such a characterization may be illustrated by the following example. Specimens of KMU-4l composite material based on LU-P carbon fiber

269 and ENFB epoxy matrix were exposed to a LEO space environment onboard MIR orbital station. Detailed information about the exposure is given in the references [1,2]. Considering the acting factors of the space environment, it was substantiated that the most important influence of the space exposure is due to a vacuum of 10-4 Pa and thermal cycling from −100° to +125°C with a period of 1.5 hours. The influence of these factors consists in postcuring of ENFB epoxy matrix collaborated with an increase in the material glass transition temperature which was measured by Dynamic Mechanical Analysis [1] (Table 3). TABLE 3. Evolution of the glass transition temperature Tg of KMU-4l carbon/epoxy composite as a function of the time of exposure to LEO space environments Time of exposure (days) Increase in Tg (°C)

102

456

839

1024

1218

1501

8

19

31

39

39

37

At the same time, thermal expansion measurements indicated that thermal cycling in LEO affected the CTE of the carbon/epoxy composite. Fig. 5 shows temperature variations of the relative thermal expansion of KMU-4l in the initial state and after 839, 1024 and 1218 days of exposure. It turns out to be that the important number of thermal cycles estimated in 13000 to 19000 with relatively a weak amplitude provokes significant relaxation in the carbon fiber. Specimens exposed to LEO thermal cycling have almost lost the ability to contract when the temperature increases. The change in the deformability of KMU-4l composite material is caused by a change in the properties of the fibers and not in those of the epoxy matrix [1-3]. ∆L/L0⋅10

-3

0.5 2

3

0.0 4

-0.5 -1.0

1

-1.5 -2.0 50

100

150

200

250

o

T, C

Figure 5. Temperature variation of the relative thermal expansion of KMU-4l carbon/epoxy composite in the initial state (1) and after exposure to space for 839 (2), 1024 (3) and 1218 (4) days

270 4.0 Conclusions The results of thermal expansion measurements prove the capacity of modern high-modulus carbon fibers to lower the level of their structural non-equilibrium and to decrease the value of negative thermal deformations under thermal cycling. This observation is important when forecasting the dimensional stability of structural elements made of carbon/epoxy composites designed for operation under variable temperatures. In particular, this study pointed out an important influence of low-amplitude thermal cycles met by materials in LEO not only upon polymer matrices properties but also upon the properties of LU-P carbon fiber provided that the number of thermal cycles is big enough. The discovered effect has to be taken into account when dimensioning composites-based space structures in order to meet strict requirements on their stability if long-term operation of the structures under LEO conditions will be involved. 5.0 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Startsev, O.V. and Nikishin, E.F. Ageing of polymeric composite materials in outer space, Mechanics of Composite Materials 29 (1993) 457-467. Startsev, O.V., Issoupov, V.V. and Nikishin, E.F. The diagnostics of micro-meteor protection for LEO stations from the results of the in-flight experiments, in Proc. of the Seventh International Symposium on Materials in a Space Environment, CNES, Toulouse, 1997, 367-373. Startsev, O.V., Issoupov, V.V. and Nikishin, E.F. The gradient of mechanical characteristics across the thickness of composite laminates after exposure in LEO environment, Polymer Composites 19 (1998), 36-42. Sorina, T.G. and Gunyaev, G.M., in: Polymer Matrix Composites, Shalin R.E. (ed.), Chapman & Hall, 1995, 132-198. Mostovoy, G.E., Kobets, L.P. and Frolov, V.I. Mech. Compos. Mater. 1 (1979) 27-33. Kalnin, L.I. Mech. Compos. Mater. 3 (1979) 397-406. Perepelkin, K.E. and Geller, A.B. Mech. Compos. Mater. 2 (1980) 350-353. Vishvanyuk, V.I., Alimov, V.T. and Vishnevsky, Z.N. Mech. Compos. Mater. 6 (1982) 1102-1104. Gurvich, M.R., Sbitnev, O.V., Sukhanov, A.V. and Lapotkin, V.A. Mech. Compos. Mater. 1 (1990) 32-36. Sukhanov, A.V., Lapotkin, V.A., Artemchuk, V.Ya. and Sobol, L.A. Mech. Compos. Mater. 4 (1990) 599604. Khristoforov, D.A., Klyushnichenko, A.B., Startsev, O.V., Suranov, A.Ya., Fizulov, B.G. and Startsev, V.O. in Proc. of the 2nd Int. Scientific and Technical Conf. "Experimental Methods in Physics of Heterogeneous Condensed Media", Altai State University, 2001, 186-194.

COMBINED EFFECT OF THERMAL AND MECHANICAL STRESSES ON THE VISCOELASTIC PROPERTIES OF A COMPOSITE MATERIAL FOR SPACE STRUCTURES V. ISSOUPOV AND O.V. STARTSEV Altai State University 656099 Barnaul, Russia C. LACABANNE , P. DEMONT L2P, CIRIMAT, Université Paul Sabatier, 31062 Toulouse, France V. VIEL-INGUIMBERT3, M. DINGUIRARD ONERA/Département Environnement SPatial, 31055 Toulouse, France E.F. NIKISHIN M.V. Khrunichev State Research and Production Space Center, 121087 Moscow, Russia

Abstract Thermal ageing of polymer-matrix composite materials of any long-life spacecraft in LEO involves evolution of mechanical, thermophysical and morphological properties, especially if the structure operates under external mechanical stress like bending. Behavior of KMU-4l carbon/epoxy composite material under combined effect of thermal profiles and static mechanical loading has been studied in a simulated LEO space environment. Thermal cycling was performed on specimens under bending stress reaching 30.4 % of the material ultimate strength. Specimens subjected to combined thermal and mechanical fatigue were examined in a series of 4 intermediate points with a maximum of 120 cycles. The Dynamic Mechanical Thermal Analysis has been employed to follow the global thermomechanical response of KMU-4l composite. The degradation produced was detected by moisture diffusion analysis, the crack density was estimated at the edges of the specimens using a scanning electron microscope. Effects such as the temperature dependence of material properties and matrix post-curing phenomena are also discussed. The results are compared with the appropriate data obtained on specimens of KMU-4l exposed for up to 1501 days to LEO space environments onboard SALYUT and MIR orbital stations.

271

272 1.0 Introduction Due to their outstanding qualities, polymer-matrix composites have gained wide use in space applications. The problem of their durability and performance in real space environments is being investigated by M.V. Khrunichev State Research and Production Space Center (Moscow, Russia), as well as by research departments of ONERA and CNES (France). This study may involve long-term exposure of materials to LEO onboard space stations and posterior laboratory examination of specimens brought back to earth. Materials response was characterized in our works [1-3] through assessment of changes taking place during space ageing by different physical methods like DMA, TMA, DSC, thermostimulated creep, SEM, as well as through erosion mass loss and matrix microhardness measurements. Evolution of physical and mechanical properties of the materials in a space environment has been found to happen mostly as a consequence of combined and competitive effects of erosion etching and thermal cycling [3-6]. The influence of harmful factors such as atomic oxygen and space debris on polymer-matrix based composite materials is observed in the form of matrix erosion etching, stripping of filler fibers and microdamage produced on uncoated surfaces. Another concern is the material bulk microcracking provoked by thermal cycling which can be also attributed to morphology and structural degradation. 2.0 Experimental 2.1. MATERIAL AND METHODS The material used in this study is KMU-4l (0°/90°) cross-ply laminated composite developed at All-Russian Institute of Aviation Materials (Moscow) [7]. It is produced from unidirectional prepreg tape and contains 9 plies of ENFB matrix reinforced with LU-3 carbon tape. The basic components of the matrix are EN-6 high-strength heatresistant resin: O O CH2 CH CH2

O O CH2 CH CH2

O O CH2 CH CH2

OH CH2

CH

CH2O

CH2

CH2

CH2

CH2

CH2

OCH2

CH

CH2 O

O O CH2 CH CH2 O

O CH2 CH CH2 O

and furfurylglycidylic ether (the major hardener): CH 2O

CH 2

CH 2

CH

O O

273 As against to epoxy matrices cured by amines, anhydrides or phenolformaldehyde oligomers, ENFB matrix is cured in autoclave by boron trifluoride curing agent BF3 added in a proportion of 4 %. The chemical composition determines the network molecular structure and main advantages of this matrix (improved stress-strain properties, high crack-resistance and glass transition temperature). Materials on the basis of ENFB matrix have high thermal stability combined with a low coefficient of thermal expansion [7]. In order to obtain the diffusion coefficient D and the maximum moisture content w∞, specimens were dried in a thermal chamber at 70±3°C until zero moisture content is reached. The specimens were weighed once per day to control the mass loss by moisture desorption. The material diffusion parameters result from mathematical treatment of data on the relative mass change with application of the second Fick’s law using FITTER non-linear regression tool [8]. The viscoelastic behavior of the composite was characterized by measurement of the dynamic bending modulus E’ and damping factor tan δ at a frequency of 2 Hz with a Polymer Laboratories Dynamic Mechanical Thermal Analyzer (PL-DMTA) in a temperature range between 20° and 250°C at 4°C/min. The material α-relaxation temperature Tα was deduced from the maximum of the variation of tan δ vs. T. Three repeated DMTA scans have been performed on each specimen. The experimental error of determining the above parameters was minimized to 3 % by data averaging over 3-5 identical specimens. 2.2. EQUIPMENT AND TEST PROCEDURE For the LEO simulation, only the effects of thermal cycling will be discussed. The testing consisted in exposure of KMU-4l specimens of 50×10×1.2 mm3 under mechanical bending stress to thermal environments (pressure less than 1.3⋅10-4 Pa) and followed temperature profiles designed to intensify and accelerate thermal ageing of the material. In this way, the exposure conditions resembled to those of materials of a space station in LEO as far as any space station structure is subjected to mechanical stresses, both static and dynamic. During the tests, the specimens were mounted on sample-holders allowing temperature variation by means of a thermal conditioning system (electrical resistance rods for heating and liquid nitrogen injection for cooling) inside the copper support. The temperature was precisely monitored with a GULTON WEST 2050 temperature programmer-controller. Several K-type thermocouples attached to the specimens provided feedback to the controller. The sample-holders, as described in [9], were specially designed to permit bending stress application and studying effects of thermal cycling combined with mechanical loading. In the configuration of thermal cycling, specimens are fixed at their extreme points on the sample-holders so that good thermal contact be ensured. The specimens surface is curved with a radius R and an apex angle α those of the sample-holder (Fig. 1). This case is the case of a rectangular girder under uniformly distributed load P which provokes displacement of each elementary unit called the arrow which is only x-

274 depended (Fig. 2). The three levels of the stress realized on the sample-holders are represented in Table 1 by the arrow f, deformation force F and bending stress σx as compared to the material ultimate strength σr. The solution of the elasticity problem as stated in Fig. 1 and 2a is given for an elementary unit G(x, y). Any cross-section of the specimen is affected by the bending moment Mz which determines the rotation ω = GZ axis and its displacement f G =

df Mz x = G of the unit G around the E' I z dx

M z 2 in the o y direction normally to the x 2 E' I z

medium deflection line. contraction/dilatation of the specimen left support

right support +σ

+σ

+σ

–σ

–σ

P

+σ –σ

f

+σ –σ

+σ –σ

P

P

+σ

–σ P

R Figure 1. Deformation of a specimen under vertical uniformly distributed load P

Y

Y

b right edge of the specimen under stress

εz > 0

G

Z MZ

a

G

h b

εz < 0

X

MZ

X

Figure 2. Rectangular girder under mechanical bending stress

For a rectangular girder (Fig. 2), the moment of inertia is given by the relationship E' I z E' b h 3 . The b h 3 , and the bending moment can be calculated as Iz = Mz = = 12 R 12 R

275 only non-zero component of the stress is σ = − M z y = − E' y , so the deformation of x Iz R the specimen can be expressed as εx = σx/E’, εy = εz = − f εx (Fig. 2b). A more detailed treatment of this development is given elsewhere [9,10]. TABLE 1. Mechanical bending stress applied to KMU-4l specimens on the sample-holders in the configuration of thermal cycling R (cm)

α (°)

f (mm)

σx/σr (%)

F (N)

−

0

0

0

−

37.5

9.2

1.2

26.0

15.2

18.7

18.4

2.4

52.1

30.4

The WEST 2050 controller was programmed to a desired temperature profile comprising a series of linear segments. A typical thermal cycle with a starting point at 0°C consisted of three stages; each stage consisted of a ramp where the set-point is being steadily increased or decreased and a dwell where the set-point is held constant for a definite time (Fig. 3). During thermal testing, the thermocouples were used to record the actual temperature profile. 160

Rate: 8 C/min. dwell at +160 C: 6 min. dwell at -160 C: 10 min.

Temperature ( C)

80

set-point value sample temperature

0

-80

-160 0

40

80 Time (min.)

120

160

Figure 3. Features of a temperature profile used for simulation of thermal cycling in LEO

A number of preliminary tests were performed in order to check the material response and sensitivity to thermal cycling. Two different thermal cycles were used in this study (Table 2). The first one was performed in the range of -90° and +130°C which is representative of a LEO exposure; the upper temperature limit of the second cycle was

276 +160°C to ensure that the material is not cycled above its glass transition temperature. Note that the Tg of KMU-4l composite measured on MDSC 2920 is 205±3°C. As seen from Fig. 3, the specimens soaked for 6 min. at the positive and for 10 min. at the negative target temperature to achieve thermal equilibrium with the sample-holder support. The ex-situ material properties examination has been made on specimens subjected to 14, 30, 60 and 120 thermal cycles and on those after 20 and 60 cycles combined with mechanical bending stress (Table 1) performed according to the experimental parameters represented in Table 2. TABLE 2. Thermal cycles parameters used for testing of KMU-4l carbon/epoxy composite Maximum temperature measured on the specimens (°C)

Average temperature rate on the specimens (°C/min.) heating

cooling

Test 1

-95 and +133

4.5

7.5

Test 2

-159 and +158

6.0

7.5

3. 0 Results and Discussion The response of KMU-4l composite material was characterized by determination of the microdamage resulting from thermal cycling and from the influence of the exposure on the thermomechanical properties. Typical PL-DMTA curves as a function of thermal ageing of the material are shown in Fig. 4 and Fig. 5. After exposure to LEO-like thermal cycling (Test 1) without mechanical stress, the properties remained constant even upon the maximum of 120 cycles. The testing did not significantly modify the modulus E’ or the position and shape of the α-transition peak (Figs. 4, 5 and 8). This is further confirmed by a dumping value tan δ which is unchanged or slightly higher at temperatures well above the material Tg. The effect of LEO space exposure on the viscoelastic properties of carbon/epoxy composites is known to involve the matrix postcuring process [1,3,11]. With the purpose of accelerating the thermal effect of the simulated LEO environment, it was decided to realize simulation with a larger cycle amplitude (Test 2). The hot halfcycle provokes postcuring phenomena in the matrix which take the consequences that the Tα value may be changing. Low temperatures contribute to the always-present residual stresses in composite laminates mostly coming from composite curing conditions. The resulting microcracks are responsible for an important decrease of the modulus E’ in the whole temperature range (Fig. 6). An increase in the Tα temperature reaching 7.3°C over 120 cycles (Test 2) of KMU-4l composite is observed in Fig. 7 from DMTA measurements of the bending tan δ. The peak intensity is reduced during the first step of 14 thermal cycles and the following evolution becomes apparent as a weak decrease of the surface and a shift of

277

Dynamic bending modulus (GPa)

Dynamic bending modulus (GPa)

the position of the peak (Fig. 9). The first effect is probably due to low-molecular substances outgassing when the material is exposed to hot temperatures in vacuum. On the whole, one can conclude that the effect of thermal ageing of materials in LEO can be reproduced and accelerated as seen from the resulting evolution of the α-relaxation temperature and mechanical properties. When acting alone with bending stress, thermal cycling leads to an important decrease in the dynamic modulus E’ of the laminate and has a specific influence upon the surface and temperature position of the tan δ peak (Fig. 10 and 11). 40 36

Cycle amplitude: -95 to +133 C

32

reference 14 cycles 30 cycles 60 cycles 120 cycles

28 24 25

50

Cycle amplitude: -159 to +158 C

30 cycles 60 cycles 120 cycles

36 32 28 24 20 16 12 25

50

75 100 125 150 175 200 225 250 Temperature ( C)

Figure 6. Typical DMTA E’(T) curves of KMU-4l 0.05

0.05 Cycle amplitude: -95 to +133 C

0.03 reference 14 cycles 30 cycles 60 cycles 120 cycles

0.02 0.01

reference 14 cycles 30 cycles 60 cycles 120 cycles

0.04 Bending tan δ

0.04 Bending tan δ

40

75 100 125 150 175 200 225 250 Temperature ( C)

Figure 4. Typical DMTA E’(T) curves of KMU-4l

reference 14 cycles

44

0.03 0.02

Cycle amplitude: -159 to +158 C

0.01 0.00

100

125

150 175 200 Temperature ( C)

225

Figure 5. Typical DMTA tan δ (T) curves of KMU-4l

250

100

125

150 175 200 Temperature ( C)

225

250

Figure 7. Typical DMTA tan δ (T) curves of KMU-4l

External stress σx(f) implies transfer of an extra energy to macromolecular chains of the epoxy matrix and, during thermal cycling exposure, may cause a rise in the residual stresses conserved in plies of different orientation. Therefore, thermal relaxation of raised internal stresses during the test had caused intensification of intermolecular friction between polymer chains which resulted in the increase of the registered mechanical loss values. The postcuring process and microcracking produced in ENFB

278

203

1st scan

2nd scan

α-relaxation temperature ( C)

α-relaxation temperature ( C)

matrix by exposing to mechanical loads and extremes in temperature act simultaneously, involving a noticeable effect of synergism. In Fig. 11, the most important increase of 13.1°C in the Tα of KMU-4l is naturally observed for specimens thermally cycled under the most important bending stress σ = σ(f2) applied.

3rd scan

201 199 197

195 Tα ref.

14 30 60120

14 30 60120

Number of thermal cycles [-95 ,+133 C]

1st scan

2nd scan

3rd scan

201 199 197

195 Tα ref.

14 30 60120

Figure 8. Influence of thermal cycling on Tα of KMU-4l

14 30 60120

14 30 60120

14 30 60120

Number of thermal cycles [-159 ,+158 C]

Figure 9. Influence of thermal cycling on Tα of KMU-4l

0.05

0.05

Cycle amplitude: -159 to +158 C

reference σ=0 σ = σ(f1) σ = σ(f2)

0.03 0.02 0.01 100

125

Cycle amplitude: -159 to +158 C

0.04

Tα (ref.) = 195.0 Tα (f = 0) = 199.1 Tα (f = f1) = 199.8 Tα (f = f2) = 195.9

150 175 200 Temperature ( C)

C C C C

225

Figure 10. Influence of 20 thermal cycles on Tα of KMU-4l specimens under bending stress

Bending tan δ

0.04

Bending tan δ

203

reference σ=0 σ = σ(f1) σ = σ(f2)

0.03 0.02 0.01

250

100

125

Tα (ref.) = 195.0 Tα (f = 0) = 205.3 Tα (f = f1) = 203.6 Tα (f = f2) = 208.1

150 175 200 Temperature ( C)

C C C C

225

250

Figure 11. Influence of 60 thermal cycles on Tα of KMU-4l specimens under bending stress

Micro-damage in the epoxy matrix of KMU-4l composite due to mechanical loading and thermal fatigue has led to a significant increase in the moisture diffusion coefficient D of the material as a function of the exposure (Fig. 12). Optical and scanning electron microscopy reveals pronounced microcracking on the long edges and in the matrix surface layer of the specimens. These microcracks vary between a few microns and a few hundred microns in size. All of them should be attributed to exposure of the composite material to low temperatures during the thermal tests. Generally, micro-cracking in composites does not lead directly to failure. The fibers and adjacent plies serve as obstruction to crack propagation preventing a

279

Relative moisture diffusion coefficient D/Dref.

dominant crack from forming in a multi-ply composite. It can, however, facilitate other modes of material degradation, such as delamination, which could cause the failure. More importantly, microcracking was seen to produce decrease in the mechanical properties, in particular, in the material stiffness, coefficient of thermal expansion, ultimate strength and failure strain. As a consequence, microcracking, growing more rapidly under bending stresses, is also responsible for the observed increase of up to 10 times in the D coefficient of KMU-4l composite depending on the ageing procedure. Through the moisture transport parameters, the diffusion analysis puts in evidence microdamage produced in the matrix by its thermal degradation. On the other hand, the maximum residual moisture content w∞ of the material is not sensitive enough to detect morphology degradation. The w∞ value was estimated in 0.35-0.40 % for each set of KMU-4l specimens and has not undergone any significant change.

10 - Test 1 (cycling without mechanical stresses) - Test 2 (cycling without mechanical stresses) - Test 2 (60 cycles with bending stress σ1, σ2)

8

6

4

2

0 14

30

60

120

thermal cycles

14

30

60

120

thermal cycles

f1

f2

mech. stress

Figure 12. Evolution of the diffusion coefficient of KMU-4l composite as a function of the exposure of specimens under mechanical bending stress to simulated thermal cycling environments

According to [2,3], the diffusion coefficient is the physical parameter most sensitive to in-space ageing of carbon/epoxy composites. The change in the D coefficient was by mistake explained by microerosion etching on the surface of the materials during LEO exposure [12] caused by impact of fast atomic oxygen, micrometeorites and space debris. The testing of KMU-4l composite in simulated thermal cycling environments shows that the principal reason for those changes is profound thermal degradation of the epoxy matrices.

280 4.0 Conclusions A modern carbon/epoxy composite KMU-4l was tested to evaluate the effects of thermal cycling combined with mechanical bending stress applied to the specimens. The knowledge of mechanisms behind the physical ageing observed on the composites after exposure to real LEO environments at space stations was used to design the experimental equipment and test procedures, to isolate and confirm the proposed mechanisms, and to accelerate them. The thermal ageing of KMU-4l composite simulated in this study noticeably affects the mechanical properties inducing a substantial decrease of the dynamic bending modulus and a rise in the material glass transition temperature. The material long-term behavior in a LEO environment with thermal cycling may be derived from the conducted accelerated testing. It will involve epoxy matrix postcuring effects accompanied by thermal relaxation of residual ply stresses and significant microcrack damage in the composite laminate. Measurements of the moisture diffusion parameters have confirmed the prevailing role of thermal cycling combined with external mechanical bending stress in the microcrack formation. Although the matrix postcuring process progressively overcomes the microdamage effects, the present study illustrates the necessity to use thermal control coatings and to improve the manufacturing technology of polymer-matrix composites in order to protect the spacecraft materials from thermal degradation and microcracking. 5. 0 References 1. 2. 3.

4. 5. 6. 7.

Startsev, O.V. and Nikishin, E.F., Ageing of polymeric composite materials in outer space, Mechanics of Composite Materials 29 (1993), 457-467. Startsev, O.V., Issoupov, V.V. and Nikishin, E.F. The gradient of mechanical characteristics across the thickness of composite laminates after exposure in LEO environment, Polymer Composites 19 (1998), 36-42. Issoupov, V.V., Startsev, O.V., Paillous, A., Viel, V., Siffre, J. and Nikishin, E.F.: Generalized conclusions made on the basis of in-flight and ground-based simulation experiments, in Proc. of the 8-th Int. Symp. on Materials in a Space Envronment / 5th Int. Conf. on Protection of Materials and Structures from the LEO Space Environment, CNES, Toulouse, 2000. Silverman, E.M. Space environmental effects on spacecraft LEO materials selection guide, Progress Report, 1995. Guillaumon, J.-C. and Paillous, A. ARAGATZ Experience COMES-ECHANTILLON, Rapport Final, CNES, Toulouse, 1992. Paillous, A. and Pailler, C., Behaviour of carbon/epoxy composites in simulated LEO and GEO environments, in Proc. of the 6th Int. Symp. on Materials in a Space Environment, ESTEC, Noordwijk, 1994, 95-102. Gunyaev, G.M., Khoroshilova, I.P., Kapitonova, O.R. and Shkolyarenko, V.P.: New structural epoxymatrix composites, in Stroganov, G.B. and Shalin, R.E. (eds.), Anniversary collection of scientific articles of the All-Russian Institute of Aviation Materials, Moscow, 1982, 177-179.

281 8. 9. 10. 11. 12.

Bystritskaya, E.V., Pomerantsev, A.L. and Rodionova, O.Ye. Non-linear regression analysis: new approach to traditional implementations, J. Chemometrics 14 (2000), 667-692. Issoupov, V.: Proposition d’une procédure pour la simulation de l’effet d’un environnement spatial d’orbite basse sur des matériaux composites, Dissertation for Doctorate, Paul Sabatier University, Toulouse, 2002. Laroze, S.: Mécanique des Structures. Poutres, ENSAE, Toulouse, 2000. Funk, J.G. and Sykes, G.F. The effects of simulated space environmental parameters on six commercially available composite materials, NASA Technical Paper 2906, 1989. Deev, I.S and Nikishin, E.F. Effect of outer space factors on the polymer composite structure under longterm staying conditions in the near-Earth orbit, in Space Forum, Vol. 1, Overseas Publishers Association, Amsterdam, 1996, 297-302.

This page intentionally left blank

HYPERTHERMAL REACTIONS OF OXYGEN ATOMS WITH SATURATED HYDROCARBONS TIMOTHY K. MINTON, DONNA J. GARTON, and HIROSHI KINOSHITA Department of Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717 USA, voice: (406) 994-5394, fax: (406) 994-6011, e-mail: [email protected] Abstract We have conducted detailed crossed-beams experiments on the reactions of O(3P) with model hydrocarbon compounds, CH4, CH3CH3, and CH3CH2CH3, at center-of-mass collision energies in the range 2.8 - 3.9 eV, which are representative of those encountered on spacecraft in low-Earth orbit (LEO) when an incident O atom interacts with a localized effective mass on a hydrocarbon surface. The hyperthermal reactions of O atoms with all three alkanes showed evidence for direct reaction with carbon, in addition to the expected (dominant) H-atom abstraction reaction to form OH. Reactions leading to Hatom elimination and the corresponding methoxy, ethoxy, or propoxy radical are possible, and reactions involving breakage of the CC bond in ethane and propane may also occur. While the H-atom abstraction channel has a modest barrier in the range 0.1 - 0.3 eV, the other observed reactions have barriers greater than 1.8 eV and therefore might become important at the high collision energies between atomic oxygen and spacecraft surfaces in LEO. 1.0 Introduction The erosion/degradation mechanisms of hydrocarbon polymers on spacecraft in low Earth orbit (LEO) involve hyperthermal atomic oxygen in complex reactions that are still not well understood. Many experiments have investigated surface oxidation resulting from O-atom bombardment of hydrocarbon polymers,1-3 while others have probed the volatile species that are ejected from the surface.4 CO and CO2 are believed to be the important volatile species that carry mass away from the surface during steady-state bombardment by O atoms.4 Initial interactions/reactions have been shown to involve atom-surface energy transfer and H-atom abstraction to form OH.4 The dynamical behavior of the OH product suggests a gas-phase-like process where the incident hyperthermal oxygen atom is viewed as interacting with a localized region of the surface. Experiments to date have not been able to verify the direct attack of a carbon atom in the initial reaction of a ground-state O(3P) atom with a hydrocarbon surface. Nevertheless, a theoretical calculation has been performed on the transition state structures and energies for potential CC bond breaking channels in the reactions of O(3P) with small alkanes.5 In the O + CH3CH3 reaction to form CH3O + CH3, for example, the barrier to reaction was found to be approximately 2 eV. This theoretical result suggests the possibility that new reaction pathways might become accessible at the high collision energies between O atoms and a spacecraft surface in LEO and might explain why the reactivity of hyperthermal (5 eV) O-atoms with a hydrocarbon surface appears to be much higher than that of lower-energy O atoms. In order to understand the important initial interactions between hyperther-

283

284 mal atomic oxygen and a hydrocarbon surface, we have conducted detailed studies of the reactions of O(3P) with small hydrocarbon molecules, CH4, CH3CH3, and CH3CH2CH3. The center-of-mass collision energies vary from 2.8 eV to 3.9 eV, depending on the alkane collision partner. An earlier study of the interaction of 5-eV O atoms with a saturated hydrocarbon surface found that the average c.m. collision energy between an incident O atom and a localized effective mass of 40 amu on the surface was 3.8 eV.4 Therefore, the gas-phase reaction between 5 eV O atoms and propane at a c.m. collision energy of 3.9 eV may be considered to be very close to the interaction of 5 eV O atoms with a hydrocarbon surface. Thus, the gas-phase reactions provide a basis upon which to model the initial reactions between atomic oxygen and polymer surfaces in LEO. 2.0 Experimental Details The experiments were performed with the use of a crossed molecular beams apparatus6 (see Figure 1). A pulsed beam containing hyperthermal O(3P) was crossed by a pulsed supersonic alkane beam. The hyperthermal O-atom beam was generated with a laser detonation source based on an original design by Physical Sciences, Inc.7 A synchronized chopper wheel was used to select a narrow portion of the hyperthermal beam pulse. The average translational energy of the hyperthermal O atoms was 5 eV, and the energy spread (FWHM) was 1 eV. The fraction of O atoms in the hyperthermal beam was -70 percent. The molecular oxygen component of the beam had an average translational energy and width roughly double that of the atomic component. The supersonic alkane beams had nominal velocities of 800 m s1, 990 m s1, and 1100 m s1 for propane, ethane, and methane, respectively. Reaction products and elastically (or inelastically) scattered products were monitored with a mass spectrometer detector, which can rotate about the crossing point of the two beams. Number density distributions of products that scattered from the crossing region of the two beams were collected as a function of their arrival time in the electron-bombardment ionizer of the detector. These time-of-flight (TOF) distributions were collected at a variety of detection angles. ROTATABLE DETECTOR

TO ION COUNTING SYSTEM

CHOPPER WHEEL

SOURCE CHAMBER PULSED VALVE

QUADRUPOLE MASS FILTER

NOZZLE

IONIZER

MIRROR

CO2 LASER

SKIMMER

MAIN SCATTERING CHAMBER

PULSED VALVE

ALKANE GAS SUPPLY LINE

Figure 1. Schematic diagram of the crossed molecular beams apparatus showing the hyperthermal oxygen atom source, the alkane source, and the rotatable mass spectrometer

285 3.0 Results, Analysis, and Discussion 3.1. O + CH3CH2CH3 The majority of our data on O-atom reactions with alkanes to date have been collected for the reaction of O + propane, with a c.m. collision energy of 3.9 eV. Figure 2 illustrates several possible reaction pathways that are accessible at this high collision energy. The different possible isomers are ignored in this figure, because the experimental resolution prohibits distinguishing between them. While the H-atom abstraction reaction to yield OH + C3H7 has a modest barrier of a few tenths of an electron volt, the other pathways are expected to have barriers greater than 1.8 eV. The time-of-flight (TOF) distributions are proportional to the number density of the detected species as a function of flight time from the crossing point of the two beams to the ionizer of the mass spectrometer detector. TOF distributions were collected at lab angles from 30( to 50(, with respect to the O-atom beam, following the interaction of the O-atom and propane beams. Representative TOF distributions collected at a lab angle of 10( for m/z = 16, 17, 29, 30, and 31 are shown in Figure 3. Very weak signals were also observed for m/z = 56 (C3H4O+) and 57 (C3H5O+). While the TOF distributions indicate a propensity for non-reactive scattering (note the large signal at m/z = 16), there is also significant reactive scattering to produce the OH radical (m/z = 17) and other oxygen-containing radicals (detected at ionizer fragments m/z = 29, 30, 31, 56, 57). The structure in the TOF distributions collected at m/z = 29 and 30 suggests the occurrence of at least two reaction pathways. (Note that the TOF distribution collected at m/z = 31 contains only one peak for reactive signal, as the peak at short times comes from mass leakage of elastically scattered O2.) Signals at m/z = 29, 30, and 31 may arise from formation (and ionizer fragmentation) of CH3O, C2H5O, or C3H7O. The latter product is likely responsible for the signal observed at m/z = 56 and 57. Figure 4 shows TOF distributions for m/z = 29 at

138.366

O( 3 P) + C 3 H 8 + E coll

(5.42 eV)

Heat of Formation / eV

115.305

E coll = 3.9 eV

92.244 69.183 46.122 23.061

C 3 H 6 O + 2H (~2.6) C 3 H 7 O + H (~2) O( 3 P) + C 3 H 8 (1.5)

OH + C 3 H 7 (1.45)

CH 3 O + C 2 H 5 (1.5) C 2 H 5 O + CH 3 (~1.3) H 2O + C 3H 6

(~0.95)

0.000

Figure 2. Energy diagram for possible reaction channels of the bimolecular reaction O(3P) + C3 H 8 .

286 three different lab angles. Lines drawn on the distributions indicate the minimum possible flight times at which CH3O, C2H5O, and C3H7O products are expected to arrive at the detector, given the nominal collision energy of 3.9 eV. In reality, there is a spread in the TOF distributions resulting from the time width of the O-atom beam pulse. This spread may lead to detected signal at times shorter than the position of each line by approximately 16 µs. The very slow signal is believed to arise from C3H7O products that crack to m/z = 29 in the ionizer. The faster signal is not resolved into two components, but at least some of it arrives too fast to be the result of a C2H5O product. Considering the large time range of the faster signal and its peak flight time, it is likely that both CH3O and C2H5O products are contributing.

Figure 4. TOF distributions for fragments detected at m/z = 29 in the reaction of O(3P) + C3H8. Three laboratory detection angles are shown. The solid, dashed, and dash-dot-dashed lines represent the minimum expected flight times for the products CH3O, C2H5O, and C3H7O, respectively.

The angular distributions (integrated TOF distributions as a function of laboratory angle) provide further inFigure 3. TOF distributions of m/z = 16, 17, formation about the reaction mecha3 29, 30, and 31 from the reaction of O( P) + nisms. Figure 5 shows an angular disC3H8 taken with a detector angle of 10 E . The peak at short times in the m/z = 31 TOF tribution for the m/z = 17 (OH+) signal, distribution is from mass leakage of elastialong with a corresponding Newton diacally scattered O2. gram, which provides the link between the laboratory and center-of-mass reference frames. The data points are connected by straight lines. The dotted line is a guess at what the angular distribution might look like as we move on to collect data at more angles. In the c.m. frame, the OH products appear

287 to scatter preferentially at an angle near 30 E , indicating slight sideways scattering. The angular distribution for the m/z = 29 (CHO+) signal (see Figure 6) exhibits a prominent narrow peak at the c.m. angle that can only come from the C3H7O radical, which is a product of an H-atom elimination channel. (Note that our experiment can not distinguish between the various possible isomers with this empirical formula.) There is also significant signal corresponding to sideways scattering in the c.m. frame. This signal must arise from CC bond breaking channels, yielding either CH3O, C2H5, C2H5O radical fragments.

3.2. O + CH4 Representative TOF distributions collected at m/z = 16 (O+), 17 (OH+), 29 (CHO+), 30 (CH2O+), and 31 (CH3O+) following the interaction of the hyperthermal O-atom and methane beams (c.m. collision energy = 2.8 eV) are shown in Figure 7. These TOF distributions were collected at a detector angle of 10( with respect to the oxygen beam. The m/z = 16 signal arises from non-reactive interactions (predominantly elastic scattering) between the oxygen atoms and methane molecules. The signal at m/z = 17 comes from hydrogen-atom abstraction by the oxygen atoms. The OH product signal is very weak, while m/z = 29 (CHO+), 30 (CH2O+), and 31 (CH3O+) showed much stronger signals. The signal at long times in the m/z = 17 TOF distribution is the result of 13CH4 which slowly effuses out of the alkane source chamber after the pulse. The OH radical signal is very difficult to detect because of the unfavorable kinematics and the relatively high background at m/z = 17. The only possible explanation for signals

288 at m/z = 29, 30, and 31 is elimination of a hydrogen atom to form CH3O or CH2OH. The m/z = 29 and 30 peaks come from cracking of the m/z = 31 parent fragment in the electron-impact ionizer. 3.3. O + CH3CH4 Figure 8 shows representative TOF distributions, collected at a laboratory angle of 10E for products detected from the reaction of atomic oxygen with ethane. Elastic scattering signals were observed at m/z = 16. Other detected masses were m/z = 15 (CH3+), 17 (OH+), 29 (CHO+), 30 (CH2O+), 42 (C2H2O+), and 43 (C2H3O+). All signals obtained were very weak in intensity, except for the elastic scattering signal. The m/z = 15 TOF distribution (not shown) was bimodal, indicating at least two unique processes, probably elastic scattering to produce CH4, which cracks in the ionizer to produce CH3+, and a reaction involving CC bond cleavage to produce CH3O and CH3. TOF distributions collected at m/z = 30 and 29 are quite broad and probably arise from two reaction pathways, H-atom elimination to produce C2H5O + H and CC bond cleavage to produce CH3O and CH3. The higher masses, m/z = 42 and 43, are likely the result of cracking of C2H5O in the ionizer. No signal was observed for the C2H5O parent at m/z = 45. 3.4 SUMMARY OF OBSERVED REACTION PATHWAYS A summary of the reaction pathways for which we have experimental evidence is presented in Figure 9. Only empirical formulas are written for the radical products, because the velocity resolution of our data does not permit resolution of the various isomers, whose heats of formation differ relatively little. The basic mechanisms that have been observed are (1) H-atom abstraction, (2) H-atom elimination, and (3) CC bond fission. Much more experimental work remains to be done in order to delineate the dynamics and relative yields for the observed reaction pathways. Nevertheless, recent direct ab initio trajectory dynamics calculations are beginning to reveal the details of the reaction mechanisms (see paper in this proceedings by Pascual, Garton, and Schatz). These calculations suggest that the H-atom elimination reaction proceeds by a substitution mechanism in which the H atom adjacent to the nascent OC bond is ejected, leaving a hot alkoxy radical. The calculations also suggest a substitution mechanism for the

Figure 7. TOF distributions of m/z = 16, 17, 29, 30, and 31 from the reaction of O(3P) + CH4 taken with a detector angle of 10 E. The peak at short times in the m/z = 31 TOF distribution is from mass leakage of elastically scattered O2.

289 CC bond fission channel. This mechanism resembles an SN2 reaction, in which the O atom approaches the methyl or methylene group, and the CC bond opposite the in coming bond breaks.

4.0 Conclusion New experiments on the model hyperthermal reactions of O atoms with CH4, CH3CH3, and CH3CH2CH3 show evidence of direct reaction with carbon, in addition to the Hatom abstraction channel to produce OH. The reaction, O + CH4, yields signals at mass-to-charge ratios, m/z, of 29 and 30, suggesting the formation of substitution products, CH3O + H. Atomic-oxygen reactions with CH3CH3 and CH3CH2CH3 also yield signals at m/z = 29 and 30, which again implies the production of a radical fragment that contains carbon and oxygen. Various reaction pathways have been identified: CH 3O + CH3, C2H5O + H, C2H5 + CH3O, C3H7O + H. Pathways other than H-atom abstraction have not been observed previously in the reaction of ground-state O(3P) with these small alkanes. While near-thermal reactions of O(3P) atoms with gaseous alkanes are known to produce exclusively OH, the analogous hyperthermal reactions apparently proceed often (although not most of the time) through substitution reactions that may lead to CC bond fission or H-atom elimination. Because these reactions have high barriers, they likely become important only when center-of-mass collision energies exceed -2 eV. Atom-surface collision energies in LEO do exceed 2 eV; therefore, the mechanisms that have been inferred for the analogous gas-phase reactions probably play a significant role in the O-atom-induced degradation of hydrocarbon materials on spacecraft in LEO.

290 5.0 Acknowledgements This work has been supported by grants from the Department of Defense Experimental Program for the Stimulation of Competitive Research (DEPSCoR), administered by the Air Force Office of Scientific Research (Grant No. F49620-01-1-0276), and from the Air Force Office of Scientific Research through a Multiple University Research Initiative (Grant No. F49620-01-1-0335). We are grateful to M. Tagawa, M. Dorrington, and J. Manso for assistance in data collection and to R. Pascual and G. Schatz for many helpful discussions and for sharing their theoretical results. 6.0 References 1. 2.

3.

4.

5. 6. 7.

Brinza, D.E., Chung, S.Y., Minton, T.K., and Liang, R.H. (1994) Final Report on the NASA/JPL Evaluation of Oxygen Interactions with Materials - 3 (EOIM-3), NASA Contractor Report 198865, NASA, Pasadena, CA. Skurat, V.E. (1997) Evaluation of Reaction Effiencies of Polymeric Materials in Their Interaction with Fast (5 eV) Atomic Oxygen. Proceedings of the 7th International Symposium on Materials in the Space Environment; 1997 June 16 - 20; C-paduesEditions, Toulouse. Tagawa, M., Yokota, K., Ohmae, N., Kinoshita, H., Umeno, M., and Gotoh, K. (1999) Atomic Oxygen-Induced Erosion of Polyimide Films Studied by QCM, AFM, XPS, and Contact Angle Measurements. Proceedings of the 6th Japan International SAMPE Symposium and Exhibition; 1999 October 26-29; Tokyo, SAMPE, Covina, CA. Minton, T.K. and Garton, D.J. (2001) “Dynamics of Atomic-Oxygen-Induced Polymer Degradation in Low Earth Orbit.” In Chemical Dynamics in Extreme Environments: Advanced Series in Physical Chemistry, R.A. Dressler, ed. World Scientific, Singapore. Gindulyte, A., Massa, L., Banks, B.A., and Rutledge, S.K. (2000) Can Hydrocarbon Chains be Disrupted by Fast O(3P) Atoms? J Phys Chem 104, 9976-9982. Lee, Y.T., McDonald, J.D., LeBreton, P.R., and Herschbach, D.R. (1969) Molecular Beam Reactive Scattering Apparatus with Electron Bombardment Detector. Rev Sci Instrum 40, 1402-1408. Caledonia, G.E. (1989) “Laboratory Simulations of Energetic Atom Interactions Occurring in Low Earth Orbit.” In Rarefied Gas Dynamics: Space-Related Studies, E. P. Muntz, D.P. Weaver, and D.H. Campbell, eds. AIAA, Washington, DC.

A REVIEW OF LUBRICATION ON THE CANADARM 2 JOSEPH ANTONIAZZI AND DON MILLIGAN MD Robotics Ltd 9445 Airport Rd., Brampton, Ontario, Canada

Abstract A review of the various issues regarding lubrication of mechanisms on the Canadarm 2 is presented. The discussion is broken down into three basic sections. The first section deals with the basic types of lubricants used for mechanisms which operate in space. The second section deals with the various effects of the space environment on these lubricants. The third section will discuss examples of mechanisms tested at MD Robotics in support of the design of the Canadarm 1. 0 Introduction The discussion of the environmental effects on materials has, in recent years, revolved around the concerns of issues such as Atomic Oxygen (AO) and Micrometeoroids/Orbital Debris (M/OD). Because lubricated components are largely contained within structures, these issues play a somewhat less important role for lubricants. Often the most important effects with respect to lubrication are those due to the vacuum environment. Temperature will also have important effects on wet lubricants. However, there are cases when some lubricants can be exposed to the external environment, in which case all of these other factors can play a role as well. 2.0 Background The selection of a lubricant system for a mechanism which operates in low earth orbit (LEO) must take into consideration the mechanical requirements for the application as well as the environmental factors which may affect performance. This selection philosophy was adopted for the Shuttle Remote Manipulator System (SRMS or Canadarm), the predecessor to Canadarm 2. The SRMS, or Canadarm, is a mechanical analogue of the human arm, consisting of a shoulder, elbow, and wrist joint. The joints are connected by composite, tubular elements known as arm booms, which provide the necessary strength and stiffness to precisely attach to and manipulate payloads. The end effector, or "hand", at the end of the wrist joint utilizes a snare mechanism to grasp a grapple fixture on the payload. The

291

292 grapple fixture is basically a flat plate with a probe that is designed specifically to interface with the end effector. The heart of the SRMS is a motor and gearbox, and therefore the principal lubrication considerations for the mechanism are gears and rolling element bearings. The mechanical requirements for such a mechanism are operating speed, load, and life (cycles to failure). The SRMS life requirement is 10 years or 100 missions. The LEO environmental factors which may affect performance are temperature, vacuum, radiation, atomic oxygen, and micrometeoroids and orbital debris. Note, however, that the latter two were not design requirements for the SRMS. Based on these requirements, state-of-the-art reviews of lubrication systems for space mechanisms, and ground testing, the principal lubrication systems selected for use on the SRMS were narrowed down to a molybdenum disulphide-based, dry-film lubricant with an inorganic binder and a perfluoropolyether-based grease. The former was selected for gears and rolling element bearings. The latter was selected for certain bearings for which the duty cycle was not amenable to a dry-film lubricant. The Canadarm 2 life requirement is 10 to 15 years of service on-orbit, and therefore more severe than that of the SRMS. In addition, the payloads are larger and resistance to atomic oxygen and micrometeoroids and orbital debris are design requirements. Selection of lubricant systems for the Canadarm 2 used the SRMS as a baseline and supplemented that with state-of-the art reviews and more extensive ground testing. The result was a more durable molybdenum disulphide-based dry film lubricant for gears, and more extensive use of a perfluoropolyether-based grease for bearings. 3. 0 Lubricants 3.1 WET LUBRICANTS In many terrestrial lubrication systems, an oil is contained within a mechanism and recirculated as the mechanism operates. Alternatively, that oil can be mixed with a thickener to create a grease. The grease will tend to stay in place, and can be used to lubricate mechanisms where it would be inconvenient or impossible to re-circulate the lubricant. This latter case is usually the situation for mechanisms operating in space. The lack of gravity makes it impossible to gather the lubricant together for recirculation. This has an impact on the life of the lubricant. This is because, although a grease will tend to hold the lubricant at a surface, there is no active replenishment of lubricant at that surface. However, there are many applications which can be well lubricated with a grease both in space and on earth. Another aspect of wet lubrication in space is the presence of the vacuum environment. In a vacuum fluids tend to evaporate. In the case of a mechanism operating in space, if the lubricant evaporates before the end of the required life, the mechanism will likely break down. Additionally, the lubricant which has evaporated can contaminate other sensitive surfaces such as camera lenses or thermal control surfaces. The property of a fluid which indicates how much it will evaporate when in a

293 vacuum is the vapour pressure. In order for a grease to be useful in vacuum, it should have a low vapour pressure. Temperature will also affect the viscosity of a fluid lubricant. Although most fluids become more viscous at low temperatures, low vapour pressure fluids can be especially prone to this problem. As such, in applications where low temperatures will be encountered, it is difficult to find a wet lubricant with the necessary optimum combination of properties. An advantage that a grease has over a dry-film lubricant is that it offers the potential for elastohydrodynamic (EHD) lubrication. This is a condition where the lubricant will form a fluid film (referred to as an EHD film) which will support the load between the two surfaces. The EHD film will tend to form when the loads are low enough and/or the speed is high enough and is dependent on the viscosity of the fluid. EHD conditions are generally preferred because it leads to less lubricant breakdown. The alternative to EHD lubrication is boundary lubrication, which is a condition where the two surfaces being lubricated come into direct contact. Depending on the situation, wet lubricants may operate under either condition. In contrast, dry lubricants always operate under boundary lubrication conditions. In the case of the Canadarm 2, all of the applications for wet lubrication use a particular Perfluoropolyether (PFPE) type oil mixed with a thickener to form a grease. The molecule which makes up the base oil is a long chain organic molecule that has a basic carbon backbone with repeated points where carbon atoms are substituted with oxygen atoms. All of the rest of the carbon bonds (other than those which form the backbone) are bonded to Fluorine atoms. Because of the chemistry of the PFPE oil, it is impossible to include any of the traditional EP (Extreme Pressure) additives with the oil. Because of this, the life of this lubricant is somewhat less than many other oils where EP additives can be included. However, the benefit of this PFPE lubricant is that it does not increase its viscosity as the temperature is lowered as much as all the other low vapour pressure lubricants. So this lubricant is particularly suited for use in space. 3.2 DRY LUBRICANTS Dry lubrication represents an alternate approach for the lubrication of mechanisms for both terrestrial and space applications. The following are examples of some common dry lubricant systems: 1) Bonded dry-film lubricants. 2) Sputtered, impingement, or manually applied powders. 3) Soft metals. 4) Polymers or polymer-based composites. Bonded dry-film lubricants represent the most widely used type of dry lubricant system, and have been used extensively at MD Robotics for the lubrication of gears and bearings. Dry-film lubricants consist of a binder and a lubricating pigment. The binder may be ceramic, polymeric, or inorganic, examples of which are silicates, phenolics, and phosphates respectively. The lubricating pigment is usually MoS2. The lubricity of dry-film lubricants is a result of the platelet-like structure of the MoS2, which allows the layers of the lubricant to slide over each other under load, much like a

294 deck of cards. These lubricants can be applied by brush, dip, or spray techniques, much like a paint. Curing of the applied lubricant may be at room or elevated temperature. In general, the ceramic and inorganic binders provide a higher load carrying capacity. Sputtered, impingement, and manually applied lubricants consist of powders of MoS2 or other compounds with similar platelet-like structures such as tungsten disulphide (WS2). These coatings tend to be very thin and do not have the load carrying capacity of the bonded dry-film lubricants. Their use is usually limited to lightly loaded applications. Soft metal coatings, such as lead and silver, have also found limited use as dry lubricants, most notably on bearings. The coatings may be applied by conventional plating or physical vapour deposition techniques such as sputtering. Silver is also a common high-temperature thread lubricant for fasteners. Polymer and polymer-based composites find application as thrust washers and bushings, where they are used as buffer or sliding interfaces between metal parts. Dupont's Vespel polyimide and Teflon fluorocarbon are examples of polymer dry lubricants, and are also used as matrix materials for polymer composites. Polymer composites are generally based on additions of dry lubricant powders such as MoS2 to enhance lubricity. Glass and graphite fibre reinforcements are also use to increase strength and stiffness. 4. 0 Environmental Effects 4.1 VACUUM The vacuum environment of LEO is perhaps one of the most important considerations with respect to lubrication, particularly when coupled with temperature, for several reasons. Materials proposed for use in LEO must meet strict NASA requirements for thermal/vacuum stability (TVS), also referred to as outgassing. These requirements place limits on the quantity of volatile and condensable products allowed from a given material after vacuum exposure at 125oC for 24 hours. The purpose of these requirements are two-fold; material performance degradation due to excessive outgassing, and contamination of sensitive surfaces such as camera lenses by the condensable portion of the outgassed products. Dry-film lubricants, in general, meet the outgassing requirement and thus do not pose a contamination problem, however, their performance can be significantly affected. The classic example of this is graphite, which becomes abrasive in vacuum and thus not suitable as a brush material for motors. This effect is due to desorption of moisture and gases, which enhance the lubricity of graphite. The performance of MoS2 is the opposite, which has led some to assume that the performance of bonded dry-film lubricants based on MoS2 in vacuum is superior to that in air. This is not necessarily the case, due no doubt to contributing factors from other constituents of the lubricant, and therefore emphasizes the importance of testing bonded dry-film lubricants in the operational environment.

295 In the case of wet lubricants, greases such as the PFPE type meet the outgassing requirement but must also consider evaporative loss over the life of the mechanism. This problem is dealt with on Canadarm 2 through the use of an evaporative loss calculation based on the Langmuir expression (Ref 1). The analysis involves the use of the vapour pressure of the base oil of the grease to estimate the total mass loss over the life. Temperature and labyrinth effects are also included. The quantity of lubricant for a given application is then adjusted to ensure that there is a sufficient quantity at the end of life of the mechanism to provide adequate lubrication. One final consideration for the vacuum environment with respect to lubrication is cold welding. In air, the natural oxide on most metals reforms within milliseconds if removed, for example, by sliding friction with another metal surface. In vacuum this is not the case, and intimate metal-metal contact can occur, the result of which is severe galling or cold welding. This failure mode further stresses the importance of life testing dry-film lubricants in their operational environment. 4.2 TEMPERATURE Exterior temperatures on the Canadarm 2 can reach approximately +/-120°C. However, most of the mechanisms which require lubrication are contained within structures like housings. There are various active and passive thermal controls used which will restrict the internal temperatures to a more limited range (approximately +/- 50°C). Dry-film lubricants are generally unaffected by the range of temperatures which a mechanism will see in on the Canadarm and Canadarm 2 in LEO. However, the viscosity of the wet lubricants can be affected by these temperatures. That is not to say that the viscosity increases are so great that the forces to operate a given mechanism will increase many times. However, there will still be some changes to the viscosity of the grease as the temperature drops and this will tend to increase the loads required to operate a mechanism. For some mechanisms within the Canadarm and Canadarm 2, precise control of the mechanism is needed. The power to the motors which drive these mechanisms is modulated to ensure that mechanism speeds are precisely controlled. If the friction within the mechanism changes as the temperature changes, the control of the power to the motor may not be able to adapt quickly enough to this as well as all the other changing conditions as the mechanism operates. For this reason, a wet lubricant may complicate the control aspect of a mechanism. 4.3 ATOMIC OXYGEN Fortunately, the majority of the lubricants used on the Canadarm and Canadarm 2 are contained within housings and structures so that they are not exposed to Atomic Oxygen (AO). Obviously if any organic wet lubricant were to be exposed to AO it would be eroded. There would also, no doubt, be some effects on any silicone-based lubricant. Because of the variety of dry-film lubricants, they can present special cases with AO. The solid plastic materials will erode at varying rates depending on the compound. Most soft metal films will form a protective oxide on the surface of the metal which will

296 tend to protect the underlying metal. The only exception to this is Silver. Silver’s oxide does not provide protection. The last category of lubricants we shall consider here are those based on compounds like MoS2. There have been a few studies which have exposed MoS2-based lubricants to simulated AO (Refs 2 and 3). In one case a sputtered MoS2 film was used and in another a bonded dry-film lubricant was used. In both cases, the surface of the MoS2 film was oxidized, but then stabilized in this state. The oxide did change the friction of the lubricant initially when tested, but the friction then returned to normal when the oxide had been removed by the sliding action of the test. 4.4 RADIATION The gamma-ray, x-ray, ultraviolet (UV), proton, and electron portion of the LEO radiation environment can produce significant damage in certain materials. Organicbased materials, such as Teflon, are particularly prone to radiation damage. Radiation analyses based on worst case integrated doses over the life of the Canadarm 2 indicated that the levels are too benign (~2 Mrad) to affect the mechanical properties of metals and most non-metallics, with the exception of Teflon. The effect on lubricants is further reduced due to shielding by the surrounding structure, and therefore radiation damage is not a design driver. 4.5 MICROMETEOROIDS/ORBITAL DEBRIS The micrometeoroid and orbital debris (M/OD) environment of LEO consists of solid particles of extraterrestrial and man-made origin respectively. The distinguishing feature between the two types of particle is their size and velocity. Micrometeoroids are small in size (<1 g) with velocities in the 10 to 70 km/sec range. The micrometeoroid distribution is also considered to be isotropic. Orbital debris, on the other hand, consists of inactive or spent spacecraft, fragments, and other particulates up to and greater than 10 cm in diameter, with lower average velocities and a non-isotropic distribution. The main concern with respect to the M/OD environment is mechanical damage. In the case of the Canadarm 2 this concern is more for the primary structure, such as the arm booms, where the damage could lead to a loss of strength or stiffness. Since the bearings and gears are buried in the mechanism, the possibility of damage due to M/OD impact is considered to be remote. 5. 0 Testing 5.1 TEST CONDITIONS Any testing of lubricant for use in space will generally require testing to be performed in a vacuum environment. This is because the friction and wear life of the lubricant will change in vacuum. Also, because most space mechanisms undergo at least some

297 qualification and/or acceptance testing in air, the performance of the lubricant in that mechanism must also be known in air as well. Therefore some testing in air is also required. For dry-film lubricants based on MoS2, it is tempting to rationalize that because the performance of MoS2 is better in vacuum, then any testing conducted in air will be conservative. As was discussed in Section 4.1 on vacuum effects, this may not be a valid assumption. 5.2 MECHANISM TESTING PERFORMED AT MD ROBOTICS The types of mechanisms tested at MD Robotics reflect the various mechanisms which are used in the various MD Robotics’ products. The type of testing described below is different from the basic qualification or acceptance testing of flight hardware. Although this testing is necessary and can be difficult, the type of testing which is the subject of this section, is testing used to produce data which can then be used to design the mechanisms. This type of testing is important in order to ensure that any flight hardware produced has a good chance of passing its respective tests without having to be overdesigned. At the heart of many of the mechanisms are a motor and a gearbox. MD Robotics typically uses DC brushless motors. The greatest issue with regard to lubrication in a motor like this is in the bearings. If we then consider the gearbox, there are both bearings and gears to lubricate. MD Robotics has conducted testing on individual pairs of gears as well as on complete gearboxes. For testing individual gears, a common type of test arrangement called “4-Square” arrangement is used. This is where two pairs of external spur gears are mated with each other on two parallel shafts. A twist is introduced to one of the shafts (using a clutch) which applies the load to the two pairs of gears. In this way each pair of gears is effectively applying the load to the other. Then one of the shafts is attached to an external motor to rotate the gears. Although this is a common arrangement for gear testing, MD Robotics has designed and built a test rig in order to allow it to be used in a MD Robotics-owned vacuum chamber. For testing gearboxes, the usual arrangement is to measure the input torque and output torque through torque sensors when applying a load to the gearbox. This will allow one to determine the efficiency of the gearbox during the entire test. Ferrofluidic feedthroughs are used to transmit the rotation from a motor outside the vacuum chamber to the gearbox inside the vacuum chamber. There are a number of ways to apply the load to the gearbox. One method used at MD Robotics is to have the output shaft from the gearbox under test backdrive another gearbox, which then backdrives a motor. The current sent to the backdriven motor will then control the level of torque applied to the gearbox under test. Although some bearing testing has been performed at MD Robotics (in support of the Canadarm Program) much of the more recent data on bearing performance comes from gearbox testing. Another mechanism which is very challenging from a lubrication point of view is the “Snare/Rigidize” mechanism. This mechanism is unique to MD Robotics. This mechanism is used in the end effectors of both the Canadarm and Canadarm 2. The

298 function of the mechanism is to capture a payload and form a rigid connection in order to be able to move it. The mechanism must also be able release a payload without imparting any momentum to it. A special test rig had to be designed and built to test this arrangement. The sliding interface between the snare cables and the grapple shaft is especially hard on lubricants because of the high contact stresses. The snare cables are made of stainless steel wire rope and they slide along the grapple probe (attached to the payload) as the mechanism operates. 6. 0 Summary A review of the issues surrounding lubrication of mechanisms for LEO is presented with examples of mechanisms tested at MD Robotics in support of the design of the Canadarm 2. 7. 0 References 1. Leger, Lubert J., Dufrane, Keith, “Space Station Lubrication Considerations” 21st Aerospace Mechanisms Symposium, Apr 29 – May 1, 1987. 2. Arita, M., et al, “Investigation of Lubrication Characteristics of MoS2 films in Space Environment” Proc. Fourth European Symposium on Space Mechanisms and Tribology, Cannes, France, 20 – 22 September, 1989 (ESA SP-299, March 1990). 3. Martin, J.A., Cross, J.B., Pope, L.E., “MoS2 Interactions with 1.5eV Atomic Oxygen” New Materials Approaches to Tribology: Theory and Applications, Symp. Nov 27 – Dec 2, 1988, Boston, Mass (Published by Materials Research Society, Pittsburg, Penn.).

DEGRADATION OF POLYMERS BY O(3P) IN LOW EARTH ORBIT ASTA GINDULYTE Department of Chemistry, Hunter College 695 Park Avenue, New York, NY 10021, LOU MASSA The Graduate School, City University of New York 365 Fifth Avenue, New York, NY 10016 BRUCE A. BANKS AND SHARON K. R. MILLER NASA Glenn Research Center 21000 Brookpark Road, M/S 309-2, Cleveland, OH 44135

Abstract O(3P) is a highly reactive species which may cause damage to materials on contact. In low Earth orbit (LEO) high energy collisions (∼4.5 eV) of O(3P) with spacecraft materials can lead to extensive degradation. We use ab initio molecular orbital calculations to investigate the possibility of chain breaking in polyethylene caused by a single O(3P) attack under LEO conditions, since the occurrence of such reactions could greatly accelerate the erosion. The smallest alkanes (N=2,3,5,7) serve as models of polyethylene. For the case of ethane (N=2), we explore the triplet potential energy surface of the following reaction: O(3P) + CH3xCH3 Æ •OxCH3 + •CH3. Analogous reactions, where O(3P) attacks a central carbon atom, are studied for the higher alkanes. Results obtained using the Hartree-Fock method, and density functional theory are reviewed. We conclude that conditions in LEO are conducive to chain breaking in polyethylene caused by a single O(3P) attack. We also address the question whether the most abundant species in LEO, viz., atomic oxygen in its ground state, O(3P), alone can cause the degradation in fluoropolymer materials. The smallest fluorocarbons CNF2N+2 (N=2,3,5) serve as models of fluoropolymers. Since electronegativity of fluorine seems to preclude Fabstraction by O(3P), we concentrate on direct O(3P) attacks on carbon-carbon bonds. For the case of fluoroethane (N=2), we explore the triplet potential energy surface of the following reaction: O(3P) + CF3xCF3 Æ •OxCF3 + •CF3. Analogous reactions, where O(3P) attacks a central carbon atom, are studied for the higher fluorocarbons. Results obtained using the Hartree-Fock method and density functional theory are reviewed. We conclude that O(3P) species in LEO possesses enough translational energy to degrade fluorocarbon materials.

299

300 1. 0 Introduction Many satellites travel around the Earth through the space region called low Earth orbit (LEO), the altitudes of 180 to 650 kilometers above the Earth’s surface. The largest component of the atmosphere [1] at these altitudes is atomic oxygen (AO). The typical O-atom number density at space shuttle altitudes is on the order of 108 cm-3. A LEO orbiting body typically traveling at 7.2 km/s relative to this density experiences a flux of 1014 O-atoms/cm2s. Oxygen atoms hit the spacecraft surface with impact energies of approximately 4.5 eV (∼100 kcal/mol) [2]. Materials, particularly affected by LEO, are organic polymers which lose weight and, depending on thickness, can be eroded away completely [3-10]. The erosion yield of materials may be influenced by factors such as AO flux, AO fluence, synergistic solar radiation, AO impact energy, AO impact angle, material temperature, etc. Phenomenological models have been developed to explain observed trends in materials degradation [11]. However, to date there is no clear understanding of elementary reaction mechanisms for material interactions with AO. LEO observed erosion rates for polyethylene are among the highest for all organic polymers (∼3.7×10-24 cm3/O-atom) [8]. We use ab initio techniques to investigate the reaction of O(3P) attack on carbon-carbon bonds of ethane, and several higher alkanes, which we consider to be models, however simple, of polyethylene. Polytetrafluoroethylene (PTFE) as well as fragments of the FEP Teflon® polymer chain has a similar structure to that of polyethylene, differing only by substitution of fluorine atoms for hydrogen atoms. The results of our work on polyethylene suggests the possibility that chain breaking by O(3P) could be a contributing cause of degradation [12]. We use the phrase “chain breaking” to mean a single attack of O(3P) directly on a polymer backbone, resulting in a broken carboncarbon bond, and thus, a broken polymer chain. Such chain breaking, if it should occur, would represent a new and efficient mechanism to explain polymer degradation, as we have argued elsewhere [12,13]. 2. 0 Computational Details Quantum mechanical calculations were carried out with the GAUSSIAN98 [14] and MULLIKEN [15] program packages. For all the calculations Gaussian-type basis sets were employed (see Table 1 and 2). The explanation and abbreviations of the basis sets can be found in the literature [16]. The geometries of all reactants, products, and transition states have been optimized [12,13]. For all the triplet species, i. e., the transition states and O(3P), an unrestricted wave function was implemented and examined for spin contamination, which was found to be inconsequential. No symmetry constraints were imposed for optimizations of the transition states. Vibrational frequencies have been calculated using the same approximation to characterize the nature of stationary points and to determine zero-point energy (ZPE) corrections [12, 13]. All the stationary points have been positively identified for either minimum energy with no imaginary frequencies or for transition states with one imaginary frequency. In the cases where it was not clear

301 (from the analysis of vibrational modes) whether a transition structure is connecting the desired reactants and products, intrinsic reaction coordinate (IRC) analysis was carried out in order to confirm that [12, 13]. TABLE 1: Structural parameters (Angstroms and degrees) for the transition state of C2H6 + O(3P) Æ CH3 + OCH3 reaction.

H2A O

C1

C2

H1A

Theorya,b rO-C1 rC1-C2 aO-C1-C2 aO-C1-H1A aC1-C2-H2A HF/3-21G 2.005 2.372 178.7 92.4 99.6 MB3LYP/6-31+G(d,p) 1.789 2.008 176.9 93.8 104.5 a See references 18-20 for HF and reference 21 for MB3LYP methods and formalism. b See reference 16 for definitions and nomenclature of basis sets. TABLE 2: Total energies (a. u.) and activation barriers (kcal/mol) for C2H6 + O(3P) Æ CH3 + OCH3 reaction.

Theorya,b HF/3-21G HF/3-21G ( 0 K ) MB3LYP/6-31+G(d,p) MB3LYP/6-31+G(d,p) ( 0 K ) a,b Same as in Table 1.

ETS -153.08712 -153.01595 -154.75243 -154.68065

EC2H6 -78.79395 -78.71389 -79.77573 -79.70123

EO -74.39366 -74.39366 -75.03772 -75.03772

Ea 63.1 57.5 38.3 36.6

3. 0 Computational Results: Alkanes Reaction of ethane with O(3P): O(3P) + CH3xCH3 Æ •OxCH3 + •CH3. The above reaction is a methyl abstraction from ethane by O(3P). The key geometric parameters computed by ab initio and DFT methods of the transition state structures for the ethane reaction with O(3P) [12] are presented in Table 1. The classical and vibrational adiabatic (zero-point corrected) reaction barriers obtained by DFT and ab initio methods [12] are presented in Table 2. There is no experimental evidence that suggests any information about the structure of the transition state. The appearance of this structure is very similar for both theoretical methods used. The oxygen atom is almost aligned with the two carbon atoms. C1 and three hydrogen atoms attached to it lie in a plane, while C2 with its three

302 hydrogen atoms form a pyramidal shape. HF calculated values for the above mentioned distances are very long as compared to DFT methods. Reactions of higher alkanes (n=1,2,3) with O(3P): O(3P) + CH2x(CnH2n+1)2 Æ •OxCH2x(CnH2n+1)+ •CnH2n+1. The key geometric parameters computed by HF and DFT methods of the transition state structures for the reactions of propane, pentane and heptane with O(3P) (where the O(3P) attacks the central carbon atom) have been calculated [12]. There are several differences in the structures of the transition states for the reactions of higher alkanes as compared to the ethane case. First, the partially formed (O−C1) and partially broken (C1−C2) bonds are longer. Second, the oxygen atom is no longer aligned with the two carbon atoms. The O−C1−C2 angle is ∼160° as opposed to almost 180° in the case of ethane. In addition, in the case of pentane and heptane there is a considerable rotation around the C1−C1A bond, which results in a dihedral angle, C2-C1-C1A-C1B, of ∼−100° as opposed to 0° in the hydrocarbon chain. The cause of these differences is certainly the repulsion between the oxygen atom and the groups attached to the carbon atom that is being attacked by O (3P). Despite the considerable differences in the geometrical parameters between the transition states of higher alkanes and that of ethane, the energy barriers for the reaction remain almost the same. HF/3-21G produces a barrier of 58.4, 57.9, and 58.0 kcal/mol for propane, pentane, and heptane, respectively, as opposed to 57.5 kcal/mol for ethane. MB3LYP/6-31+G(d,p) calculated values for the barrier are 38.5, 36.3, and 36.5 , in the longer alkanes for propane, pentane, and heptane, respectively, as opposed to 36.6 kcal/mol for ethane. Our best estimate of the energy barrier for longer alkanes [12], bracketed from above by moderate basis HF and from below by moderate basis DFT values, is 40-45 kcal/mol. 4. 0 Computational Results: Fluorinated Alkanes Chain breaking reactions of O(3P) with C2F6, C3F8, and C5F12, where in the case of C3F8 and C5F12, O(3P) attacks the central carbon atom, were also studied [13]. For C2F6 the key geometric parameters, computed by HF and DFT methods of the transition state structure, are presented in Table 3. The classical and vibrational adiabatic (zero-point corrected) reaction barriers obtained by HF and DFT methods are presented in Table 4. As is the case with the alkane analogs, there is no experimental evidence that would suggest any information about the structures of the transition states or the energy barriers associated with these reactions.

303 TABLE 3: Structural parameters (Angstroms and degrees) for the transition state of O(3P) reaction with C2F6.

F

(2A) F

(2)

F O

(1)

F

F

(1A)

F

Theorya,b HF/3-21G MB3LYP/6-31+G(d,p) a,b Same as in Table 1.

rO-1

r1-2

aO-1-2

2.914 1.765

2.746 1.920

174.2 176.7

aO-1-1A

a1-2-2A

81.9 92.5

102.7 107.8

TABLE 4: Total energies (a. u.) and activation barriers (kcal/mol) for the transition state of O (3P) reaction with C2F6.

Theorya,b HF/3-21G HF/3-21G ( 0 K ) MB3LYP/6-31+G (d,p) MB3LYP/6-31+G (d,p) ( 0 K ) a,b Same as in Table 1.

ETS -742.98149 -742.95561 -749.95470 -749.92822

Emodel -668.74490 -668.71264 -675.04639 -675.01763

EO -74.39366 -74.39366 -75.03772 -75.03772

Ea 98.6 94.6 81.2 79.8

The appearance of the transition state structure for O(3P) reaction with C2F6, is similar to that of O(3P) reaction with C2H6. The oxygen atom is almost aligned with the two carbon atoms. C1 and three fluorine atoms attached to it lie in a plane, while C2 with its three fluorine atoms form a pyramidal shape. Turning our attention to HF results only, a major difference between the hydrogenated and fluorinated analogs is observed for the partially formed (O−C1) and partially broken (C1−C2) bonds of the transition state structure. The O-C1 bond length is 2.914 A and 2.005 A for C2F6 and C2H6 cases, respectively. Also, the C1-C2 bond length is 2.746 A and 2.372 A for the same respective cases. Now considering the results of MB3LYP DFT only, the bond length values calculated for C2F6 are very similar to those in the case of C2H6, i. e., 1.765 A for O-C1 bond and 1.920 A for C1-C2 bond. The fluorinated case differs from the hydrogenated case in two ways. First, highly electronegative fluorine atoms withdraw some electron density from the carbon

304 atom that is being attacked by oxygen and therefore, the oxygen atom would have to be slightly closer to the carbon atom in order for a bonding interaction to take place. Second, there is a repulsive interaction between highly electronegative fluorine atoms and the oxygen atom. These two interactions have opposite effects on the bond lengths. The latter effect seems to be strongly favored by the HF method which does not account for electron correlation. These results emphasize that electron correlation is especially important in complex systems, such as transition states which involve bond breaking and forming. Since DFT methods do account for electron correlation the results obtained by the use of these methods are likely to be more trustworthy in this case. The two effects described above offer a possible explanation for the high activation energy for the fluorinated case, i.e., 94.6 kcal/mol and 79.8 kcal/mol as obtained with the use of HF and DFT methods, respectively. Based on our previous experience12 with C2H6, we believe that these two numbers are high and low bounds for the activation energy of the reaction. The transition state structures for the reactions of O(3P) with C3F8 and C5F12, obtained with the use of the MB3LYP DFT method, are similar to what we would expect based on our experience with the O(3P) reactions with alkanes [12]. The bonds that are being broken and formed are slightly longer than in the C2F6 case, and the OC1-C2 angle is slightly smaller due to the repulsion between the oxygen atom and the groups attached to the carbon atom that is being attacked by O(3P). We were unable, however, to obtain the result for the C5F12 case with the use of HF method due to convergence problems which, we believe, arose because of the unrealistically long partially broken and formed bonds. The activation energy values for the reactions of O(3P) with C3F8 and C5F12, obtained with the use of MB3LYP DFT method, are 67.2 kcal/mol and 67.1 kcal/mol, respectively. The activation energy value for the reaction of O(3P) with C3F8 obtained with the use of HF method, is 74.1 kcal/mol. The similarity of the two DFT results encourages us to estimate the activation energy value for longer fluoropolymer chains at ∼67-75 kcal/mol. 5. 0 Conclusions In summary, we have brought into concordance the results of molecular orbital calculations on certain model alkanes and their fluorinated analogues. What emerges is the possibility of a chain breaking mechanism in these compounds, caused by a single O(3P) attack under LEO conditions. We have studied four small alkanes as models of polyethylene. The magnitude of the activation barrier, for O(3P) attack, is similar for all these cases, making it a probable magnitude for longer polyethylenes too. The estimated barriers (∼40-45 kcal/mol) are substantially less than the O(3P) kinetic energy available in LEO. Thus chain breaking, pictured in Figure 1, is an important candidate mechanism for contributing to polyethylene degradation in LEO. Since our estimate of the activation energy for the chain breaking reactions of O(3P) with longer fluoropolymer chains is ∼67-75 kcal/mol, we conclude that such reactions are possible under LEO conditions where the oxygen atoms possess

305 O

O

O CH2 H2C

O CH2 H2C

O CH2 H2C

O CH2

H2C

Figure 1. Proposed hydrocarbon erosion mechanism via chain breaking reactions with O(3P).

translational energy relative to a spacecraft on the order of ∼100 kcal/mol. Not all of the total kinetic energy is available to surmount the reaction barrier. However, even after the total kinetic energy is reduced by the center-of-mass kinetic energy (∼25 kcal/mol), there remains sufficient kinetic energy (∼75 kcal/mol) associated with the interparticle distance coordinate to make the reaction possible[17]. An implication of our computational results is the (previously dismissed) possibility that degradation by AO alone could be the most important contributor to fluoropolymer erosion in LEO.

306 6. 0 Acknowledgement We acknowledge the Maui High Performance Computational Center for the allocation of computational time. One of us (A.G.) thanks the Burrroughs-Wellcome Company for funding a Gertrude Elion Scholarship, administered by the Hunter College Chemistry Department. L. M. acknowledges an IBM Shared University Research (SUR) grant, a CUNY Research Award, a CUNY Collaborative Award, and a NASA JOVE/JAG grant. 7. 0 References and Notes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21.

U. S. Standard Atmosphere, 1976, NOAA-S/T 76-1562; Published by the National Oceanic and Atmospheric Administration, National Aeronautics and Space Administration, and the United States Air Force, 1976. Banks, B. A.; de Groh, K. K.; Rutledge, S.; DiFilippo, F. J. “Prediction of In-Space Durability of Protected Polymers Based on Ground Laboratory Thermal Energy Atomic Oxygen.” NASA TM107209, 1996. Leger, L. J. “Oxygen Atom Reaction With Shuttle Materials at Orbital Altitudes" NASA TM58246, 1982. Leger, L. J.; Visentine, J. T. Aerospace America 1986, 24, 32. Hunton, D. E. “Shuttle Glow.” Sci. Am. 1989, 261, 92. Murr, L. E.; Kinard, W. H. “Effects of Low Earth Orbit.” American Scientist 1993, 81, 152. Reddy, M. R. J. Mat. Sci. 1995, 30, 281 and references therein. Banks, B. A. “The Use of Fluoropolymers in Space Applications” In Modern Fluoropolymers; John Wiley & Sons: New York, 1997 and references therein. Gregory, J. “On the Linearity of Fast Atomic Oxygen Effects.” NASA CP-3257. Dooling, D.; Finckenor, M. M. “Material Selection Guidelines to Limit Atomic Oxygen Effects on Spacecraft Surfaces.” NASA TP-209260, 1999 and references therein. Iskanderova, Z. A.; Kleiman, J. I.; Gudimenko, Y. I.; Tennyson, R. C. “Influence of Content and Structure of Hydrocarbon Polymers on Erosion by Atomic Oxygen.” J. Spacecraft and Rockets, 1995, 32, 878. Gindulyte, A.; Massa, L.; Banks, B. A.; Rutledge, S. K. “Can Hydrocarbon Chains be disrupted by Fast O(3P) Atoms?” J. Phys. Chem. A, 2000, 104, 9976. Gindulyte, A.; Massa, L.; Banks, B. A.; Rutledge, S. K. “Direct C-C Bond Breaking in the Reaction of O(3P) with Flouropolymers in Low Earth Orbit”, J.Phys. Chem. A ( In Press). Gaussian 98, Revision A.6; Gaussian, Inc.: Pittsburgh PA, 1998. MULLIKEN is IBM proprietary software. Foresman, J. B.; Frisch, A. Exploring Chemistry with Electronic Structure Methods, 2nd ed.; Gaussian, Inc.: Pittsburgh, 1996. The total kinetic energy (KE) of the interacting particles may be transformed into a sum of centerof-mass KE and reduced mass KE. The center-of-mass KE is conserved in the collision. Therefore the total KE is reduced by that amount. Assuming an effective surface mass of 40 amu for the oxygen atom collision with the fluoropolymer surface, the center-of-mass collision energy of 4.5 eV oxygen atom would be approximately 75 kcal/mol. Roothan, C. C. J.; Rev. Mod. Phys. 1951, 23, 69. Pople, J. A.; Nesbet, R. K. J. Chem. Phys. 1959, 22, 571. McWeeny, R.; Dierksen, G. J. Chem. Phys. 1968, 49, 4852. Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623. MB3LYP is very similar to B3LYP defined in this paper, except it uses the local correlation functional of Perdew and Wang (Perdew, J. P.; Wang, Y. Phys Rev. B 1992, 45, 1324) instead of the Vosko, Wilk and Nusair functional.

IRIDIUM METAL AS POTENTIAL SUBSTRATES FOR EXPERIMENTS IN SPACE LI YAN JOHN A. WOOLLAM Center for Microelectronic and Optical Materials Research, Department of Electrical Engineering, University of Nebraska-Lincoln Abstract Due to its unique properties, Iridium is currently being considered by NASA as a substrate (optically thick Ir on ultra-smooth fused silica) for the study of contamination layers deposited in the space environment. Extremely smooth thin films of iridium were deposited in this work onto super-polished fused silica substrates, using DC magnetron sputtering in an argon plasma. The influence of deposition process parameters on film micro-roughness was investigated. Moreover, Ir film optical constants were determined using variable angle spectroscopic ellipsometry, over the spectral range from vacuum ultraviolet to middle infrared (140 nm—35 µm). Because the Ir films were optically thick and the surface roughness were measured by atomic force microscopy, and then accounted for in the optical model, the as-determined film optical constants are expected to be the best available for Ir bulk metals, minimally affected by surface over-layers or microstructure. 1.0 Introduction The need for durable, corrosion-free, reproducible iridium (Ir) thin films with smooth surfaces and good adhesion to substrates has drawn attention recently for various applications [1,2]. Ir is of interest because of unique properties, including high melting point (~2713K), low oxygen permeability, high chemical stability, and good electric conductivity, to name a few [1-4]. In addition, it is currently being considered as a potential substrate (Ir on fused silica) for use in space contamination studies. Ir has virtually no oxide and can be extremely smooth, thus minimizing the difficulty of contamination analysis [5,6]. For use in space, Ir films with excellent surface qualities are necessary, including extremely low surface roughness and superb stability in the adverse environment in space. A variety of techniques have been used to prepare Ir films, including metal-organic chemical-vapor deposition (MOCVD) [1-9], CVD [10], DC and RF magnetron sputtering [1-3,11]. Because of the strong influence of morphology on properties, thin film microstructure is important to characterize for both optical and microelectronic applications. Moreover, environmental stability also depends strongly on film

307

308 morphology. Surface roughness of sputtered Ir metal films is influenced by the surface roughness of the substrate and coating microstructure. Super polished fused silica substrates minimized roughness. Magnetron sputtering is currently the most widely commercially practiced sputtering method [12]. It features high sputtering rate at the target, high deposition rate, and superior adhesion of sputtered films [13,14]. In the present study, Ir films with smooth surfaces were prepared by DC magnetron sputtering. The effects of the processing conditions, including gas pressure in the deposition chamber, deposition duration, etc., on film surface roughness were investigated and optimized. In addition, Ir film optical constants over the spectral range from vacuum ultraviolet (VUV) through the middle infrared (MIR) were determined using variable angle spectroscopic ellipsometry (VASE£). Previous measurements of Ir optical constants can be found in the literature [15,16]. The present work covers a much wider spectral range and includes atomic force microscopy (AFM) characterization of surface roughness. 2.0 Experimental details 2.1 SAMPLE PREPARATION Ir films were prepared by DC magnetron sputtering in a four-gun cryopumped

Output Light Beam

Sample

Input Light Beam

Plasma Sputter Gun Polarizer

Analyzer/ Detector

2 1

Sputter Guns

3

4

Optical Ports Figure 1 Four-gun cryopumped magnetron sputtering deposition chamber.

309 deposition chamber, as illustrated in Figure 1. Each of the four guns can be powered separately by either rf or dc source, and eight substrate holders are placed overhead on a rotating platen. The substrate-target spacing is 10 cm. Commercial fused silica discs (Esco Products Inc.) of 1″ in diameter and 1/4″ in thickness were used for this experiment. The discs were subsequently cleaned ultrasonically with acetone and methanol, and then blow-dried with nitrogen gas. The coating target iridium was 99.8% pure, in the form of a disc of 2 inch diameter and 1/8 inch thick. To help the Ir films to better adhere to the fused silica surface, a chromium (Cr) layer of about 24nm thickness was first deposited as a buffer. Before any deposition, the target (either Cr or Ir) was pre-sputtered for ~10 min while keeping the fused silica substrates covered by shutters. The coating system was first cryopumped to a working pressure of ~5 × 10-7 Torr before introducing ultra high purity argon (Ar) sputtering gas. Next, Cr layers were deposited under 5mTorr of gas pressure in the chamber, 20 sccm of Ar gas flux flow, and 40W of power; the entire buffer layer deposition lasted about 2min, with an approximate deposition rate of 0.2 nm /sec. Finally, the Ir film deposition was systematically investigated at a sequence of gas pressures in the chamber ranging from 2 to 5 mTorr; dc power of 35 W, and with deposition durations of 20, 30, or 40 minutes. All films were made optically thick; that is, light could not reach the bottom of the film. X-ray diffraction data of the as-deposited Ir films show a preferred (111) orientation, regardless of the deposition parameters employed. 2.2 AFM AFM was used to examine the rms (root mean square) surface micro-roughness under ambient conditions. Data were taken over areas of 2µm × 2µm using a DI (Digital instruments) AFM DimensionTM 3100 in the Tapping Mode. 2.3 SPECTROSCOPIC ELLIPSOMETRY Spectroscopic ellipsometry (SE) is a well-known surface sensitive, non-destructive optical technique widely used to determine film thickness and optical constants. Reflection ellipsometry measures change in polarization state of light upon reflection from a sample surface. The measurement is expressed as psi (Ψ) and delta (∆), which are related to the Fresnel reflection coefficients by [17]: ρ ≡ tan(ψ )e i∆ = R p R s (1) where p- and s- correspond to directions parallel and perpendicular to the plane of incidence, respectively. In this work, measurements were performed over a wide spectral range, using two separate ellipsometers. The first covers the vacuum UV to the NIR (140 to 1700nm), and the second is an infrared ellipsometer covering a spectral range from 8000 to 250 cm–1 (1.25 to 40 microns). All spectroscopic ellipsometric data were taken at three angles of incidence (50°, 55°, 60°).

310 3.0 Results and Discussions 3.1 MICROROUGHNESS OF THE AS-DEPOSITED IR FILMS A series of deposition process parameters were employed to examine their effects on the smoothness and microstructure of coatings as determined using AFM. Table 1 lists six different samples prepared under six different deposition conditions. TABLE 1. A list of six Ir/Cr/fused silica samples prepared under different deposition conditions. Sample #

Ar gas flow (sccm)

DC power (W)

Gas pressure (mTorr)

Deposition time (min)

RMS Roughness (AFM) (nm)

1

20

35

5

20

0.84

2

20

35

4

20

0.44

3

20

35

3

20

0.41

4

20

35

2

20

0.3

5

20

35

5

30

1.0

6

20

35

5

40

1.03

At a dc power of 35W and an Ar gas flow of 20sccm, the variation of gas pressure

(a)

(b)

(c)

(d)

Figure 2: AFM images of Ir surface morphology with an Ar gas pressure (P) of (a) 5, (b) 4, (c) 3, and (d) 2 mTorr.

311 in the chamber (P) was in the range of 2-5mTorr. The AFM images in Figure 2(a)-2(d) show iridium films deposited at a pressure of 5mTorr, 4mTorr, 3mTorr, and 2mTorr, labeled as sample #1, 2, 3, and 4, respectively. Clearly, the Ir films consist of closely packed grains with various shapes and very fine grain size. The RMS roughness values calculated from AFM were only 0.84 nm, 0.44 nm, 0.41 nm, 0.3 nm, respectively. Apparently, the average roughness decreased slightly with decreasing gas pressure in the chamber. This was somewhat expected, because lower gas pressure could mean less entrapment of the incident working gas (Ar) in the film, and higher energy of the sputtering particles (due to fewer collisions with the sputter inert gas) when they strike the substrate, thus resulting in better film adhesion.18 The film thickness, assumed linear to the deposition time, was also found to play a role in film topography, as illustrated in Figure 3. At a dc power of 35W, an Ar gas flow of 20sccm, and a gas pressure of 5mTorr, Ir films were sputter deposited for 20min [Fig. 3(a)] for sample #1, 30min [Fig. 3(b)] for sample #5, and 40min [Fig. 3(c)] for sample #6. The RMS roughness values for these three sample films were 0.84 nm, 1.0 nm, and 1.03 nm, respectively. Results indicate an increase in roughness of sputtered Ir films as a function of increased film thickness. In general, for a given materialsubstrate combination under a given set of deposition conditions, the average grain size

(a)

(c)

(b)

Figure 3 AFM images of Ir surface morphology with a deposition duration of (a) 20, (b) 30, and (c) 40 min.

312 of the deposited film increases slightly as its thickness increases. Bigger grain size, eventually, will lead to a rougher surface. Since our main goal was to prepare as smooth surfaces as possible Ir films, relatively thin films were favored. By relatively thin, we mean the films still had to be optically thick (opaque, meaning thickness greater than 100 nm for metals). This was important because we determined the optical constants on as-deposited films using spectroscopic ellipsometry and light reaching the back surface would complicate analysis. 3.2 IRIDIUM FILM OPTICAL CONSTANTS

Extinction Coefficient, k Index of refraction, n

As a general rule, film density increases with increasing film thickness until reaching a bulk density value where it saturates. Film optical constants for very thin films can be somewhat different than those of bulk metals. In this study, Ir metal films were deposited in an inert argon gas atmosphere, at relatively low operating pressures, without heating the substrates (i.e., at room temperature), and long enough to be optically thick. As a result, the resulting films showed extremely clean, smooth surfaces (recall the small RMS values from AFM). They were also free from oxides or

5.0 4.0

(a) sample #1 sample #2 sample #3 sample #4 sample #6

3.0 2.0 1.0 0.0 0

30

12

60

90

120

Wavelength (nm)

150

180

10 (b) 8 6 4 2 0 0

sample #1 sample #2 sample #3 sample #4 sample #6

30

60

90

120

150

180

Wavelength (nm)

Figure 4 Ir film optical constants obtained from five different samples (samples #1, 2, 3, 4, and 6, see table I) in the VUV-Visible-NIR (140 to 1700 nm). (a) n. (b) k.

313

Extinction Coefficient, k

Index of refraction, n

contaminants, as evidenced by Energy dispersive x-ray data (EDX) taken on the asdeposited Ir sample films, which showed the Ir peak with nothing else. Thus the optical constants acquired from these films are representative of Ir bulk metals. The Ir film was represented by a classical Drude dispersion layer in the optical model along with a few Gaussian oscillators, to account for both free carrier absorption and inter-band absorption, separately [17]. Surface roughness was modeled by a Bruggeman effective medium approximation (EMA) layer using the RMS thickness values taken directly from AFM results, assuming 50% material and 50% void. Since the films were optically thick, the thicknesses of what were underneath the Ir films— both the fused silica substrate and the Cr adhesion layer—didn’t matter. A nearly perfect fit was thus achieved. Note that only one parametric model set was employed to cover the entire spectral range. Shown in Figures 4 and 5 are comparisons of Ir optical constants (n and k), determined from the VASE data taken on five different samples (see table 1), over the VUV-Visible-NIR (140-1700 nm) and MIR (8000-250 cm–1), respectively. Despite slight differences, the Ir optical constants obtained from these sample films are indeed

80

(a)

sample #1 sample #2 sample #3 sample #4 sample #6

60 40 20 0 0 80

2000

4000

6000

Wave Number (cm

(b)

–1

8000

)

sample #1 sample #2 sample #3 sample #4 sample #6

60 40 20 0 0

2000

4000

6000

Wave Number (cm -1)

8000

Figure 5 Ir film optical constants obtained from five different samples (samples #1, 2, 3, 4, and 6, see table I) in the MIR (8000-250 cm–1). (a) n. (b) k.

314 very close to each other. n (and k)’s are basically lying on top of each other with only slight deviations, mainly seen in the lower wavelengths. No appreciable differences between samples were detected in the middle IR, as shown in Figure 5. This indicates low scattering of infrared light by roughness, and further justifies the fact that these films are optically thick and the optical constants correspond to bulk values. By modeling we remove the effects of surface roughness, and therefore determine the true optical constants of the metal; as a result, there shouldn’t be any differences in the optical constants between samples. The very small differences seen in Figure 5 are likely due to the fact that AFM and optical spectra don’t measure quite the same “roughness”. Ir optical constants provided by other sources, taken on different sample forms (either bulk or thin films), and under different ambient conditions, can be found in the literature [15,16]. There are differences, and these are likely due to different surface

5.0



4.0 3.0 t_roughness=0 nm t_roughness=0.5 nm t_roughness=1 nm t_roughness=1.5 nm

2.0 1.0 0.0 0

300

12

600 900 1200 1500 1800 Wavelength (nm)



10 8 6

t_roughness=0 nm t_roughness=0.5 nm t_roughness=1 nm t_roughness=1.5 nm

4 2 0 0

300

600 900 1200 1500 1800 Wavelength (nm)

Figure 6 A simulation of surface roughness effects on Ir film optical constants (n and k), assuming the roughness thickness is 0, 0.5, 1, and 1.5 nm, respectively. (a) n. (b) k.

315 roughnesses on samples evaluated in each case. We believe the present work represent the best optical constants available for intrinsic Ir material. 3.3 ROUGHNESS AND OVERLAYERS EFFECTS ON APPARENT OPTICAL CONSTANTS Simulations of surface overlayer effects (both roughness and potential hydrocarbons adsorbed onto the sample surfaces) on Ir film optical constants, n and k, were performed using the analysis software. These were based on the acquired true Ir optical constants discussed earlier, and done by adding the overlayers explicitly in the optical model. Figure 6 shows variations of calculated n and k due to a change in roughness layer thickness, assuming the roughness to be 0, 0.5, 1, and 1.5 nm thick, respectively. Clearly, the roughness effects on n and k are substantial; tenths of nm change in surface roughness has a pronounced effect on Ir film optical constants. Notice also increasing roughness decreases n and k.

5.0



4.0 3.0

t_hydrocarbons=0 nm t_hydrocarbons=0.5 nm t_hydrocarbons=1.0 nm t_hydrocarbons=1.5 nm

2.0 1.0 0.0 0

300

12

600 900 1200 1500 1800 Wavelength (nm)

10

8 6

t_hydrocarbons=0 nm t_hydrocarbons=0.5 nm t_hydrocarbons=1.0 nm t_hydrocarbons=1.5 nm

4 2 0 0

300

600 900 1200 1500 1800 Wavelength (nm)

Figure 7 A simulation of hydrocarbon overlayer effects on Ir film optical constants (n and k), assuming the hydrocarbon is polyethylene (PE) and its thickness is 0, 0.5, 1, and 1.5 nm, respectively. (a) n. (b) k.

316 Likewise, another simulation was made of hydrocarbon overlayer effects on Ir film optical constants, as illustrated in Figure 7. A common hydrocarbon—Polyethylene (PE)—was employed to account for the possible hydrocarbon overlayers present, with a thickness of 0, 0.5, 1, and 1.5 nm, respectively. As can been seen, the changes in n and k are similar to changes due to roughness, as shown in Figure 6. Overall, Ir film optical constants determination is very sensitive to surface overlayers, including surface roughness and possible adsorbed hydrocarbons. With that in mind, care must be exercised by future users if they are to employ our reported Ir optical constants. If possible, potential contamination and roughness need to be either removed physically or accounted for in the optical modeling. For space applications (such as the PEACE experiments5-6), these simulations show that Ir provides a highly sensitive base for detecting contaminations. 4.0 Conclusions Ir films with extremely smooth surfaces (RMS < 1 nm, for most cases) were deposited by DC magnetron sputtering onto fused silica substrates at room temperature. Cr was employed as a buffer layer, which served to decrease residual stresses in the Ir films and promote the adhesion between films and substrates. The surface morphologies and microstructures were examined under various deposition conditions of Ar gas pressure and deposition duration, using optical microscopy, AFM, X-ray diffraction, and EDX. Results indicate that, the average surface roughness decreased slightly with decreasing gas pressure in the chamber, and increased as a function of increased film thickness. Variable angle spectroscopic ellipsometry was employed to determine Ir film optical constants from the VUV through middle IR (140 nm to 35 µm). Because the Ir films were optically thick and the surface roughnesses were measured by AFM then accounted for in the optical model, the as-determined film optical constants are expected to be the best available for Ir bulk metals, minimally affected by surface overlayers or microstructure. 5.0 Acknowledgements This work is supported by the NASA Glenn Research Center, Grant No. NAG3-2219. The authors would like to thank Dr. Samir M. Aouadi, Ms. Denise Hornyak, and Professor Suzanne L. Rohde for enlightening discussions and suggestions. 6.0 References 1.

Mumtaz, K., Echigoya, J., Hirai, T., and Shindo, Y. (1993) RF magnetron sputtered iridium coatings on carbon structural materials, Mater. Sci. Eng. A 167, 187-195.

2.

El Khahani, M. A., Chaker, M., and Le Drogoff, B. (1998) Iridium thin films deposited by radiofrequency magnetron sputtering, J. Vac. Sci. Technol. A 16, 885-888.

317 3.

Mumtaz, K., Echigoya, J., and Taya, M. (1993) Preliminary study of iridium coating on carbon/carbon composites, J. Mater. Sci. 28, 5521-5527.

4.

Sun, Y.-M., Endle, J. P., Smith, K., Whaley, S., Mahaffy, R., Ekerdt, J. G., White, J. M., and Hance, R. L. (1999), Iridium film growth with iridium tris-acetylacetonate: oxygen and substrate effects, Thin Solid Films 346, 100-107.

5.

Banks, B. A., DeGroh, K. K., Rutledge, S., Baney-Barton, E., Sechkar, E., Hunt, P., Willoughby, A., Beamer, M., Hope, S., Koo, J., Kaminski, C., and Youngstorm, E. (1999), A space experiment to measure the atomic oxygen erosion of polymers and demonstrate a technique to identify sources of silicone contamination, NASA TM 1999-209180.

6.

DeGroh, K. K., Banks, B. A., Clark, G. C., Hammerstrom, E. E., Kaminski, C., Fine, E. S., and Marx, L. M. (2000) A sensitive technique using atomic force microscopy to measure the low earth orbit atomic oxygen erosion of polymers, NASA TM 2001-211346.

7.

Gerfin, T., Hälg, W. J., Atamny, F., and Dahmen, K.-H. (1994), Growth of iridium films by metal organic chemical vapour deposition, Thin Solid Films 241, 352-355.

8.

Vargas, R., Goto, T., Zhang, W., and Hirai, T. (1994), Epitaxial growth of iridium and platinum films on sapphire by metalorganic chemical vapor deposition, Appl. Phys. Lett. 65, 1094-1096.

9.

Gelfond, N. V., Tuzikov, F. V., and Igumenov, I. K. (1993), Effect of the deposition temperature on the iridium film microstructure produced by metal-organic chemical vapour deposition: sample characterization using X-ray techniques, Thin Solid Films 227, 144-152.

10.

Hamilton, J. C., Yang, N. Y. C., Clift, W. M., Boehme, D. R., McCarty, K. F., and Franklin, J. E. (1992) Diffusion mechanisms in chemical vapor-deposited iridium coated on chemical vapor-deposited rhenium, Metall. Trans. A 23, 851-855.

11.

Kovacs, G. T. A., Storment, C. W., and Kounaves, S. P. (1995), Microfabricated heavy metal ion sensor, Sens. Actuators B 23, 41-47.

12. 13.

Ohring, M. (1991), The materials science of thin films, Academic Press, San Diego, CA. Parsons, R. (1991), in J. L. Vossen and W. Kern (eds), Thin film process II, Academic Press, San Diego, CA, Part II-4.

14. Wasa, K. and Hayakawa, S. (1991) Handbook of sputter deposition technology, Noyes, Park Ridge, NJ. 15.

Lynch, D.W. and Hunter, W. R. (1991) in E. D. Palik (eds), Handbook of optical constants of solids II, Academic Press, Boston, p. 296.

16.

Weaver, J. H., Olson, C. G., Lynch, D. W. (1981) in J. H. Weaver, C. Krafka, D. W. Lynch, and E. E. Koch (eds), Optical properties of metals, Fachinformationszentrum Energie, Physik, Mathematik GMBH, Karlsruhe, Germany, p. 263.

17. Azzam R. M. A. and Bashara, N. M. (1977), Ellipsometry and Polarized Light, North-Holland, New York. 18.

George, J. (1992) Preparation of thin films, Marcel Dekker, Inc., New York.

This page intentionally left blank

THE INFLUENCE OF THE ATOMIC OXYGEN PLASMA ON THE SURFACE AND ON THE PHOTOELECTRIC PROPERTIES OF SOLAR ARRAYS B.G. ATABAEV, L.F. LIFANOVA AND F. RAKHIMOVA Arifov Institute of Electronics Uzbek Academy of Sciences, Tashkent, Uzbekistan A.V. MARKOV AND I.V. TCHOURILO Rocket Space Corporation "Energia", Korolev, Russia

Abstract The results of measurements of the parameters of photoelectric converters as well as the estimates of the effect of low temperature oxygen plasma erosion on commuting busses materials are presented. Exposure of photoelectric converter systems containing metal commuting busses to oxygen plasma is affecting severely their operating characteristics that include the photoelectrodriving power (Unl) and the short-circuit current (Isc). As a result, the spectral sensitivity of the photo converters is reduced. The subsurface doped regions and the changes in the substrate layers that are caused by appearance of additional recombination centers may be held responsible for these effects. The interaction of the oxygen plasma with the metal parts of solar cells, like the contacts of the commuting busses, results in erosion and sputtering of the silver protective layer. The sputtered away silver can be deposited on the cold surfaces of the photoelectric converters (PEC) that, in turn, may also reduce the operating characteristics. Similar experiments with solar arrays (SA) and with the commuting busses that are protected by glass show an increase of the short-circuit current and the efficiency after exposure to oxygen plasma. This is possible because of a decrease in power losses on contacts and elimination of bridging.

1. 0 Introduction A need in photoelectric converters (PEC) with increased energy, longevity and reliability for spacecraft applications triggered intensive efforts in research and development in this important field and their testing and ground-based simulators in conditions of space environment. A significant decrease of the power output of PEC’s and solar arrays (SA) is observed as a result of the influence of positive protons and

319

320 electrons from the Earth radiation belts, atomic oxygen in low Earth orbit and also the multiple temperature cycling of the spacecraft surface due to orbiting the Earth [1]. The solar cells and arrays are among a few semiconductor devices operating in open space. They, therefore, could be heated above 800 °C during solar exposure and could be cooled down to –150 °C when in Earth’s shadow [1]. Chemical-thermal and photon degradation of PEC were observed [1]. The first happens due to the influence of the residual atmosphere and the exhaust gases of the rocket engines as well as to the gaseous contaminants produced by the alkaline and lead-acid constituents of the residual atmosphere of the cities, all of which could interact with the spacecraft. Similar degradation effects were attributed also to interaction with photons. It is difficult in many cases to separate the influence of a particulate radiation from the chemicalthermal degradation. It was understood, that the atomic oxygen in space might destroy Kapton polymeric films that are used as the substrates in PEC manufacturing for the majority of spacecraft. It was also noticed that electrical discharges might occur due to a considerable difference in potentials between dielectric coatings of the upper and the back side of PEC’s that, in turn, can disrupt the operation of the PEC’s or even destroy them. During previous spacecraft flights it was found that silver coatings of commuting busses are prone to oxidation, leading to modification of the surface structure of the films and to an increase in their electroresistance [2]. These effects were causing changes of physical and mechanical characteristics of the coatings as a result of surface erosion, causing the silver protective layer to peel off [3]. If the molecular oxygen, as a result of adsorption on the surface of silver, will interact with the silver forming a thin film of protoxide, the atomic oxygen can oxidize the protoxide up to an oxide [4]: Ag2O + O = 2AgO

(1)

Since the PEC surfaces in space orbits are exposed to the effects of solar radiation and ionospheric streams while on the sun side of the orbit and to the effects of only ionospheric streams when in the shadow, it is of interest to investigate the sequential effects of light irradiation and of oxygen plasma on PEC structures. The purpose of the present paper, therefore, is to compare the PEC parameters before and after irradiation of their surfaces by low temperature oxygen plasma. An additional goal of this paper is also to investigate the processes of erosion of the commuting busses in PEC structures that occur during such interactions. 2. 0 Experimental Methods The photoelectric parameters, namely the short-circuit current Isc, the voltage-current characteristics and the no-load voltage Unl were compared when exposed to natural solar radiation with power P0 = 90 mW / cm2 on the test desk "Ray - C" [5] and also in a solar radiation simulator on the test desk "NUR-1" [6]. After that, the samples were exposed to low temperature oxygen plasma, produced in a vacuum chamber facility, using a high frequency generator and the photoelectric parameters were again recorded.

321 The background pressure in the plasma generator was about 10-6 Torr, and the chamber was filled with oxygen up to pressure of 10-2 Torr. A high-frequency generator of cylindrical geometry was used for the plasma formation [7]. The current-voltage characteristics of all PEC’s were recorded using an automatic recording instrument [6]. The samples exposure in the simulator was conducted in such a way that the Isc of the sample was equal to Isc obtained in exposures to the Sun. The fluence of solar radiation was monitored and measured (PP-7 meter). A high-speed film termoresistance sensor with a linear operating characteristic and a support of reliable thermal contact was used for measurements of the solar cell’s (SC) temperature. Samples were attached to termoinsulating substrates that were provided either with or without forced cooling. Running water was used for forced cooling. The analysis of the surfaces of the PEC’s commuting busses and their erosion by oxygen plasma was done by optical and interference microscopy (MII-4, magnification 500x), with the microphotographs of samples taken before and after the plasma exposure. The sample size of the PEC was approximately 5x5 cm2. The ion irradiation dose (cm2) was determined from equation (2) multiplying the ion fluence by time (s) F=Nit, (2) For the calculation of the fluence of ions (cm-2s-1) arising from a measured ionic current, the following expression was used: Ni=6 1012 i/s, (3) where i - ionic current in µA, s - area of the target or electrode, cm2. 3.0 Experimental Results The photoelectric parameters of the investigated PEC are presented in Table 1. The noload voltage Unl – parameter allows estimating the entire range of the forbidden zone potential, i.e. its increase leads to proportional increase of the efficiency and at the same time is in inverse relation to the level of recombination losses. As can be seen from the table, the values Unl for unirradiated samples exposed to Sun on the bench "Ray - C" on the average exceed by 6 % the passport values for Unl. A similar analysis of a short-circuit current indicates at a large decrease in Isc (on average 24 %.) concerning nameplate data at Sun exposure. It is necessary to note that the power of the luminous flux during the trials for nameplate data differ somewhat from data during our analysis. This is indicative of possible differences in the solar spectrum depending on seasonal changes of height and latitude of the Sun as well as the composition of the atmosphere, etc.

322 Table 1: The Photoelectric Parameters of PEC The original data, before solar radiation

At natural solar exposure on the test desk “Ray-C”

Under the simulator of After exposure solar radiation on the in plasma test desk "Nur-1"

Sample ʋ

12

13

22

12

13

22

12

13

22

12

13

Isc, mA

950

950

900

715

722

689

715

723

689

705

714

0,546

0,544

0,533

0,594

0,593

0.592

0,585

0,588

0,4

0,4

0,4

0,6

0,6

0,61

0,65

0,62

0,64

0,65

0,62

10,4

10,5

10,0

12,3

11,8

11,6

12,1

11,6

0,16

0,18

0,19

0,16

0,18

20 90

20 90

20 90

20 90

20 90

Unl, V

0,52

0,51

0,5

Iopt, mA

900

850

830

Uopt, V

0,365

0,365

0,36

Filling. Factor ff, relat. Units Efficiency η, %

Resist Consis Rc, om Temper., 0C Pow. Of Stream, MW/Cm2

AMI

AMI

AMI

50 90

50 90

50 90

The values of the sample exposure levels in the simulator, Unl on average are 9.6 % higher than obtained in exposure by the Sun on the bench "Ray - C". This is explained by the differences of the spectral content of natural solar radiation and the tungsten bulb of the simulator. This observation does not prevent one from obtaining the information about changes in photoelectric parameters of PEC’s due to the influence of the oxygen plasma, as the changes in PEC parameters are fixed in each of separately carried out preliminary exposure experiments. Accordingly filling factors “ff” and efficiency of the artificial exposure, ηat are higher than in the case of at solar irradiation. In exposure experiments in the simulator there is almost no difference in parameters Isc and Unl in comparison with exposure to the Sun. To evaluate the influence of oxygen plasma on photoelectric parameters, the PEC samples were exposed to the plasma for 4 hours that corresponds to an equivalent simulated annual exposure in space. The no-load voltage Unl after the oxygen plasma exposure has decreased on average by 1.2 %, the short-circuit current on average has decreased by 1.3 %. The PEC efficiency on average has decreased by 1.5 %. The parameters ff, Rc after the exposure in plasma did not changed. Considering the obtained results, it is necessary to note that the chemical reactions happening at the elevated temperature on the surfaces of the solar cells can cause a short circuit of the electron-hole transitions, a corrosion of the contacts and a darkening of the coatings [1]. Due to a photon degradation mechanism, the current is decreasing after several hours of intensive exposure, apparently, due to the release of

323 entrapped point defects by dislocations. In flight experiments it was shown [8] that the spectral characteristics of PEC’s irradiated by electrons are changing as a result of the influence of light causing changes in the spectral sensitivity. It is known [1] that the spectral sensitivity linearly depends on the wavelength. It was also shown [8] that the onset of degradation is stipulated by "red" light penetrating deeper into the semiconductor material and is determined by the width of the forbidden band in semiconductors. The lower edge of sensitivity for exposed to light PEC surfaces depends mainly on the speed of surface recombination and is associated with the ultraviolet region. There is an inversely proportional dependence of the influence of irradiation wavelength on the intensity of charged particles’ flux in conditions of simultaneous irradiation. It follows that in space conditions PEC’s should be damaged more by a short wavelength UV radiation than in the accelerated laboratory tests. Apparently, in our experiments the degradation is affecting the volume bulk regions in the base area of the device that eventually results in the decrease of the diffusion length of minority carriers. These processes in the subsurface-doped layers then, under the influence of oxygen plasma, change both the Unl, and the Isc. Thus, after processing the PEC’s in oxygen plasma, the decrease in parameter values is possible to attribute to chemical and photon degradation. The investigation of the influence of oxygen plasma on PEC’s has shown that the morphology of the surfaces of the plating layers on the commuting busses is changing [9]. The surfaces of solar cells covered by protective glass plates are less sensitive to erosion than the ~50 :m thick copper busses (contacts) covered with a 6 :m thick layer of silver. Figure 1 presents a microphotograph of such copper contact at a magnification of x500.

Figure 1. A microphotograph of a site of the silvered copper bus of sample N 13, ×500.

324 As can be seen from Fig. 1, a quite uniform surface with a fine grain structure and without any cracks is visible. It was shown, that even at short exposure times in plasma of ~ 7200 s that correspond to a accelerated simulated semi-annual exposure in space conditions, the protective layer of silver used as the coating on the commuting busses is exposed to the oxidation process. After exposure in plasma, just by visual examination, it is possible to define, from the blackening of the surfaces of the busses, that the surface has been exposed to an oxidation process. Analyzing the exposed buss surfaces in an interference microscope, black and light-brown discoloration regions corresponding to different oxides of silver – AgO and Ag2O were observed. Upon increasing the exposure time of the sample in plasma twice that corresponds to the duration of the annual exposure in space conditions, the surface was found to almost completely oxidize. At a smaller irradiation dose, the number of lightbrown sites was found to increase. The protoxide of silver Ag2O in vacuum conditions is transparent, but during handling of the irradiated samples in atmosphere they change the color to light brown. The most dramatic results in oxygen plasma experiments were obtained from the edge region of the commuting buss of sample ʋ 13 (Fig. 2). It was found visually that the upper edge of the sample had a distinctive red discoloration, indicating that the copper was exposed in this region. This allowed concluding that at the edges of the commuting copper buss, where the erosion was found to be the strongest, the etching of the protective layer up to the base copper was observed. The erosion and chemical sputtering off of the protective layer of silver is explained by formation of oxides of silver and their subsequent layer-by-layer etching. During the sputtering of the silver, a resputtering process is possible, i.e. the deposition of sputtered away silver onto the surfaces of PEC’s, with the probability of such process increasing upon cooling in the shielded areas.

Figure 2. A microphotograph of an edge of the sample N 13 after exposure in oxygen plasma for 4 hours.

This assumption is supported by data from reference [10], where it was shown that up to 15 % of materials removed from a spacecraft surface is resettling back, being scattered by surrounding atmosphere. Figure 3 represents the load volt-ampere characteristics of an SA, obtained under the exposure in the solar irradiation simulator with a flux around 70.4 W / cm2

325 before and after the exposure to oxygen plasma. The duration of plasma exposure was estimated to correspond to three-years in LEO space environment. It can be seen, that the short-circuit current increased approximately by 6 %, the no-load voltage remained without change, the value of the filling factor also practically did not change and was at a level of 0.62. The efficiency increased by 0.8 % and became equal to 12.9 %. Our data is in good corellation with data in reference [10] where a degradation of the efficiency of an SA was reported with a subsequent improvement of the characteristics after longer flight times. In addition, the influence of the ultraviolet radiation on the degradation of the coefficient of solar absorption and the role of atomic oxygen in presence of the ultraviolet radiation was also discussed and it was shown that the AO/UV synergistically interacts with the contaminants, oxidize them and reduce the absorption factor for PEC’s with protective glasses, i.e. increases the efficiency that is also in good correlation with our results. Busses, unlike the situation of samples shown in table 1. I, mA

2

1 690

645

ff r1 r2

= = =

0,62 12.1

%

12.9 %

1.62

U, V

Figure 3. A load volt-ampere characteristic of SA, obtained under the simulator of solar radiation: 1 – before exposure in oxygen plasma; 2 – after the effect of plasma, P0=70,4 W/cm2.

In the results shown in Fig. 3, a PEC system was utilized with a protective glass covering it’s surface, including the commuting busses. It should be noted that in interpretation of the results it is necessary also to take into account that the artificially created oxygen plasma is a very complex mixture of different ionic species that contains at least 10 different types of ions [11]. The ambiguities of interaction of charged particles with metals and dielectrics, i.e. the commuting busses covered with a layer of silver and glass should be taken into account. Presently, an investigation of the PEC’s coming back from space is being conducted that will allow making more valid conclusions and estimating the appropriate choice for an ion source as the simulator.

326 3. 0 Conclusions Exposure of photoelectric converters with metal commuting busses to oxygen plasma is affecting severely their operating characteristics that include the photoelectrodriving power (Unl) and the short-circuit current (Isc) that reduces the spectral sensitivity of the photoelectric converters. Additional recombination centers that are created in the subsurface-doped regions may be held responsible for these effects. The silver protective layer used on the contacts of the commuting busses eroded in experiments with oxygen plasma exposures. The sputtered away silver was found to be deposited back on the cold surfaces of the photoelectric converters (PEC) that reduced further the operating characteristics. Experiments with solar arrays with the commuting busses that are protected by glass, had shown an increase of the short-circuit current as well as in the efficiency after exposure to oxygen plasma. This finding was explained by a possible decrease in power losses on contacts and elimination of bridging. 4. 0 Acknowledgements The authors express deep gratitude to B.M. Abdurahmanov for critical reading of the paper and involvement in the discussion of results. 5. 0 References 1. Koltun M.M. Solar cells. Moscow, Science (1987) 191 p. 2. Akishin A.I., Teplov I.B. PCOM, The Simulation of effect of space radiations on materials 2 (1992) p. 47 - 57. 3. Akishin A.I., Guzhova S.K., The Interaction of ionospheric plasma with materials and equipment of space vehicles, PCOM 3 (1993) pp. 40 - 47. 4. BSE, M., Soviet encyclopedia 23 (1976) p. 638 5. Abdurahmanov B.M., Soloveychik V.I., Hayrullin I.I., The test desk "Ray - C", Geliotehnika 5 (1991) pp. 34 - 37. 6. Abdurahmanov B.M., Baydakov S.G., Soloveychik V.I., Chirva V.P., The Units and elements of solar photoelectric stations with concentration of the radiation, Tashkent, FAN (1993) p . 200 7. Radjabov T.D., Lifanova L.F., Invention certificate ʋ 758798, The device for obtaining multilayer structures, (1977). 8. S.N. Vernov, The model of space., M 2 (1983) p. 771 9. Atabaev B.G., Kareev M.CS, Saidhanova N.G., etc., The Influence of plasma of atomic oxygen onto the surface and photoelectric properties of solar cells, Materials of XIII International conference “ Interaction of ions with surface”, v. 2 (1997) pp. 104 - 107. 10. Eremin P.A., Zayavlin V.R., The Influence of rocket drives of space vehicles onto the battery of photoconverters, Geliotehnika 3 (1998) pp. 3 - 12. 11. Parhutik V.P., Labunov V.A., The Plasma anodization, Minsk: Nauka i technika. (1990) p. 280

SOME ASPECTS OF SIMULATION OF OUTGASSING PROCESSES IN THERMAL VACUUM EXPOSURE OF COATINGS APPLIED TO SPACE VEHICLES R.H. KHASSANCHINE, A.V.GRIGOREVSKIY, Y.P. GORDEEV Close joint-stock company “Institute “Kompozit-Test” 4, Pionerskay street, 141070 Korolev, Moscow region, Russia Phone: + 7 095 513 20 20 Fax: + 7 095 513 20 75 E-mail: [email protected] [email protected]

Abstract The mathematical model describing the outgassing processes under thermal vacuum exposure to coatings applied to space vehicles is presented. The model is based on assumptions that changes of outgassing product concentrations in material under investigation are due to its destruction under environmental influence, desorption from surface on the materials-vacuum boundary, first order chemical reactions, and diffusion specified by the aforesaid processes. Analytical expressions for the space-time distribution of outgassing products in material were obtained from the model equations for the influence of individual processes on outgassing kinetics to be investigated numerically. According to the obtained results, the analysis of how the particular model parameters and their ratio may influence or alter the outgassing product concentrations and outgassing kinetics has been carried out. 1. 0 Introduction. Outgassing of materials working in space environment are one of the main sources that contaminate sensitive surfaces of optical systems, cells of solar batteries, etc. This is, in turn, a factor that potentially limits serviceability and service life of space vehicles. Consequently, production of prediction models describing the process of outgassing of coatings is a present-day problem to predict the sensitive surfaces contamination. Simulation of this complicated physical-chemical process is possible under some assumptions. For example, in [1] the author suggested that outgassing is a first order reaction [2], i.e. flux density of each component emitting through the material-vacuum boundary is proportional to its amount remaining in material at the given moment. We think this assumption is not always correct because flux density of each outgassing component is proportional to its concentration in the near-surface layer at the given moment that is far from being proportional to its amount remaining in material. Our model takes into account the fact that outgassing is a result of several processes occurring inside the material and on its surface, as a response on external influence. When choosing processes determining the phenomenon under investigation one should to identify the components taking place in it and estimate levels of factors exerting influence on specified material being in service.

327

328 2.0 Mathematical Model. Usage of mathematical formulation permits to assign certain mathematical symbols to each aspect of the process under investigation. As a result, interaction between different parameters of the process become more evident. In addition, mathematical model gives a basis for numerical analysis that allows to obtain not only descriptive but also prognostic data. We propose the model describing changes of outgassing products concentrations in the material under study and outgassing kinetics of those through the surface formed by the material-vacuum boundary, under thermal vacuum exposure. The change of concentration C n ( x, t ) of n-component (molecules or atoms) of outgassing process in material applied onto the hermetic substrate is stipulated mainly by the following processes: destruction of material;- desorption from the surface on materialvacuum boundary; chemical reactions in material; evaporation of material through the surface on material-vacuum boundary; diffusion caused by the aforesaid processes. This model is based on the following assumptions: 1. Thickness of a sample of material is significantly less as compared to other linear dimensions so one may neglect edge effects scrutinizing only one-dimensional problem; 2. Sample temperature is fixed; 3. Coefficients of diffusion, desorption, and thermal destruction depend by no means on time, they are defined by temperature of the given sample of material. Dependence of two primary coefficients on temperature is described by Arrenius correlation; 4. Some components of outgassing process can be produced as a result of destruction of other ones; their other interactions is not considered in the model; 5. Outgassing components in material are involved only in the first-order reactions; 6. Material is evaporated from total surface with steady rate that depends for the given material only on temperature; 7. Outgassing occurs through the materal-vacuum boundary only. For the present, the macroscopic approach is the only possible one to study of the cited processes in composites and on their surfaces. Hence here we are obliged to speak only about effective diffusion and desorption coefficients that are the parameters by means of which we describe processes that are observed in laboratory and on-board. Let’s express the concentration of outgassing components that are not the components that could be generated in the course of destruction of other ones as ɋ i ( x, t ) (i=1,2,3…N). Then change of their concentrations in a sample in the network of made assumptions can be described by the following differential equations:

(1)

∂C i ( x, t ) ∂ 2 C i ( x, t ) M = Di − ¦ σ i → j C i ( x, t ) − χ i C i ( x, t ) + S i ( x , t ) ∂t ∂x 2 j =1 at

∀x ∈ (0, h − υ ⋅ t ) , t > 0 , υ ⋅ t < h

329

(2) C i ( x, t ) (3) Di at t > 0 ,

t =0

= Ri

∂ɋ i ( x, t ) ∂x

x = h −υ ⋅t

at

∀x ∈ [0, h]

+ k i C i ( x, t )

x = h −υ ⋅t

=

∂C i ( x, t ) ∂x

x =0

=0

σ i→ j is a weighting coefficient of thermal destruction of i-component through jchannel; χ i is chemical reaction rate with involvement of i-component; Di, ki are

where

effective coefficients of diffusion and desorption of i-component respectively; Ri is the concentration of i-component in material at initial moment; S i ( x, t ) is a i-component source function; h is the thickness of material; υ is an evaporation rate of material. The second and the third terms of the second member of equation (1) may be combined with the source function but not all of variants will be included then. The functions ɋ i ( x, t ) obtained when solving the equations (1) - (3) with the weighting coefficients

σ i→ j

can enter the analogous equations describing changes of

concentration of j outgassing component (j=1,2,3…L) that can be possibly generated during the destruction of corresponding i-component. Of course, there may arise cases when R j , σ j →l and σ i → j =0 (where σ j → l - is the weighting coefficient of thermal destruction of j-component through l-channel). Thus, examining all possible outgassing components, we have constructed the differential equations system describing changes of C n ( x, t ) in a sample of material. For the most practical tasks solving this system is not very difficult. Because of their unhandiness here we give only some of the tasks that are, in our opinion, of practical interest. While solving the equations (1)-(3), one may determine the outgassing rate of icomponent from the unit of surface

(4)

dFi (t ) = (υ + k i ) ⋅ C i (h − υ ⋅ t , t ) . dt

Dependence of flux of i-component emitted from material through the unit of surface on time t is received as soon as the equation (4) was integrated. Temperature of materials in-situ and, as a consequence, coefficients of diffusion and desorption is usually a time depending function. It is difficult to solve the task (1)-(3) with such coefficients. That is why the “observation” period was divided into finite intervals to perform computational investigation of influence of variable temperature on outgassing kinetics. Here was also assumed that inside each interval sample temperature is fixes undergoing a certain jump at the end. If before the time point t0 the sample was at the temperature T1 , then, from the moment t0 +0 it became T2 , values of coefficients Di, ki,

σ i→ j , χ i

ɢ S i ( x, t ) change as well, together with temperature, that leads to

330 change of ɋ i ( x, t ) in material. We got analytical solutions of the equations (1)-(3) for computational analysis of processes taking place with temperature changes in material (here it is given for the case when S i ( x, t ) = 0 ɢ υ = 0 ):

[ (

∞

)]

2

Ci ( x, t < t 0 ) = 2Ri ¦ Ak0 exp − t β i0 + λk Di0 cosλk x k =1

∞

(5) ɋ i ( x, t > t 0 ) = 2 Ri exp[ −τ ⋅ β i − β i0 t 0 ]¦ Bn exp( −τ ⋅ λ2n Di ) ⋅ cos λ n x n =1

, 0 k

A =

Bn =

2

2

sin λk h

ki0 + λ2k Di0

λk

ki0 Di0 + h ki0 + λ2r Di0

(

2

2

)

∞ § sin(λk −λn )h sin(λk +λn )h· ki2 +λ2nDi2 ¸¸exp(−λ2k Di0t0 ), + ⋅ Ak0 ⋅¨¨ ¦ 2 2 2 λk +λn ¹ ki Di + h(λnDi + ki ) k=1 © λk −λn

where the upper index 0 is attributed to values of parameters existing before the time point t0 , τ = t − t 0 ;

λk , λn

are solutions of appropriate transcendental equations:

tgλ k h =

k i0 λDi0

and

tgλ n h =

ki ; λDi

β i0 , β i are effective first-order reaction rates that occur in material n

n

β i0 = ¦ σ 0 i → j + χ i0 j =1

,

β i = ¦ σ i→ j + χ i . j =1

Analytical expressions of dependence of individual outgassing product fluxes and their concentrations during this process in a sample of material on time were obtained from the mathematical model. Omitting analytical study of these functions we give only results of computational experiments.

3. 0 Results of Numerical Computation and Discussion. To determine influence of individual processes on outgassing here were performed numerical experiments, their results are given and discussed below. We confined ourselves to i-component alone, calling it as a volatile component (VC) and omitting indexes of parameters. To simplify perception of results here was chosen R=1, h=100 µm (characteristic thickness of temperature control coatings) whereas other parameters were being varied.

331 To show separate processes and determine their influence on outgassing, fig.1 gives plots of time dependence of VC concentration in material with different model parameters. With the same values of parameters D [µm2/s], k [µm/s] and β [s-1], the

plots 1ɚ and 1ɜ represent behaviour of change of C ( x , t ) for υ = 0 and υ = 0.015 [µm/h] respectively. Here is seen that in the second case C ( x , t ) , on the substrate side, with time becomes significantly greater then in the first one. This is caused by the timeindependent evaporation of the desorption depleted near-surface layer that, in turn, reduces gradient of VC concentration in it. The value of parameter D is five times greater on the plot 1ɫ as compared with the plot 1ɚ that is appreciably affected on behavior of the function C ( x , t ) . The plot 1d shows the influence of augmentation of values of parameters k, D and

β

on distribution of concentration at time point t0 =400

hrs.

Figure 1. Distribution of VC concentration in material with different model parameters:1ɚ – D=0.001, k=0.01,

β =10E-7, υ =0; 1ɜ – D=0.001, k=0.01, β =10E-7, υ =0.015; 1c – D=0.005, k=0.01, β =10E-7, υ =0; 0 – D =0.001, k =0.01, β =1.0E-7, D=0.002, k=0.024, β =1.80E-7, υ =0. 0

0

1d

332 The plot 2ɚ shows dependence of change of C ( x , t ) on non-dimensional parameter

Nu = k ⋅ h / D during 1000 hrs in the case of fixed boundary. Here is seen that desorptive or diffusive inhibitions of outgassing process occur when values of Nu are ≤0.01 or ≥50 respectively. We are so detailed in study of influence of value of parameters on C ( x , t ) in order to give more clearness in comprehension of outgassing process. Having examined relationships between C ( x , t ) and model parameters, we examined their influence on outgassing kinetics. If we know

C (h − υ ⋅ t , t ) , we may

Figure 2. Influence of parameter Nu on outgassing process: 2ɚ – distribution of VC concentration in depth during 1000 hrs in dependence on the value of parameter Nu for υ =0; 2ɜ – dependence of VC outgassing kinetics from the unit of surface in dependence on the value of parameter Nu for υ =0 .

determine the outgassing rate from (5). When integrating it by time we get the value of VC flux emitted through the unit of surface during the time interval t . In laboratory envorinments is often realized the event when sample thickness and temperature are considered to be fixed, then experession for flux from the unit of surface has a simplier view:

( 6)

1 − exp(−bt ) , where b = λ2n D + β . 2 ª k λD º n =1 b ⋅ «1 + h( + n )» D k ¼ ¬ ∞

F (t ) = 2kR ¦

The plot 2ɜ shows dependence of integral VC flux emitting through the unit of surface on the material-vacuum boundary on the parameter Nu and time for υ =0. This plot confirms validity of detailed numerical analysis of the function C ( x , t ) because desorptive or diffusive inhibitions of outgassing process become apparent more distinctly on the plot 2ɚ.

333

Figure 3. Plots of VC outgassing rate (3a) and integral flux (3b) through the unit of surface in dependence on time and evaporation rate of material υ with fixed values of other model parameters: D=0.001; k=0.01;

β =10E-7.

When making numerical analysis of influence of outgassing rate of material on outgassing kinetics this parameter has been chosen in such a way that evaporation of material was 530% while “observing” the process. The plot 3ɚ shows the outgassing rate dF / dt in dependence on time and evaporation rate of material. At first, as is shown on the plot, the outgassing rate is maximum, the value of parameter υ making insignificant influence on the rate. It is due to the outgassing process at this time interval is mainly defined by desorption of the near-surface layer. Influence of the parameter υ , with time, became noticeable. It is stipulated by two reasons. Firstly, increasing the value brings to more rapid evaporation of the VC desorption depleted near-surface layer. Secondly, evaporation gives contribution to outgassing process which is equal to υ ⋅ C ( h − υ ⋅ t , t ) dt within any given small time interval from t till t+dt. The aforesaid about influence of evaporation rate can be also referred to the integral flow function F (υ , t ) emitting through the unit of surface on the material-vacuum boundary. This is shown on the plot 3ɜ where its dependence on time and evaporation rate of material is given.

4.0 Conclusions The goal of the work is an attempt to create the mathematical model that could form the basis for an engineering prediction method of outgassing under influence of space environment by coatings that are applied to space vehicles. Using the model, numerical

334 calculations on how different processes exert influence on outgassing have been carried out. Analys of results shows that each of accountable by the mathematical model physical-chemical phenomena give significant contribution. They helped us to highlight correctly key point while preparing and carrying on experiments concerning investigation of outgassing kinetics that, in turn, so far qualitatively confirm correctness of principal model’s aspects. It is obvious that experimental determination of all parameters of the model with adequate accuracy is a complicated problem. But having a good model, the present-day nonlinear regression analysis methods (NRAM) may help to determine missing parameters using sets of experimental points. Having determined the parameters for the prediction model with the help of NRAM you have to carry on control experiments on outgassing kinetics to make them more precise. Taking into account the peculiarities of the source function S i ( x, t ) under exposure to UV and ionizing radiation, essential principles of the model may be used in these cases. Later on the model will be spreaded to solve applied engineering tasks.

5.0 References 1. Delphine FAYE: «Calculation approach for outgassing curves of PU1 paint and molecular contamination modeling: ground testing and computer simulation», 8th International Symposium on «Materials in a space environment», Arcachon-France, 5 - 9 June, 2000. 2. J.Guillin : «Evaluation of isothermal outgassing kinetics for some materials used in space», Proceedings of the Third European Symposium on Spacecraft Materials in Space Environment, ESA SP-232, The Netherlands, October 1985, pp. 35-38. 3. Fitter Add-Inn. (1998), http://polycert.chph.ras.ru/fitter.htm

VACUUM ULTRAVIOLET RADIATION CHARACTERIZATION OF RF AIR PLASMA AND EFFECTS ON POLYMER FILMS JOYCE DEVER NASA Glenn Research Center, 21000 Brookpark Rd., Cleveland, OH 44135 CARA McCRACKEN Cleveland State University, 1983 East 24th St., Cleveland, OH 44115 ERIC BRUCKNER QSS Group, Inc., 21000 Brookpark Rd., Cleveland, OH 44135

Abstract Radio-frequency (RF) plasma ashers have been extensively used in atomic oxygen (AO) durability screening of candidate spacecraft materials. Because RF excitation/deexcitation of a gas produces ultraviolet radiation, samples exposed to AO in the plasma asher are also exposed to ultraviolet radiation. In an effort to quantify the vacuum ultraviolet environment in the asher, this paper will describe measurements of the intensity of vacuum ultraviolet (VUV) radiation over the 115-200 nm wavelength range for a plasma asher operated on a feed gas of room air. Measurements were made in narrow bands (9-27 nm) within the 115-200 nm range and for the full 115-200 nm range. VUV intensity in the asher was compared to space solar intensity in the same wavelength ranges. Preliminary tests were conducted to examine the use of the asher as a VUV source while examining general VUV degradation trends for common polymer materials. Mechanical properties of polymer films of Kapton HN, Teflon FEP, Tefzel ZN, Mylar, and polyethylene were evaluated as a function of asher VUV exposure. For these tests, samples were wrapped in aluminum foil with the gage area visible under magnesium fluoride windows in an effort to protect them from atomic oxygen while still allowing VUV radiation of wavelengths down to 115 nm to reach the samples. 1.0 Introduction Erosion of polymer materials observed during the first Space Shuttle flights in the early 1980’s was attributed to material interactions with low Earth orbit (LEO) atomic

335

336 oxygen. [1]. These observations led to development of many different methods for ground based evaluation of the LEO atomic oxygen durability of materials. One method for AO ground testing uses commercially available radio-frequency (RF) plasma chemistry reactors, referred to as plasma ashers, which operate by applying RF power between two electrodes which surround a Pyrex reaction chamber. When a feed gas is introduced into the reaction chamber and the chamber is evacuated to a low pressure, a thermal plasma is produced with an energy of approximately 0.04 eV [2]. Relevant to materials interactions, the plasma contains electrons, ions, excited and metastable atoms and molecules, and photons [3]. Plasma ashers are generally used to simulate the atomic oxygen environment of LEO, however, near ultraviolet (NUV) radiation (200-400 nm) and vacuum ultraviolet (VUV) radiation (below 200 nm), molecular oxygen and electrons, also present in space, are produced in the plasma [4]. This paper will confine its scope to measurements and discussion of the VUV and atomic oxygen components of the RF plasma. The plasma produced in the asher is considered a low-pressure glow discharge. The low-pressure glow discharge produces mainly spectral lines, dependent on the composition of the feed gas, although band and continuous spectra can also be produced [5] This paper describes characterization of the vacuum ultraviolet (VUV) emission in the wavelength range between 115 and 200 nm, in the entire range and within 9-27 nm wavelength bands, for an RF plasma asher operated using a feed gas of room air. Because the intensity of light produced by the plasma asher can be adjusted by varying the power and tuning of the RF source, use of this device as an adjustable-intensity vacuum ultraviolet radiation source was investigated while gathering data to examine general trends of VUV-induced mechanical properties degradation for common spacecraft polymer materials. This paper will provide results of exposure of polymer films to the plasma asher VUV while attempting to protect them from AO. Issues and problems related to use of the plasma asher as a VUV source will be discussed. 2.0 Experimental Equipment and Procedures 2.1 DESCRIPTION OF THE RF PLASMA ASHER A Plasma-Prep II plasma chemistry reactor manufactured by Structure Probe, Inc. was used for these experiments. This instrument uses a radio-frequency (13.56 MHz) generator to apply RF power (variable, up to 100 watts) between 2 semi-circular electrodes that surround its cylindrical Pyrex reaction chamber. A tube fed into the reaction chamber delivers the feed gas at a constant flow rate. In the experiments described in this paper, a feed gas of room air was used. The asher’s reaction chamber was evacuated to approximately 0.15 torr (20 Pa), and then RF power was applied to produce a low-pressure gaseous discharge. In addition to the RF power adjustment, the asher instrument provides a tuning adjustment to assure matching of the output impedance of the RF generator and the capacitive load of the reaction chamber. In these experiments, RF power was at the maximum value of 100 watts, and the tuning

337 adjustment was used to vary the intensity of visible light in the asher’s glow discharge between bright and dim conditions to examine the VUV intensity limits. 2.2 MEASUREMENT OF VUV IN THE PLASMA ASHER The apparatus used to measure the VUV content of the RF plasma is shown in Figure 1. A special stainless steel flange was fabricated in order to position a VUV detector in the plasma environment to measure the VUV intensity. An o-ring vacuum coupling was brazed to the flange. This coupling included a Viton o-ring compression seal around the glass VUV phototube. The detector used was a Hamamatsu model R1187 phototube, which contained a cesium iodide (CsI) photocathode and a MgF2 input window. This phototube provided a spectral response in the 115-200 nm wavelength range. The phototube was wired via the manufacturer’s recommended circuitry to a 15 V power supply and to a Keithly model 617 electrometer in order to measure the VUV signal. Special provisions were needed to shield the detector and its leads from RF noise in order to prevent the noise from contributing to the measured signal. Inside the reaction chamber, an aluminum tube with a 36% transmitting stainless steel mesh screen cover was secured to the flange covering the phototube/coupling assembly. Figure 1a shows the asher flange with the detector/coupling covered by this shield. The flange was sealed to the asher’s Pyrex reaction chamber with a urethane rubber gasket when a rough vacuum environment was applied to the chamber. On the exterior of the asher chamber, the detector leads were fed through an aluminum tube about 15 cm in length secured to the flange. The detector leads were twisted to minimize the loop area, and two grounding straps of copper braiding were fastened from edges of the flange to the chassis of the asher instrument as shown in Figure 1b. The asher, assembled with the detector flange, shielding, and an exterior cover in place, is shown in Figure 1c. In order to determine VUV signal within narrow spectral bands (9-27 nm bandwidth), 2.54 cm diameter narrow bandpass filters were used. Nine different filters were used with peak wavelengths ranging from 122 nm to 200 nm. These filters were obtained from Acton Research Corporation (ARC). The characteristics of the VUV narrow bandpass filters are given in Table 2. To measure the signal from the plasma in each of the VUV wavelength bands, the filter was positioned over the input window of the phototube and underneath the protective screen. The asher chamber was evacuated to approximately 20 Pa (0.15 torr), and RF power was applied to generate the plasma. RF power was adjusted to the maximum level. Then, the visual intensity of the plasma was adjusted to a medium level using the asher’s tuning control. After approximately 3 minutes of operating at a medium visual intensity, the intensity was adjusted to the desired level (bright or dim), and a signal reading was taken. The plasma was then extinguished and then re-adjusted to the desired level (bright or dim) using the tuning adjustment, and a second reading was obtained for the same brightness level.

338

Mesh screen Asher flange, interior (plasma-facing) surface

Interior shielding tube for phototube

(a) Exterior shielding tube for phototube leads

Grounding strap (1 of 2)

(b)

Exterior cover

(c)

Figure 1. Apparatus used to measure VUV signal from the RF plasma asher including the asher flange with detector shielding tube/screen installed (a), the flange installed on the asher chamber with the shielding tube for the detector leads installed and grounding straps attached (b), and the completely assembled apparatus to include exterior RF protective shielding (c).

The “bright” condition represented maximum visual intensity of the plasma achievable by making tuning adjustments, and the “dim” condition represented the lowest level of intensity achievable such that any further de-tuning resulted in extinguishing the plasma. For each filter, 2 or 3 readings were taken at each level of visual intensity to determine the intensity limits of the plasma. The signal was also obtained with no filter in place for bright and dim conditions to determine VUV signal for the full sensitivity range of the VUV phototube, 115-200 nm. 2.3 CALIBRATION OF THE VUV PHOTOTUBE DETECTOR The VUV phototube used for the asher VUV measurements did not provide a direct measure of intensity, rather, it provided signal in current which was assumed to be proportional to intensity. It was necessary to obtain calibration factors so that intensity

339 of VUV radiation from the RF plasma could be directly calculated from measurement of the detector signal. The VUV phototube used to measure the RF plasma was calibrated in a separate vacuum facility using a deuterium lamp calibrated by the National Institute of Standards and Technology (NIST). NIST calibrated this lamp, referred to herein as the VUV standard source, for spectral irradiance at a source-to-detector distance of 25.4 cm within the 115-200 nm wavelength range using a 1.6 nm bandpass. This calibration is described in detail in reference 6. The NIST calibration data were used along with the bandwidth and transmittance data for the VUV narrow bandpass filters, shown in Table 1, to calculate the output intensity of the VUV standard source within the bandwidth of each filter at a 25.4 cm calibration distance. TABLE 1. Properties of VUV Narrow Bandpass Filters and Calibration Factors for VUV Phototube within the Filter Wavelength Ranges

ARC Filter Label 122-N-1D 130-N-1D 140-N-1D 145-N-1D 150-N-1D 170-N-1D 180-N-1D 190-N-1D 200-N-1D No filter

Peak wavelength (nm) 121.6 129.5 138.5 145.5 151 170.1 182 188 198 --

Calibration Factors for VUV phototube bandwidth, full-width at using VUV filters Peak half-maximum (W⋅cm-2⋅A-1) transmittance (nm) Transmittance (%) 9 13 4849 14.5 18.4 512 20 17 572 21.7 14.6 506 15.5 15 435 26.9 17.8 2483 21.5 19 1233 18 15.5 1518 20 17.2 5116 --134

The VUV standard source was installed in the calibration facility along with the VUV phototube. Signal current from the VUV standard source was measured with the VUV phototube as a function of distance between 51 cm and 96 cm (20 in and 38 in), the available source-to-detector distance range in the calibration facility. In order to determine calibration factors, it was necessary to determine VUV detector signal for 25.4 cm, the distance used by NIST for the VUV standard source calibrations. For each filter, the signal vs. distance data was fit to an exponential decay function, which produced an excellent fit for each data set. Then, the signal at a 25.4 cm distance was calculated from this function. Dividing intensity of the VUV standard source by signal for the same calibration distance and the same filter wavelength range provided a calibration factor, in W⋅cm-2⋅A-1, for the detector in that wavelength range. Calibration factors calculated using this method are shown in Table 1. These calibration factors were multiplied by signal values obtained from the measurements of the plasma to determine VUV intensity, in W⋅cm-2, in the filter wavelength ranges and in the full 115200 nm wavelength range of the VUV phototube for dim and bright visual intensities of the plasma. Intensities were corrected to account for the 36% transmitting mesh used to cover the phototube. The VUV signal measurements were taken approximately 3 minutes after the plasma was turned on; however, near the conclusion of these experiments, it was found that the VUV signal was proportional to operating pressure, and the plasma reached a steady operating pressure in approximately 90 minutes. Additional experiments determined that the 90-minute VUV signal was approximately

340 0.38 times the 3-minute VUV signal . These data were, therefore, corrected using this correction factor in order to estimate the steady-state signal after 90 minutes of plasma operation. 2.4 PLASMA ASHER VUV EXPOSURE OF POLYMER MATERIALS In order to investigate the use of the plasma asher as a VUV source for space simulation testing, polymer film samples were exposed to plasma VUV while attempting to protect them from AO. Descriptions of polymer materials exposed in these tests are given in Table 2. Other than polyethylene, these polymers are widely used in the aerospace industry. Polyethylene was included for comparison because of its simple chemistry and susceptibility to ultraviolet radiation degradation. Tensile test specimens were die-cut from the polymer sheets using a die manufactured according to ASTM Standard Test Method D-638 for Type V specimen dimensions. [7] These tensile samples had a gauge area of 3.18 mm wide and 7.62 mm long and an overall length of 63.5 mm. Tensile samples were not pre-conditioned prior to VUV exposure. TABLE 2. Polymer Materials Exposed to VUV in Plasma Asher Material Name (Type) Manufacturer Thickness Chemical Description of Polymer (µm) DuPont 50.8 aromatic polyimide Kapton (200 HN) fluorinated ethylene propylene - a DuPont 50.8 copolymer of hexafluoropropylene and Teflon FEP (200 A) tetrafluoroethylene a modified copolymer of  DuPont 63.5 Tefzel (250 ZN) tetrafluoroethylene and ethylene A polyester - poly(ethyleneterephthalate) or DuPont 25.4 Aluminized Mylar PET Consolidated Polyethylene 50.8 polyethylene Thermoplastics

Figure 2 shows the aluminum foil sample holder mounted on the asher flange with samples installed. Individual sample assemblies installed on the sample holder consisted of 3 tensile samples of the same material with a 2.54 cm diameter MgF2 window covering the gage areas of the samples. For aluminized Mylar, the only metalized polymer, the samples were installed so that the polymer side, not the metal side, faced the plasma through the MgF2 window. The sample/window sets were wrapped tightly with thin aluminum foil leaving a view to the gage areas of the tensile samples through the MgF2 window, but covering the grip areas of the tensile specimens. Four sample assemblies were included for each asher exposure run. One of these sample assemblies is identified in Figure 2 showing the foil wrapped area and the MgF2 window. Sample assemblies were mounted to the holder using wire. The holder was installed on the asher flange such that the samples surrounded the detector in the same plane as the detector.

341

MgF2 window, through which gage areas for one set of 3 tensile specimens are visible Foil wrapped grip areas of tensile specimen set

VUV phototube installed in vacuum coupling (interior shield with mesh not installed)

Figure 2. Asher VUV detector flange with sample holder and samples installed shown from the top view. Detector shielding tube/screen is not installed so that the VUV phototube and coupling are visible in the center of the sample holder. One sample group is highlighted to show the foil wrapped grip areas and the view through the MgF2 window to the gage area of one of the sample assemblies.

Samples and the VUV phototube, with the shielding tube/screen installed, but without a VUV filter, were exposed to the plasma for a brief period to obtain signal, for the 115200 nm wavelength range, for plasma adjusted to a medium visual intensity. Plasma was run for approximately 3 minutes prior to taking the signal reading. Then, the plasma was extinguished, and samples were brought to atmosphere in order to place a protective aluminum foil cover over the phototube. Then, without readjusting the level or tuning, the asher was evacuated to 20 Pa, and RF power was turned on to generate the plasma for the VUV exposure of the samples. At various exposure intervals, the asher was turned off and one of each type of sample was retrieved from the sample assemblies for tensile testing. VUV signal of the plasma was measured prior to the start of each increment. For each exposure, VUV signal was multiplied by the calibration factor to get intensity. Intensity was corrected for the 36% transmitting screen installed on the detector shielding tube and for transmittance through the magnesium fluoride window. The total integrated transmittance over the 115-200 nm range was calculated to be 80% assuming constant intensity. As described earlier, the data were then corrected to obtain the 90-minute steady-state VUV intensity by multiplying by 0.38, the ratio of 90-minute signal to 3-minute signal. The 90-minute VUV intensity value was divided by the air mass zero (AM0) solar intensity [8] in the same wavelength range (115-200 nm), in order to determine the number of equivalent VUV suns for the exposure. The average number of VUV suns for an exposure was multiplied by the exposure duration, in hours, to calculate equivalent sun hours (ESH) for the exposed samples.

342 Because samples were in contact with aluminum foil during exposure, and it is known that metal surfaces become heated when exposed to the RF plasma, temperature was measured in two separate runs dedicated to this purpose. A set of temperature measurement labels, with dots that darken when they are heated beyond the indicated temperature, were used to indicate the maximum exposure temperature within 5.5 °C intervals between 40.6 °C and 110 °C. These temperature labels were assembled in the same manner as the tensile samples and placed in the plasma. A maximum temperature of 54.4 °C was measured for both tests. 2.5 TENSILE TESTING OF POLYMER FILM SAMPLES A bench-top tensile tester, Model 200Q manufactured by DDL, Inc., was used to tensile test the plasma asher VUV-exposed tensile samples and pristine controls. Loaddisplacement data was downloaded from the instrument to a personal computer, and data were analyzed to obtain stress as a function of strain, ultimate tensile strength, and elongation at break for pristine and plasma VUV-exposed polymer film materials. For all samples, a test speed of 6.35 cm/min was used. Although the specimen die prepared samples to the ASTM D 638 Type V specified dimensions, samples were installed in the tensile tester so that their initial grip distance, also used as the gage length, was 1.27 cm, rather than the ASTM D 638 specification of a 2.54 cm initial grip distance. The reason for this departure from the standard practice was that, for each sample, a length of only approximately 1.9 cm had a view of the plasma during asher exposure, and, for best interpretation of results, it was important that only exposed material be in the gage area for tensile testing. Prior to tensile testing, sample thickness was measured with digital calipers to the nearest 0.01 mm. Ultimate tensile strength was calculated using this measured thickness. 3. 0 Results And Discussion 3.1 CHARACTERIZATION OF THE VACUUM ULTRAVIOLET RADIATION ENVIRONMENT IN THE AIR-OPERATED RF PLASMA ASHER Table 3 lists intensity values for room air plasma for visually bright and dim conditions for nine different VUV wavelength ranges isolated using the narrow bandpass filters described in Table 1. Two separate sets of measurements labeled Data Set 1 and Data Set 2 were conducted on different days, for both bright and dim conditions. For Data Set 1, three data points were taken at each condition, and for Data Set 2, two data points were taken at each condition. Repeatability of measurements is good within each of the data sets as indicated by the standard deviation for the individual data sets, however, there is a large variation in values obtained on different days as evidenced by the high standard deviations on the averages of Data Sets 1 and 2 shown in Table 3. Average values of Data Sets 1 and 2 are plotted in Figure 3, although the standard deviation values for the averages of the data sets, often equal to the measurement values, were not plotted in this figure for simplicity. X-error bars on each data point indicate the range of wavelengths transmitted by the filter, in other words, the bandwidth of the filter. The

343 significant scatter between the data taken on different days indicates wide variability in the plasma conditions between different tests. Despite variability in plasma VUV output over time, these data indicate that there is significant intensity throughout the 115-200 nm range that is not dominated by a single narrow wavelength range. Although intensity was not measured beyond 200 nm, it is expected that the plasma output continues into the 200-400 nm NUV wavelength range [4], the wavelength range where radiation can be absorbed deeper into polymer film materials. TABLE 3. Intensity Measured in Air Plasma for Two Measurement Sets Data Set 1 Bright 1 Dim 1 Peak wavelength (nm) 121.60 129.50 138.50 145.50 151.00 170.10 182.00 188.00 198.00

Avg. (3) 7.0 4.4 8.2 11.4 9.2 27.9 7.2 4.2 6.8

Std. Dev. 1.06 0.57 1.25 0.31 2.29 3.02 1.99 0.85 1.64

Avg. (3) 0.44 0.17 0.41 0.50 0.42 1.25 0.22 0.33 1.76

Std. Dev. 0.03 0.02 0.13 0.02 0.03 0.37 0.05 0.23 0.60

Intensity (µW/cm2) Data Set 2 Bright 2 Dim 2 Avg. (2) 257.9 58.3 53.9 57.0 32.5 92.8 32.3 0.34 29.4

Std. Avg. Dev. (2) 17.9 4.52 39.7 1.30 13.2 1.26 14.3 2.93 10.4 0.97 12.9 4.05 5.7 1.36 0.04 0.09 2.7 2.07

Std. Dev. 0.22 0.08 0.34 0.45 0.31 1.29 0.28 0.00 0.34

Averages of Data Sets 1 & 2 Bright Dim Avg Std. Avg Std. Dev. Dev. 132.5 31.3 31.1 34.2 20.8 60.3 19.7 2.3 18.1

177.4 38.1 32.3 32.3 16.5 45.9 17.8 2.7 15.9

2.48 0.73 0.83 1.72 0.69 2.65 0.79 0.20 1.92

2.88 0.80 0.60 1.72 0.39 1.98 0.81 0.17 0.22

1000.0 Avg. Bright Plasma Avg. Dim Plasma

Intensity (µW/cm 2)

100.0

10.0

1.0

0.1 100.00

120.00

140.00

160.00

180.00

200.00

220.00

Peak Filter Wavelength (nm) Figure 3. Vacuum ultraviolet intensity of room air plasma in the RF plasma asher.

It is relevant to compare the asher VUV intensity to that of the LEO solar intensity in the same wavelength ranges to determine the equivalent VUV suns. These data are useful, because the number of equivalent suns can be multiplied by asher exposure hours to determine equivalent space exposure in equivalent sun hours (ESH) for a given

344 wavelength range. Figure 4 shows a plot of number of suns as a function of the filter peak wavelengths calculated from the intensity data in Table 3. As with Figure 3, error bars indicating standard deviation on the y-data are not plotted for simplicity in viewing the average data. Figure 4 indicates a greater number of suns for shorter wavelengths than for longer wavelengths. Based on these data, the asher is not expected to produce spectral output of greater than one sun in wavelengths above approximately 200 nm, although wavelengths above 200 nm are expected to be present. Measurements of the plasma in the 200-400 nm region would verify this trend. 1000.0 Avg. Bright Plasma Avg. Dim Plasma

Number of Suns

100.0

10.0

1.0

0.1

0.0 100.00

120.00

140.00

160.00

180.00

200.00

220.00

Peak Filter Wavelength (nm) Figure 4. Number of equivalent suns produced in room air plasma as a function of filter peak wavelength between 115 and 200 nm.

In order to use the RF plasma asher for quantitative assessments of space VUV durability of materials, control of the plasma tuning would be necessary to assure a repeatable VUV output. Also, use of a feed gas other than room air, such as dry nitrogen or dry air, may eliminate some of the instability in output due to variations in room humidity and air impurities. The room air plasma was also measured for intensity and number of suns within the overall 115-200 nm range, the full range of sensitivity of the VUV phototube, as shown in Table 4. TABLE 4. Intensity and Number of Suns in the 115-200 nm Range for Room Air Plasma Number of Suns Intensity (µW/cm2) Dim Plasma Bright Plasma Dim Plasma Bright Plasma Average 2.92 46.8 0.27 4.36 Standard Deviation 4.04 59.0 0.38 0.55 No. of Measurements 7 4 7 4

345 The number of suns in this large wavelength range falls between less than one sun for dim plasma and more than 4 suns for bright plasma, although, as shown in Figure 4, the number of suns is higher at shorter wavelengths and lower and longer wavelengths. 3.2 EFFECTS OF VUV RADIATION IN THE ROOM AIR PLASMA ON MECHANICAL PROPERTIES OF POLYMER FILMS Table 5 shows the number of VUV suns for each of the four polymer VUV exposure tests based on intensity measurements made at various exposure intervals. Number of measurements used to calculate the average is shown. Tables 6a and 6b show the results of tensile testing of asher VUV exposed polymer films and pristine controls along with the average and standard deviation on VUV ESH and on mechanical properties measurements for pristine control samples. Elongation at break and ultimate tensile strength are shown as a function of asher VUV exposure ESH in Figures 5a and 5b. For simplicity in identifying general degradation trends, Figures 5a and 5b do not show error bars; however, Tables 6a and 6b provide standard deviation data. Note that only one sample was tested for each of the exposures. TABLE 5. Measured Number of Suns for Asher VUV Exposure Tests Number of Equivalent VUV suns, 115-200 nm range Number of measurements (Avg. ± Std. Dev.) Test Set 1 2 1.8 ± 1.1 2 2 2.2 ± 1.3 3 3 1.1 ± 0.1 4 2 1.7 ± 0.9

Figures 5a and 5b indicate general qualitative trends including reduction in elongation at failure as a function of asher VUV exposure. The only material which retained its elongation as a function of asher VUV exposure is Kapton HN, although it underwent reduction in tensile strength as shown in Figure 5b. The most degraded material was polyethylene. Two of the asher-exposed polyethylene samples were so brittle upon removal from their holders that they broke prior to being tested (shown as 0% elongation and 0 MPa tensile strength in Figures 5a and 5b). Teflon FEP showed reduction in elongation as a function of asher VUV exposure, but only slight degradation in tensile strength. Mylar showed significant degradation in both tensile strength and elongation. Tefzel ZN showed slight degradation in tensile strength, and the largest pristine value of elongation, and it still possessed high elongation after over 2000 ESH VUV exposure. Tables 6a and 6b also show measured thickness for pristine controls and for each asher exposed sample. It is important to note that most samples lost a significant amount of thickness as a result of the asher VUV exposure. Tefzel ZN is the only material that did not lose measurable thickness in any of the asher exposure tests. Despite best efforts to tightly wrap the samples within the sample assemblies to avoid atomic oxygen exposure, it is evident that plasma, and therefore, atomic oxygen was being formed within the small volume of air present inside the sample assemblies.

346 TABLE 6a. Tensile Test Data for Teflon FEP and Kapton HN exposed to Asher VUV

Material FEP FEP FEP FEP FEP FEP FEP FEP FEP FEP

Test Set -1 1 -2 2 -3 3 3

Asher Exposed Hours 0 8.47 18.82 0 43.42 109.81 0 64.18 168.88 334.95

ESH, Avg. 0 8.18 34.5 0 54 257 0 69.6 175.2 363

ESH Std. Dev. 0 9.5 21 0 17 43 0 7.1 18.6 37

FEP FEP FEP Kapton HN Kapton HN Kapton HN Kapton HN Kapton HN Kapton HN Kapton HN Kapton HN Kapton HN Kapton HN

4 4 4 -1 1 -2 2 -3 3 3

74.35 170.22 280.89 0 8.47 18.82 0 43.42 109.81 0 64.18 168.88 334.95

97.4 356 554 0 8.18 34.5 0 54 257 0 69.6 175.2 363

5.2 8.4 -0 9.5 21 0 17 43 0 7.1 18.6 37

Kapton HN Kapton HN Kapton HN

4 4 4

74.35 170.22 280.89

97.4 356 554

5.2 8.4 --

Number Measured of Thickness controls (mm) 8 0.05 0.03 0.03 9 0.05 0.05 0.05 6 0.05 0.03 0.03 0.03

9

9

6

Elongation at Break (%) 441.3 ± 49.3 482 431 381.9 ± 7.1 372 335 386.3 ± 15.3 313 264 135

Ultimate Tensile Strength (MPa) 32.0 ± 5.0 39.6 34.2 29.2 ± 10.6 23.5 17.8 44.9 ± 1.2 31.0 26.1 25.9

0.03 0.03 0.03 0.07 0.03 0.03 0.07 0.07 0.06 0.07 0.03 0.03 0.03

308 198 106 102.3 ± 9.7 104 87 87.6 ± 8.0 89 104 85.2 ± 5.4 83 73 64

32.9 27.2 23.3 354.0 ± 90.0 397 350 259 ± 113 184 204 395 ± 11 353 297 182

0.03 0.03 0.03

92 95 75

361 350 219

Therefore, these exposures represent combined atomic oxygen/VUV exposure. The atomic oxygen fluence accompanying VUV exposure was not quantified by mass measurement for the tested samples. Following these tests, it was attempted to estimate the AO fluence using a 2.54 cm diameter 127 µm thick Kapton HN witness coupon placed within a foil wrapper where the sample had a view to the plasma through the magnesium fluoride window. The resulting mass loss for the witness coupon was insignificant compared to that which would be predicted by the thickness losses measured for the exposed tensile samples. Closer contact between the samples and the magnesium fluoride window appeared to reduce available volume within which a plasma could form, and it was easier to make good contact between the window and the witness sample, which was a round sample of the same diameter as the window and thick enough to hold a flat shape, than with the groups of 3 thinner tensile specimens. Because of the variation in tightness of wrapping, thickness loss was not uniform from test to test. Fortunately, elongation at failure is not dependent on sample thickness, and ultimate tensile strength was calculated based on measured sample thickness, so the tensile test data are not in significant error due to the thickness loss; however, atomic

347 oxygen erosion is likely to have influenced the mechanical properties to some extent. For example, an embrittled surface layer could be removed by atomic oxygen erosion. 



TABLE 6b. Tensile Test Data for Mylar , Tefzel ZN and Polyethylene exposed to Asher VUV Asher Number Measured Test Exposure ESH , ESH, of pristine Thickness Elongation at Ultimate Tensile Material Set Hours Avg. Std. Dev. controls (mm) Break (%) Strength (MPa) Mylar/Al -0 0 0 9 0.02 146.1 ± 57.4 208 ± 8.7 Mylar/Al 1 8.47 8.18 9.5 0.01 102 119 Mylar/Al 1 18.82 34.5 21 0.01 118 120 Mylar/Al -0 0 0 8 0.02 93.1 ± 13.0 254 ± 104 Mylar/Al 2 43.42 54 17 0.02 87 224 Mylar/Al 2 109.81 257 43 0.02 27 148 Tefzel ZN 4 74.35 97.4 5.2 0.05 491 40.3 Tefzel ZN 4 170.22 356 8.4 0.05 451 38.5 Tefzel ZN 4 280.89 554 -0.05 401 35.1 Polyethylene -0 0 0 9 0.06 483.5 ± 156.4 30.9 ± 4.4 Polyethylene 1 8.47 8.18 9.5 0.03 412 28.3 Polyethylene 1 18.82 34.5 21 -0* 0* Polyethylene -0 0 0 9 0.06 384.2 ± 44.2 16.9 ± 7.7 Polyethylene 2 43.42 54 17 0.03 304 10.5 Polyethylene 2 109.81 257 43 0.03 305 10.1 Polyethylene -0 0 0 6 0.06 410.7 ± 31.3 29.4 ± 1.3 Polyethylene 3 64.18 69.6 7.1 0.03 255 19.5 Polyethylene 3 168.88 175.2 18.6 0.03 60 11.7 Polyethylene 3 334.95 363 37 -0* 0* Polyethylene 4 74.35 97.4 5.2 0.03 178 20.4 Polyethylene 4 170.22 356 8.4 0.03 197 20.5 Polyethylene 4 280.89 554 -0.03 34 4.0 *Broke during removal from holder after asher exposure

The polymer degradation observed in these experiments is not expected to be due to VUV wavelengths alone, but rather, to the entire range of VUV-NUV wavelengths present in the plasma. The depth of ultraviolet radiation absorption in polymers varies as a function of wavelength, and the shorter wavelengths are absorbed in a shallower surface layer than the longer wavelengths [6]. It is expected that the shorter wavelengths produce a highly degraded surface layer and that ultraviolet radiation of longer wavelengths produces more bulk damage, and the degree of damage lessens with increasing depth into the polymer. This combination of surface and bulk damage can lead to stress concentrations at the surface facilitating polymer failure during tensile testing. 4.0 Summary and Conclusions The intensity of vacuum ultraviolet (VUV) radiation in an RF plasma asher operated on room air was quantified over the 115-200 nm wavelength range and within narrow bands (9-27 nm) isolated using narrow bandpass filters. Results indicated that by adjusting the RF tuning, the VUV intensity in the plasma could be adjusted with a range of nearly two orders of magnitude.

348 2 mil FEP, #1 2 mil FEP, #2 2 mil FEP, #3 2 mil FEP, #4 2 mil Kapton HN, #1 2 mil Kapton HN, #2 2 mil Kapton HN, #3 2 mil Kapton HN, #4 2 mil Polyethylene, #1 2 mil Polyethylene, #2 2 mil Polyethylene, #3 2 mil Polyethylene, #4 1 mil Mylar/Al, #1 1 mil Mylar/Al, #2 2.5 mil Tefzel, #4

900 800

Elongation at Break (%)

700 600 500 400 300 200 100 0 0

100

200

300

400

500

600

VUV ESH

(a) 2 mil FEP, #1 2 mil FEP, #2 2 mil FEP, #3 2 mil FEP, #4 2 mil Kapton HN, #1 2 mil Kapton HN, #2 2 mil Kapton HN, #3 2 mil Kapton HN, #4 2 mil Polyethylene, #1 2 mil Polyethylene, #2 2 mil Polyethylene, #3 2 mil Polyethylene, #4 1 mil Mylar/Al, #1 1 mil Mylar/Al, #2 2.5 mil Tefzel, #4

600

500

UTS (MPa)

400

300

200

100

0 0

100

200

300

400

500

600

VUV ESH

(b) Figure 5. Effect of plasma asher VUV exposure on (a) elongation at break and (b) ultimate tensile strength (UTS) of polymer films.

349 For the brightest plasma, nearly 300 VUV suns were able to be produced in a band around 122 nm, whereas approximately 0.4 VUV suns were able to be produced in a band around 190 nm. Use of the RF plasma asher as a stable VUV source would require feedback controls on RF tuning and and use of a feedgas with greater purity than room air. To investigate the feasibility and issues associated with using the plasma asher as a VUV source for space simulation, and to compare VUV durability of some common polymer films, tensile test specimens were exposed to the plasma in assemblies consisting of MgF2 windows and aluminum foil wrapping that were intended to allow VUV exposure without AO. However, plasma VUV could not be completely isolated from atomic oxygen as evidenced by significant thickness loss for many of the tensile specimens. Tensile test results indicated a general trend in reduction in elongation as a function of VUV exposure for all samples except Kapton HN. Kapton HN maintained its elongation, although tensile strength was found to decrease with increasing VUV exposure. Teflon FEP and Tefzel ZN samples showed significant degradation in elongation, but only slight degradation in tensile strength. Mylar showed degradation in both elongation and tensile strength. The greatest degradation was observed for polyethylene, for which two specimens became embrittled enough to break during removal from their sample holders. 5.0 Acknowledgements The authors gratefully acknowledge the support of George Readus, Ron Smith, Mike DePauw, and Bruce Banks (NASA Glenn Research Center), and Tom Stueber, Ed Sechkar, Scott Panko, and Russ Messer (QSS Group, Inc.). 6. 0 References 1.

Visentine, J., ed., “Atomic Oxygen Effects Measurements for Shuttle Missions STS-8 and 41-G, NASA TM 100459, Vol. I, Sept. 1988, p. 5-1.

2.

Banks, B., Rutledge, S., Paulsen, P., and Stueber, J., “Simulation of the Low-Earth Orbital Atomic Oxygen Interaction With Materials By Means of An Oxygen Ion Beam,” in Proceedings of the 18th Annual Symposium on Applied Vacuum Science and Technology, American Vacuum Society, 1989. Also appears as NASA TM 101971, February 1989.

3.

Llewellyn-Jones, F., The Glow Discharge, Spottiswoode, Ballantyne & Co. Ltd., London, 1966, p. 14.

4.

Koontz, S., Albyn, K., Leger, L., “Atomic Oxygen Testing with Thermal Atom Systems: A Critical Evaluation,” Journal of Spacecraft and Rockets, Vol. 28, No. 3, May-June 1991.

5.

Llewellyn-Jones, F., The Glow Discharge, Spottiswoode, Ballantyne & Co. Ltd., London, 1966, p. 118.

6.

Dever, J., Pietromica, A., Stueber, T., Sechkar, E., Messer, R., “Simulated Space Vacuum Ultraviolet (VUV) Exposure Testing for Polymer Films,” AIAA Paper No. 2001-1054, American Institute of

350 Aeronautics and Astronautics, January 2001. Also, NASA/TM-2002-211337, National Aeronautics and Space Administration, Glenn Research Center, January 2002. 7.

Standard Test Method for Tensile Properties of Plastics, D 638-95 Annual Book of ASTM Standards, American Society for Testing and Materials, 1995.

8.

Standard Solar Constant and Air Mass Zero Solar Spectral Irradiance Tables.”American Society for Testing and Materials ASTM-E 490-73a (Reapproved 1992), Annual Book of ASTM Standards, American Society for Testing and Materials, 1992.

STUDIES OF THE SURFACE OXIDATION OF SILVER BY ATOMIC OXYGEN. M.L. ZHELUDKEVICH, A.G. GUSAKOV, A.G. VOROPAEV, A.A. VECHER, E.N. KOZYRSKI Department of Chemistry, Belarus State University Skoriny Avenue 4, Minsk 220080, Republic of Belarus. S.A. RASPOPOV Department of Chemistry, University of Toronto 80 Saint George St., Toronto, Ontario M5S 3H6, Canada

Abstract Interaction of silver surface with atomic oxygen has been studied at low oxygen pressures (1015 - 1017 atoms⋅cm-2s-1) and in a wide temperature range from 373 to 673 K. The only product of surface oxidation observed at these conditions was silver (I) oxide Ag2O. Parabolic rate law was normally observed under such experimental conditions with the parabolic constant increasing linearly with incident flux of oxygen atoms. Parabolic rate constant also shows unusual temperature dependence. It increases with activation energy of 33 kJ/mol between 423 and 523 K. reaches maximum at about 573 K and then decreases until oxidation stops around 673 K. Thermodynamical analysis of chemical reactions in the system silver - atomic oxygen shows possibility of the ’recombinative reaction’ of silver by atomic oxygen which offers an explanation of the unusual temperature dependence. 1.0 Introduction Silver is a popular material in space technology due to its high electrical/thermal conductivity and high reflectivity. It was among the materials for which testing was performed on board of spacecraft in LEO environment [1]. Some laboratory studies of the oxidation of silver by atomic oxygen have been reported as well [2-4]. In the temperature range 273 to 358 K interaction of silver surface with hyperthermal atomic oxygen results in the formation of silver peroxide Ag2O2. Linear oxidation was observed initially which changes to parabolic rate law as thickness of the oxide film exceeds 25 nm [2]. It was found that cracking and flaking of oxide film takes place during oxidation, which facilitates diffusion of oxygen to the surface of silver. Diffusion occurs by two parallel

351

352 mechanisms: through the point defects and through micro-cracks or pores in the oxide film. Above 373 K, Ag2O becomes the main product of oxidation because it is more stable thermodynamically under such conditions [3]. When thermal oxygen atoms interact with silver surface, oxidation product is also Ag2O and parabolic rate law is observed [4]. Oxidation only takes place at temperatures below 588 K, no oxide formation occurs on the surface at higher temperature. 2.0 Experimental Experiments were carried out in a high-vacuum setup described in previous publications [6,7]. Atomic oxygen was produced in a microwave discharge (2450 MHz, 60 W). Oxygen-nitrogen or oxygen-argon mixture was passed through the discharge zone just before entering the vacuum chamber. Changing composition of the gas mixture allowed variation in the dissociation degree of oxygen by an order of magnitude – from 5 to 50%. Concentration of atomic oxygen was measured by a chemi-luminescent titration with nitrogen dioxide, NO2. Kinetics of the oxidation of silver by atomic oxygen were studied by measuring time dependence of the electrical resistance of silver filaments (dia. 0.5 mm). Filaments were heated resistively and resistance was measured as a ratio of the voltage drop across central (evenly heated) part of specimen to electrical current through the filament. Because there is an uncertainty in the distance between the potential contacts, data on electrical resistance were normalized by the initial resistance before oxidation (R/R0). Electrical conductivity of silver oxide formed on the surface of specimen is negligible compared to that of silver metal, therefore simple equation can be used to calculate thickness of oxide film from the resistance data:

· § h0 ρ Ag M Ag 2O ¨ 1 ¸ , × ¨1 − h= R ¸¸ 2 ρ Ag 2O M Ag ¨ R0 ¹ ©

(1)

where h is the thickness of oxide layer (mm); h0 is the initial diameter of silver filament (mm); ρAg is the density of silver (g⋅cm-3); MAg and MAg2O are molecular weights of Ag and Ag2O; R/R0 is the electrical resistance of the specimen normalized by initial value. Oxide film formed on silver surface was analyzed by X-ray diffraction at grazing incidence

353 3.0 Results and Discussion Experimental studies of the interaction of silver surface with atomic oxygen beam (1.8⋅1015 to 2.7⋅1016 atoms⋅cm-2⋅s-1). Were carried out at temperatures between 423 and 673 K. X-ray diffraction studies of the oxidized samples showed that Ag2Ois the only product of surface oxidation under given conditions. Rate of oxidation follows the parabolic time dependence (Eq.2) indicating that process is limited by diffusion through the oxide film.

h2 = k p ⋅ t

,

(2)

where h is the thickness of oxide film, kp is the parabolic rate constant and t is time. Plots of the square of oxide thickness vs time are given in Fig. 1 for temperatures from 423 to 623 K and incident flux of oxygen atoms of 2.7⋅1016 atoms⋅cm-2⋅s-1. At these conditions parabolic time dependence is observed throughout the experimental run.

-4

1.0x10

-4

8.0x10

-5

6.0x10

-5

4.0x10

-5

2.0x10

-5

423 K 473 K 523 K 573 K 623 K

2

h / mm

2

1.2x10

0.0 0

50

100

150

Time / min Figure 1. Square of oxide thickness as a function of oxidation time for the reaction of silver with atomic oxygen at incident flux of 2.7⋅1016 atoms cm-2s-1.

Parabolic kinetics was also observed previously for interaction of silver with atomic oxygen at temperatures between 293 and 373 K [3].

354 At higher temperatures time dependence changes to linear, as Fig.2 shows for 648 K. Linear kinetics indicates that mechanism of oxidation has changed and ratelimiting step is no longer the diffusion but chemical reactions on the surface of silver. This is due to the fact that oxide film loses its protective properties, flaking and cracking of oxide is observed above 623 K. Kinetics of oxidation also changes at lower incident flux of atomic oxygen. At 2.1-2.7⋅1016 atoms⋅cm-2⋅s-1 and 523 K parabolic time dependence was observed during the whole experimental run (Fig. 3). However at a lower incident flux of 8.6⋅1015 atoms⋅cm-2⋅s-1 initial linear section with a lower oxidation rate is observed on the kinetic curve. Lower oxidation rate is probably due to high recombination coefficient of atomic oxygen on the surface of silver metal (order of 0.1) [8]. As the surface of silver becomes covered with oxide, rate of oxidation increases because the recombination coefficient is about an order of magnitude less at the surface of oxide. Low initial rate of silver oxidation by atomic oxygen was observed earlier in Ref.2 and was attributed to consumption of a substantial amount of atomic oxygen for reactions with surface impurities at the early stages of oxidation. As the oxidation progresses, kinetic curve becomes parabolic. Lowering of the incident flux of atomic oxygen results in a linear decrease of the parabolic rate constant (Fig. 4).

-3

2.0x10

-3

1.0x10

-3

h / mm

3.0x10

0.0 0

50

100

150

Time / min

Figure 2. Kinetic curve for silver oxidation by atomic oxygen at 648 K.

355

1.2x10

-4

1.0x10

-4

8.0x10

-5

6.0x10

-5

4.0x10

-5

2.0x10

-5

15

-2

-1

2

h / mm

2

1.8x10 atoms cm s 15 -2 -1 8.6x10 atoms cm s 16 -2 -1 2.1x10 atoms cm s 16 -2 -1 2.7x10 atoms cm s

0.0 0

20

40

60

80

100

120

140

Time / min Figure 3. Square of oxide thickness as a function of oxidation time at 523 K and different fluxes of atomic oxygen.

Figure 4. Parabolic rate constant as a function of atomic oxygen flux for oxidation of silver at 523 K.

356

1E-6 8E-7 6E-7

Ea=33 kJ/mol

kp

4E-7

2E-7

1E-7 450

500

550

600

T/K Figure 5. Arrhenius plot of the parabolic rate constant for silver oxidation by atomic oxygen

We turn now to the temperature dependence of the oxidation process. In the temperature range 423-523 K parabolic rate constant increases according to Arrhenius equation with activation energy of 33 kJ/mol (Fig.5). As the temperature increases further, oxidation rate grows slowly to a maximum at about 577 K and then decreases. This behavior is similar to that found for interaction of copper with atomic oxygen [9]. As already mentioned above, kinetics changes from parabolic to linear above 648 K. When temperature reaches 673 K, oxidation of silver by atomic oxygen stops. At these conditions, oxide film formed at a lower temperature can be reduced to silver metal. Reduction of the silver oxide can be observed as decreasing resistance of the previously oxidized specimen when it is exposed to atomic oxygen at 673 K. X-ray diffraction studies show that surface oxide film disappears after such treatment. This unusual temperature dependence of the kinetics of silver oxidation by atomic oxygen can be explained by considering chemical reactions (R1-R3) that occur in the silver-oxygen system under given conditions. In addition to the oxidation process (R1), counter-acting reactions may take place namely oxide dissociation (R2) and ‘recombinative reduction’ by atomic oxygen (R3), which was suggested previously for the system copper-atomic oxygen [9]. 2 Ag + O → Ag2O

(R1)

Ag2O → 2Ag + 1/2O2

(R2)

357 Ag2O + O → 2Ag + O2

(R3)

In order to further understand the experimental observations Gibbs energies of reactions R1-R3 were calculated using the following equations R1a-R3a: oxidation of silver:

∆ r G = ∆ r H 0 − T∆ r S 0 − RT ln PO

(R1a)

dissociation of Ag2O:

∆ r G = ∆ r H 0 − T∆ r S 0 +

RT ln PO 2 2

(R2a)

PO PO2

(R3a)

‘recombinative reduction’ of Ag2O:

∆ r G = ∆ r H 0 − T∆ r S 0 − RT ln

Temperature dependence of the Gibbs energies is presented in Fig. 6. All the three reactions are allowed thermodynamically in the given interval of temperature. These calculations show that at the atomic oxygen incident flux of 2.7⋅1016 atoms⋅cm2 -1 ⋅s formation of Ag2O is thermodynamically possible for temperatures up to 1100 K, however in experiment oxidation of silver stops already at 673 K and even reduction of oxide is observed at these conditions. This observation can be explained if we consider temperature dependencies of the Gibbs energies (thermodynamic probabilities) for reactions R1-R3, shown in Fig.6. At temperatures below 700 K Gibbs energy of the oxidation reaction (R1) is much lower (thermodynamic probability is higher than for the reaction of oxide dissociation (R2), therefore rate of oxidation (R1) is higher than rate of reduction (R2 and R3). However, the thermodynamic probability of R1 decreases with temperature whereas that of R2 increases. Around 700 K, Gibbs energies of R1 and R2 become equal and at higher temperature R2 prevails. This temperature is very close to 673 K where oxidation stops according to experiment. Decreasing oxidation rate in the temperature range 573 to 673 K and change of time dependence from parabolic to linear at 648 K are probably also due to the temperature dependence of the thermodynamic probabilities of reactions R1 and R2. In Ref.5 it was found that oxidation of silver by atomic oxygen stops at 588 K. This lower temperature value is probably due to the lower pressures of molecular and atomic oxygen used in Ref.5, which changes Gibbs energies of the reactions R1-R3 according to Eqns. R1a-R3a.

358

0

-1

-50

∆

G / kJ mol

-100

-150

R1 2 Ag + O --> Ag2O R2 Ag2O --> 2Ag + 1/2 O2 R3 Ag2O + O --> 2Ag + O2

-200

300

400

500

600

700

800

900

1000

T/K Figure 6. Gibbs energy for the oxidation (R1), oxide dissociation (R2) and ‘ recombinative reduction’ (R3) reactions as a function of temperature.

Process of the ‘recombinative reduction’, R3 also has a large negative Gibbs energy under our experimental conditions and it is almost independent of temperature (Fig.6). This reaction can too play an important role in the lowering of oxidation rate at temperatures above 523 K and reduction of the previously formed oxide film at 673 K. It should have an effect on the other features of silver surface interactions with atomic oxygen. 4.0 References 1. 2. 3. 4. 5. 6. 7. 8. 9.

De Roij, A., Proceedings of the Third European Symposium on Spacecraft Materials in Space Environment, 1985, ESA SP-232, p. 99. Oakes John C., Krech R.H, Upschulte B. L., and G.E. Caledonia G.E. , J. Appl. Phys., 1995, 77(5), 2166. De Roij A, European Space Agency Journal, 1989, 13, 363. Bhan M.K., Nag P.K. , G.P. Miller and Gregory J.C. , J. Vac. Sci. Technol. A, 1994, J.C. (3), 699. Dickens P.G., Heckingbottom R. and J.W. Linett J.W. , Trans. Faraday Soc., 1969, 65, 2235. Raspopov S.A., A.G. Gusakov A.G., Voropaev A.G, A.A. Vecher, M.L.Zheludkevich, Lukyanchenko O.A, A.G. Gritsovets and V.K. Grishin, J. Chem. Soc. Faraday Trans., 1997, 93 (11), 2113. Raspopov S.A., Gusakov A.G. , Voropaev, A.G. , A.A. Vecher and V.K. Grishin, J. Alloys Comp., 1993, 201, 67. Myerson A.L. , J. Chem. Phys. 1969, 50, 1228. Gusakov A.G. , Voropaev A.G, Zheludkevich M.L. , A.A. Vecher A.A. and Raspopov S.A. , Phys. Chem. Chem. Phys., 1999, 1, 5311.

DETERMINATION OF THE ENERGY LEVEL OF THE ATOMIC OXYGEN FLUX GENERATED BY THE SPACE SIMULATION APPARATUS USING A THERMAL MODELING METHOD XIN-XIANG JIANG, LAURA LUCIER, DARIUS NIKANPOUR AND STÉPHANE GENDRON Advanced Materials and Thermal Group, Spacecraft Engineering, Space Technologies, Canadian Space Agency 6767 route de l'aeroport, St-Hubert, Québec J3Y 8Y9 Canada

Abstract This study presents a thermal modeling method to characterize the energy level of the atomic oxygen flux generated by a ground-based space environment simulation system in lieu of using a relatively sophisticated retarding energy analyser instrument. It involves both theoretical and experimental work. As a case study, the method was used to characterize the energy level of the atomic oxygen produced by the Canadian Space Agency's Space Simulation Apparatus (SSA), the value of which is determined to be 0.063 eV. 1. 0 Introduction The degradation and protection of materials exposed to the space environment has been studied extensively since the human race's first excursions into orbit demonstrated that the space environment is hazardous and detrimental to most materials used for spacecraft structures and subsystems [1,2]. Although flight experiments, such as NASA's Long Duration Exposure Facility (LDEF) [3], are considered to be the definitive source of information regarding the effects of such an environment on specific materials and protective coatings, they are very costly, inconvenient and particularly time-consuming. For example, operating in real-time, LDEF required 5 years on orbit to expose the test materials to 5 year's worth of environmental effects. For these reasons, ground-based space environment simulators, which can perform accelerated testing, have been developed and are used at major space agencies and laboratories to explore the issue [4]. The extent to which the space environment is hazardous and detrimental to materials is dependant on the length of time of the exposure, as well as the attitude of the orbit in which the spacecraft is placed. Of great concern is the environment of low Earth orbit (LEO), in which a key detrimental factor is atomic oxygen bombardment. Flight experiments have proven atomic oxygen to be very erosive to material through its

359

360 energetic collision with the surface of spacecraft and a chemical reaction with the material itself [3]. To conduct studies of atomic oxygen effects on materials in order to develop new protective coatings, researchers at the Canadian Space Agency have developed a Space Simulation Apparatus (SSA), which can produce a dense atomic oxygen flux in a vacuum chamber to perform accelerated testing. Despite successful use of the SSA for several studies [5], one important parameter, the energy level of the atomic oxygen flux generated by the facility, remains to be accurately characterized, although a recent effort made by Lucier [6] has determined its approximate range to be 0.039–0.068 eV. A number of studies show that the erosion rate of a material under the attack of atomic oxygen is not only dependent on the flux of the atomic oxygen, but also on its energy level. Because the atomic oxygen has a high velocity relative to the spacecraft, the energy level of a collision between the two is typically in the range of 4-5 eV [7], whereas the energy level of the atomic oxygen flux generated by most ground-based facilities is typically lower than 1 eV. A study by Morrison [8] suggests that 1 eV is a threshold, below which the erosion rates of most materials are sensitive to the energy level of the atomic oxygen. For this reason, it is important to know the energy level of the atomic oxygen generated by a ground-based facility in order to be able to interpret ground-based testing results properly and more meaningfully. One type of instrument previously used to determine atomic oxygen energy values is a retarding energy analyser as described by Tagawa [9]. However, a detailed description of the characterization technique, including its accuracy and limitations, was not reported. Assuming that the kinetic energy of the atomic oxygen is fully converted into heat upon its collision with the test sample, it should be possible to deduce the energy level of the atomic oxygen through thermal Modeling if the flux is known. Following this approach, this paper presents the detailed theoretical and experimental work required to obtain an accurate characterization of the energy level of the atomic oxygen flux produced by the CSA's Space Simulation Apparatus. 2. 0 Space Simulation Apparatus (SSA) Fig.1 is a photograph of the CSA's SSA, in which a material sample can be tested for its resistance to atomic oxygen. The apparatus can also be equipped with barometers, thermocouples, a flash lamp, an oxygen lamp, Krypton and Xenon continuum lamps for generation of UV/VUV radiation, and an irradiance profiler. Vacuum is generated using a Danielson Tribodyn 200/57 three-module molecular drag pump, with typical operating pressure on the order of 10-1 to 10-2 Torr inside the test chamber (dependant on the gas flow rate), although before the introduction of oxygen for producing atomic oxygen, pressures on the order of 10-4 Torr can be achieved as an absolute base. See [6] for a more detailed description.

361

Figure 1. The Pyrex Chamber of the Space Simulation Apparatus

3. 0 Thermal Modeling of the Test Sample When a sample coupon is exposed to the afterglow of an atomic oxygen flux, it is equivalently subjected to a heat injection because the kinetic energy of the atomic oxygen is lost and converted into heat upon collision with the sample surface. As a result, the temperature of the test sample will gradually rise to a point at which a new balance between heat injection and dissipation is reached. By performing a thermal analysis of this process, the power of heat injection can be calculated and consequently the average kinetic energy of the atomic oxygen can be obtained if the flux of the atomic oxygen (fAO, unit: m-2s-1) is known.

AO Conduction Tube

Test Sample

Flux of AO Particles

Stainless Steel Support Stage

Figure 2. Physical Model of the Test Sample Subjected to Atomic Oxygen Bombardment

Assuming that the power of heat injection from the atomic oxygen flux into the test sample is P/A (A is the area of the surface subject to the heat injection), and the average kinetic energy of an atomic oxygen particle is EAO and that the kinetic energy of the

362 atomic oxygen is completely converted into heat upon collision with the sample surface, then: (1)

P / A = E AO ∗ f AO E AO =

P f AO ∗ A

(2)

The atomic oxygen flux can be determined through a standard technique, which will be discussed in the Experiment section of this paper. To obtain a solution for Equation (2), the only parameter left to be known is the power of heat injection (P/A) from the atomic oxygen flux, which can be solved using a thermal analysis method. The experiment uses a thin Kapton film as the test sample, represented in Fig.2. Its thickness is 0.527 mm and it is placed directly, but loosely, on the surface of the sample stage, which is made from stainless steel with high surface flatness and finish. Because the physical contact between the sample and the stage can be considered to be limited, the heat conduction between them in the vacuum condition may be neglected. In this case, heat dissipation results primarily from radiation. The change in temperature of the test sample subject to the heat injection can be divided into two stages, as shown later in Fig.3 of the Experimental Results section of this paper. The first or beginning stage primarily features a rise in temperature as the result of net heat injection. This is the so-called unsteady state. A general mathematic expression [10] for this state is: P dT = a∗ + b ∗ ε ∗ (T 4 − Ta4 ) dt A

,

(3)

where a and b are constants: a = d ∗ Cp ∗ Hc

(4)

b = 2 ∗ F ∗σ

(5)

A – the area of the top surface of the test sample ε – the emissivity of the test sample (Kapton) d - the density of the test sample Cp – the heat capacity of the test sample Hc – the thickness of the test sample T, t – general expression for the temperature (K) and time parameters (s) Ta – temperature of the surrounding environment, which is presumed to be constant F – view factor, which is 1 in this scenario σ -Stefan-Boltzmann constant, 5.669×10-8 W⋅m-2⋅K-4

363 By measuring the change in the temperature of the test sample, values for dT/dt and T in Equation (3) can be obtained. ε can be considered to be a constant, but needs to be experimentally determined in this case because heat radiation from the back of the test sample will be partially reflected back to the sample, making the total emissivity of the test sample a complex parameter. To determine ε, the thermal balance of the test sample in the steady state needs to be considered. In steady state, the temperature of the test sample is no longer increasing (dT/dt =0) and the heat injection (P/A) from the atomic oxygen flux is simply equal to the heat loss of the sample through radiation. Equation (3) thus can be modified into: P = b ∗ ε ∗ (Tc4 − Ta4 ) A

(6)

Where Tc is the temperature of the test sample in steady state, which can be experimentally determined. Equation (6) can be converted into: ε=

P 1 ∗ A b ∗ (Tc4 − Ta4 )

(7)

Substituting Equation (7) into Equation (3), the following expression is obtained: 1 P dT P = a∗ +b∗ ∗ ∗ (T 4 − Ta4 ) A dt A b ∗ (Tc4 − Ta4 )

(8)

Equation (8) can be furthermore simplified into: P dT (Tc4 − Ta4 ) = a∗ ∗ A dt (Tc 4 − T 4 )

(9)

Now it can be seen that by experimentally obtaining dT/dt, T and Tc and inputting the values into Equation (9), the power of the heat injection (P/A) from the atomic oxygen flux can be obtained since a = d*Cp*Hc is a constant. 4. 0 Experimental Procedure The experiment needs to solve two parameters: the flux of atomic oxygen (fAO) and the change in the temperature of the test sample with time. To determine the nominal flux of the apparatus, Kapton H 500 film coupons were measured for height using a Mitutoyo electronic stylus profiler with an accuracy of 1 µm both before and immediately after exposure, in accordance with ASTM Standard E 2089-00. The effective fluence was calculated using the relation:

364 F=

(10)

∆h Y

Where ∆h is the change in height (cm) and Y is the erosion yield of the material (3*10-24 cm3/oxygen atom for Kapton H) [11]. The effective flux (fAO) was then calculated by dividing the total fluence by the number of seconds of exposure time (t): f AO =

F t

(11)

To measure the change in temperature of the sample subject to the atomic oxygen flux, a foil thermal couple was placed and bonded onto the top surface of the test sample. The temperature readings were recorded in 30-second intervals. 5. 0 Experimental Results Kapton film coupons of 25 mm diameter were exposed to the atomic oxygen flux produced by the 13.56 MHz RF plasma generator operating at 0-400 W forward power and 0 W reflected power. Height measurements of the Kapton H samples for 10 trials were used to determine the fluence (and hence the flux) of the exposures using Equations (10) and (11). The obtained result for the average flux (fAO) of the 10 tests is 2.309×1021 oxygen atoms/m2/s. Temperature of the Kapton H 500 sample

Temperature (K)

410 390 370 350 330 310 290 270 0

100

200

300

400

Time (min.) Figure 3. Typical Change in the Temperature of a Kapton H 500 Sample During a Test

The temperature change of the test sample during atomic oxygen exposure was recorded, as shown in Fig.3. The temperature profile vs. time shows exactly a type of

365 result, which is expected. The temperature in steady state, (Tc), is measured to be 400.55 K. Ambient temperature, (Ta), is 296.15 K. Table 1 shows the thermal properties of Kapton H 500 film. Using Equation (9), the P/A of the atomic oxygen flux at different temperature points throughout the measurement period is calculated and shown in Fig.4. The average value for the power of heat injection from the atomic oxygen is further calculated to be 23.37 W/m2 with a deviation range of 21.51 W/m2 to 27.15 W/m2 for 95% accuracy. TABLE 1 Thermal Properties of Kapton H 500 0.47 W/m⋅K

Density (d)

1430 kg/m3

Heat capacity (Cp)

1500 J/kg⋅K

Power of heat injection from AO (W/m 2)

Thermal conductivity

40 35 30 25 20 15 10 5 0 290

310

330

350

370

390

410

Te m pe rature (K)

Figure 4. Power of Heat Injection due to Atomic Oxygen Flux, which is calculated using Equation (9) and experimental data from Fig.3.

Using Equation (2) and substituting the values of both P/A and fAO into the function, the average energy level of the atomic oxygen (EAO) is calculated to be 10.21×10-21 J with a deviation range of 9.40×10-21 to 11.86×10-21 J for 95% accuracy. Because

1 eV = 1.602 x 10-19 J and 1 (J) = 0.624×1019 eV, therefore EAO = 0.0631 eV.

The standard deviation range for 95% accuracy of this value is 0.058 to 0.074 eV. This result is close to the upper limit of the atomic oxygen energy range previously estimated

366 by Lucier [6] using third-order polynomial equations and the Boltzmann relation (0.039–0.068 eV). This study furthers that study and confirms the range of its results, demonstrating that the thermal modeling method can be an effective and accurate way to determine the energy level of the atomic oxygen flux produced by a ground-based facility. 6. 0 Conclusions This study presented a thermal modeling method to characterize the energy level of the atomic oxygen flux generated by a ground-based space environment simulation system in lieu of using a relatively sophisticated retarding energy analyser instrument. Such characterization requires essentially three steps: the flux level of the atomic oxygen must be determined, which can be achieved by exposing a standard reference sample to the atomic oxygen environment and using the technique outlined in ASTM standard E 2089-00. The second step involves measurement of the change in temperature of the test sample with time while subject to the atomic oxygen flux. Such measurement must occur over a sufficient amount of time to obtain the temperature of the sample in steady state. The third step involves the establishment of an appropriate thermal model to obtain a mathematical solution that calculates the energy level of the atomic oxygen flux using the experimentally obtained data. As a case study, the method was used to characterize the energy level of the atomic oxygen produced by the Canadian Space Agency's Space Simulation Apparatus (SSA), the value of which is determined to be 0.063 eV. 7. 0 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Tennyson, R.C., and Morison, W.D. Proceedings: TMS Symposium on “Space Environmental Effects on Materials,” Anaheim, CA, 1990. Banks, B.A., de Groh, K.K., Rutledge, S.K., and DiFilippo, F.J., “Prediction of In-Space Durability of Protected Polymers Based on Ground Laboratory Thermal Energy Atomic Oxygen Testing” ICPMSE-3 Conference, Toronto, ON, April 25-26, 1996. LDEF experiment Banks, B.A., Rutledge, S.K., and Brady, J.A. “The NASA Atomic Oxygen Effects Test Program,” Presented at the 15th Space Simulation Conference, Williamsburgh, VA, Oct. 31-Nov. 3, 1988. Poiré, E., Amyot-Richard, C., Lachance, J., and Nikanpour, D. “Atomic Oxygen Ground Testing Facilities: Significance and Reliability of the Reference Sample” Presented at the 8th International Symposium on Materials in a Space Environment, Arcachon, France, June 5-9, 2000. Lucier, L., Jiang, X. and Nikapour, D. "Characterization of the Canadian Space Agency’s Space Simulation Apparatus – A Simple Method for Determining the Energy Level of an Afterglow Plasma Facility", submitted to the Canadian Aeronautics and Space Journal, 2002 Leger, L. J., "Oxygen atom reaction with Shuttle materials at orbital attitude – data and experiment status", AIAA paper 85-0476, Jan. 1985 Morrison, W. D., Tennyson, R.C., French, J.B and Braithwaite, T. "Atomic oxygen studies on polymers" 15th Space Simulation conference, Oct. 31-3 Nov. 1988, NASA, p89-109 Tagawa, M., Tomita, M., Umeno, M. and Ohmse, N. "Atomic oxygen generators for space studies in low earth orbit" AIAA Journal, Vol.32 No.1 January 1994, p95-100 Thermal modelling book ASTM, “Standard Practices for Ground Laboratory Atomic Oxygen Interaction Evaluation of Materials for Space Applications,” Designation E 2089-00, June 2000.

DESIGN AND TESTING OF A MINI-SPECTROMETER SYSTEM FOR ONORBIT DEGRADATION STUDIES OF OPTICAL MATERIALS

MAGDELEINE DINGUIRARD ONERA DESP, BP 4025, 31055 Toulouse, Cedex 04, France MARC VAN EESBEEK, ADRIAN P. TIGHE ESA/ESTEC 2200AG Noordwijk, The Netherlands

Abstract The degradation of optical materials exposed to the near earth orbit space environment is of primary interest for spacecraft designers and must be accurately predicted. Optical materials generally have highly polished and accurately formed surfaces so that even low levels of degradation may have a significant effect on the long-term performance of the component. This paper describes the design and testing of a miniature spectrometer system which will be used to actively monitor the on-orbit degradation of optical materials placed on the exterior of the International Space Station. 1. 0 Introduction Optical materials are used in many space systems in earth orbit, such as scientific instrumentation, solar arrays, and thermal control systems. As technology advances and applications become more critical, the effects due to degradation of the materials must be taken into account. This is especially so for missions which may last for several years or more. For near earth orbiting satellites, the environmental effects, which have traditionally been of concern, are UV radiation, atomic oxygen, high-energy electrons and protons, and more recently orbital debris. However, for the implementation of sensitive optical components on the exterior of larger manned space structures, such as the International Space Station (ISS), account must also be taken of environmental effects induced by the structure itself and by the manned activities carried out on board and in the surrounding area [3]. The ISS contamination environment is currently not well defined, and it is also very difficult to predict. This is especially so during the early utilisation phases, when new materials are continuously being added and there is a high frequency of “construction traffic”. Contamination can take the form of gaseous products and particulate and will be produced by a variety of sources, such as venting from airlocks,

367

368 outgassing of materials, thruster firings and water dumps. On the ISS, there will be quiescent periods, when direct flow of outgassed material is the only major contribution to contamination, and there will be non-quiescent periods, when specific events occur which produce additional contamination sources. Degradation studies can be carried out simply by exposing materials to the environment for a period of time, and then returning them to earth for analysis. This passive approach only gives partial information about the degradation processes; some of the degradation, particularly in the near infrared can be cured by re-exposure to oxygen, and also the dynamics of the degradation is unknown. Therefore, to provide this missing information, and to unravel the synergistic effects involved, it is necessary to have active detectors which can return data about the optical properties of the materials during the time they are exposed to the environment [4]. In the case of the ISS, this will be particularly important for monitoring the effects of contamination during the nonquiescent periods. In this context, a miniature spectrometer system is being developed to fly on an exterior platform on the ISS. The system will form part of the Materials Exposure and Degradation Experiment (MEDET) and will be located on the outside of ESA’s Columbus Module. It is due to be launched by the Space Shuttle in early 2005, and will remain in orbit for approximately 3 years. Several different instruments will be incorporated onto MEDET, to investigate different materials properties, and to monitor the external environment. An overview of the different instruments has been described previously [1] and the calibration of the environmental monitors is the subject of a different paper in these proceedings [2]. The present paper focuses on the design and testing of the mini-spectrometer system. The main scientific objectives of the experiment are to: •

• •

Evaluate the synergistic effects of the complex space environment on optical materials and coatings, with a special emphasis on degradation induced by molecular contamination from the ISS Follow the degradation dynamics on-orbit Compare and correlate the on-orbit degradation results with the observed condition of the samples before and after they are returned to earth

2. 0 Instrument Design 2.1 SYSTEM OVERVIEW The on-orbit degradation of the samples will be monitored by continuously measuring their optical spectral transmission over the wavelength range 190 to 1100 nm. The sun will be used as the spectral source, and the light will be transmitted through the samples, and then along optical fibres, to the entrance slits of commercial miniature spectrometer modules. Two modules will be used to cover the required wavelength range, and each

369 module will be duplicated for redundancy, giving four modules in total. Only relative measurements will be made, comparing the spectrum of the direct sunlight with the spectrum of the sunlight transmitted through the samples. For integration onto MEDET, the modules will be mechanically and electrically modified. Each pair of modules will then be mounted in a sealed container filled with nitrogen. This will protect the modules from the space environment and it will also prevent contamination of surrounding systems due to out-gassing of the commercial components in the modules. The samples will be placed on a rotating wheel on the ram face of MEDET [Figure 1]. The wheel will contain 24 apertures. Two apertures will remain empty to view the direct sunlight and two apertures will be opaque, to measure the spectrometer “dark signal”. Therefore, the experiment will accommodate 20 optical samples. A stepper motor will be used to rotate the wheel. At each step of the motor, two samples diametrically opposite each other on the wheel will be positioned above the entrance to two fibre optics. Diffusers will be used to increase the acceptance angle of the fibres. An encoder system, consisting of photodiodes positioned under the wheel, will be used to identify the location of the wheel with respect to the fibre optic at any given instant. To ensure that enough light is available to make the measurements, the experiment will only be activated when the sun is within an angle of ±40° to the normal of the ram face. To accomplish this, an illumination sensor will be used. The sensor will consist of a photodiode placed at the bottom of a blackened cone. An electronics interface and associated software will control the operation of the spectrometer experiment, and the readout of the data. The total mass of the system will be approximately 2 kg, and the maximum power consumption will be about 5W, when the wheel is operating. A functional schematic is given in Figure 2, and then the spectrometer modules, the sample wheel assembly and the control electronics are described in more detail in the following sections.

370

Figure 1. Location of spectrometer samples on the MEDET ram face

Figure 2. Functional schematic of the spectrometer experiment

371 2.2 SPECTROMETER MODULES The function of the spectrometer modules is to spectrally resolve and detect the light transmitted through the samples. For use on MEDET, the basic requirements are for low mass and power, low cost, fibre optic input, electronic signal output and a robust and compact design. The chosen modules are adapted versions of a commercial laboratory device called the MMS (Monolithic Miniature Spectrometer), which is manufactured by Zeiss. The as-bought module consists of a grating, fibre optic, photodiode array and control electronics all rigidly housed within the same case. A functional schematic is shown in Figure 3.

Figure 3. Functional schematic of Miniature Spectrometer Module

Two models of the MMS will be used, each one sensitive to a different wavelength range. The main specifications are given in TABLE . The models differ slightly in their construction, in that the grating for the NIR-enh is bonded onto a glass cylinder, whereas the grating for the UV-VIS is not. Apart from this, their function is the same.

372 TABLE 1. Miniature Spectrometer Module specifications. MMS UV-VIS / MMS NIR-enh Optical entrance

Fibre cable, consisting of approx 15 quartz fibres with 70 µm core diameter each Total diameter 400 µm NA = 0.22 (12.5O) SMA connector

Grating

Flat-field, 366 lines/mm (centre), blazed for approx 220 nm

Diode array

S3904-256Q, Hamamatsu

Spectral range

190 – 735 nm / 300- 1100 nm

Wavelength accuracy

0.2 nm (absolute)

Temperature induced drift

0.006 nm / OC

Spectral distance of pixel

∆λ = 2.2 nm

Resolution

7 nm (Rayleigh critereon)

Sensitivity

1012- 1013 counts / Ws (with 14 bit conversion)

Dimensions (with case)

70 x 60 x 40 mm

0.1 nm (relative)

2.3 SAMPLE WHEEL ASSEMBLY The purpose of the sample wheel assembly is to rotate the samples one after the other over the entrances to the fibre optics, and to hold them in position whilst the measurements are being made. The assembly consists of the following units : -

Sample wheel Stepper motor Mechanism to couple sample holder to rotating device (coupling shaft and support housing) Encoder

The sample wheel consists of an aluminium disc, with 24 apertures for the samples. Each sample fits into a recess, and a spacer underneath can be used to account for samples of different thickness. The wheel is connected to the stepper motor via a shaft. A bearing housing is fastened directly onto the motor and the underside of the ram face, and holds the shaft. The encoder consists of 6 photodiodes fixed directly onto the ram face underneath the spectrometer wheel. A pattern of holes is drilled into the sample wheel. Each time a sample moves into place directly above the fibre optic entrances, a unique combination of the photodiode sensors detect sunlight, depending upon the pattern of holes above the sensors for that particular sample.

373 2.4 ELECTRONICS Command and communication for the experiment will be controlled by the MEDET logic unit and power will be provided by the MEDET power supply [1]. In addition, interface units have been designed to read out the signal from each of the spectrometer photodiode arrays and pre-amplify the signal, and to energize the stepper motor in a given driving sequence. 3. 0 Sample Selection The development phase of MEDET will be two years long, with an extra storage and integration period of at least one year after the payload has been delivered for integration onto EuTEF. Therefore, the samples will be selected and procured as late in the development phase as possible, to allow up to date materials to be used, and to minimise the storage period. The requirements for selection and the type of samples, which will be used, are given below and further details of the sample storage and integration procedures are given in a following section. 3.1 REQUIREMENTS The samples will be selected according to the following requirements: - Samples will only contain low outgassing materials: typical RML < 1.0%, CVCM < 0.1% - Some novel samples will be proposed plus some well characterised samples for comparison - Samples will be mechanically and chemically stable enough to survive the ground and on-orbit environments. Obviously some degradation will be expected, but the samples should not suffer total failure, such as shattering (for the substrate material) or flaking off the substrate (for the coatings) i.e. coatings should be well adhered to the substrate. - Samples which may constitute a safety hazard will be avoided, due to the requirements for manned space missions 3.2 PROPOSED SAMPLES The following are typical types of sample which will be proposed : - Uncoated windows – e.g. SiO2 (crystalline quartz or fused silica), MgF2, Sapphire, LiF, other optical glasses - Coated windows – e.g. single layer or multi-layer dielectric (for use as antireflection coatings or optical filters) - Coated and uncoated plastic films – e.g. FEP, Kapton for second surface mirrors

374 4. 0 On-orbit Operations 4.1 OPERATING MODES For the majority of the time, the experiment will be in a nominal mode. In this mode, the 4 spectrometer modules (i.e. the photodiode arrays and electronics) will be switched on continuously, for thermal stability, and the 2 illumination sensors will acquire data continuously, at a rate of 1 measurement every 5 seconds. To record the spectra, the experiment will enter a data acquisition mode. During the first few months after launch, this mode will be activated once per orbit. However, it will also be possible to change the rate, at which data is acquired, using a telecommand from the ground, if it is found that the transmittance of the samples varies significantly at a different rate. During the data acquisition mode, sunlight must be detected by the illumination sensors within a cone angle of ±40O for a length of time long enough for the all of the spectra to be recorded. Due to the orbital parameters of the ISS, there may be times when the sun does not remain within the field of view of the sensors for a sufficient period of time (indeed for some orbits, the sun may not be visible at all). In these cases, the measurement cycle will be terminated (or never started) and the data for that orbit will not be used. At the start of the data acquisition mode, the photodiode encoders will check that the sample wheel is in a pre-determined “sample 1” location, so that data is acquired in the same order each time the mode is activated. During the measurement and control cycle, the illumination sensors will continue to acquire data, to check that the sun remains within their field of view. The mode will be de-activated when the predefined sequence of measurements has been completed, or the sun moves out of the field of view of the illumination sensors before this occurs. The sample wheel holder will have two fully blocked holes and two fully open holes, and the remaining 20 holes will hold transparent samples. The measurement sequence will consist of acquiring a combination of the spectra of the light shining through each of the 20 transparent samples, the spectra of the light shining through the fully open holes and the dark signal. The wheel will be moved to the nearest fully open or blocked hole to take the reference spectra, and then back to the next sample in the sequence. 4.2 TELEMETRY AND TELECOMMANDS The following data will be sent to the ground, via the MEDET logic unit. Scientific data -

Spectra of sunlight passing through each sample and through open apertures Time at which each measurement (i.e. a single spectrum) is made Sample identification (encoder readings)

375 Housekeeping data -

Temperatures at critical points e.g. on spectrometer modules Illumination sensor readings

It will also be possible for the user to adjust the following parameters by sending telecommands from the ground once MEDET is in orbit : -

Illumination sensor threshold Data acquisition rate for spectra Integration time for recording spectra

5. 0 Testing and Calibration 5.1 QUALIFICATION TESTS The complete spectrometer system is an integrated part of the MEDET payload and full qualification will only occur during the acceptance tests at MEDET level. However, the spectrometer modules and the sample wheel assembly have been identified as critical items and will therefore undergo a series of delta qualification tests before the flight hardware is manufactured. For the sample wheel assembly, a stepper motor has been purchased and the functional performance has been tested in conjunction with the electronics driver and a dummy sample wheel. A complete engineering model of the assembly is now being manufactured, and this will undergo a vibration test, a thermal cycling test and further functional performance tests. Partial vibration tests have been performed only on the spectrometer, inside a dummy cylinder, as we feared for the grating and fibres. Figure 4 and Table 2 below illustrate the set up and the applied vibration levels as first specified by EuTEF compared to the new specifications. TABLE 2. Specified random vibration levels. FREQUENCY

LEVEL Applied (old spec.) New spec.

20 – 30 Hz

+3.2 db/oct

+3.2 db/oct

30 – 100 Hz

1.0 g2/Hz

0.08 g2/Hz

100 – 2000 Hz

~ 3 dB/oct

~ 3 dB/oct

Composite

19.44 g rms

7.4 g rms

A He-Ne laser and a very stable sun simulator spectra were recorded before and after the vibration tests in order to control the eventual changes. No differences

376 were noted (see for example Figure 5) which qualify these spectrometers as the applied vibration levels were almost 3 time higher than required

XMEDET ZMEDET YMEDET

ZMEDET XMEDET

YMEDET

Figure 4. Vibration test set up

4000 before

3500

after

laser before

Digital Counts

3000

laser after

2500 2000 1500 1000 500

740

720

700

680

660

640

620

600

580

560

540

520

500

480

460

440

420

400

360 380

340

320

300

280

260

220 240

200

0 lambda (nm)

Figure 5. UV-Vis spectrometer module results

5.2 CALIBRATION The system will be self-calibrated, as each sample measurement will be coupled to a direct sun view. Nevertheless, as soon as the in-flight electronic card is ready, test regarding signal dynamics, linearity and noise will be done.

377 6. 0 Sample Storage and Integration Sample conditioning before and after flight is a critical aspect of on-orbit materials studies. Therefore, precautions will be taken to ensure that the samples are protected from unwanted degradation during their time on the ground, both prior to launch and after retrieval. Some of these precautions are described in the following sections. 6.1 LATE INTEGRATION The sample wheel has been designed so that it can easily be attached to and removed from MEDET without having to dismantle the whole experiment. Therefore, it will be possible to integrate the spectrometer samples onto MEDET after the end of the acceptance test campaign and possibly after MEDET has been delivered to KSC for launch on the Shuttle. This will minimise the time that the samples have to remain in storage on the ground. 6.2 GROUND ENVIRONMENTS During ground based activities the aim is to store the integrated MEDET in a controlled environment whenever possible. The following environmental requirements have been specified: Relative humidity: 40 – 60 % Atmosphere: Dry Nitrogen Temperature: 20 ± 5 OC In addition, a protective red-tag cover will remain in place over the samples until just before flight, and this will be replaced immediately after the return to earth and retrieval from the Shuttle. 7. 0 Conclusions A mini-spectrometer system, which will actively monitor the degradation of optical materials on the exterior of the International Space Station, has been described. The system is currently under development and testing. For qualification, the spectrometer modules have successfully completed a random vibration test and a thermal cycling test is currently underway. A electronics model is also being manufactured to test the functional performance of the system. The spectrometer system will form part of the Materials Exposure and Degradation Experiment (MEDET) and is due for launch on the Space Shuttle in early 2005.

378 8. 0 References 1. 2. 3. 4.

Dinguirard, M, Van Eesbeek, Tighe, A., Durin, C., Gabriel, S., Goulty, D., Mandeville, J.C. (2001) Materials Exposure and Degradation Experiment (MEDET), AIAA 2001-5070, International Space Station Utilisation Conference, Florida. Goulty, D., Calibration of environmental measurement experiments for MEDET, these proceedings. Mikitarian, R. (2000) International Space Station external contamination control, 8th ISMSE/5th ICPMSE, Arcachon, France. Wilkes, D. (1999) Optical properties monitor (OPM): in-situ experiment flown on the MIR station, SPIE 3784, 72-83.

KAPTON AS A STANDARD FOR ATOMIC OXYGEN FLUX MEASUREMENT IN LEO GROUND SIMULATION FACILITIES: HOW GOOD IS IT? E. GROSSMAN, I. GOUZMAN, G. LEMPERT, Y. NOTER AND Y. LIFSHITZ Space Environment Division, Soreq NRC Yavne 81800, Israel

Abstract A common method for assessing the Atomic Oxygen (AO) flux in LEO ground simulation facilities is measuring the mass loss of attached Kapton coupon and assuming a known erosion yield of 3x10-24 cm3/O atom. However, in most ground simulation facilities additional components, e.g., UV radiation, excited and ionized oxygen atoms and molecules or reactive volatile products of degraded samples might be involved in Kapton-AO interaction. The present work studies the effect of simultaneous irradiation of AO and other reactive species on the Kapton etching rate. The Kapton mass loss was measured in-situ using Quartz Crystal Microbalanace (QCM). The etching rate was assessed as a function of AO and either VUV radiation, ions, various feed gases, sample temperature or reactive volatile products. X-ray photoelectron spectroscopy (XPS) and Atomic Force Microscopy (AFM) were also applied in order to understand the erosion mechanism involved in the etching rate alteration. The study shows that reactive species (other than AO) present in the ground simulation facilities could lead to substantial errors in the AO flux evaluation. The validity of the use of Kapton as a reference material for ground simulation testing (of spacecraft candidate materials for LEO applications) should be critically reassessed in the context of the present findings. 1.0 Introduction The space environment of low earth orbit (LEO), at altitudes ranging from 200 to 700 km, is considered hazardous to materials and especially polymers. The main constituent of the space environmental hazards affecting polymers' properties is atomic oxygen (AO). The interaction of AO with the outer surfaces of a spacecraft may result in materials degradation, affecting their chemical, electrical, thermal, optical or mechanical properties [1-4]. The high cost and very limited availability of the in-flight experiments as well as the demands for accelerated tests simulating long duration missions, result in development of ground simulation systems to study the space environment. A variety of

379

380 ground simulation facilities is used for studying materials degradation under AO attack including RF and DC plasma sources, ion neutralization, electron stimulation desorption, photo-dissociation, supersonic and laser detonation sources [5,6]. Although the main purpose of these sources is generating AO flux, the produced AO could be associated with one or more of other source components such as UV radiation, ions, electrons, excited species, as well as varying the sample temperature. All these sources share also a common problem of how to assess the 5eV equivalent AO flux. A common and recommended [7] way of evaluating the AO flux is by measuring the mass loss of Kapton coupon exposed simultaneously with the tested samples. By assuming a reaction yield of 3x10-24 cm3/O atom (as measured for Kapton exposed in space) an equivalent of the 5eV AO flux could be calculated. However, the Kapton erosion rate could be affected by other simulation facility factors such as one of the associated components (e.g. UV, ions, etc.), sample temperature or volatile products of tested materials, thus resulting in an erroneous AO flux calculation. In the present work we study the effect of several AO flux associated components on Kapton erosion rate. We used an RF plasma asher which contains in addition to AO also UV radiation, ions, electrons, molecular oxygen and excited species. Although we used a specific simulation facility in this study, the conclusions drawn for Kapton etching rate dependence on the AO accompanying components are generalized for the use of Kapton as a reference material for determining the AO flux in various simulation facilities. 2.0 Experimental Kapton samples were spin coated on Quartz Crystal Microbalance (QCM) crystals using a Dupont procedure for deposition of polyimide (Pyralin® PI 2545). The deposited polyimide films were tested to be similar to Kapton HN films in their chemical structure and their sensitivity to AO irradiation. The Kapton erosion rate measurements were carried out using a QCM (Sycon Model STM-100) mounted in an RF plasma system (Litmas Model LB1200) equipped with an automatic matching unit and a power of 75-1200 W (see Figure 1). The coated crystals were located on a sample holder enabling positioning of the QCM in a selected downstream position. The samples were exposed to different environments (AO, AO+VUV or VUV alone). This was achieved by using a specially designed target holder assembly causing the plasma afterglow flow to reach the sample sideways while the top is covered by MgF2 window, as shown in Figure 1. In such a configuration, the sample is exposed to all RF plasma afterglow components, while the VUV radiation only available pathway is through the MgF2 window. Shielding the MgF2 window with Al foil eliminates the sample direct VUV irradiation while keeping its exposure to all other RF plasma components. Blocking of the sample holder sideways prevents the sample exposure to all RF plasma components except for the VUV radiation penetrating through the MgF2 window. At this configuration it is also possible to cover the MgF2 window, thus studying the net effect of vacuum environment on the sample erosion. The cooling system of the QCM was used for adjusting the crystals temperature to the

381 desired ones in the range of 0-55°C. The samples were also exposed to different RF plasmas (O2, Ar and N2) as well as to the environment of volatile products obtained from AO exposure of attached fluoropolymers, silicones, as well as reference Kapton samples.

MgF2 window

Litmas RF 75-1200W Power Supply

Sample holder

RF Reactor ATOX QCM

Gas inlet

Test Sample & Mounting

Figure 1. Schematic diagram of the AO simulation system.

At operating pressure of 120mTorr and sample location of 100 mm downstream the 5eV equivalent AO flux was calculated (by measuring Kapton mass 15 2 loss and assuming erosion yield of 3x10-24 cm3/atom) to be about 5x10 atoms/cm sec. The VUV flux was assessed by a Phototube sensor (Hamamatsu Model R1187) positioned at the location of the exposed sample. The measurements were performed using a 1% transmittance filter to reduce the intensity to the working range of the sensitive sensor. The VUV flux was 1.6x1016 photons/cm2sec at 300W, 100 mm away from the reactor. Chemical changes, occurring at the surface due to the various irradiations were measured by X-ray Photoelectron Spectroscopy (XPS). The XPS spectra were derived using a non-monochromatized Mg Kα radiation (1253.6 eV) and a hemispherical CLAM 2 (VG Microtech) analyzer operating at a pass energy of 100 eV for survey scans and 20 eV for high-resolution scans. The binding energy scale was calibrated using the Ag 3d5/2 line at 368.25 eV as a reference [11]. The surface morphology of the irradiated Kapton was analyzed under ambient conditions by Atomic Force Microscopy (AFM), using a Digital Instruments NanoScope II. The surface vertical roughness, Rq, (Rq = [Σ(Zi-Zave)2/N]1/2 where Zave is the average Z height value within a given area, Zi is the current Z value, and N is the number of points within the given area) was measured and averaged by the system.

382 3.0 Results 3.1 ENVIRONMENTAL EFFECTS The effect of various environmental parameters on the Kapton etching rate and chemical composition is described below. The parameters investigated were AO flux, VUV flux, combined AO and VUV irradiation, sample temperature, gas composition and type of neighboring samples. 3.1.1. AO, VUV and AO+VUV effect Erosion rate. Figure 2 shows the thickness loss of eroded Kapton polyimide coating under irradiation of VUV flux (a), AO flux (b) and combined AO+VUV (c). It is clearly seen that adding VUV radiation to the AO flux increased the etching rate by about a factor of 2. Some etching of the Kapton was observed also under VUV radiation alone. The exposure was done using a sample holder covering the QCM and preventing all plasma active species from reaching the sample except for VUV radiation penetrating through a MgF2 window. It is believed that the observed erosion is an artifact and it results from interaction of the VUV radiation with residual oxygen molecules exist in the sample holder cavity at operating pressure of 120 mTorr. This interaction forms oxygen atoms that are reactive and cause the observed etching. A similar experiment was done with Ar plasma and the results (not shown) indicate that VUV alone did not cause any etching of the exposed Kapton. 500

(c) Thickness Loss (A)

400

AO+VUV

o

300

(b)

200

AO

100

(a)

VUV

0

0

10

20

30

40

50

60

70

80

Exposure Time (min) Figure 2. QCM measurements of the Kapton thickness loss under VUV (a), AO (b) and AO+VUV (c) irradiation.

383 Chemical composition change. The pyromellitic dianhydride-oxydianiline (PMDAODA) polyimide (Kapton) structure is shown in Figure 3. The chemical species and assigned binding energies (BE) correspond to the molecular sites are labeled in Figure 3 and their values are given in Table 1. As a result of the 6 different carbon-bonding states, the carbon C1s XP spectrum is composed of 6 peaks having different binding energies (marked as1 to 6 in Table 1). In addition there is a π-π* shake-up peak, originating from the aromatic group, at a higher BE of 291.2 eV. Figure 4 shows highresolution C1s core-level line of Kapton pristine sample and its curve fitting with C1s lines originating from carbon with the different chemical states. A similar curve fitting was done for Kapton C1s spectra after exposure to net AO flux, VUV irradiation and combined AO and VUV flux. The relative intensity of the shake-up peak as a function of the exposure conditions is shown in the inset in Figure 4. The shake-up peak contribution of 9.2% for the pristine sample was not affected by the AO exposure. Exposure to VUV caused reduction of the shake-up peak concentration to 7.7%. Exposure to the combined environment of AO+VUV caused further reduction to 6.6%. O1

N

C 6 6 C

O1 4 4

2

2

O1

4

C 6

4

6 C

1 N

3

1 2 5 O 5

3

n

O1 ODA

PMDA

Figure 3. Molecular structure of poly (ether imide) (Kapton HN) showing labeled atomic sites.

TABLE 1. Poly(ether imide) (Kapton HN) Binding Energies (BE). C 1sa

π→π* 1 2 3 4 5 6 BE (eV) 284.7 285.6 285.7 285.8 286.3 288.6 291.2 % 36 10 10 19 10 17 a C 1s percentages add to 102% as stated in the original reference. Morphology modification. The effect of the irradiation type on the Kapton morphology is demonstrated in Figure 5 showing AFM images of 2x2 µm scan size. Pristine Kapton (Fig. 5a) has an initial roughness of 1 nm and no specific features on the surface. After AO exposure (Fig. 5b) the surface is eroded and is characterized by higher roughness of 5 nm and grains of about 300 nm in diameter. Addition of VUV to the AO resulted in slightly higher surface roughness (Rz=7nm) with the main effect on the grain size. The AO+VUV exposed Kapton is characterized by “needle like” surface with small grains

384

Shake-up peak intensity (%)

of about 100 nm in diameter (Fig.5c).

Intensity (arb. units)

C1s

C-C

9.5 9.0 8.5 8.0 7.5 7.0 6.5 6.0

Unexposed AO VUV AO+VUV Exposure environment

C=O

275

280

285

π−π* shake-up

290

295

300

Binding Energy (eV) Figure 4. XPS C 1s spectrum obtained from a solvent-cleaned, Kapton film before exposure to AO. The inset shows XPS C 1s shake-up peak intensity as a function of exposure environment.

3.1.2. Sample temperature The effect of the sample temperature on the Kapton erosion rate is shown in Figure 6. The Kapton was exposed to oxygen plasma while keeping all plasma parameters constant except for the sample temperature. Increase of the sample temperature results in higher erosion rate. The erosion rate follows Arhenius behavior as can be seen by the linear curve in the inset of Figure 6. The activation energy calculated from this curve was found to about 17 kJ/mole (~0.2 eV). 3.1.3. Plasma gas feed Other plasmas using gas feed of argon or nitrogen showed (see Figure 7) existence of etching, although lower than that observed for oxygen plasma. Using an inert gas such as argon, one would expect to observe no etching at all. The etching measured (Fig. 7a) indicates that the Ar ions produced in the RF plasma are energetic enough in order to cause a noticeable erosion yield. Using nitrogen (Fig. 7b) has the same effect of energetic ions in addition of higher reactivity compared to argon, thus resulting in higher etching rate.

385

Figure 5. AFM images of unexposed (a), AO irradiated (b) and AO+VUV irradiated (c) Kapton films. Rz – surface roughness. Images size 2x2 µm, Z-scale 40 nm.

3.1.4. Volatile products Different types of accompanying samples were located in the vicinity of the Kapton coated QCM crystal in order to study the effect of volatile products resulted from neighboring samples erosion on the Kapton etching rate. The results are shown in Figure 8. Each type of accompanying samples, i.e. poly(dimethyl)siloxane-based coating (PDMS), Teflon FEP film and Kapton film (reference material) was exposed in a separate experiment keeping similar dimensions and location of the sample in order to insure similar flow conditions. The results show that the presence of PDMS doesn’t affect significantly the Kapton etching rate, whereas the presence of fluoropolymer results in a drastic increase of the Kapton etching rate after a short incubation period. It can be suggested that Teflon forms reactive volatile products, such as HF and CxHyFz [12] that can react with Kapton increasing its erosion yield.

386

45

3000 ln(k)

35

Thickness Loss (A)

2500 o

(c)

k=d(Thick.Loss)/d(time) -Ea/RT k=Aexp

40 30 25 20 15

o

55 C

10

2000

3.0

3.1

3.2

3.3

3.4

(1/T)x1000

3.5

3.6

3.7

(b)

1500 o

1000

20 C

500

0C

0

(a)

o

0

10

20

30

40

50

60

70

80

Exposure Time (min) Figure 6. QCM measurements of the Kapton thickness loss as a function of sample temperature: 0°C (a), 20°C (b) and 55°C (c). The inset shows Arhenius plot of erosion rate constant as a function of temperature.

5500 5000

Oxygen Argon Nitrogen

Thickness Loss (A)

4500 o

4000 3500

(c)

3000 2500 2000

(b)

1500 1000

(a)

500 0

0

10

20

30

40

50

60

70

Exposure Time (min) Figure 7. QCM measurements of the Kapton thickness loss as a function of plasma gas feed: Ar (a), N 2 (b) and O2 (c).

387

9000

(c)

8000

Silicone paint Kapton Teflon

Thickness Loss (A)

7000 o

6000 5000

(b)

4000 3000

(a)

2000 1000 0

0

5

10

15

20

25

Exposure Time (min) Figure 8. QCM measurements of the Kapton thickness loss as a function of volatile products of co-exposed samples: Kapton (a), Polydimethylsiloxane (PDMS) (b) and Fluoropolymer (Teflon FEP) (c).

4.0 Discussion The space environment effects on spacecrafts’ materials are of obvious interest resulting in many in-flight experiments. However, these experiments are expensive, have limited availability and are very short in exposure terms. All that leaded to development of a variety of ground simulation facilities for studying the effect of the main space environment hazard constituent (AO) on materials degradation. Such facilities include plasma sources, ion neutralization, electron stimulation desorption, photo-dissociation, supersonic and laser detonation sources [5,6]. A major problem in applying an AO simulation facility for materials degradation studies is the way of assessing the AO flux. A common way is exposing a Kapton coupon in adjacent to the tested materials. By measuring its mass loss, and assuming a reaction yield of 3x10-24 cm3/O atom an equivalent of the 5eV AO flux could be calculated [7]. The generated AO beam in most AO sources is associated with accompanying reactive elements such as UV radiation, ions, electrons, and other energetic species that could affect the etching rate of the Kapton reference coupon. The AO source reactor atmosphere could also contain reactive volatile products (obtained by degradation of tested materials) and could also affect the Kapton coupon temperature, both might influence the Kapton erosion yield. The purpose of this study was to evaluate the validity of the method of assessing the AO flux by measuring the mass loss of Kapton coupon exposed

388 simultaneously to AO beam with other tested material. The study was done by exposure of Kapton coated QCM crystal to AO beam with different accompanying reactive elements and compare it erosion rate to AO exposure alone. As an AO source we used an RF plasma system that although is widely used as a common source for material screening with respect to AO degradation in LEO, it produces oxygen atoms at thermal energies (~0.04 eV) [8]. However, in addition to AO, other species are also present in the RF plasma environment including molecular oxygen, atomic and molecular oxygen ions and electrons at energies of tens eV, excited neutral and ionic species and also ~130 nm VUV radiation with flux of 1013-1016 photons/cm2sec [9,10]. By using a special sample holder design, we could expose the QCM crystals coated with Kapton polyimide films to AO, VUV or AO with VUV radiation. The system design allows also exposure to various gases plasmas and to change sample temperatures independently on the other plasma parameters. Let us discuss first the effect of the VUV radiation on the Kapton erosion rate. As is shown in Figure 2, the addition of VUV radiation (flux of 1.8x1016 photons/cm2sec) to the AO beam results in a significant increase in the Kapton etching rate. The phenomenon can be explained by studying the chemical changes occur at the surface measured by XPS. XPS spectrum of polymers with aromatic groups (such as Kapton) is characterized by the presence of a π-π* shake-up peak which is a result of electronic excitation of the π-electron system on an aromatic ring by a photoelectron [11]. The shake-up peak can be observed in the C1s and O1s spectra of unexposed Kapton samples. AO exposure causes surface reactions, hence modifying only the first atomic layers (~1 nm) [5]. The XPS probing depth is about 5-8 nm, which means obtaining a combined spectrum of damaged and undamaged layers. Kapton samples that were exposed to AO in LEO environment revealed a π-π* shake-up peak in the XPS spectra [13]. However, exposure of Kapton in an oxygen RF plasma asher caused diminishing of the π-π* shake-up peak, indicating a much deeper damage compared to space exposed Kapton. Since the RF plasma oxygen atoms have thermal energies, the only plasma component that could penetrate and react with the bulk material well below the XPS probing depth is the VUV component [13]. Pristine Kapton sample shows (Figure 4) a π-π* shake-up peak intensity of 9.2%. Exposure to AO beam reveals no change in the π-π* shake-up relative intensity and it remains 9.2%. Exposure to VUV originated from the main beam exhibits a decrease in the π-π* shake-up peak intensity of the XPS C1s spectrum (from 9.2% to 7.7%), indicating a destruction of some of the aromatic groups by the VUV radiation. A spectrum obtained after Kapton exposed to combined VUV+AO irradiation shows again the effect of the VUV radiation. In this case the π-π* shakeup peak intensity degradation was even larger, from 9.2% to 6.6%. The suggested erosion mechanism based on these results, is a destruction of the aromatic groups in underneath layers, and as a consequence the AO is reacting with already damaged layer thus resulting in higher etching rate (see Figure 2). The effect is also shown in the morphology modification (see Figure 5). A VUV irradiation simultaneously with the AO causing the modified damaged layer to have an increase in the oxygen reactive sites. The result is smaller grain size (100Å compared to 300Å for AO exposure alone) and higher surface roughness.

389 The temperature effect on the Kapton erosion rate (Figure 6) is showing Arhenius behavior with activation energy of about ~17 kJ/mole. Similar activation energy was measured by Tagawa [14] for thermal oxygen atoms reacting with Kapton. The high sensitivity of Kapton erosion rate to sample temperature could be a result of phenomena such as: (i) volume enlargement of the polymer matrix affecting its susceptibility to reactive species, (ii) increase in the volatile products desorption rate, or (iii) increase in the oxygen atoms surface diffusion. Note that the obtained activation energy of about 17 kJ/mole is too low to be attributed to the desorption processes. However, further experiments are needed to clarify the detailed mechanism of the observed effects. Other reactive elements in the RF plasma environment, such as, ions (Figure 7) and reactive volatile degradation products (Figure 8) were found also to affect the Kapton erosion rate. It should be emphasized at this point that although the study is based on one specific simulation facility the conclusions drawn are general in nature and apply to all other relevant facilities and especially to the use of Kapton as a reference material for assessing the AO flux. It was found that evaluation of the simulated AO flux based on a Kapton coupon sample could result in a massive error. The results shown indicate a flux measurement error of about one order of magnitude for relatively short exposure times. Higher AO fluences will result in much higher deviations from the correct fluence value. Kapton is used as a reference material for evaluating AO fluxes due to its extensive exposure in space and accurate erosion yield calculated from these experiments. It is, however, shown that its erosion yield could be significantly affected by other reactive elements in the AO source leading to inaccurate AO flux measurements. The study shows that there is a need for a better and more accurate method for evaluating the AO flux of ground simulation facilities. 5.0 Summary and Conclusions The most common way for assessing the AO flux in ground simulation facilities is by measuring the mass loss of a Kapton coupon attached to the tested materials. However, most LEO environment simulation sources generate AO beam associated with other reactive elements such as UV radiation, ions, electrons, and other reactive fragments that could affect the etching rate of the Kapton reference coupon. The AO source reactor atmosphere could also contain reactive volatile products (obtained by degradation of tested materials) and could also affect the Kapton coupon temperature, both might influence the Kapton erosion yield. The study includes in-situ measurements erosion rate of Kapton exposed to AO simultaneously with other simulation facility components, such as, VUV, ions, coexposed samples reactive volatile products, and sample temperature. The results indicate that the erosion rate could be severely affected by the presence of AO accompanying components. It is concluded that the Kapton erosion rate is system dependent. Thus, a simple and unambiguous method for measuring the AO flux is needed.

390 6.0 Acknowledgment This work was partially supported by ISA (Israeli Space Agency). 7. 0 References 1.

Reddy, M.R. (1995) Review: Effect of low earth orbit atomic oxygen on spacecraft materials, J. of Materials Science, 30, 281-307.

2.

Packirisamy, S., Schwam, D. and Litt, M.H. (1995) Atomic oxygen resistant coatings for low earth orbit space structures, J. Materials Science, 30, 308-320.

3.

Koontz, S.L, Albyn, K and Leger, L.J. (1991) Atomic oxygen testing with thermal atom systems: a critical evaluation, J. Spacecraft, 28, 315-323.

4.

Reddy, M.R., Srinivasamurthy, N. and Agrawal, B.L. (1992) Effect of low earth orbit atomic oxygen environmnet on solar array materials, ESA Journal, 16, 193-208.

5.

Minton, T.K., Garton, D.J. (2001) Dynamics of atomic-oxygen-induced polymer degradation in low earth orbit, in “Chemical Dynamics in Extreme Environments: Advanced Series in Physical Chemistry”, ed. Dressler, R.A., World Scientific, Singapore.

6.

See for example: 13th Space Simulation Conference, NASA CP 2340, Orlando, Florida, 8-11 October 1984.

7.

ASTM E1208.

8.

Banks, B.A., Rutledge, S.K., de Groh, K.K., Stidham, C.R., Gebauer, L. and LaMoreaux, C.M. (1995) Atomic oxygen durability evaluation of protected polymers using thermal energy plasma systems, NASA Technical Memorandum 106855, 1-15.

9.

Townsend, J.A. (1996) A Comparison of atomic oxygen degradation in low earth orbit and in a plasma etcher, Proc. 19th Space Simulation Conf., Baltimore MD, Oct. 29-31, 249-258.

10. Kearns, D.M., Gillen, D.R., Voulot, D., McCullough, R.W., Thompson, W.R., Cosimini, G.J., Nelson, E., Chow, P.P., and Klaassen, J. (2001) Study of the emission characteristics of RF plasma source of atomic oxygen: measurements of atom, ion, and electron fluxes, J.Vac. Sci. Technol. A 19, 993-997. 11. Nakayama, Y., Persson, P., Lunnell, S., Kowalczyk, S.P., Wannberg, B. and Gelius, U. (1999) High resolution X-ray photoelectron spectroscopy and INDO/S-CI study of core electron shakeup states of pyromellitix dianhydride-4,4’-oxydianiline polyimide, J. Vac. Sci. Technol. A , 17, 2791-2799. 12. Grossman, E., Gouzman, I., Lempert, G., Shiloh, M., Noter, Y., and Lifshitz Y. (2000) Advanced simulation facility for in-situ chracterization of space environmental effects on materials, Proceedings of the 8th International Symposium on Materials in a Space Environment, 5th International Conference on Protection of Materials and Structures from the LEO Space Environment, 5-9 June, Arcachon, France. 13. Golub, M.A., Wydeven, T. and Cormia, R.D. (1988) ESCA study of Kapton exposed to atomic oxygen in low earth orbit or downstream from a radio-frequency oxygen plasma, Polymer Communications, 29, 285-288. 14. Tagawa, M., Tagawa, Y., Kida, T., Yokota, K. and Ohmae, N. (2002) Temperature and Incident Angle Dependences of Atomic Oxygen Induced Erosion Rates of Polyimide and Polyethylene Films Measured by Quartz Crystal Microbalance, Proceedings of 6th International Conference on Protection of Materials and Structures from the LEO Space Environment, 1-3 May 2002, Toronto, Canada.

TEMPERATURE AND IMPINGEMENT ANGLE DEPENDENCES OF ATOMIC OXYGEN-INDUCED EROSION OF POLYIMIDE AND POLYETHYLENE FILMS MEASURED BY QUARTZ CRYSTAL MICROBALANCE M. TAGAWA, K. YOKOTA, T. KIDA and N. OHMAE Department of Mechanical Engineering, Kobe University, Rokko-dai 1-1, Nada, Kobe, Hyogo 657-8501 Japan Email: [email protected]; Voice and Facsimile: +81-78-803-6126 Abstract Herein, the ground-based experimental results of temperature and impingement angle dependences of erosion rate of polyimide and polyethylene films under hyperthermal atomic oxygen beam exposures are reported. The in-situ mass loss measurement was made during the atomic oxygen exposure by using a quartz crystal microbalance (QCM). A 5 eV atomic oxygen beam was generated with a laser detonation-type atomic oxygen beam source. The translational energy of the atomic oxygen beam was approximately 5.1 eV, and the typical flux of the beam was 6.5 x 1014 atom/cm2/s at the sample position. The polyimide and polyethylene films used in this study were spin-coated on QCM sensor crystals and the shifts in resonance frequency of polymer-coated QCMs were recorded with a mass resolution of 2 ng. It was observed that the erosion rates of both polyimide and polyethylene followed cosine function with incident angle of atomic oxygen. These results were explained by considering microscale roughness at the surface. The temperature dependence of erosion rate of polymers was also examined in the temperature range of 253 K to 353 K. The Arrhenius plots gave the activation energies of gasification reaction of 10-3 to 10-4 eV for both materials. It is, thus, concluded that the erosion rate of polymers in low Earth orbit is temperature independent due to the high collision energy of atomic oxygen. 1. 0 Introduction It has been recognized that atomic oxygen is one of the most important factors that influence the erosion of polymeric materials in a low Earth orbit space environment. Although a number of polymeric materials are utilized in space systems, polyimide is one of the most widely used polymeric materials in spacecraft applications. The erosion rate of polyimide films due to atomic oxygen attack in low Earth orbit have been established to be 3.00 x 10-24 cm3/atom and, hence, polyimide film has been used as one of the reference materials to evaluate the erosion rate of other materials. Polyethylene is also a candidate as a reference material because of its simple structure [1]. For reference

391

392 materials, erosion properties of polyimide and polyethylene at various exposure conditions need to be well understood. However, the existing basic knowledge of the erosion of polyimide and polyethylene under atomic oxygen exposures is not sufficient for predicting the erosion under various exposure conditions. Major factors that influence the erosion rate of polymer films for example are the impingement angle of atomic oxygen and the sample temperature. However, accurate measurements of these factors on the erosion rate have not been carried out. In this paper, we are reporting on ground-based experimental results of the impingement angle and temperature dependences of the erosion rates of polyimide and polyethylene films under a 5 eV hyperthermal atomic oxygen beam exposures. The in-situ mass loss measurements were made during the atomic oxygen exposure by using a quartz crystal microbalance (QCM), so that any possible disturbance influencing the reliability of the post-process erosion measurements (e.g., moisture absorption, contamination or unexpected change in the beam conditions during the exposure) could be eliminated. 2. 0 Experiments The laser detonation atomic oxygen beam source, which was originally invented by Physical Sciences Incorporation (PSI), was used in this study (Figure 1) [2]. Details of the experimental apparatus are reported elsewhere [3]. The translation energy of the atomic oxygen beam used in this study was approximately 5.1 eV, while the beam flux at Q-mass for TOF measurement

Rotatable QCM

Pulsed valve

Carbon dioxide laser Figure 1

Photograph of the laser detonation atomic oxygen source used in this study.

Figure 1. Atomic oxygen beam experimental facility

393 the sample position was measured at 6.5 x 1014 atoms/cm2/s by using a silver-coated QCM. The polyimide film used in this study was the pyromelliticanhydride (PMDA)-oxydianiline (ODA) polyimide supplied by Toray Industries Inc. (Semicofine SP-510). The polyimide film was spin-coated on a QCM sensor crystal, and annealed at 423 K and then at 573 K. Details of the sample preparation are reported in [4]. The polyimide film, thus prepared, was examined by X-ray photoelectron spectroscopy and it was confirmed that the surface structure was similar to that of Kapton-H which is a commercially available polyimide film. The polyethylene film was also prepared by a spin-coating on QCM sensor crystals. A solution which contains 0.3 g of low-density polyethylene (average molecular weight of 6500) in 40 ml of solvent was used for the spin-coating. The erosion rates of the polymer films were calculated from the change in the resonant frequency of the QCM during the atomic oxygen beam exposures. The frequency of the QCM was measured every 10 s with a frequency resolution of 0.1 Hz, which corresponds to a mass resolution of 2 ng. The temperature of the film was controlled with an accuracy of 0.1 K by the temperature-controlled recirculating water system. The impingement angle dependence was measured at the sample temperature of 311 K, where as the temperature dependence was observed in the sample temperature range between 253 K and 353 K with the impingement angle of 0° (normal incidence). Before the measurements, the polyimide and polyethylene films were exposed to atomic oxygen (6 x 1017 atoms/cm2) to saturate the surface oxygen content of the sample. This is in order to avoid the effect of the non-linear mass loss phenomenon observed at the beginning of the atomic oxygen exposure of pristine polymer surfaces [5]. 3. 0 Results and Discussion 3.1. IMPINGEMENT ANGLE DEPENDENCE Figure 2 shows the frequency shift of the polyethylene-coated QCM during the atomic oxygen beam exposures at the impingement angles from 0 to 90°. The impingement angle was taken with respect to the surface normal. A good linear relationship between the frequency shift and the exposure time, i.e., mass loss and atomic oxygen fluence, was observed at all impingement angles. The good linearity of the mass loss with atomic oxygen fluence was also identified for a larger time scale. The slope of the mass loss rate at every impingement angle was calculated by a least squares fit, and plotted against the impingement angle. The results are presented in Figure 3. It is clear that the rate of the frequency shift, or erosion rate, of polyethylene depends on the impingement angle and the dependence follows the cos0.87 law as indicated by the solid line in Figure 3. A similar result was obtained with a polyimide film; it also followed the cosine law (Figure 4) [6]. Note that the data points at the impingement angle of 80° in Figures 3 and 4 are affected by the QCM holder which blocks a part of the incoming atomic oxygen beam.

394 impingement angle: 0 deg. impingement angle: 10 deg. impingement angle: 20 deg. impingement angle: 30 deg. impingement angle: 40 deg. impingement angle: 50 deg. impingement angle: 60 deg. impingement angle: 70 deg. impingement angle: 80 deg. impingement angle: 90 deg. Plot 1 Regr

25

Rate of frequency shift of the

Frequency Shift (Hz/s)

20

15

10

5

0

0

100

200

300

400

Time (s)

Figure 2. Resonant frequency shift of polyethylene-coated QCM under the 5.1 eV atomic oxygen exposures at impingement angles from 0° to 90°.

Rate of Frequency Shift (Hz/s)

0.07 0.06

y = 0.059 cos0.87θ

0.05 0.04 0.03 0.02 0.01 0.00

0

10

20

30

40

50

60

70

80

90

100

Impingement Angle (degree)

Figure 3. Rate of frequency shift of the polyethylene-coated QCM as a function of impingement angle.

Banks and co-workers reported that the impingement angle dependence of the erosion of FEP Teflon in the LDEF flight experiment followed the cos1.5θ law rather than a cosine law [7]. An analysis of the flight data of Kapton-H and Mylar aboard STS-8 concluded that the impingement angle dependence followed the cos1.5θ law [8]. However, their conclusions were based both on a small number of data points obtained from flight experiments, which data incorporated a wide range of error spoiling the accuracy of the analysis. Furthermore, no physical explanation was provided for the cos1.5θ dependence.

395

Rate of Frequency Shift (Hz/s)

0.035 0.030 0.025

y = 0.031 cos θ

0.020 0.015

Experiments cos θ

0.010 0.005 0.000

0

20

40

60

80

100

Impingement Angle (degree)

Figure 4 Rate of frequency shift of the polyimide-coated QCM as a function of impingement angle.

Figure 5. Tapping mode atomic force microscope image of atomic oxygen exposed polyimide surface. Atomic oxygen fluence 8.8 x 1017 atoms/cm2, scan area 500 nm x 500 nm.

The cosine law of the impingement angle dependence observed in this experiment was physically explained as follows: the effective flux of atomic oxygen at the sample surface decreases with the increasing of the impingement angle; the effective flux of atomic oxygen is in proportion to the cosine of the impingement angle. The fact that the impingement angle dependence of the erosion rate follows a cosine law clearly indicates that the erosion rate is proportional to the effective flux of atomic oxygen; i. e., the reaction yield of oxygen atom is independent of the impingement angle. Figure 5 shows the atomic force microscope image of the polyimide film that was exposed to atomic oxygen with a fluence of 8.8 x 1017 atoms/cm2. Note that all experimental data

396 shown in Figures 2 and 3 were obtained using the same sample, so that the atomic oxygen fluence at the sample surface reached 1018 atoms/cm2, including pre- exposure of 6 x 1017 atoms/cm2, when the mass loss data were taken. Although, the atomic oxygen fluence is relatively small compared with many in-flight experiments, viewed at the microscopic level, the surface of the polyimide was already roughened due to the atomic oxygen attack. The peak-to-valley height of the surface was larger than 10 nm which is approximately 100 times larger than the size of a carbon atom. Therefore, on the microscopic scale, the impingement angle of oxygen atoms incident to the polyimide surface is widely distributed due to the presence of microscale roughness even though the macroscopic impingement angle is fixed. In addition, the multiple bounce effect, which is a key to the high reaction yield of atomic oxygen at the rough graphite surface [9], which also promotes the independence on the impingement angle in the reaction. Therefore, the microscopic roughness and the multiple bounce effect at the polymer surfaces erase the impingement angle dependence of atomic oxygen reactivity. Thus, the macroscopic erosion phenomena of polyimide simply reflects the effective fluence of atomic oxygen which follows the cosine law with the macroscopic impingement angle. 3.2. TEMPERATURE DEPENDENCE Figure 6 shows the frequency shift of the polyethylene-coated QCM during the atomic oxygen exposures at the sample temperatures from 253 K to 353 K. Since the frequency Temp: 253K Temp: 263K Temp: 273K Temp: 283K Temp: 293K Temp: 303K Temp: 313K Temp: 333K Temp: 343K Temp: 353K Least square fits

Frequency Shift (Hz/s)

30 25 20

-1 Activation Energy = 6.1x10-3 eV

loge (Hz/s)

35

15

-2

10 5 0

-3 0

100

200

300

400

Time (s)

Figure 6. Resonance frequency shifts of polyethylene-coated QCM during 5.1 eV atomic oxygen beam exposure at sample temperatures ranging 253 K to 353 K.

0.0028

0.0032

0.0036

0.0040

1/T (K-1)

Figure 7. Arrhenius plot of the erosion rates of polyethylene film during 5.1 eV atomic oxygen beam exposure.

397 shift of QCM, which corresponds to the erosion rate Re, was considered to be expressed in the Arrhenius-type function (Re = A exp(-Ea/kT), where Re is the reaction rate, A the pre-exponential factor, Ea the activation energy, k the Boltzmann’s constant and T the temperature), the relationship between 1/T and Re was plotted (Arrhenius plot). The results are shown in Figure 7. From the slope of the Arrhenius plots in Figure 7, the activation energy of mass loss reaction, Ea, was calculated to be 6.1 x 10-3 eV for the 5.1 eV atomic oxygen beam. The same experiment was carried out with a polyimide-coated -2

Loge(Hz/s)

Activation Energy : 4.71x10-2 (eV)

-3

-4

0.0028

0.0030

0.0032

0.0034

0.0036

0.0038

1/T (K-1)

Figure 8. Arrhenius plot of the erosion rates of polyimide film during 1.1 eV atomic oxygen beam exposure.

QCM, and the activation energy of 5.7 x 10-4 eV was obtained [10]. It was, thus, concluded that the activation energies of gasification reactions of hydrocarbons with a 5 eV atomic oxygen in LEO are in the order of 10-3 to 10-4 eV. These small activation energies are responsible for the erosion property of polymeric materials flown on space shuttle missions; no temperature dependence was obvious in LEO [11]. In contrast, some ground-based experiments using thermal atom system reported relatively large activation energies ranging 0.13-0.29 eV [12-14]. The inconsistency of the activation energies between space and ground-based experiments was considered to be due to the translational energy of impinging atoms. In order to clarify the origin of the discrepancy, activation energy was measured with a 1.1 eV atomic oxygen beam. Figure 8 shows the Arrhenius plot of gasification reaction of polyimide with the 1.1 eV atomic oxygen beam. From the slope of the Arrhenius plot, the activation energy of 4.7 x 10-2 eV was obtained [10]. This is an activation energy two orders larger than that with 5 eV atomic oxygen beam. The activation energy of 4.7 x 10-2 eV under the 1.1 eV atomic oxygen beam exposure was consistent with that reported by Cross [15]. The effect of translational energy on the temperature dependence in polyimide erosion, which has only been speculated in reports to date, was evident in this study.

398 4. 0 Conclusions The impingement angle and temperature dependences on the erosion of polyethylene and polyimide films by 5eV atomic oxygen were investigated. The in-situ mass loss measurements during atomic oxygen beam exposure clearly indicated that the erosion rate of polymers follows a cosine law regarding the impingement angle of atomic oxygen. The physical explanation of this phenomenon was made by considering the microscopic impingement angle at the polymer surfaces under the presence of the microscale roughness. Temperature dependences on the erosion rates of polymeric films in hyperthermal atomic oxygen exposures showed that the activation energies of erosion by 5 eV atomic oxygen beam were 5.7 x 10-4 eV for polyimide, whereas 6.1 x 10-3 eV for polyethylene. The presence of translational energy effects on the temperature dependence of polyimide erosion in LEO was experimentally verified. The basic knowledge of polymer erosion by 5 eV atomic oxygen beam is invaluable for the quantitative analyses of the synergistic effects upon the material erosion phenomena in a LEO space environment. 5. 0 Acknowledgments The authors acknowledge H. Kinoshita of Kobe University for his help in the experiments. A part of the study was supported by the Grant-in-Aid for scientific research from the Ministry of Education, Sports, Culture, Science and Technology, Japan, contract numbers 13750842 and 14350511. 6. 0 References 1. Minton, T. K., “Protocol for Atomic Oxygen Testing of Materials in Ground-Based Facilities, Version Number 2,” Jet Propulsion Laboratory, Publication 95-17, California Inst. of Technology, Pasadena, CA, 1995. 2. Caledonia, G. E., Krech, R. H., Upschulte, B. L., Sonnenfroh, D. M., Oakes, D., and Holtzclaw, K. W., “Fast Oxygen Atom Facility for Studies Related to Low Earth Orbit Activity,” AIAA Paper 92-3974, July 1992. 3. Tagawa, M., Yokota, K., Ohmae, N., and Kinoshita, H, “Volume Diffusion of Atomic Oxygen in α-SiO2 Protective Coating,” High Performance Polymers, Vol.12, No.1, 2000, pp. 53-63. 4. Kinoshita, H., Tagawa, M., Umeno, M., and Ohmae, N., “Surface Reaction of a Low Flux Atomic Oxygen Beam with a Spin-Coated Polyimide Film: Translational Energy Dependence on the Reaction Efficiency,” Transaction of the Japan Society for Aeronautical and Space Science, Vol.41, No.132 (1998), pp.94-99. 5. Tagawa, M., Yokota, K., Ohmae, N., Kinoshita, H., “Effect of Ambient Air Exposure on the Atomic Oxygen-exposed Kapton Films,” Journal of Spacecraft and Rockets, Vol.39, No.3 (2002) 447-451. 6. Yokota, K., Tagawa, M., Ohmae, N., “Impingement Angle Dependence of Erosion Rate of Polyimide in Atomic Oxygen Exposures,” Journal of Spacecraft and Rockets, Vol.39, No.1 (2002) 155-156.

399 7. Banks, B. A., Dever J. A., Gebauer, L., and Hill, C. M., “Atomic Oxygen Interactions with FEP Teflon and Silicones on LDEF,” LDEF-69 months in space, NASA CP-3134, 1991, pp.801-815. 8. Visentine, J. T., Leger, L. J., Kuminecz J. F., and Spiker, I. K., “STS-8 Atomic Oxygen Effects Experiment,” AIAA Paper 85-0415, Jan. 1985. 9. Kinoshita, H., Tagawa, M., Umeno, M., and Ohmae N., “Hyperthermal Atomic Oxygen Beam-induced Etching of HOPG (0001) Studied by X-ray Photoelectron Spectroscopy and Scanning Tunneling Spectroscopy,” Surface Science, Vol. 440, No.1, 1999, pp.49-59. 10. Yokota, K., Tagawa, M., and Ohmae, N., “Temperature Dependence in Erosion Rates of Polyimide under Hyperthermal Atomic Oxygen Exposures”, Journal of Spacecraft and Rockets, in press. 11. Peters, P. N., Gregory, J. C., and Swann, J. T., “Effects on Optical Systems from Interactions with Oxygen Atoms in Low Earth Orbits,” Applied Optics, Vol.25, No.8, April 1986, pp.1290-1298. 12. Koontz, S. L., Albyn, K., and Leger, L. J.,“Atomic Oxygen Testing with Thermal Atom Systems: A Critical Evaluation” Journal of Spacecraft and Rockets, Vol.28, 1991, pp.315-323. 13. Golub, M. A., Lerner, N. R., Wydeven, T., “Reactions of Atomic Oxygen (O3P) with Polybutadienes and Related Polymers,” Chemical Reactions on Polymers, Edited by Benham J. L. and Kinstile, J. F., ACS Symposium Series No.364, American Chemical Society, Washington, DC, 1988, Chapter 25. 14. Chou, N. J., Parazczask, E. B., Chaug, Y. S., and Goldblat, R., “Mechanisms of Microwave Plasma Etching of Polyimides in O2 and CF4 Gas Mixtures,” Microelectronic Eng., Vol.5, 1986, pp.375-386. 15. Cross, J. B., Koontz, S. L., and Gregory, J. C., “Laboratory Investigations Involving High-Velocity Oxygen Atoms,” Proceedings of the Advanced Aerospace Materials Symposium, 119th TMS Annual Meeting and Exhibit, TMS, Warrendale, PA, Feb. 1990, pp. 1-14.

This page intentionally left blank

INTEGRATING SPHERE UNIT FOR PRECISION MEASUREMENT OF ABSOLUTE REFLECTANCE AND TRANSMITTANCE OF SPACECRAFT MATERIALS IN A VACUUM CHAMBER V.V.EREMENKO Inst. for Low Temperature Physics & Engineering Nat.Acad.Sci of Ukraine, 47, Lenin ave, Kharkov,61103, Ukraine V.M.NAUMENKO Inst. for Low Temperature Physics & Engineering Nat.Acad.Sci of Ukraine 47, Lenin ave, Kharkov,61103, Ukraine Tel. -38-(0572)-30-03-41; -38-(0572)-32-09-91 Fax. -38-(0572)-32-23-70; -38-(0572)-33-55-93 E-mail:[email protected] V.N. FENCHENKO Special Research and Development Bureau of Inst. for Low Temperature Physics & Engineering Nat.Acad.Sci of Ukraine 47, Lenin ave, Kharkov,61103, Ukraine V.G.TYKHYY M.K.Yangel's Yuzhnoe Desigh Office 3,Krivorozhskaya str. Dnepropetrovsk,49008,Ukraine Abstract The unit for high absolute accuracy (1%) measurements in situ of the reflectance (total, specular and diffuse) and regular transmittance is described. The measurements of reflectance and transmittance we performed on the same spot of the sample, permitting, first, to calculate absorptance with a high accuracy and, second, to exclude errors induced by structural inhomogeneities of the sample and treatment causes surface detects, the structural inhomogeneities of the sample being due to radiation beam inhomogeneities in the space environment simulator. The high absolute accuracy does not necessitate the use of reference samples, simplifying considerably the unit design, the experiment preparation and the measurement process and hence, reducing the cost of experiment as a whole. The measuring optical block is based on the integrating sphere connected to the outside systems through fiber bifurcated and trifurcated optic cables equipped with collimating and collecting probes. It has the small dimension and is placed easy in a space simulator.

401

402 1. 0 Introduction The optical methods of investigation provide essential information on physico-chemical properties of materials and their changes under the influence of space factors. From the methodology standpoint, conceptual are the problems of determination of material stability in space and qualitative and quantitative analysis of contaminations. They are the most important because the industry is now being confronted with an acute problem of producing spacecrafts with a service life of 10 years or more. Ground-based investigations of material optical properties by using space environment simulators (below referred to as "space simulators") are economically more profitable then those in space (though they do not completely replace the latter tests). Considerable advances have been made in the ground tests described in [1]). But the ground facility in service has been developed a ling time ago, while the latest advances in optical technologies and the specifics of space simulator operation open up new opportunities. First, it should be noted that the exposure time for samples in space simulators is frequently in the range of several hours or several tens of hours, and the effect, as a rule, is only several percent. Hence, it is essential that the measurements should be of high accuracy (~1%). One of the most important optical characteristics of materials is the absorptance which is calculated by the reflectance and transmittance measured. Since the case in point is the high accuracy of measurement, account must be taken of even minor inhomogeneity of the samples studied, which may be of technology nature or resulted from low-quality surface treatment or induced by considerable inhomogeneity of radiation beams of space simulators. This necessitates that transmittance and reflectance be measured on the same spot of the sample under identical conditions (sample temperature and configuration, angle of incidence of probing optical beam, etc.). These requirement are usually violated, although meeting them is quite essential at all times and is obligatory in studies of inhomogeneous samples. Such violations also occur frequently when reference samples are used. Moreover, the use of reference samples makes the measurements more difficult and the cost of experiment more expensive. It is also well to bear in mind that reference samples are not necessarily reliable. This, for instance, especially true for those materials which have long put through filed test in space and then are returned to the earth for physical (in particular, optical) measurements. Absolute measurements eliminate the necessity for reference samples. The importance of optical measurements without removing samples from the simulator vacuum chamber after their irradiation is shown in [1] and [2]. Below we offer a brief description of the unit, which allows the above problems to be solved and which is reliable and efficient in operation. The optical block of the unit can be readily placed even inside a small-sized space simulator.

403 2. 0 Description of the Unit The unit (Fig.1) consists of a monochromator with a light source, a detector and a measuring optical block placed in the space simulator vacuum chamber. Outside facilities (controlling block, data processing, recording and visualization block and power systems) are not shown in the figure. The optical block is based on the integrating sphere connected to the outside systems through bifurcated and trifurcated fiber optical cables equipped with collimating and collecting probes. The collecting probe is located in the detector port of the integrating sphere and allows for a hemispherical signal pick-up.

All the assemblies in the simulator vacuum chamber (except 13,14,19) are mounted on a small platform or fixed directly on the integrating sphere, which is on the

404 same platform too. All this is called an optical block. During irradiation the optical block is taken away of the exposure zone with electric drive 19. The integrating sphere ~80mm in diameter has four ports. Flat samples 10mm in dia are mounted on a 100mm turret with is rotated with a step motor. The same turret may hold, if necessary, standards (one of them is simply an orifice in the turret) which are used in testing the unit at interval of several months. Temperatures of the turret and the samples are controlled by electronic control system 14 which incorporates a small liquid nitrogen vessel placed inside the vacuum chamber accessibly. Shutter 11 in port 8 of the integrating sphere has an absorbing coating and is set in one of two "open-closed" positions with electromechanical relay 18. Shutter 12 in port 10 has absorbing and diffuse reflecting coating and is set in one of three "diffuse reflecting-closed-open" positions. Light chopper 5 activated with a motor modulates probing 6 and reference 7 beams in antiphase. The optical block overall dimensions are approximately 140x120x220mm and, if necessary, may be optimized. 3. 0 Measurement of Optical Coefficients and Working Trends The unit makes it possible to measure total, specular and diffuse reflection and regular transmission. We shall describe briefly some methods of measuring to illustrate the unit operation. Let us consider first the measurement of total reflectance. To measure it, shutter 11 in the integrating sphere port is set in the position ensuring absorption, shutter 12 in the specular component outlet port in the position ensuring diffuse reflection of incident radiation, and the sample under study on the turret is in port 8. The probing and reference beams of the monochromator enter radiation inlet port 4 through the bifurcated light guide, the chopper and the collimating probes ensuring the beams quasiparallelism (the divergence being less then 10). Probing beam 6 is incident on the sample at a small angle (~100), then it is reflected with its surface, repeatedly rereflected with the sphere inner surface and enters the detector through the light guide with probe collected the beams from hemisphere (exclusive of the beams reflected from the sample). Here incident out of phase is also the reference beam repeatedly rereflected from the sphere inner surface. The total reflection coefficient, So , of the sample depends on the ratio of signals between the working, Bw , and reference, Be , channels. In the first approximation this dependence is of the form:

So =

Bw Be a − b ⋅ Bw Be

,

(1)

where the constants a, b are determined by the least-square method when calibrating the unit by measuring the total reflectance of some standards. There constants permit making allowance for the ports-induced inhomogeneity of the sphere inner surface, the departure of the sample surface from sphericity, the existence of the sample-sphere clearance (~1mm) and that between the shutter in the specular component outlet port

405 and the sphere (~0.5mm) and a possible nonidentity between the working and reference channels. The estimations show that the application of eq. (1) and some apparature provisions (a low value of the stray light inside the sphere and high values of linearity, dynamical range and sensitivity of the outside facilitics of the measuring channel and some other measures) provides the required absolute accuracy (1%) within the reflectance range 1-0.01. To measure the diffuse reflectance of the sample requires shutter 12 in the specula component outlet port be set in the position ensuring absorption of the incident light. The diffuse reflectance, S d , depends on the ratio of signals between the working, Bw , and reference, Be , channels and, in the first approximation, is of the form:

Sd =

Bw Be −k , a − b ⋅Bw Be

(2)

where k is the reflectance of the absorbing shutter in the specular component outlet port; it is determined when calibrating the unit by measuring the diffuse reflectance of standard, a standard with a high reflectance being chosen to improve the accuracy. When measuring the regular transmittance, shutter 11 in the sample port is set in the position ensuring light transmission and shutter 12 in the specular component outlet port in the position ensuring diffuse reflection of incident light. In this case the light from the working and reference channels enter the detector through the collecting probe in the detector port and then travel through one of the bundles of the trifurcated light guide. Besides, other light beams producing a signal proportional to the regular transmittance of the sample enter the detector through the collimating probe in the sample port and a second bundle of the trifurcated light guide. The collimating probe provides the incident of only that component of the probing beam with is transmitted through the sample. The beams incident at other angles (i.e. reflected from the sphere inner surface and the transmitted through the sample) become so weak that their influence may be neglected. The regular transmittance, St , of the sample is dependent on the ratio of signals between the working, Bw , and reference, Be , channels and can be given by expression:

St = ( Bw Be − So ) c ,

(3)

where ɫ is the constant determined when calibrating the unit, and S o is the total reflectance of the sample measured before (it is significant that S o was measured on the same spot of the sample). It should be mentioned that the constant ɫ takes account of the fact that incident on the detector is only a part of the light beam reflected from the sphere surface which is within the filed of view of the collecting probe (the constant c being equal to the ratio of signals between the working and reference channels when the probing beam passed through the turret orifice). The estimations made demonstrate that the application of Eq.(3) and the above apparature provisions provide the absolute accuracy (1%) within the regular transmittance range 1-0.05. When the high accuracy is required within the factor range 0.05-0.01, to Eq.(3) must be added the second-order terms of smallness the

406 determination of which requires additional calibration measurements on two or three standards. All calibration data are in the computer memory. Here also enter current data, and some time later (from several seconds to several minutes depending on the accuracy required and the spectral resolution) the diagrams of spectral coefficients are displayed and, if desired, printed. Some more specific features should be mentioned. Until recently, of all optical coefficients the emphasis was on the integral absorption coefficient normalized to the Sun spectrum. But to our opinion, important information may be carried by the spectrum of materials or its contaminants, their nature and reason of their occurrence. Therefore, a modern monochromator of small overall dimension and rather high spectral resolution (~0.1mm) is used in the unit described. As mentioned above, during irradiation of the samples the optical block is removed from the radiation zone and this appears to by sufficient for the optical block to remain intact and to have a long service time. But there are some space simulators in which the stray radiation is rather high. In such cases protecting shield are provided (of special rubber in the simple case) which are placed and removed with an additional electric drive mounted inside a simulator arbitrarily. For extreme cases with powerful space simulators and frequent severe contaminations inside simulators, we provide a version of the unit in which the optical block is placed inside a little vacuum chamber which is outside the simulator chamber and is connected to the simulator chamber with a sealed flange date valve joint. The turret with samples is inserted into the simulator for the time of irradiation, but after completion of the exposure it is removed on command to the optical block to perform measurements. This makes it possible not only to keep the optical block clean but to replace samples while the simulator is operating. We may produce any the optical variations. The unit specifications: - spectral range 250-2000nm (extended, if necessary)

- spectral resolution - transmittance, absorptance and reflectance interval - absolute error of measurement - angle of incidence of beam - angle of divergence of beams - turret temperature

0.1-5 nm 1-0.01 1% 100 10 80-430K

4. 0 References 1. Smith C.A., Dever J.A. and Jaworske D.A. (1997) Advances in optical property measurements of spacecraft materials, in: Proc. 7th Intern. Symp."Materials in Space Environment", Toulouse, France, 531-536. 2. Dever J.A. et al. The effects of the simulated low Earth orbit environments on the spacecraft thermal control coating, 38th Intern. SAMPE Symp. and Exhibition, Anaheim, California, 1993.

NUMERICAL SIMULATION OF THERMAL STRESS INDUCED BY THERMOCYCLING IN HOT ROLLED 1420 AL-LI ALLOY

HONGBIN GENG, SHIYU HE, DEZHUANG YANG Space Materials and Environment Engineering Lab., Harbin Institute of Technology Harbin 150001 Tel: 86-451-6412462, E-mail: [email protected]

Abstract A model of grains under plane stress state has been established. According to the grains model, thermal stress induced by thermal cycling (77K—393K) in the alloy is numerically simulated by finite element method. The numerical analysis results show that the difference in coefficient of thermal expansion and elastic modulus for grains along different crystal directions can produce alternate thermal misfit stresses and strains near boundaries due to thermal cycling. At the temperatures of the upper and lower limit (393K and 77K), thermal stress nearby grain boundary reaches maxima.

1. 0 Introduction It is well known that the difference of the thermal expansion coefficients, as well as the elastic moduli of matrix and the precipitate phases is of obvious importance in the thermal misfit stress and strain. The cyclic internal stress may induce a degradation of tensile properties or dimensional stability, and even thermal fatigue [1-9]. The factor may limit the service life of materials. For the time being, there is no experimental method that is able to quantitative analysis of the kind of thermal stress and strain. This study reports a model of grains

407

408 under plane stress state. According to the grains model, thermal stress induced by thermal cycling in the alloy is numerically simulated by finite element method.

2. 0 Experimental Procedure The 5mm thick plate of 1420Al-Li alloy hot rolling state consisted of grains (Fig.1). The specimens were cut normal to the rolling direction with a gage length of 70, 5mm width, and 2mm thickness. The final size of tensile specimens is shown in Fig.2. Tensile tests were carried out on an MTS with crosshead speed of 0.5mm/min and a fixed testing temperature within 77 to 393K. The temperature-time curves of thermal cycling are shown in Figure 3.

3. 0 The 2D Model for Thin Plate Fig.4 is schematic diagram of the plate with grains under tension and a two-grain model for the alloy. Fig.5 is finite element meshes of the model for the two-grain model. The grain 1 is defined as a rectangle, which consists of 180 cells of quadrilateral four nodes. Cell property is defined as the physical property of aluminium single crystal in hard orientation (Table 1). It is assumed that the yield stress of the cell is about 240 Mpa, and don’t depend on temperature.

The grain 2 was divided into 716 cells, and their

properties are defined as the physical property of aluminium single crystal in soft orientation on the assumption that the physical property keeps a constant with temperature changes. But the yield stress depends on temperature, and approximates the yield stress of the hot rolled alloy with polycrystal grains (Table 2). The cells obey Misses yield criteria, and their Poisson ratio equals 0.33. The whole model is consisted of 953 nodes. The reference temperature for zero thermal stress is 25. Simulation analysis in first thermal cycling period of 480 seconds is finished.

409

Fig.1. Optical

cr mi ostruc ture o thf

etho rolled alloy

Fig.3. Temperature vs. time curve of thermal cycling

Fig.2. Geometry of tensile specimen

Fig.4. Schematic diagram of the grains and their model under tension (a) the alloy plate with grains under tension (b) a model for the grains

Fig.5. Finite element meshes of the model for grains

Fig.6. Temperature and equivalent stress vs. time curves for node

410

Table 1 The physical properties of Aluminium single crystal along soft and hard orientation

[9]

orientation factor

Young's Modulus, Mpa

Coefficient of Thermal Expansion

0. 5

63700

24×10−6

0

76100

5×10−6

Table 2 Testing temperature and the yield strength of 1420Al-Li alloy Temperature, K

77

203

298

393

Yield strength, Mpa

320

300

280

220

4. 0 Results The yield strength of hot rolled alloy at different temperatures: Table 2 shows that the yield strength of hot rolled alloy vs. testing temperature from 77 to 393K. The yield strength of the alloy decreases with increasing temperature. The numerically analysis of the distribution of thermal stress: Temperature and equivalent stress vs. time of a period curves for node 214 near grain boundary and node 205 far from grain boundary are shown in Fig.6. The equivalent stress for node 214 and node 205 increases with decreasing temperature, and stress concentration nearby grain boundary increases too. At lower and upper temperature, the stress concentration reaches an extreme. The equivalent stress contour bands of the model at 0 s is shown in Fig.7. There is no thermal stress but external load now. Kt is defined as a factor of stress concentration at time t, Kt= ımax/ıa; ımax is the maximum of equivalent stress at time t; ıa is the average stress, which equals 40Mpa according to the model for the grains in Fig.4. So, Kt=0s§1.08, it is obvious that there is little stress concentration now. Fig.8 shows that the equivalent stress contour bands of the model at 144s, and Kt=144s§5.63, so, the factor of stress concentration reaches a maximum at lower temperature. Fig.9 shows that the equivalent stress contour bands of the model at 444s , and Kt = 444s§3.68, so, the factor of stress concentration also reach a maximum at upper temperature. A thermocycling period is finished when the temperature fall to initial point. Fig.10 shows that the equivalent stress contour bands of the model at 480s.

411

Fig.7. The model with equivalent stress contour bands Fig.8. The model with equivalent stress contour bands at 0 s

Fig.9. The model with equivalent stress contour bands at 444s

at 144s

Fig.10. The equivalent stress contour bands at 480s

5. 0 Conclusions The difference in coefficient of thermal expansion and elastic modulus for grains along different crystal directions can produce alternate thermal misfit stresses and strains near boundaries due to thermal cycling. At the temperatures of the upper and lower limit (393K and 77K), thermal stress nearby grain boundary reaches maxima.

412 6. 0 References 1. Yajun Wang and Zhiyong Mao, Sinica J. , Aeronautical Manufacturing Technology, 1998, 6, 12. 2. Jixiong Deng, Yan Li, Wenyang Zhang, Sinica J. Materials Engineering, 1998, 11, 24. 3. Bratukhin A.G, Densisov B.S, Sotnikov V.S, Sinica J. Materials Engineering, 1997, 2, 38. 4. Le Flour J.C. and R.Locicero, Scripta Metall., 1987, .21, 1071. 5. Siegmund T., Werner E. and F.D. Fischer, Mater.Sci. Eng., 1993, A169, 125. 6. Daehn G.S. and T.Oyama, Scripta Metall., 1988, 22, 1097. 7. ViVek Ratna and D.S.Sarma, Scripta Metall., 1993,29:467 8. Tsuji K., Y.Takegawa and K.Kojima, Mater. Sci. Eng., 1991, A136, L1. 9. Zhongxi Cui, Metallography and Heat Treatment, Mechanical Industry Publishing House, 1989

CHANGES IN MICROSTRUCTURE AND TENSILE PROPERTIES OF HOT ROLLED 1420 AL-LI ALLOY SUBJECTED TO THERMOCYCLING

Hongbin GENG, Shiyu HE, Dezhuang YANG Space Materials and Environment Engineering Lab. Harbin Institute of Technology, Harbin 150001 Tel: 86-451-6412462, E-mail: [email protected]

Abstract The thermal cycling tests, which simulate the cyclic temperature events in space, between 77 and 393K with a period of 480s by a constraint thermal cycling apparatus were performed to hot rolled 1420 Al-Li alloy. The tensile properties before and after thermal cycling were evaluated. The microstructure of the alloy was studied by means of TEM. Experimental results show that the strength and ductility of the alloy decrease obviously after 1000 or 3000 thermal cycles. Thermal stress induces changes in microstructure are characterized by emitting dislocations from boundaries to matrix, forming dislocation pile-ups against grain boundaries and an increase of dislocation density. The accumulation of changes in microstructure may lead to stress concentration at grain boundaries, and result in an obvious degradation in tensile properties.

1. 0 Introduction 1420Al-Li alloy, by virtue of its low density, higher modulus-to-density, strength-to-density and favorable weldability, is attractive for structural applications in spacecraft [1-5]. However, these structural materials are often subjected to thermal cycling. It is well known that the difference of the thermal expansion coefficients, as well as the elastic moduli of matrix and the precipitate phases is of obvious

413

414 importance in the thermal misfit stress and strain. Thermal cycling can influence the deformation and failure behavior of the alloy to a considerable extent. The cyclic internal stress may induce a degradation of tensile properties or dimensional stability, and even thermal fatigue [6-10]. The factor may limit the service life of materials. There is little information about degradation of 1420 Al-Li alloy induced by thermal cycling between 393 and 77K. This study reports the degradation in tensile properties of 1420 Al-Li alloy by focusing on the microstructural changes under the conditions of thermal cycling and an applied external load.

2. 0 Experimental Procedure The 5mm thick plate of 1420Al-Li alloy hot rolling state consisted of lathy grains (Fig.1). The specimens were cut normal to the rolling direction with a gage length of 70, 5mm width, and 2mm thickness. Thermal cycling tests were carried out in a constraint thermal cycling facility in which the specimens applied 40 MPa external loads were heated by applying 2-volt voltage and then cooled by liquid nitrogen. The temperature was controlled by thermocouples welded to the specimen surface. The temperature-time curves of thermal cycling are shown in Figure 2. The final size of tensile specimens is shown in Figure 3. Tensile tests were carried out on an MTS with crosshead speed of 0.5mm/min at room temperature. Thin foils for electron microscopy (TEM) observations were twin jet electropolished at a maximum temperature of 233K and at a potential of 25V(DC) in a solution consisting of 10 parts methanol, 7 parts n-butanol, and 1 part perchloric acid.

Figure 1. Optical microstructure of the hot rolled alloy

Figure 2. Temperature vs. time curve of thermal cycling

Figure 3. Geometry of tensile specimen

415 3. 0 Results

3.1 EFFECT OF THERMAL CYCLING ON MICROSTRUCTURE OF 1420 AL-LI ALLOY A representative TEM micrograph of the specimen before thermal cycling is shown in Figure 4. The microstructure of the specimen before thermal cycling consisted of lathy grains. The subgrains also are lathy, in which there are few dislocations. No precipitate phases can be observed in the grains. After thermal cycling, the metallurgical structure of the specimen has no change, but the microstructure has obviously change. With increasing thermal cycles, the density of dislocations of the specimen increases. The dislocations are sent from the boundaries to interior grains. The dominant feature is the dislocation pile-ups against sub-grain boundaries and dislocation tangles.

Figure 4. TEM micrographs showing the sub-grains and dislocation configuration in the alloy before and after thermal cycles: (a) N=0; (b) N=1000; (c) N=3000

3.2 EFFECT OF THERMAL CYCLING ON TENSILE PROPERTIES OF WELDED JOINT OF 1420 AL-LI ALLOY

416

417

418 9. ViVek Ratna and D.S.Sarma, Scripta Metall., 1993, 29, 467 10. Tsuji K., Y.Takegawa and K.Kojima, Mater. Sci. Eng., 1991, A136, L

PHOTOSIL¥ SURFACE MODIFICATION TREATMENT OF POLYMERBASED SPACE MATERIALS AND EXTERNAL SPACE COMPONENTS

Yu. GUDIMENKO, R. NG, J.KLEIMAN, Z. ISKANDEROVA Integrity Testing Laboratory, Inc., Markham, Canada 80 Esna Park Dr., Units 7-9, Markham, Ontario, L3R 2R7, Canada P.C. HUGHES, R.C. TENNYSON University of Toronto Institute for Aerospace Studies, Toronto, Canada 4925 Dufferin St., Toronto, Ontario, M3H 5T6, Canada D. MILLIGAN MacDonald Dettwiler Space and Advanced Robotics Ltd. 9445 Airport Road, Brampton, Ontario L6S 4J3, Canada Abstract Results of a space program that involved ground-based testing, characterization, and durability evaluation of advanced thermal-control polymer-based materials and black and white thermal control paints with organic binders are presented. The materials have been treated by different variants of Photosil¥, a patented surface modification technology. Some of the investigated in this work materials are similar to the surfacemodified materials included in “Materials on International Space Station Experiment” (MISSE), being tested now in flight conditions. The Photosil™ process has also been successfully applied to a variety of external space components that included flat braided fiber lacing tape and painted surfaces of grapple fixtures, target plates and rods, upper and lower housings, camera baffles. The polyurethane paints used on these components consisted of white paint (Aeroglaze A276), black paints (Aeroglaze Z306 and Aeroglaze Z302), and a gray paint (a mixture of Aeroglaze A276 and Z306). In all cases, the Photosil™ treatment was able to effectively modify the treated surfaces. Fast atomic oxygen beam and plasma asher testing results revealed the Photosil™ treatment significantly reduced the erosion of the polymer films, paints, and the lace, while leaving their outgassing, thermal cycling, and for some of them, thermaloptical characteristics almost unchanged, that is essential in space applications. 1.0 Introduction Polymers and polymer matrix composites have become vital engineering materials for spacecraft design, being used in spacecraft structures, payloads, thermal control components, and power subsystem applications because of their light weight and excellent strength, and the fact that they can be used in applications where other

419

420 materials could not, such as a flexible, thin thermal insulation blanket and flexible second surface mirrors. The drive to use polymeric materials and composites is quite simple; they allow the engineer to explore new applications and clearly offer a way for weight savings, while improving performance. One of the limits to using polymers and polymer matrix composites in low Earth orbit (LEO) is the effect of Atomic Oxygen (AO) erosion. Unprotected polymeric materials exposed to the AO environment undergo accelerated erosion, limiting their long-term use capability. The erosion effect of AO on polymeric materials has been widely published [1-9]. The patented Photosil™ surface modification process [10,11] has been developed to provide a solution to the severe problem of AO erosion of polymer materials and composites used on spacecrafts in low Earth orbit. The Photosil™ process is a surface modification technology that substantially alters the surface structure and chemistry of a polymer, incorporating silicon-containing chemical groups into the sub-surface layer (i.e. up to 1 µm in depth) of the polymer structure. The surface becomes a new material and attains new properties. In essence, the Photosil™ is a three-stage surface treatment consisting of: (1) photo-activation, (2) liquid-phase silylation, and (3) stabilization, allowing effectively modify a wide variety of polymers and composite materials, from polyethylene to complex polyimides [8,10,11]. This surface modification technology produces a uniformly graded subsurface region without an abrupt transition boundary that thereby resists cracking and spalling caused by thermal and physical stresses. Under certain circumstances, the modified surface structure also has a unique self-healing capability. A key advantage of this proces is that only a thin layer near the surface is modified, resulting in little or no change to the bulk material properties. The Photosil™ surface treatment either significantly reduces or eliminates the AO erosion of polymer materials used in spacecraft applications. The Photosil™ process has shown to provide exceptionally low erosion yields of =10-26 g/atom under fast AO exposure, i.e. two orders of magnitude lower than exhibited by most polymer materials used presently for space applications, for a variety of materials including: Kapton® polyimide, Mylar® PET, PEEK, polyethylene, PVC, polyamide, graphite fiber reinforced PEEK, epoxy composites, and polyurethane-based paints [10,11]. In many cases, the process does not significantly affect the thermal-optical properties or mechanical properties of the treated materials, a critically important factor for many space applications. 2. 0 Experimental 2.1. SURFACE MODIFICATION AND AO TESTING OF POLYMER-BASED SPACE MATERIALS Ten selected space materials used for thermal control and other spacecraft applications (Table 1) have been treated by different versions of Photosil™ surface modification technology. These materials represent the clear and the metallized thin polymer films, polymer-based lacing tape, and three Aeroglaze paints (white and two black ones, flat

421 and glossy) used for thermal control and other space applications on different components of the International Space Station Mobile Servicing System. TABLE 1. Specification of materials, selected for surface modification

No. 1 2 3 4 5 6 7 8

Material Kapton 100 H Lot# 92-02-1-1 Polyester film Mylar Type 500D-1 Kapton 50 EAg Spec# 3390 Kapton 50 EAu Spec#33902 PEEK Film

Chemical Composition Polyimide film Poly(ethylene terephthalate) film Polymidide film/silverized (for inflatables) Polymidide film/goldized (for inflatables) Polyetheretherketone film

Kapton 500 HN Kapton 500 H Lacing Tape HT-30-TVS Black Paint Aeroglaze Z302

Polyimide film Polyimide film Resin-impregnated Nomex fiber braid Aromatic polyurethanebased thermal control paint (glossy)

10

White Paint Aeroglaze A276

11

12

9

Supplier

Sample Characteristics

Sheldahl Corp. DuPont

1 mil 5 mil

L’Garde, Inc

0.5 mil

L’Garde, Inc

0.5 mil

Westlake Plastics Company DuPont DuPont MDRobotics

3 mil 5 mil 5 mil Lace (tape)

Lord Corporation

Painted space-grade Al alloy disks. d ~ 40 µm

Aliphatic polyurethanebased thermal control paint

Lord Corporation

Painted space-grade Al alloy disks. d ~ 60 µm

Black Paint Aeroglaze Z306

Aromatic polyurethanebased thermal control paint (flat)

Lord Corporation

Kapton E

Polyimide film

L’Garde, Inc

Painted space-grade Al alloy disks. d ~ 30 µm 0.5 mil

The paints have been applied over special primers onto space-grade Al-alloy disks, following Aeroglaze (Lord Corporation) specifications for painting and curing. The thickness of the dry paint films was ~25-38 µm for Z306, ~51-76 µm for A276, and ~35-42 µm for Z302. The variable parameters in Photosil™ treatments included: the type and the energy of the UV or corona irradiation source for pretreatment (surface activation) and post-treatment (stabilization), the temperature and the time of silylation and the type of silylation agent. The material samples listed in Table 1 were tested before and after the treatment in a ground-based AO simulator and a plasma asher to determine the erosion behavior, i.e. the degree of achieved protection. The AO testing was performed in recently modernized UTIAS/ITL space simulator, under fast Atomic Oxygen (AO) beam exposure, similar to testing described in [4,8,10]. Each sample was placed in a holder that oriented the sample at 45° to the trajectory of AO beam. The samples were exposed to fast atomic oxygen (FAO) with energy ~2.5-3.0 eV at average flux of 1x1016 atoms/cm2·s for a minimum period of 8-10 hours and were held at a constant temperature for the duration of the test. More details

422 on the FAO system can be found in [4,8]. Similar samples were placed in the same chamber for the duration of the test outside the FAO beam, to account for mass loss due to outgassing. The plasma exposure tests were conducted in a low-temperature, inductively coupled radio frequency plasma asher. The asher operated at 13.56 MHz with the following settings: RF power ~ 200 watts, oxygen pressure ~ 100 mTorr ± 5%, oxygen input ~ 100 sccm ± 5%, minimum average equivalent fluence ~ 1.9×1020 atoms/cm2. The test and witness samples were placed in a holder and positioned in the middle of the 8 L asher reactor. All samples have been left in the plasma facility under vacuum for ~24-48 hours, for outgassing. The samples were exposed to the plasma for a minimum period of 5 hours that is equivalent to a total effective fluence of about 1.9×1020 atoms/cm2, as established from the mass loss for a control Kapton 500 HN sample that was equal to about 820 µg/cm2. Mass loss of the surface-modified samples was also measured by an electronic microbalance after the same time of plasma exposure. To normalize the averaged “erosion rates” in the Plasma facility, the rate of erosion (by mass loss) for untreated Kapton 500HN was set to 1.0. The results are presented in table 2. The erosion yield cited for the treated materials represents the lower limit of measurable mass loss established through weight measurements using an electronic microbalance, i.e. when no mass loss could be measured within the accuracy of the microbalance, the erosion yield limited by the resolution of the equipment is cited. The results of mass loss for all tested materials are presented in Table 3. The “averaged erosion rate” was not calculated, since it is impossible to distinguish the mass loss due to reduced erosion (if any) and some mass loss (mass change) due to conversion process, i.e. final stabilization of the treated samples under AO beam. The final conclusion about durability improvement required high magnification scanning electron microscopy (SEM) analysis (see Results and Discussion). As can be seen from Table 3, the mass loss for all materials after the Photosil™ treatment decreased in FAO exposure tests by about an order of magnitude, with best results achieved for all tested Aeroglaze paints and the lacing tape. 2.2 SURFACE MODIFICATION OF EXTERNAL SPACE COMPONENTS Various external space components used on the International Space Station (see table 4 for the list of components), painted, in most cases, by the same paints, have been Photosil™-treated for LEO space durability improvement. The description of the components and the lacing tape as well as their visual images and the details of the silylation process are given in the sub-sections below.

423 TABLE 2. Process identification for Photosil™ treatment and oxygen plasma testing results. (C) is used for silylation with corona pre-treatment.

No. 1

Material

Chemical Composition

Kapton H (C)

=0.16

Poly(ethelene terephthalate) Kapton 50E/silverized

Mylar (C)

=0.25

Kapton Ag (C)

=0.16

Kapton 50E/goldized

Kapton Au (C)

=0.16

Polyetheretherketone Polyimide film

PEEK (C) Kapton 500HN (C) Kapton 500HN Kapton 500 H (C) Kapton 500 H Kapton E (C) Lacing Tape Black Paint

=0.25 =0.14

5 6 7 8

Kapton 500 HN Kapton 500 H

Polyimide film Polyimide film

9 10 11 12

Kapton 500 H Kapton 50 E Lacing Tape Black Paint Aeroglaze Z302 White Paint Aeroglaze A276 Black Paint Aeroglaze Z306 Kapton 500HN

Polyimide film Polyimide film Braided fibres, synthetic resin Aromatic polyurethane Based paint Aliphatic polyurethane based paint Aromatic polyurethane-based paint Polyimide film

3 4

13 14 15

Averaged relative erosion rate

Polyimide film

Kapton H 1 mil Lot# 92-02-1-1 Mylar Polyester Mylar Type 500D-1 Kapton EAg Spec# 3390 Kapton EAu Spec# 33902 PEEK Natural Film Kapton 500 HN

2

Photosil™ Process ID

=0.16 =0.14 =0.16 =0.16 n/a =0.1 =0.1

White Paint Black Paint

=0.1

Pristine

1.00

2.2.1. PHOTOSIL™ TREATMENT OF PAINTED SPACE COMPONENTS In the application of the Photosil™ process to the painted surfaces, the activation stage was conducted in custom-made UV chamber under normal atmospheric conditions. This chamber is equipped with UV low-pressure quartz-mercury vapor lamps that generate the UV emission in UV-C region with a maximum at 254 and 185 nm. Four lamps of 27 feet in length each are stationary mounted, and six portable one-foot long lamps are used as well. At this stage in process, the functional groups with active hydrogen atoms, such as hydroxyl (OH), are created in the surface region of the paint coating. The activation process is a result of simultaneous excitement of polymer molecules and attack by molecular oxygen, as well as ozone, atomic oxygen, and singlet oxygen generated from molecular oxygen by the UV radiation. Maximum shelf life of the activated parts is about 30 min. The liquid-phase silylation stage was carried out in a specially designed chamber (See Figure 1). The custom-made silylation chamber allows the treatment of three-dimensional parts under low humidity (< 10%) and temperature between 30-100 °C. Depending on the chemical structure of the surface to be treated, a number of different silylating agents, diffusion promoters, and solvents can be used to prepare the silylating solution.

424 TABLE 3. Results of FAO testing for materials treated by Photosil™ technology (FAO fluence 1.5x1020 at·cm-2)

Material Kapton 500HN Kapton 500HN Kapton 500H Kapton 500H(C) PEEK PEEK Mylar Mylar Kapton 50E Kapton 50E Kapton 50EAg Kapton 50EAg Kapton 50EAu Kapton 50EAu Kapton 500HN Kapton 500HN Kapton 500HN (C) White Paint A276 White Paint A276 Black Paint Z302 Black Paint Z302 Black Paint Z306 Black Paint Z306 Lacing Tape Lacing Tape

Treatment Pristine (witness) Silylated Pristine Silylated Pristine Silylated Pristine Silylated Pristine Silylated Pristine Silylated Pristine Silylated Pristine Silylated Silylated Pristine Silylated Pristine Silylated Pristine Silylated Pristine Silylated

Mass loss after FAO exposure, µg/cm2 660 60 650 70 570 +50 860 70 190 70 340 10 340 30 720 30 +50 480 0 550 0 280 0 200 20

The silylating solution can be applied by one of painting techniques, such as dipping, brushing, spraying, depending on the size and geometry of the parts. After the silylation at the prescribed temperature and time, the parts were rinsed with a solvent and dried under the conditions prevailing in the silylation chamber. The goal of the silylation step is to replace the reactive hydrogen atom from the functional groups formed in the activation stage with silicon groups from the silylating agent, creating an organo-silicon molecule derived from the original polymer. Stabilization of the silylated surface is required to convert the organosilicon surface layer to a more stable silicon containing structure. The same set-up as for the photo-activation was used for stabilization of the silylated space components.

425 TABLE 4. External space components protected by Photosil™.

Space Component

Grapple Fixture

Target Plate & Rod

Upper Housing

Lower Housing

Baffle, Camera

Lacing Tape

Description

Image

The grapple fixture is attached to a satellite or payload for the end effector of the Remote Manipulator System (RMS) robot arm to grasp it and maneuver it. The grapple fixture is painted with polyurethane-based gray (Aeroglaze A276:Z306) and black (Aeroglaze Z306) paints. Target plate/rod is used as a visual alignment aid to assist in positioning the RMS robot arm end effector over the grapple spike for capture. The visual cues of the target and rod are painted with black (Aeroglaze Z306) paint with white (Aeroglaze A276) markings.

Socket extension tool, external component of the Special Purpose Dexterous Manipulator. Painted with Gray paint (Aeroglaze A276:Z306).

Socket extension tool, external component of the Special Purpose Dexterous Manipulator. Painted with Gray paint (Aeroglaze A276:Z306). Camera baffle built for the Special Purpose Dexterous Manipulator. Painted with Black paint (Aeroglaze Z306). Aromatic polyamide (Nomex) flat braided lacing tape impregnated with a synthetic resin. The braided fiber lacing tape is suitable for use with electrical wire harness assemblies.

Note: -Aeroglaze Z306 is a flat black absorptive paint, which consists of fumed silica and carbon black pigments in a polyurethane binder.; -Aeroglaze A276 is a white reflective paint made from titanium dioxide pigment in a polyurethane binder. Gray polyurethane-based paint is a mixture of the Z306 and A276.

426

Figure 1. Photograph of the silylation chamber.

2.2.2. PHOTOSIL™ SURFACE TREATMENT OF THE LACING TAPE Aromatic polyamide (Nomex®) fiber flat braided lacing tape, suitable for spacecraft wire assemblies, was treated in a continuous feed Photosil™ process (See Figure 2). In this set up, the lace was continuously fed over a number of rollers mounted along the treatment stations of the three-stage process. The photo-activation stage of the lace was conducted in an UV/Ozone chamber. Following activation, the lace was fed through a cylindrical aluminum vessel, where the liquid phase silylation stage took place. The vessel, filled with the silylating solution, was wrapped around with a heating tape. The heating tape maintained the liquid bath inside the vessel at the prescribed temperature. After exiting the silylation bath, the lace was rinsed in a solvent bath and air-dried before being rolled onto a take-up spool. The lace was fed through the system at a speed of approximately 3 m per hour. Following a complete activation and silylation run, the lace was reloaded onto the feed spool and carried once again through the UV/Ozone system, where the silylated lace underwent stabilization.

Figure 2. Schematic diagram of the continuous Photosil™ process lace treatment facility.

427 3.0 Post-Process Analysis of Photosil™ - Treated Components To study the changes in composition, structure, and surface morphology after the surface treatment, a number of complementary surface analysis methods have been used. X-ray photoelectron spectroscopy (XPS) analysis was carried out using a modified SSL SSX-100 X-ray photoelectron spectrometer. Survey scans of all samples were obtained using a 600 µm X-ray spot size and 150 eV pass energy. The samples were analysed under a nickel grid to minimize peak distortion due to charging effects. Scanning electron microscopy (SEM) analysis of all samples was performed using a JEOL JSM-T300 model microscope. The surfaces of the polymer samples were vacuum coated with a thin layer of carbon to prevent charging. An energy dispersive spectroscopy (EDS) x-ray microanalysis system URSA manufactured by Mektech Inc. was used for elemental analysis in some cases.

4.0 Results and Discussion 4.1. THIN POLYMER FILMS AND POLYAMIDE FIBER LACING TAPE. XPS surface analysis was used to characterize and verify the surface modification of the thin polymer films, lacing tape, and paints. The results for polymer films and the lace are presented in Tables 5 and 6. The elemental compositional analysis of the upper layer of the Photosil™-treated lacing tape revealed that the fluorine content dropped sharply and the content of silicon and oxygen had increased (Table 6). As can be seen from table 6, as a result of the Photosil™ treatment, the silicon content has increased dramatically to 20 at.%, with oxygen making up 23 at.% of the surface content. This data confirms the modification of the uppermost surface region of the lace. For all Photosil™-treated thin polymer films, the amount of carbon was significantly reduced, the oxygen amount increased, and ~15-25 at% of Si appeared at the surface. Such changes, based on previous results [8,10,11], clearly indicate at a successfully completed Photosil™ treatment. Scanning electron microscopy (SEM) was used to study the surface features of AO exposed polymer films and lace and the erosion protection capability of the Photosil™ process. The SEM micrograph in Figure 3 shows an untreated lace sample exposed to oxygen plasma for a total effective fluence of 2.0 x 1020 atoms/cm2, estimated from the mass loss of a control Kapton specimen that was exposed to the environment along with the tested lace. The left section that represents the untreated lace masked from AO shows no change in surface morphology as expected. In contrast, the exposed area (right section) shows the typical highly eroded surface texture and the development of surface pores

428 TABLE 5. Surface composition of pristine and silylated materials by XPS.

Materials/Treatment

Elemental Composition, at %

C

O

N

Si

Kapton 500HN pristine Kapton 500HN silylated Kapton 500HN silylated (C)

76.62 37.71 36.15

16.64 37.61 40.25

6.33 0.76 0.85

0.36 23.92 22.75

Kapton 500H pristine

76.63

16.67

6.44

0.10

Kapton 500H silylated Kapton 500H silylated (C)

33.96 39.71

41.46 37.48

0.72 0.82

23.86 22.00

Kapton 50E pristine

77.31

16.09

6.28

0.24

Kapton 50E silylated (C)

53.94

26.91

2.85

16.30

Kapton 50E Ag pristine

77.31

16.09

6.28

0.24

Kapton 50E Ag silylated (C)

45.23

30.12

0.79

23.86

Kapton 50E Au pristine

77.47

15.99

6.22

0.24

Kapton 50E Au silylated (C)

54.30

27.54

2.91

15.26

Mylar 500-1 pristine

73.13

26.62

0.00

0.07

Mylar 500-1 silylated (C)

36.85

39.22

0.47

23.46

PEEK pristine

81.45

15.89

0.00

2.56

PEEK silylated (C)

42.11

35.91

0.78

21.20

Kapton 100 H (1 mil, Sheldahl) pristine

77.65

15.65

6.53

0.10

Kapton 100 H (1 mil, Sheldahl) silylated

50.10

32.59

2.46

14.85

TABLE 6. XPS surface composition (at.%) of lace.

Specimen Lacing Tape

Treatment Control (untreated) Photosil™-treated

C 52.50 43.70

XPS chemical content, at.% O Si N 13.86 23.34

2.01 20.39

0.87 0.05

F 30.76 12.52

The Photosil™-treated lace in Figure 4 exhibits no visual deterioration features after exposure to oxygen plasma. For comparison purpose, the left section of the Photosil™-treated lace was also masked from plasma and the right section was exposed to plasma. No change in surface morphology was evident between these two sections, as noted in the SEM micrograph (Figure 4). These visual results have demonstrated the protection effectiveness of the Photosil™ process. It is also important to mention that similar results have been obtained after FAO beam testing.

429

(a)

(b)

Figure 3. SEM micrograph of oxygen plasma exposed untreated lacing tape (effective fluence ~ 2.0 x 1020 atoms/cm2). Magnification 2000x: (a) masked section; (b) exposed section. The black scale bar in the lower right corner of the SEM image corresponds to 10 m

(a)

(b)

Figure 4. SEM micrograph of oxygen plasma-exposed Photosil™-treated lacing tape (effective fluence ~ 2.0 x 1020 atoms/cm2). Magnification 2000x: (a) masked section; (b) exposed section The black scale bar in the lower right corner of the SEM image corresponds to 10

4.2 POLYURETHANE-BASED PAINTS FOR PAINTED SPACE COMPONENTS The chemical composition of the surface, in relative atomic percentage, for control (pristine) and Photosil™-treated paints were obtained from the XPS analysis. The results are shown in Table 7. Only survey scans for the elements of interest (C, O, Si and N) were conducted. The XPS data indicates that considerable changes have occurred in surface layers of the Photosil-treated paints. Carbon and nitrogen content, for the gray paint, for instance, have dropped to 40.29 at% and 0.46 at% respectively.

430 The oxygen content increased to 32.92 at% with the silicon making up 26.33 at% of the surface content. These comparisons between the control and treated paints show the reduction of the carbon content, and verify the incorporation of silicon and oxygen into the surface region of the original paint material. These results confirm the modification of the sub-surface region of the polyurethane-based paints. To demonstrate the detrimental effect of atomic oxygen exposure and the protective effectiveness of the Photosil™ process, one half of the painted coupons that were selected for AO testing (left-half) was treated by the Photosil™ process and the other half was left untreated. Photographs of gray (A276:Z306) and black painted coupons after a 6-hrs and a 12-hrs oxygen plasma exposures are shown in Figures 5 and 6 respectively. Visual examination of the treated coupons indicates that the untreated half of the gray paint specimen exhibited considerable surface erosion following the 6hrs plasma exposure (effective fluence ~ 2.4 x 1020 atoms/cm2) while the protected half was not eroded. The AO eroded the polyurethane resin portion of the paint, leaving behind the pigments. The untreated gray surface acquired a white powdery appearance (see Figure 5a). The loosely bounded fine white particles on the paint surface can be identified with the exposed titanium dioxide pigments. The surface of coupons with the untreated black paint turned to porous and powdery dark-gray surface (see Figure 5b). After the 12-hrs plasma exposure (effective fluence ~ 4.3 x 1020 atoms/cm2), the Photosil™-treated half remained almost intact with minor color fading. However, the untreated half was completely eroded, in particular the black (Z306) paint exposing the primer (yellow) (see Figure 6b). TABLE 7. XPS surface composition (at.%) of polyurethane-based paints

Specimen Gray Paint A276:Z306 Black Paint Z306 (flat) White Paint Black Paint Z302 (glossy)

XPS chemical content, at.% Treatment

C

O

Si

N

Control (untreated) Photosil™ Control (untreated) Photosil™ Control (untreated) Photosil™ Control (untreated)

75.07 40.29 70.38 41.32 75.78 40.77 72.33

20.47 32.92 27.77 32.23 19.47 31.15 23.45

1.44 26.33 0.35 26.17 0.56 27.45 1.62

3.02 0.46 1.50 0.28 4.20 0.63 2.6

Photosil™

30.64

42.85

26.15

0.36

The untreated black paint that contains carbon black pigment in a polyurethane binder was significantly eroded due to the reactive nature of both the carbon pigment and the polyurethane to AO. The gray (A276:Z306) paint, which contains polyurethane resin, titanium dioxide and carbon pigments, also was severely eroded. Both the resin and the carbon black were eroded away from the surface leaving behind the titanium dioxide pigment particles that is not affected by the AO. The degradation effects of AO on these paints are well known [1-9].

431

(a)

(a)

(b)

(b)

Photosil™-treated Untreated (Control) Photosil™-treated Untreated (Control) Figure 5. Painted test coupons exposed to 6-hr oxygen plasma, effective fluence ~ 2.4 x 1020 atoms/cm2: (a) gray (mixture of Z306 and A276) paint and (b) black (Z306) paint.

In addition to the plasma asher exposure, FAO beam testing was conducted as well. The SEM image in Figure 7 clearly demonstrates the severe degradation effect of the FAO irradiation onto the pristine black (Z306) paint. The polyurethane binder was eroded from the surface leaving exposed the loosely bounded carbon black pigment particles. In contrast, no binder erosion was detected in the SEM micrograph of the AO exposed Photosil™-treated paint (Figure 8). Similar AO stability was confirmed by SEM studies for the white (A276) and gray (A276:Z306) paints.

(a)

Photosil™-treated

Untreated (Control)

(b)

Photosil™-treated

Untreated (Control)

Figure 6. Painted test coupons exposed to 12-hr oxygen plasma, effective fluence ~ 4.3 x 1020 atoms/cm2: (a) gray (mixture of Z306 and A276) paint and (b) black (Z306) paint.

432 2000x

Figure 7. SEM micrograph of untreated black (Z306) paint after FAO exposure (effective fluence ~ 3.6 x 1020 atoms/cm2). Magnification 500x. The insert shows at higher magnification the morphology of the surface. The black scale bar in the lower right corner of the SEM image corresponds to 100 :m

2000x

Figure 8. SEM micrograph of Photosil™-treated black (Z306) paint after AO exposure (effective fluence ~ 3.6 x 1020 atoms/cm2). Magnification 500x. The insert shows at higher magnification the morphology of the surface. The black scale bar in the lower right corner of the SEM image corresponds to 100 :m TABLE 8. Erosion yields of polyurethane-based paints used on external space components.

Material Black paint Z306 White paint A276 Gray paint Z306/A276 Black paint Z302

Treatment

Erosion yield* (g/atom) Plasma asher Fast AO beam

Control (untreated) Photosil™ Control (untreated) Photosil™ Control (untreated) Photosil™ Control (untreated)

8.90E-24 0.60E-24 7.92E-24 0.25E-24 4.69E-24 0.41E-24 7.6E-24

0.62E-24 0.06E-24 0.51E-24 0.003E-24 1.00E-24 0.07E-24 0.52E-24

Photosil™

0.5E-24

0.05E-24

433 Table 8 summarizes the erosion yields for untreated and Photosil™-treated paints following a ground-based accelerated AO exposure that are averaged by mass loss and time of exposure. These relative values obtained as upper limits are used only as a means of comparison and, for the treated paints, are not the absolute erosion yields that can be compared to those in real LEO environment. Photosil™-treated paints show erosion yields of about 1-2 orders of magnitude lower than the erosion yields for respective untreated samples in both plasma asher and FAO beam testing. These results demonstrate the AO erosion resistance of Photosil™-treated paints and, in doing so, confirm the successful surface modification of the painted space components. The adhesion of all pristine and silylated paint samples to the space-grade Al alloy substrates and other space-grade metals that are used for the external space components has been tested following the ASTM method [12]. The above samples also have been tested for thermal stability by exposing them to 15 alternating cycles consisting of a one-minute immersion in liquid nitrogen and one minute exposure to 150 °C in a preheated oven. All pristine and silylated samples successfully passed both adhesion and thermal cycling tests. These results indicate, that Photosil™ surface modification treatment does not affect the adhesion and thermal cycling durability of the treated organic-based paints. It should be noted that a set of six materials that have shown best performance in the ground-based testing, together with a set of six specimens of similar materials, treated by another patented surface modification process Implantox™ [13] were selected and silylated in this program and have been provided for flight exposure to LEO space environment to the two-stage “Materials on International Space Station Experiment” – MISSE-1 and MISSE-2 [14]. First set of six samples is now being exposed in space for over a year as part of MISSE-1, and is going to be brought back in 2003 for analysis. A second set of samples is planned to be exposed in LEO for 3 years as part of the MISSE-2 experiment.

5.0 Conclusions The Photosil™ process has been successfully used in surface treatments of space-related thin polymer films and thermal control materials, a number of spacecraft component surfaces painted with polyurethane-based paints and polyamide fiber-based lacing tapes. Using the Photosil™ surface modification technique, the relative reaction yields of all treated components and materials were reduced by, at least, 1-2 orders of magnitude. The Photosil™ technology has proven to be a practical solution to a dramatic reduction in the AO erosion of these polymer-based materials. In addition, the Photosil™ process has opened the door to the potential use of polymer-based materials, such as organicbased paints and others that were previously considered as unacceptable in the AO environment.

434 6.0 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Reddy, R. “Review: Effects of low earth orbit atomic oxygen on spacecraft materials”, J. of Mat. Sci., 30, 281-307, (1995). Silverman, E.M. Space Environmental Effects on Spacecraft: LEO Materials Selection Guide, NASA CR 4661, part 1-2, (1995). J.I. Kleiman, Z.A. Iskanderova , Y.I. Gudimenko, W.D. Morison, and R.C. Tennyson, “Polymers and Composites in the Low Earth Orbit Space Environment: Interaction and Protection”, Canadian Aeronautic and Space Journal, 45, No.2, (June 1999) 148-160 Tennyson, R.C. “Atomic Oxygen Effects on Polymer-Based Materials”, Can. J. Phys., 69, 11901208, (1991). Banks, B.A. “Atomic Oxygen”, LDEF Materials Data Analysis Workshop, NASA CP 10046, 191217, (1990). Whitaker, A.F., Kamenetzky, R.R., Finckenor, M.M. and Norwood, J.K., in Proceedings of the Second LDEF Post-retrieval Symposium, edited by A.S. Levine, NASA Conference Publication 3194 (3), 1125, (1993). Golden, J.L., in Proceedings of the Second LDEF Post-retrieval Symposium, edited by A.S. Levine, NASA Conference Publication 3194 (3), 1099, (1993). Kleiman, J.I., Gudimenko, Y., Iskanderova, Z. Tennyson, R.C., and Morrison, W.D. “Modification of Thermal Control Paints by Photosil™ Technology”, Proc. of ICPMSE-4, Fourth International Space Conference, 23-24 April 1998, Toronto, Canada, 243-252. Tennyson R.C., in “Gossamer Spacecraft: Membrane and Inflatable Structures Technology for Space Applications”, Ed. Cr. Jenkins, AIAA, 2001. Gudimenko, Y., Iskanderova, Z., Kleiman, J.I., Cool, G., Morison, D. and Tennyson R. “Erosion Protection of Polymer Materials in Space”, Proc. 7th International Symposium “Materials in a Space Environment”, 16-20 June 1997, ENSAE-SUPAERO, Toulouse, France, 403-410. Gudimenko, Y., Kleiman, J.I., Iskanderova, Z.A., Tennyson, R.C. and Cool, G.R. US Patent #5,948,484, issued September 7, 1999. Annual Book of ASTM Standards, “Standard Test Method for Measuring Adhesion by Tape Test”, D-3359-95. Z. Iskanderova, J. Kleiman, Y. Godimenko, R.C. Tennyson, G.Cool, “Surface Modification of Polymers and Carbon-based Materials by Ion Implantation and Oxidative Conversion”, US Patent 5,683,757, issued Nov. 4, 1977. MISSE home page, http://www.Misse1.larc.nasa.gov/index.html, 2001.

ATOMIC OXYGEN RESISTANT, LOW ALPHA, ANTI-STATIC POLYIMIDES FOR POTENTIAL SPACE APPLICATIONS ARTHUR J. GAVRIN, SIU WA AU-YEUNG, BOB MOJAZZA Triton Systems Inc KENT A. WATSON ICASE JOSEPH G. SMITH, JR., AND JOHN W. CONNELL NASA Langley Research Center Abstract Ongoing efforts to develop organic polymers for potential applications on future ultralight weight spacecraft have focused on the incorporation of specific combinations of properties into polyimides. The target properties include space environmental durability, low solar absorptivity (α), and sufficient electrical conductivity to mitigate static charge build-up. Anew aromatic diamine, [2,4-bis(3-aminophenoxy)phenyl]diphenylphosphine oxide, was developed and used to prepare polyimides that exhibited improved atomic oxygen (AO) resistance and low α while maintaining good UV resistance. Electrical conductivity was achieved by the addition of carbon nanotubes at relatively low loading levels (<0.5 weight percent). An increase in solar absorptivity and electrical conductivity was observed with increased loading of the carbon nanotubes. In this study, three different types of carbon nanotubes were placed into an AO resistant, low α polyimide and the effects of carbon nanotube type and loading on electrical conductivity and α were investigated. The results indicate that a good balance of electrical conductivity, α and space environmental durability can be achieved by this approach.

1. 0 Introduction Large, deployable, ultra-light weight Gossamer spacecraft will require film-based polymeric materials that possess a unique combination of physical, mechanical, and optical properties [1]. The requirements are dependent on the intended mission and orbital environment of the spacecraft. Some of the desired properties may include atomic oxygen (AO) resistance [for spacecraft in low Earth orbit (LEO)], low solar absorptivity (low color), high thermal emissivity, resistance to UV and vacuum UV radiation, good mechanical properties, good processability and handling, and sufficient electrical conductivity for static charge dissipation. Typically the conductivity values necessary for static charge mitigation fall between 1x10-6 to 1x10-8 S/cm. Recently the preparation and characterization of [2,4-bis(3-

435

436 aminophenoxy)phenyl]diphenylphosphine oxide (3-APPO) and space environmentally durable polyimides therefrom was reported [2]. These polyimides contained a unique combination of properties including AO and UV radiation resistance, low solar absorptivity, high thermal emissivity, solubility in common organic solvents, good mechanical properties, and formed colorless to near colorless thin films (~25 µm thick). Preliminary results from work on imparting electrical conductivity into space environmentally durable polyimides has been reported [3,4]. Samples of polyimides containing 3-APPO were submitted to the Materials International Space Station Experiment (MISSE, http://misse1.larc.nasa.gov/) program for in-space flight-testing. The current research investigates the effect on AO resistance on a polyimide containing varying amount of phosphine oxide content. 2. 0 Experimental 2.1 STARTING MATERIALS 3-APPO was prepared as previously described [2] 1,3-Bis(3-aminophenoxy)benzene (1,3-APB, melting point (mp): 107 - 108.5 °C, Mitsui Chemicals America, Inc.) was used as received. 4,4'-Perfluoroisopropylidiene dianhydride (6FDA, mp: 241 - 243 °C, Hoechst Celanese Inc.) was purified via sublimation. N,N-Dimethylacetamide (DMAc) and N-methyl-2-pyrrolidinone (NMP) were obtained from commercial sources and used as received. 2.2 PREPARATION OF POLYMERS AND COPOLYMERS Phenylphosphine oxide (PPO) containing polyimides were prepared by reacting stoichiometric quantities of 3-APPO with 6FDA. Copolymers were prepared by the addition of 1,3-APB. The following procedure is representative for the preparation of all polymers. Into a 100 mL three neck round bottom flask equipped with a mechanical stirrer, nitrogen gas inlet, and drying tube filled with calcium sulfate was placed 3APPO (2.3768 g, 4.8258x10-3 mole) and NMP (7.0 mL). Once dissolved, 6FDA (2.1436 g, 4.8258x10-3 mole) was added along with additional NMP (8.0 mL) to give a solution with a concentration of 20% (w/v) solids. The light yellow solution was stirred overnight at room temperature under a nitrogen atmosphere. A 0.5% (w/v) solution in NMP exhibited an inherent viscosity (ηinh) of 0.39 dL/g at 25 °C. The remaining polymer solution was chemically imidized by the addition of 1.49 g of acetic anhydride and 1.15 g of pyridine. The solution was stirred at room temperature overnight. The polymer was precipitated in a blender containing water, filtered, washed with excess water and dried in a vacuum oven at 110 °C overnight to afford a white, fibrous material. Copolymers were prepared by using 25, 50 and 75 mole percent of 1,3-APB.

437 2.3 FILMS Thin films were cast from polyimide solutions in DMAc (20% solids). The polymers and copolymers were centrifuged, the decantate doctored onto clean, dry plate-glass and dried to a tack-free form in a low humidity chamber. The films then were treated in a forced air oven and staged to 250 °C for one hour. Thin film tensile properties were determined according to ASTM D882 using four specimens per test conditions. 3. 0 Results and Discussion 3.1 POLYMER SYNTHESIS Polyimides containing PPO units were prepared from 3-APPO and 6FDA in NMP under a nitrogen atmosphere. Typically the poly(amide acid) was prepared by stirring the reactants at room temperature for 24 hours, followed by chemical imidization/dehydration using a solution of acetic anhydride and pyridine. The resulting polymers were isolated in > 95% yield. Copolymers were prepared by the addition of 1,3-APB in various amounts as shown in Figure 1. 3.2 POLYMER CHARACTERIZATION Polymer and copolymer characterization is presented in Table 1. The ηinhs ranged from 0.39 to 0.73 dL/g. Viscosities of Polymer 1 were increased from 0.39 to 1.01 dL/g with improvements in monomer purity; however, data in the following tables was obtained on the lower viscosity sample. The glass transition temperatures (Tgs) ranged from 211 to 228 °C and decreased with increased percentages of 1,3-APB. All of the polymers exhibited solubility in DMAc, dimethylsulfoxide, dimethylformamide, diglyme, ethyl lactate and NMP in the imide form. The polymers exhibited temperatures of 5% weight loss ranging from 491 to 502 °C in nitrogen as determined by thermogravimetric analysis. TABLE 1. Polymer characterization

1 2

mole percent 1,3-APB 0 25

3

50

4

75

polymer

ηinh, dL/g

Tg, °C

0.39 0.41 0.46 0.73

228 221 216

211

temperature of 5% weight loss in nitrogen, °C 491

496 501 502

438 H2N

O

O Ph

O

NH2

+

O

F3C CF3

O

P O Ph

O

O

+

H2N

O

O

NH2

O

23 °C, N2 Polar aprotic solvent H O N

Ar

F3C

CF3

O H N OH

HO O

O

-H2O O Ar

F3C

CF3

N

N

O

O

Where Ar is

O

O

O Ph

and/or

O

O

P O Ph

Figure 1. Preparation of polymers

3.3 THIN FILM TENSILE PROPERTIES Thin films were cast from polyimide solutions and thermally treated in a forced air oven to remove residual solvent. All of the films prepared were nearly colorless. Room temperature thin film mechanical property data is presented in Table 2. The films exhibited room temperature tensile strengths and moduli from 71 to 94 MPa and 3.3 to 3.4 GPa, respectively. The elongations to break ranged from 3 to 4%. TABLE 2. Thin film tensile properties

polymer 1 2 3 4

mole percent 1,3-APB 0 25 50 75

tensile strength, Mpa 81 71 80 94

tensile modulus, GPa 3.4 3.3 3.3 3.3

elongation @ break, % 3 3 3 4

3.4 OPTICAL PROPERTIES Thin films were measured for optical transparency using UV/visible spectroscopy. Percent transmission at 500 nm (solar maximum) of the films ranged from 86 to 88%.

439 Typically a low colored film exhibits a low solar absorptivity (α) value. Thin films were measured for α and thermal emissivity (ε) and the results are shown in Table 3. The α value for all films was 0.07 and ε values ranged 0.77 to 0.85. TABLE 3. Optical properties of polymers polymer 1 2 3 4

percent 1,3-APB 0 25 50 75

%T @ 500 nm

α

ε

α/ε

film thickness, mil

88 86 87 88

0.07 0.07 0.07 0.07

0.81 0.83 0.85 0.77

0.09 0.08 0.08 0.09

1.5 1.5 1.5 1.0

3.5 ATOMIC OXYGEN EXPOSURE Ground-based AO exposure tests were performed on thin film samples at NASA Glenn Research Center, Cleveland Ohio. AO exposure tests were done according to ASTM E 2089-00 using a directed atomic oxygen facility with an electron cyclotron resonance plasma source. The ability of a material to undergo minimal changes in α, ε and other properties upon exposure to radiation in space is of significant importance. The trend with regards to AO resistance clearly follows the amount of 1,3-APB used in the copolymer preparation. Polymer 1, with 0 mole % of 1,3-APB, exhibited the highest AO resistance as evidenced by the lowest total mass loss per area. Polymer 4, containing 75 mole % of 1,3-APB, exhibited the lowest AO resistance. After exposure to AO fluences of ~1.8 x 1021 atoms/cm [2], polymer 4 was nearly completely degraded and no further measurements could be obtained. Kapton® was used as a standard to determine AO fluences. Other phosphine oxide containing polyimides have exhibited non-linear erosion rates after longer duration AO exposure [4,10]. The preliminary results observed with the polyimide tested herein are consistent with previous results [2]. After exposure to AO fluences of ~1.8 x 1021 atoms/cm2, α values for all samples increased from 0.07 to 0.10 and ε values decreased slightly. No appreciable trends can be observed with regards to varying the amount of 1,3-APB in the polyimide backbone and the effect on α and ε values. All films were visibly hazy after AO exposure and % transmissions at 500 nm dropped to below 1%. 3.6 COMPARATIVE AO EXPOSURE STUDIES For comparison purposes, AO exposure data obtained from two other facilities is presented in Tables 4 and 5. Table 4 presents data obtained from NASA Marshall atomic oxygen beam facility on a polymer structurally similar to Polymer 1 only that it was prepared from oxydiphthalic dianhydride instead of 6FDA [2]. Alpha, ε, and mass loss data is presented in Table 4. Alpha values increase slightly with initial exposure

440 but no further changes were evident with continued exposure. Changes in ε with continued AO exposure were insignificant. TABLE 4. AO studies performed at NASA Marshall Space Flight Center

Exposure

α

ε

α/ε

mass loss, x 10-3 g/cm2

None

0.19

0.74

0.26

0

1.71 x 1020 AO/cm2

0.22

0.73

0.30

1.1

2.00 x 1020 AO/cm2

0.22

0.74

0.30

0.8

2.13 x 10 AO/cm2

0.23

0.74

0.31

1.4

20

Table 5 lists AO exposure data obtained from Physical Sciences, Inc. on the same polymer used to obtain data in Table 4. Percent transmission at 600 nm is presented to show changes in optical properties with continued AO exposure. TABLE 5. AO studies performed at Physical Sciences, Inc.

Exposure 19

2

3.79 x 10 AO/cm 1.07 x 1020 AO/cm2 2.69 x 1020 AO/cm2

total mass loss, x 10-3 g/cm2

% T @ 600 nm

0.05 0.06 0.13

89 88 79

3.7 CONDUCTIVITY OF POLYIMIDE/NANOCOMPOSITE FILMS Based on previous studies [3,4], single wall carbon nanotubes (SWCNT) can be incorporated into the AO resistant polyimides presented in this paper to incorporate conductivity sufficient for static charge mitigation. Results indicated that weight loadings of 0.1 to 0.2% SWCNTs are sufficient to incorporate enough conductivity for static charge mitigation without significantly decreasing the optical properties of the films. Preparing SWCNT/polyimide nanocomposites from the polyimides presented herein is currently being investigated. 4. 0 Summary Polyimides prepared from 3-APPO, 6FDA and various amounts of 1,3-APB were characterized for mechanical and optical properties and AO resistance. Copolyimides were prepared using 25, 50 and 75 mole % of 1,3-APB. Increasing the amount of 1,3APB decreased the AO resistance of the polyimides. Polymer 4 containing 75 mole % of 1,3-APB was nearly completely eroded after exposures to AO fluences of ~1.8 x 1021 atoms/cm2. The data suggests that the amount of 3-APPO can be lowered by 25 or possibly even 50 mole % (and thus reducing the cost of the material) without

441 detrimental changes in AO resistance. Optical properties of the copolymers were largely unchanged with regards to varying amounts of 1,3-APB. Different ground-based AO exposure studies from three different testing centers (NASA Glenn, NASA Marshall and PSI, Inc.) were also presented for comparison purposes only. Differences in mass loss for similar AO fluences were observed for nearly identical materials in these tests. Samples that receive 1.5 year exposure to AO in LEO aboard the International Space Station will be retrieved and the results will be compared with ground-based AO exposure studies. 5. 0 Acknowledgements The authors thank Sharon Miller, Bruce Banks (NASA Glenn Research Center) and Deborah Waters (QSS, Inc.) for AO testing of the copolymers. The authors also thank PSI, Inc. and NASA Marshall Space Flight Center for additional AO exposure data. 6. 0 References 1.

Jenkins, C. H. M. “Gossamer Spacecraft: Membrane and Inflatable Structures”, Technology for Space Applications, Volume 191, American Institute of Aeronautics and Astronautics, 2001.

2.

Watson, K. A.; Palmieri, F. L.; Connell, J. W. Macromolecules, 2002, in press.

3.

Watson, K. A.; Smith, J. G. Jr.; Connell, J. W: Science for the Advancement of Materials and Process Engineering Technical Conference Series 2001, 33, 1551.

4.

Park, C.; Watson, K. A.; Ounaies, Z.; Smith, J. G. Jr.; Connell, J. W. "Polyimide/SWCNT

5.

Smith, C. D.; Grubbs, H.; Webster, H. F.; Gungor, A.; Wightman, J. P.; McGrath, J. E. High

6.

Smith, J. G. Jr.; Connell, J. W.; Hergenrother, P. M. Polymer (1994), 35, 2834.

7.

Connell, J. W.; Smith, J. G. Jr.; Hergenrother, P. M. Polymer (1995), 36, 5.

Composite Films for Potential Spacecraft Applications", NanoSpace 2001 Conf. Proceedings, 2001. Perform. Polym. (1991), 3 (4), 211.

8.

Connell, J. W.; Smith, J. G. Jr.; Hedrick, J. L.; Polymer (1995), 36, 13.

9.

Connell, J. W.; Smith, J. G. Jr.; Kalil, C. G.; Siochi, E. J. Polym. Adv. Technol. (1998), 9, 11.

10. Connell, J. W.; Smith, J. G. Jr.; Hergenrother, P. M. J. Fire Sci. (1993), 11, 137. 11. Connell, J. W.; Watson, K. A. High Perform. Polym. (2001), 13, 23.

This page intentionally left blank

DEPOSITION AND CHARACTERISTICS OF ATOMIC OXYGEN PROTECTIVE COATINGS USING PLASMA POLYMERIZED HMDSO

JINGYI WANG, YUNFEI WANG, XIANKAI ZHOU, ZEGENG JIN AND ZHIZHAN YU Lanzhou Institute of Physics, Chinese Academy of Space Technology, Lanzhou 730000 (China)

Abstract Plasma

polymerized

films

can

be

deposited

on

many

materials

from

hexamethyldisiloxane (HMDSO)/oxygen (O2) mixtures. The device for plasma polymerization is described. Different HMDSO/ O2 ratios in the das mixture provide varying degree of corrosion and erosion protection. The composition and structure of the deposited films were analyzed. The protective coatings on Kapton and polyester have been evaluated by testing in atomic oxygen beam equipment and in thermal energy plasma system. The HMDSO/O2 ratio in mixture has a significant effect on the composition and characteristics of polymerization films. Tests proved that HMDSO polymerization films deposited with appropriate HMDSO/ O2 ratio (3/1) in mixture can protect efficiently substrates from atomic oxygen attack, without changing significantly the substrates optical properties. If there is no sufficient oxygen amount in gas mixture, polymerized films have a high rate of degradation due to attack by atomic oxygen. If oxygen is in excess, the polymerized films become rough, and can not sufficiently protect the substrates.

1. 0 Introduction Atomic oxygen is the principal constituent of at LEO environmental conditions. It can oxidize the materials on spacecraft surface, change the properties of the materials, and

443

444 finally affect the performances and lifetime of spacecraft. In order to withstand the adverse effects of atomic oxygen, protective coatings had been prepared by various methods [1,2,3]. Coatings can be deposited onto many kinds of substrates by plasma polymerization [4,5]. The coatings of plasma polymerization HMDSO can protect effectively substrates from atomic oxygen errosion, and did not change the original properties of substrates. The homogeneous coatings could be deposited continually by plasma polymerization on larger-area substrates, so the method could be used for depositing atomic oxygen protective coatings onto solar array blanket, large-area thermal-control materials, etc. The facility for plasma polymerization was introduced briefly in this paper, the variation of gas composition and construction of polymerization coatings with respect to HMDSO/O2 ratios in mixture was analyzed. Finally the results of atomic oxygen experiments was described, together with the characteristics of the coatings, deposited with different ratio of gas components.

2. 0 Device of Plasma Polymerization Polymerized coatings were deposited in a homemade reactor, shown in Fig.1. It is composed of a chamber (500 mm diameter, 500 mm high), a pair of stainless steel parallel electrodes (300 mm diameter) and the space between two electrodes is adjustable in the range 40~150mm. In the chamber there is an r.f. generator (13.56MHz) providing a power in the range 0~1000W, discharge gas inlets, a vacuum system, and a temperature controller is placed under the lower grounded electrode. The substrates are put on lower grounded electrode. The gasses for the discharge are introduced into the chamber through mass flux controllers, disturbed in the cone-shaped upper electrode and then get into discharge chamber through apertures on the upper electrode, so the gases are well distributed over whole discharge chamber. Prior to the discharging, a vacuum of the order of 10-4 Pa was achived in the chamber. The working pressure was then held by flux of discharge gases and valve. The polymerization monomer is hexamethyldisiloxane (HMDSO).

445

Figure 1.Schematic of plasma reactor system for polymerization

3. 0 The Composition and Chemical Structure Analysis of Deposition Coatings The coatings have been deposited on substrates such as silica, Kapton, polyester, carbon fiber-epoxy composite, etc. The elemental composition and chemical structure of the coatings were analyzed by a PHI-5702 XPS/AES. Table 1 lists the elements in deposited films with respect to HMDSO/O2 ratios in discharge mixture, the other discharge parameters keeping the same. It is shown in Table 1 that the amount of C in the deposited films decreases with increasing O2 ratios in the mixture, and the amount of O and Si in the films increase. The XPS spectra for deposition films are shown in Fig.2. The curves indicate that contents of Si and O in films increase intensely with the increase of the ratio of O2 in the das mixture. Table 2 lists functional groups in deposited films on sample 2 (25% O2 in the mixture) and sample 3 (50% O2 in mixture). Besides a small quantity of -CHx-Si-O-Si-CHx- bonds, there are mostly SiOx and -Si-O-Si- bonds in sample 2. Sample3 was entirely composed of SiOx-like and -Si-O-Si- bonds. The AFM graphs of sample 2 and sample 3 are shown in Fig.3. If there is 25% O2 ratio in the mixture, the films are smooth and dense. If 50% O2 ratios in the mixture, they become rough and uneven on the surface of films. Figure 4 represents the depth distribution of Si, O, and C in polymerization films in the sample 2. It is shown that there are more oxygen atoms and less carbon on surface, carbon atoms increase with depth.

446 TABLE 1. Atomic ratios in coatings as a function of their volume percent of O2 in the mixture

No.

Volume percent of O2 in

Atomic ratio in coatings (%)

the mixture

Si

O

C

1

0%

12.2

16.9

70.8

2

25%

16.1

23.9

59.9

3

50%

28.7

41.1

20.3

4

75%

34.3

42.2

23.5

Figure 2. PS results of the elements Si2p and O1s spectra of samples

TABLE 2.

No.

2

3

Binding of Si2P/ eV

103. 6 102.7 101.7 103.6 102.8

Functional groups in sample 2 and sample 3

Compound

FWHM

SiOx Si O Si

1.40 n

CH2 Si O Si nCH2

SiOx Si O Si

14.5

1.65

66.7

1.51

18.8

1.49 n

Peak area ratio(%)

1.44

64.5 35.5

447

Figure 3. AFM graphs of sample 2(left) and sample 3 (right)

Atomic concentration£¬AC%

70

Si2p O1s C1s

60 50 40 30 20 10 0 0

20

40

60

80

100

ș

Figure 4. Atoms in-depth distribution of sample 2

4. 0 Experiments of Atomic Oxygen Durability of Deposition Films The oxidation durability of HMDSO polymerization coatings deposited onto polyester films was determined by thermal energy oxygen plasma system. The results of experiments are present in Table 3. The films deposited from pure HMDSO had the similar degradation characteristics as pristine polyester. Sample 3 deposited with 50% O2 ratios in the mixture didn’t show any visible appearance change, but indicated some mass loss. The transparent appearance of sample 4 deposited with 75% O2 ratios in the mixture was converted to a translucent milky after oxygen plasma exposure, and it had the

448 greatest mass change in these samples. Only sample 2, that was deposited at 25% O2 ratio in the mixture, not only had no any noticeable appearance change, but also had a neglidjable mass loss, such as just 0.1mg. TABLE 3. Experimental results of atomic oxygen durability of the deposition films

Mass 1 (mg)

Mass 2 (mg)

(Before Experiment)

(After Experiment)

1

43.5

2

ǻm(mg)

(mg/cm2h)

41.3

2.2

0.5

45.3

45.2

0.1

0.023

3

44.7

44.4

0.3

0.068

4

44.0

40.8

3.2

0.73

40.1

38.1

2.0

0.45

No.

Control sample

The atomic oxygen durability of HMDSO polymerization coatings deposited onto aluminized Kapton was tested by our atomic oxygen equipment. The atomic oxygen equipment has been presented in details in the literature [6]. The bare aluminized Kapton sample was attacked by atomic oxygen at an integral flux 3×1020O Atoms/cm2, its mass loss was about 1.4mg/cm2. It lost its gloss and became diffuse, specular reflection on its surface almost disappeared, and solar spectral reflectivity changed to 0.610 from former 0.693. After depositing HMDSO polymerization coatings onto aluminzed Kapton, solar spectral reflectivity of the sample was 0.687. After atomic oxygen exposure (at an integral flux 3×1020O Atoms/cm2), the mass loss of the sample was only 0.03mg/cm2, it had no appearance change, and solar spectral reflectivity of the sample became 0.680. After atomic oxygen exposure, the total reflectance of bare Kapton changed slightly, but diffuse reflectivity increased largely. Especially at shorter wavelength, the curve of diffusion reflection overlaps almost with that of total reflection, as proved that specular reflection at shorter wavelength almost disappeared. The total reflectance of aluminized Kapton had no obviously change after HMDSO polymerization films were deposited on it, and diffusion reflectance decreased slightly. After atomic oxygen exposure, total reflectance of the sample had no obviously change yet, and diffuse reflectance further decreased.

449 5. 0 Analysis and Discussion From the forenamed XPS analysis, there are many Si-O-Si bonds and SiOx in HMDSO polymerization coatings, Si-O-Si bonds in the surface can be converted to SiOx by atomic oxygen attack [7]. SiOx has quit strong atomic oxygen durability. SiOx can reduce the arriving flux of atomic oxygen at the coated substrates efficiently, and can decrease the speed to that of heat motion of atomic oxygen that reached substrates. Therefore, SiOx can protect substrates efficiently from atomic oxygen erosion. If O2 ratio in discharge mixture is appropriate, polymerization films are even and dense, and can protect substrates efficiently. If there is no sufficient oxygen in the mixture, that polymerization films have a high rate of degradation due to attack by atomic oxygen. If oxygen is excessive, polymerization films are rough, and cannot protect substrates. The specular reflectance of HMDSO/O2 polymerization films increased a little bit due to more SiOx in it. After reaction with atomic oxygen, Si-O-Si bonds in the surface of the material was converted into SiOx. The more SiOx were formed, the more slick and dense the surface became, this may be the reason that the specular reflectivity of Kapton coated with HMDSO/O2 polymerization films increase further after the atomic oxygen exposure.

6. 0 Conclusions Under certain conditions, during the preparation of HMDSO plasma polymerization coating, the rate of O2 in the mixtures has prominent influence on the composition and structure of polymerization coatings. If there is no oxygen in the mixture, there are too much C and H in polymerization coatings. If oxygen is excessive, polymerization coatings are rough, and emerge granular-like structures on the surface. When oxygen in the mixture is provided in an appropriate ratio, polymerization coatings are smooth and dense, and contain more Si-O-Si bonds, forming finally the SiOx strucrure. In additional, the coatings are colorless and transparent, and they do not change obviously the intrinsic optical properties of substrates. After the reaction with the atomic oxygen, Si-O-Si bonds in the surface of the polymerization films were further converted to SiOx. The contacts

450 between atomic oxygen and substrate were held back efficiently, so SiOx can protect substrates from atomic oxygen attack. After reacting with atomic oxygen, diffuse reflection of aluminzed Kapton coated with HMDSO/O2 polymerization films decreases. The reason may be that Si-O-Si bonds were converted into SiOx, then more SiOx were formed, which make the surface of the material more dense and slick.

7. 0 References 1.

Banks B. A., Mirtich M., Rutledge S. K., and Nabra H. K, Protection of Solar Array Blankets from Attack by Low Earth Orbital Atomic Oxygen, 18-th Photovoltaic Specialists Conference, 1985, p.381-385.

2.

Gulino D. A., Solar Dynamic Conference Durability in Atomic Oxygen and Micrometeoroid Environment, Journal of Spacecraft and Rockets, (1988), 25, 3, 244-249.

3.

Nabra H. K. and Rutledge H. K., Technologies for Protection of Space Station Power System Surface in Atomic Oxygen Environment, Proceedings of the Twenty-Fifth Space Congress, Poster 5-1.

4.

Janka J. and Sodomka L., Selective Gas Sorption of Plasma-polymerized Organosiloxane Thin Films, Thin Solid Films, (1992), 216, 235-238.

5.

Biederman H., Polymer Films Prepared by Plasma Polymerization and their Potential Applications, Vacuum, (1987), 137, 367-373.

6.

Wang J., Yu Z. et al, Coaxial Atomic Oxygen Simulation Facility and its Properties, Chinese Space Science and Technology, (1998), 18, 5, 50-55.

7.

Rutledge S. K, Cooper J. M., and Olle R. M., The Effect of Atomic Oxygen on Polysiloxane-Polyimide for Spacecraft Application in Low Earth Orbit, Forth Annual Workshop on Space Operations Application and Research (SOAR-90), 1991, p.755-762.

DEVELOPMENT OF PROTECTIVE AND PASSIVE THERMAL CONTROL COATINGS ON CARBON-BASED COMPOSITE MATERIALS FOR APPLICATION IN SPACE M. FRANCKE, B. FRITSCHE, A. MOC, R.B. HEIMANN Freiberg University of Mining and Technology, Department of Mineralogy, Brennhausgasse 14, D-09596 Freiberg, Germany Z. ISKANDEROVA, J.I. KLEIMAN Integrity Testing Laboratory Inc. 80 Esna Park Drive, Markham, Ontario, L3R 2R7, Canada 1. 0 Introduction Improvement of the environmental stability of materials subjected to harsh outer space conditions is an important objective of modern materials science research. In particular, many attempts were made to develop protective coating designs for low Earth orbit (LEO)-environment sensitive polymeric and/or other carbon-based composites [1, 2]. Among these designs thermally sprayed coatings are being investigated as a possible option (for example Ref. [3]). Research was carried out in a project sponsored by the German Federal Ministry of Research and Technology (BMBF) under the auspices of the Bilateral German-Canadian Agreement on Cooperation in Science and Technology [4]. The objectives were to investigate the potential of thermally sprayed thin metallic coatings to protect various polymeric and composite compounds considered as space construction materials against the erosive chemical action of fast atomic oxygen (FAO), hard ultraviolet radiation, and charged particle radiation, as well as the mechanical erosion by impact of micrometeorites and man-made space debris, and deterioration by frequent thermal cycling. 2. 0 Materials The metallic coating materials consisted of gas-atomized spheroidal 88Al-12Si powder of two grain sizes –90+45 µm (Metco 52C-NS) and –45 µm (AMDRY 355), supplied by Sulzer Metco (Deutschland) GmbH. Three different substrate materials were investigated: Kapton, poly(N,N’(P,P’ oxydiphenylene)pyromellit)imide (DuPont de Nemours Cie.), CFC (graphite fiberreinforced epoxy) composite material, and C-C (graphite fiber-reinforced graphite) composite material. Since Kapton was only available as thin foil several layers of foil

451

452 were bonded together to yield a laminate sheet of 1.5 mm thickness (courtesy: Peter Wong, P.Leo & Co.(B.C.) Ltd., Richmond, B.C., Canada). From the substrate sheet materials coupons of size 50x20x1.5mm (Kapton) and 30x30x3mm (CFC) and 25x25x1.5mm (C-C), respectively were cut, roughened by grit blasting with corundum, cleaned in an ultrasonic bath in Tickopur at 60 °C, rinsed in distilled water and dried. Before spraying the coupons were kept in a dust-proof environment. 3. 0 Preparation of Coatings The metallic powders were applied to the substrate coupons by atmospheric plasma spraying (APS) using an F4 MB plasmatron (Sulzer Metco) in conjunction with a PT M-1000 (Plasma-Technik, Wohlen, Switzerland) control unit. Some test coatings were also deposited by acetylene/oxygen combustion flame spraying (FS) (courtesy: Department of Iron and Steel Technology, Freiberg University of Mining and Technology) and electric wire-arc spraying (courtesy: Eutectic Canada Inc., Mississauga, Ontario, Canada). 3.1 KAPTON SUBSTRATES For APS and FS spraying statistical experimental designs were applied to optimize the plasma parameters. For APS coatings a fractional factorial 25-1 design of resolution V was selected whose five variables were the hydrogen gas flow rate (x1) at levels of 1.5 and 2.0 SLPM, the powder feed rate (x2) at 15 and 25% of maximum feed rate, the spraying distance (x3) at 120 and 135 mm, the powder grain sizes (x4) at –90+45µm and –45 µm and the powder carrier gas (argon) flow rate (x5) at 3.0 and 3.5 SLPM. The plasma gas (argon) flow rate, the rotation speed of the powder feeder screen, and the plasma current were kept constant at, respectively 30 SLPM, 40% of maximum, and 300 A corresponding to a plasma power of 15 kW. For FS coatings a second order design of type 2231 was used whose variables were the spraying distance (x1) at three levels of 27, 29 and 31 mm, the compressed air pressure (x2) at 2.0 and 4.0 bar, and the number of passes (x3) at 1 and 2. The powder grain size used for FS coatings was – 90+45µm.

Figure 1. Scanning electron microscopy analysis of a Kapton TM sample. Magnification 350x.

453 3.2 CFC SUBSTRATES Because of the thermal sensitivity of the epoxy matrix plasma power was kept at a relatively low level, and varied between 9 and 13 kW. The number of traverses was varied between 1 and 2, the traverse speed between 3 and 10 m/min, the argon gas flow rate between 30 and 40 SLPM, and the rotation of the powder feeder screen between 20 and 30% of maximum. All other parameters were held constant at the following levels: hydrogen gas flow rate at 1.5 SLPM, powder gas flow rate at 3 SLPM argon, rotation of powder feeder stirrer at 40 % of maximum, and the spray distance at 120 mm.

Figure 2. Optical microscopy analysis of a CFC substrate sample. The scale bar at the bottom of the picture is 1 mm.

3.3 C-C SUBSTRATES The materials consist of a graphite matrix and a graphite fibre reinforcement. It was chosen because it combines many of the desirable high temperature properties of conventional carbon and graphite Hence it is claimed to be useful at temperatures as high as 2800 °C. Its important properties are high strength, high modulus, and low creep. Furthermore, the C-C composite is a material with low sensitivity to thermal shock.

Figure 3. Scanning electron microscopy analysis of a C-C substrate sample. Magnification 100x.

454 This is caused by its high thermal conductivity and low coefficient of thermal expansion, coupled with high bending strength. High fracture toughness and pseudoplasticity are characteristic as well. The C-C composite substrate is produced as a 2Dlaminate of plain-weave cloth fabricated fibres, so-called twill. The cloths are stacked up to provide a 0°/90° composite of approximately 1.5 mm thickness. The C-C composite plate was cut into coupons of size 25 x 25 mm. In addition, small coupons with a dimension of approximately 5 x 20 mm for the cross-sections were cut. Six plasma spray parameters were varied: plasma power (9-13 kW), argon (plasma) gas flow rate (30-40 SLPM), hydrogen (auxiliary) gas flow rate (0.5-1.5 SLPM), spray distance (100-120 mm), rotation speed of powder feeder screen (7-20% of maximum), number of traverses (1-3). Three parameters were kept constant: plasma gas (argon) flow rate at 3 SLPM, rotation speed of powder feeder stirrer at 20% of maximum, and traverse speed at 6 m/min. For details cp. Ref. [6]. 4. 0 Characterization of Coatings 4.1 OPTICAL AND SPECTROSCOPIC MEASUREMENTS Surfaces and cross-sections of as-sprayed coatings were investigated by conventional light optical microscopy and scanning electron microscopy in conjunction with energy dispersive X-ray spectroscopy (EDAX). Spectral absorptance α(λ) of as-sprayed and AO exposed coatings were determined according to ASTM E903-96 with a Beckman DK-2A spectrometer between 2450 and 250 nm wavelengths. The solar absorptance αs(λ) was calculated by integrating the α-values over the wavelength range and weighted against the solar standard spectral irradiance E(λ) taken from ASTM E490-73a. Infrared reflectances were measured with a Gier-Dunkle infrared reflectometer DB 100 against standard gold and black body surfaces. The thermal emittance ε was calculated from the measured reflectances using ε = 1-R*, where R* is the arithmetic mean value of eight measurements made to account for inhomogeneities produced by the rough coating surfaces. X-ray photoelectron spectra (XPS) of as-sprayed and AO treated surfaces were recorded with a Leybold MAX 200 spectrometer with monochromatized AlKα (1486.6 eV) and non-monochromatized MgKα (1253.6 eV) sources operated at 15 kV/30 mA and 15 kV/20 mA, respectively. Typical FWHM values for the Ag 3d5/2 peak were 0.67 eV (Al source) and 0.88 eV (Mg source), respectively. The energy range of the spectrometer was calibrated to place Cu 2p3/2 and Cu 3p3/2 at 932.67 and 75.16 eV, respectively. 4.2 MECHANICAL AND THERMAL TESTS Surface roughness of the coatings were determined according to DIN 4768 with a diamond stylus system (Surftest 500, Mitutoyo, Japan). Coating adhesion was estimated by a scribing test according to ASTM B571-84. Thermal cycling was performed by

455 repeatedly cooling the samples with the coatings attached in liquid nitrogen and subsequent heating to 125 to 140 °C in an oven. The samples were immersed in liquid nitrogen for one minute and then transferred to the preheated oven for one minute. This cycle was repeated 15 times. During the experiment the temperature of the oven decreased on average by 10 °C. 4.3 ATOMIC OXYGEN BEAM FACILITY Atomic oxygen (AO) was produced in the University of Toronto’s Institute for Aerospace Studies (UTIAS) SURFATRON facility using a plasma torch. The plasma was generated by propagation of an electromagnetic surface wave along a plasma column containing helium (98.5%) and oxygen (1.5%) [5]. The plasma power was set to 200 W and the total gas flow rate was 3 SLPM. Under these conditions the degree of oxygen dissociation was about 64%, and the fraction of exited oxygen was < 1.2%. The maximum oxygen atom energy was about 2.2 eV, only 50% of the collisional energy of 4.5 to 5 eV of oxygen atoms impinging at a spacecraft’s surface in orbit. The as-sprayed

Figure 4. UTIAS FAO beam facility [5]

coatings were exposed to an atomic oxygen beam for 6 hours to yield an AO fluence of 1.8 1020 atoms/cm2. 4.4 IC-RADIO FREQUENCY PLASMA FACILITY An inductively coupled radio frequency cold plasma generator (“plasma asher”) was operated at a frequency of 13.56 MHz and a power of 200 Watts. Figure 5 shows details of the plasma generator (top) and the sample chamber (bottom).

456

Figure 5. Schematic drawing of the „plasma asher“ used in the testing.

After placing up to eight samples in the sample holder, low pressure (1.3 to 290 Pa) was applied for up to 42 hours to outgas the samples (ASTM E 595-93). The edges of the samples were protected with thin aluminium foil. The oxygen plasma environment consisted of atomic oxygen, ionised oxygen molecules, and small amounts of ozone. Another possible component is singlet oxygen, the electronic state of molecular oxygen. After testing the specific mass losses (mass loss rate per exposed unit area) were determined according to the equation for the effective flux Re in (cm-2 s-1)

Re =

∆m Aexp⋅ texp⋅ ρKapton⋅ ES,Kapton

with ∆m mass loss, Aexp exposed area in cm², texp exposure time in seconds, ρKapton specific mass of Kapton = 1.42 g/cm³, ES,Kapton erosion yield of Kapton = 3 10-24 cm³/atom [7]. 5. 0 Results and Discussion In general, atmospheric plasma spraying is suitable of depositing sufficiently dense and adherent 88Al-12Si coatings on polymeric and composite material surfaces. For each substrate material used several combinations of plasma spray parameters were applied in a statistical design fashion and the resulting coatings were studied with regards to their microstructure, surface roughness, thermo-optical properties, resistance to erosion by fast atomic oxygen (FAO), adhesion to the substrate, and thermal cycling to simulate environmental conditions likely to occur in a LEO (low Earth orbit).

457

(a)

(b)

Figure 6. Scanning electron microscopy analysis of the microstructure of the produced coating: (a) typical splats, magnification 350x; (b) coating sprayed with large powder, magnification 750x.

The microstructure of the coatings is characterized by layers of splats produced by deformation on impact of the molten metallic material. However, larger powder particles do not melt completely to their core and hence will be incorporated into the coatings as semi-solid spherical particles resulting in the development of coating porosity. The coatings thickness varies strongly among the samples and also within the samples. The spray parameters that affect mainly the coating thickness are powder grain size, number of traverses, and powder feed rate. The degree of melting in the plasma jet is essentially controlled by the hydrogen gas flow rate and the powder feed rate. Hence degree of melting is also an important parameter that controls coating thickness.

Figure 7. Cross-sectional scanning electron microscopy analysis of a coating ( seen as a bright contrast layer). Magnification 200x.

The surface roughness size and the surface treatment of the substrates. The degree of melting of the powder particles may affect the surface roughness too. The metallic coating changes the optical properties of the composite materials from flat absorber to solar absorber [8]. The ratio of the solar absorptance and thermal emittance (αs/ε) is mainly affected by the powder grain size and number of traverses,

458 dependence of α and ε for solar absorber α = 1.5 ε, α / ε >1 dependence of α and ε for a flat absorber α = ε, α / ε = 1

1,0

solar absorptance α

0,8

0,6

0,4

0,2

0,0 0,0

0,2

0,4

0,6

0,8

thermal emittance ε

Figure 8: Correlation between solar absorptance α and thermal emittance ε. From ref. [6].

and appear also to be dependent on the surface roughness. In particular, Į/İ ratios were found to increase with increasing precursor powder grain size, presumably through increased diffuse light scattering at the rougher surfaces. Also, coatings deposited on Kapton showed a slightly higher Į/İ ratio than those deposited on CFC substrates. Atomic oxygen (AO) does not erode the surface of the coatings. Photoelectron spectroscopic (XPS) measurements before and after AO as well as IC radio frequency plasma exposure reveal the outermost extremely thin surface layer to consist of strongly adhering oxidized film that act as protective barrier for AO diffusion and limit further

Figure 9 Scanning electron microscopy analysis of AO plasma exposed coating sample. Magnification 1,000x.

459 AO attack. The most important findings concern the chemical reactions occurring between the coating material and the reactive atomic oxygen beam. The topmost coating layer interacted with atomic oxygen during FAO exposure by forming an aluminum silicate solid solution [9] of the type Al2[Al2+2xSi2-2xO10-x] in which the stoichiometric factor x varies between 0 (sillimanite Al[AlSiO5]), ¼ (3:2 mullite 3Al2O32SiO2 ) and 2/5 (2:1 mullite 2Al2O3SiO2). Since these materials have a high scratch and erosive hardness they may effectively counteract erosive abrasion by impinging micrometeorites in outer space.

4.5

x 10

C-COMPOSIT-13/2

4

2.8

361

x 10

C-COMPOSIT-13/2

4

361

4 2.6

Intensity (arb. units)

3

2.5

2

1.5

1

0.5 -85

5

x 10

-80

-75

-70 Binding Energy (eV)

-65

-60

2.2

2

1.8

1.6 -115

-55

C-COMPOSIT-13

4

3.8

351

3.6

4

3.4

Intensity (arb. units)

2.5 2 1.5 1 0.5 -85

-80

-75

-70 Binding Energy (eV)

-65

-60

-55

-100 -105 Binding Energy (eV)

-95

-90

C-COMPOSIT-13

4

(d) Si2p peak: after AO exposure with a binding energy of 100.5 eV

3.2 Intensity (arb. units)

(c) Al2p peak: after AO exposure with a binding energy of 74.3 eV

3

x 10

-110

351

4.5

3.5

(b) Si2p peak: before AO exposure with a binding energy of 99.1 eV

2.4 Intensity (arb. units)

(a) Al2p peak: before AO exposure with a binding energy of 74.2 eV

3.5

3 2.8

2.6 2.4

2.2 2 1.8 -115

-110

-100 -105 Binding Energy (eV)

-95

-90

Figure 10. XPS analysis of the coating. Single Al2p and Si2p peaks before (a, b) and after(c, d) AO plasma exposure

The coated samples appear to resist well spalling due to thermal cycling. No micro-cracking and loss of adhesion were detected, i.e. the strength of the coating as well as the adhesion to the substrate are higher than the thermal stresses introduced during thermal cycling. The adhesion of the coating to the substrate varies strongly among the samples and appear to be affected by the degree of overheating of the Kapton and the carbon composite materials. The CFC in particular, but also the C-C composite are sensitive to the high heat input during the plasma spray process. Volatile compounds released by decomposition of the epoxy resin but also, to a lesser degree, by residual binders in the C-C composite material tend to accumulate at the interface between the coating and the substrate and disrupt the mechanical anchorage of the coating. Hence spray parameters that control the heat flow to the substrate such as hydrogen gas flow rate, plasma power, number of traverses, and powder feed rate are determining factors in the adhesion as well. Pre-treatment of the substrate surface by polishing reduces the original surface roughness and consequently results in reduced coating adhesion.

460 In conclusion, it can be expected that on exposure to atomic oxygen, as experienced during long-term LEO space activities, a continuous protective layer is being formed that adheres well to the substrate and thus limits oxygen diffusion into the polymeric or carbon-based composite material. Since reactions with atomic oxygen will result in materials deterioration through carbon backbone scission and side-chain abstraction as well as destruction by hard UV radiation, protection from space environment hazards by the thin metallic layers will enhance the service lifetime of polymeric composite space construction materials. A further beneficial effect of a continuous protective layer may be the formation of a hard erosion-resistant barrier against the abrasive action of micrometeorites and, maybe more important, man-made small space debris such as alumina particles from solid-propellant motors [10]. 6. 0 References 1.

Kleiman, J. and Iskanderova, Z., Technological aspects of protection of polymers and carbonbased materials in space. Proc. 8th Intern. Symp. ‘Materials in a Space Environment’, Arcachon, France, June 5-9, 2000.

2.

Reddy, M.R. Effect of low earth orbit atomic oxygen on spacecraft materials, J. Mater. Sci. 30 (1995), 281-307.

3.

Heimann, R.B., Fritsche, B., Kleiman, J.I., Tkachenko, A. and Tennyson, R.C. Thermally-sprayed protective coatings for polymer-based materials in space environment. Proc. 8th Intern. Symp. ‘Materials in a Space Environment’, Arcachon, France, June 5-9, 2000.

4.

WTZ project ‘Thermally sprayed protective coatings for polymeric materials in a space environment’, International Bureau of BMBF, Project CAN 98/022, June 1, 1999 to December 31, 2001.

5.

Tennyson, R.C. Atomic oxygen effects on polymer-based materials, Can. J. Physics 69 (1991), 1190-1208.

6.

Francke, M. Deposition and evaluation of thin plasma-sprayed 88Al/12Si coatings on C-C composite substrates for space application. Unpublished Master thesis, Freiberg University of Mining and Technology, in preparation.

7.

Packirisamy, S., Schwam, D. and Litt, M.H. Atomic oxygen resistant coatings for low Earth orbit space structures. J. Mater. Sci. 30 (1995) 308-320.

8.

Silvermann, E.M., Space Environmental Effects on Spacecraft: LEO Materials Selection

9.

Guide. NASA Contractor Report 4661. Part 2. August 1995, p. 10-6 ff.

10.

Handbook of X-ray Photoelectron Spectroscopy: A reference book on standard spectra. PerkinElmer Corp., Physical Electronics Div., Eden Prairie, 1992.

11.

Meyer, R.X., Elements of Space Technology, Academic Press, 1999, p.280.

STRUCTURE AND COMPOSITION OF NON-METALLIC SOLAR ARRAY MATERIALS RETRIEVED AFTER LONG-TERM EXPOSURE OVERBOARD THE “MIR” ORBITAL SPACE STATION V. A. LETIN, L. S. GATSENKO SPRE “KVANT”, 3-d Mytischinskaya Str., 129626 Moscow, Russia Phone: 7(095) 287-97-42, Fax: 7(095)287-18-71 e-mail: [email protected] I.S. DEEV, E. A. BAKINA, Ⱥ.V. MALENKOV All-Russian Institute of Aviation Materials 17, Radio Str., Moscow, 107005, Russia Phone: +7 095 261-86-77, Fax: +7 095 267-86-09, E-mail: [email protected] E. F. NIKISHIN M.V. Khrunitchev State Space Scientific Production Center 18, Novosavodskaya Str., Moscow, 121087, Russia Phone: +7 095 142-50-36, Fax: +7 095 956-24-41

Abstract Extended studies of the structure and elemental composition of non-metallic materials were carried out, and the major structural changes were defined after long-term exposure (10,5 years) of solar arrays (SA) overboard the “MIR” orbital station (OS). The studies were conducted by scanning electron microscopy (SEM) and X-ray spectral dispersion microanalysis (SDMA). The results (SEM-photo and SDMA data) had shown the dinamics of the changes of structural characteristics and elemental composition of materials for 10,5 years of SA operation on OS “MIR”. It was detected that multylayer films of condensed particles having various forms and sizes appeared at front and rear surfaces of protective glasses on solar cells (SC). For the first time it was detected that traces of conic form remained in the film of condensed deposit of microparticles flown in the same direction. 1. 0 Introduction Extended studies and analysis of SA that has undergone operation in space is extremely important. The results of the studies of selected identical SA before and after prolonged exposure may give a unique opportunity to reveal new information about influence of

461

462 space environment on the SA materials and components, to determine the degree and to understand the mechanisms of their properties degradation. Visual control and instrumental analysis of electrical, optical, mechanical and thermo-physical parameters were conducted with SA retrieved to the Earth after 10,5 years of exposure in LEO environment [1-3]. Photoelectrical properties degradation of SA can be caused by many chemical and physical processes at different structural scale: thermal- and radiation-induced chemical decomposition of non-metallic materials, mainly silicone polymer materials due to prolonged interaction with LEO environment and due to the thermal processes inside the SA and components; surface contamination by products of polymer material decomposition and the space station’s own atmosphere, erosion and impact effects of meteoroids, debris and atomic oxygen which deteriorate the optical performance of SC. Elaborated studies of the effects of these physical and chemical processes had shown that significant degradation of the SA materials and components, along with power and current losses, is caused by a local thermal process in SC blocks, known as hot-spotting, which is caused by strip shadowing of SC by space station structural elements. These processes led to thermal degradation of some of SC. However the majority of SC (∼90%) weren’t exposed to partial shadowing and the degradation of their parameters happened under the effect of natural space environment. Recently the interest to microparticles and debris with sizes ∼1 µm [4] has grown, so it was decided to study carefully the surface of protective covers of SC and some SC non-metallic materials from this particular point. 2. 0 Research Objects • • • • • •

Surface (front and rear) of coverglass (K-208 type). Surface (front and rear) of white acrylic enamel AK-573 on metal substrate. Surface of filaments made from aluminum boro-silicate glass fibers impregnated with adhesive BF-4. Surface (front and rear) of filaments aramide-T impregnated with adhesive BF-4. Surface of cotton filament. Surface of the silicone adhesive before and after its operation in space

3. 0 Analitical Devices and Methodology Micro-structural analysis was conducted on the scanning electron microscope JSM-840 made by “JEOL” (Japan) and optical microscope MBS-10 (USSR) at various magnifications (from 200 to 5 000 times). Before the structure is studied in the scanning electronic microscope the surface of samples is coated with metal by the method of ion deposition of gold layer with the thickness of 15-20 nm in a vacuum unit JFC-1100 FINE COAT (“JEOL”).

463 To make automatic analysis of computer picture of the structure of the analyzed samples a special methodology was applied. The chemical composition of surface layer of materials after space flight testing was defined by the method of X-ray microanalysis on micro-analyzer JXA-840 made by “JEOL” (Japan) with powerdispersing detector of “LINK” (England), that provides the ability to analyze chemical elements from Na (N11) to U (N92). Lighter chemical elements, such as H, C, O, N, which are always present in polymers, have not been analyzed. Before the analysis a thin layer of carbon ( ∼ 100-200 Å) was deposited on the surface by thermal deposition in a vacuum unit JEE-4X/5B manufactured by “JEOL”. 4. 0 Study of Coverglass Surface Analysis of K-208 type protective glass had show that the surface structure of initial glass is uniform and flat, without well-defined features. After 10,5 years of space exposure the deposit of condensed particles is formed at rear and front surfaces. These particles have both rounded (diameter to 1 µm) and anisotropic (with length to 10 µm) shape. At different areas of the surface of the glass, particles with various size and packing density are located - from single particles, arranged randomly, to compact packed, with special isotropy and orientation, and they form multi-layer films (Figures 1-4).

Figure 1. SEM photo of coverglass K-208 rear surface after 10,5 years exposure: a – section without condensed deposit (below) and with it (after top), x200; b – spalling of butt-end of multi-layer condensed deposit, x2000.

Stratified film structure with thickness of about 3 µm appeared, and the figures indicate on some spallation of condensed deposit from the surface of the glass. Thickness and irregularity of formation of a condensed deposit at the rear surface of coverglass is considerably more that at the front one. Besides this, the flow of hard micro-particles that arrived at the film with slipping traces of length to 50 µm in similar direction is well visible at front and rear surfaces in the thin film of condensed deposit. Apparently, these particles are either micro-debris [4] or particles of the spacecraft own atmosphere.

464 Consequently the film of condensed deposit is present at the surface of protective glass on which the traces of the flow of flying particles are fixed and they are detected using SEM, that can be used for registration and detection of parameters (concentration, size, form, composition, speed, etc) of hard micro-particles arriving at the film from its own atmosphere of OS and/or space. It is also useful for the studies of the process of these particles interaction and the decay process of non-stable particles. SDMA of the surface of initial glass indicated (Figure 5) that silicon dioxide (the basic part) and Na and K (falls far short) are present in its composition.

Figure 2. SEM photo of coverglass K-208 front surface after 10,5 years exposure, x2000: a – section with isotropy arrangement of rounded particles; b – section with anisotropy particles; c – section with space arranged oriented particles; d – section with compactly packed, flight-oriented particles.

465

Figure 3. SEM photo of coverglass K-208 front surface after 10,5 years exposure: a – rounded forms in condensed deposit, 2000; b – the same, x5000; c – section with sparse arrangemented oriented sliding particles, x5000; d – section with compact packed up oriented particles, x5000.

Figure 4. SEM photo of coverglass K-208 front surface after 10,5 years exposure: a, b – surface section of condensed deposit with sliding particles, x2000; c – the same, x5000; d – butt-end and condensed deposit surface with sliding particles, x5000.

466

Figure 5. SDMA of coverglass K-208 before exposure.

After 10,5 years of operation in space a significant increase of silicon dioxide was defined by analysis of front and rear surfaces of glass (Figures 6, 7).

Figure 6. SDMA of coverglass K-208 front surface after 10,5 years exposure.

467

Figure 7. SDMA of coverglass K-208 rear surface after 10,5 years exposure. Taking into account the presented above microanalysis data, it can be concluded that the condensed deposit and the flow of microparticles involved in it are of the same origin and they consist mostly of silicon dioxide. The mechanism of the formation of siliceous condensed deposit hasn’t been studied extensively. The most probable reason of its appearing is the effect of atomic oxygen and ultraviolet radiation on part of surface with silicone adhesive SKTNF located on SA joints of SC and destructured under the effect of space environment to formation of molecular silicon dioxide. Later silicon dioxide is atomized and condensed on adjacent surfaces (front and rear) of array as a result of erosion and temperature processes. 5. 0 Studies of Enamel AK-573 It is well known that the surface of white acrylic enamel AK-573 on a metal substrate (alloy AMG-6) before operation in space is practically flawless (free of large pores and splits) and it’s rather embossed with protruding spherical particles of pigment-filler which sizes reach to 10 µm (Figure 8).

468

Figure 8. SEM photo of white acrylic enamel AK-573 surface on metallic substrate before exposure: x200; b – x1000; c – x2000; d – x5000.

a–

Polymeric film covering the surface of particles filler and among these particles has the uniform and well-defined structure. According to data of SDMA (Figure 9) the composition of the initial surface, except for light cells, consists of Zn, somewhat below silicon and Al, S, Cl to a small extent. After 10,5 years of operation in space a rear side of coating’s surface has become less embossed (Figure 10).

469

Figure 9. SDMA of white acrylic enamel AK-573 surface before exposure.

Figure 10. SEM-photo of white acrylic enamel AK-573 rear surface on metallic substrate after 10,5 years exposure: a – x200; b – x1000; c – x2000; d – x5000.

470 as the quantity of protruding of filler tend to decrease and their surface is cleaned of polymeric film through erosion effect of space. It will cause to decrease of Zn content combination at the surface of enamel and to a sharp increase of siliceous combinations (Figure 11).

Figure 11. SDMA of white acrylic enamel AK-573 rear surface on metallic substrate after 10,5 years exposure. After 10,5 years of being in space there are micro-splits at the enamel on the front surface (Figure 12), and its relief is smoothed practically throughout all the surface as the particles of filler were removed from the surface. Besides this, the polymeric film of enamel between the particles was etched and partially eroded. As a result of erosion process Zn pigment is removed from the surface layer practically throughout that lead to a sharp decrease of Zn content, whereas the formation of condensed siliceous deposit at the surface of enamel in space lead to the increase of siliceous compositions (Figure 13).

471

Figure 12. SEM photo of white acrylic enamel AK-573 front surface on metallic substrate after 10,5 years exposure: a – x200; b – x1000; c – x2000; d – x5000.

Figure 13. SDMA of white acrylic enamel AK-573 front surface on metallic substrate after 10,5 years of exposure.

472 6. 0 Research on Other Non - Metallic Materials It was shown that the exposed surface of filaments from aluminum boron-silicate impregnated with glue BF-4 was modified to a greater extent than the surface enclosed by SC after 10,5 years of being in space. In both cases glass filaments remain covered by glue but the netting of micro-splits appeared at the exposed surface, and this netting is absent at the surface of filaments being under SC. After 10,5 years of being in space the filaments aramid -T impregnated with glue BF-4 also keeps its structure of front and rear sides, except for a number of microsplits that appeared in the glue film which is more visible at front surface. After 10,5 years of being in space the structure of cotton filament was strongly destroyed owing to its capability to become fragile that causes a lot of breaks of cotton filaments. The differences of the silicone adhesive structure before and after its operation in space were found during the study process of the surface of silicone adhesive film SKTNF located on the boundary between SC (Figure 14).

Figure 14. SEM photo of silicone adhesive SKTNF film surface on the boundary among solar cells before (a) and after (b, c, d) 10,5 years exposure, x2000. In the starting state (Figure 14a) the surface structure of silicone adhesive film is uniform, doesn’t contain any embedded particles and micro-defects (splits, pores etc). After the operation of the array in space essential changes can be seen at the silicone

473 adhesive surface (Figure 14 b, c, d) owing to the destruction - a thin (thickness from 0,01 to 0,3 µm) and fragile layer of degraded silicone adhesive with micro-splits and rounded shape embedded particles (diameter from 0,1 to 5 µm) appeared in some places. Probably appearing of the embedded particles is related to the falling of microsplits from substance, and their penetration into a superficial silicone adhesive layer. These micro-splits easily penetrate and fix themselv in a superficial silicone adhesive layer due to its high elasticity. Referring to SDMA-data, only siliceous combinations are present in a superficial silicone adhesive layer, and the quantity of these silicon combinations was significantly increased after 10,5 years of operation in space. The formulation of the phase of embedded particles of substance micro-particles detected isn’t practically different from the formulation of main silicone adhesive material.

7. 0 Conclusions Extended studies of the structure and cell’s formulation were carried out, and their main structural changes and modifications were defined after 10,5 years of operation outboard of OS “MIR”. It was detected that multi-layer film of condensed particles having various forms and shape (round, anisotropical) and size (from 0,1 to 10 µm) appeared at front and rear surfaces of K-208 type protective glass of SC. At different sections of the glass surface these particles form both isotropical and oriented structures with different packing density. For the first time it was detected that traces of conic form remained in the film of condensed deposit by flow of micro-particles flown in one direction. The length of slipping traces of the micro-particles in the condensed film ranges up to 50 µm. Cell formulation of these particles was similar to condensed deposit formulation and largely consists of silicon atoms. A possible mechanism of condensed deposit formation was proposed and discussed. It was proposed to use the film of condensed deposit at the surface of cover glass of SC for registration and detection of hard micro-meteoritic and other particles of a microscale range in space. As a result of research on modification in the structure and composition of front and rear surfaces of acrylic enamel AK-573 on a metal substrate after 10,5 years of operation in space it became clear that the cover remains safe (intact), but erosion of a thin superficial layer of polymeric matrix and particles of powdery filler in it that lead to the enamel topography has become less relief, and the Zn pigment composition has decreased. There wasn’t detected any macro- and micro-defects on the cover. There was made out that structure of front and rear surface of glass filaments and arimid-T impregnated by glue BF-4 was kept without changes after theirs exposition in space, but micro-splits nettings were formed in glue film at front surface of filaments.

474 The structure of cotton filaments was destroyed after exposition in space and it led to significant decrease of its strength. It was shown that the structure of silicone adhesive SKTNF located on the boundary between SC was drastically changed after the exposition in space. Thin superficial layer becomes fragile and, besides this, micro-splits and embedded particles appears in it. 8. 0 Acknowledgements The authors express their gratitude to T. I. Bogdan and V. V. Guseva for their assistance in translation and editing. 9. 0 References

1. Chertok B. E. et al., (1999) Report on after light studies of a returned to the Earth fragment (wing) of the Mountable Solar Array 17KS5810-0, RSC “Energia”, SPRE “Kvant”, Korolev, Moscow Region. 2. Letin V. A. et al., Basic results of post-flight researches of solar array returned to the Earth from the orbital space station “MIR” after 10,5 years of operation in: Proc. of the 28th IEEE Photovoltaic specialists conference, Alaska, Ancourage, Sept. 15-22, 2000. 3. Letin V. A. and Babayevsky P.G., (2001) Comparative analysis of solar array fragments degradation at different structure scales during and after long-term exposure overboard the “MIR” space station, High Performance Polymers, 13. 453-460. 4. Baritean M. et al., Erosion of spacecraft materials by microparticles and atomic oxygen: implications for the environment, Proceedings of the 8th International Symposium on "Materials in a space environment", Proceedings of the 5th International Cjnference on "Protection of Materials and Structures from the LEO Space Environment". Arcachon, France, 5-9 June 2000.

MICRO- AND MACROTRIBOLOGICAL PROPERTIES OF SOLID LUBRICANTS IN 5 ELECTRONVOLTS ATOMIC OXYGEN EXPOSURES M. TAGAWA, M. MUROMOTO, H. KINOSHITA AND N. OHMAE Department of Mechanical Engineering, Kobe University, Rokko-dai 1-1, Nada, Kobe 657-8501 Japan K. MATSUMOTO AND M. SUZUKI National Aerospace Laboratory, Jindaiji-Higashimachi 7-44-1, Chofu, Tokyo 182-8522, Japan Abstract Macro-scale and micro-scale tribological properties of MoS2 and diamond- like carbon (DLC) films were evaluated in hyperthermal atomic oxygen beam exposure conditions. A friction coefficient at the beginning of sliding increased with increasing atomic oxygen fluence. Appearance of the atomic oxygen-induced high friction on the in-situ macroscopic tribological properties depends on the exposure and tribological conditions such as atomic oxygen fluence, pin material, repetition rate and normal load. A DLC film that contains 45 % of hydrogen showed a low friction in vacuum, but atomic oxygen exposure led to a high friction. An elongation of the wear life of DLC film was suggested by atomic oxygen exposures.

1. 0 Introduction Molybdenum disulfide (MoS2) is one of the most widely used lubricants in space systems. Its low friction coefficient (typically 0.02-0.05) and durability in vacuum are suitable for space applications. MoS2 lubrication film has successfully applied in many LEO satellites. Application of MoS2 lubrication film to the Exposure Facility of Japanese Experimental Module of the international space station, however, requires high reliability and much longer life of MoS2 lubricants even after a long exposure to space environment. In order to evaluate the effect of 5 eV atomic oxygen (AO) bombardment to MoS2-based lubricants, which is one of the most important space environment factors, some research groups have studied the tribological properties of the AO-exposed MoS2 lubricants. However, such research results were often contradictory. Results of the previous studies are summarized in Table 1 [1-6]. As shown in this table, low energy atomic oxygen seems to be less effective to the degradation in tribological properties of MoS2 sputtered film. Therefore, it is important to simulate translational energy of the impinging AO in order to access the tribological properties of MoS2 lubricant in LEO environment. Also, most of the tribological tests regarding AO effect have been carried out after the AO exposure was completed; i.e., ex-situ testing. Since the effect of AO

475

476 bombardment to the MoS2 surface is restricted only in the surface region of MoS2 film (6-10 nm in depth), in-situ exposure tests may provide different results from the ex-situ TABLE 1. Previous results on tribological properties of AO-exposed MoS2 lubricants

Researcher Martin (Ref: 1, 2)

MoS2 Sample RF sputtered on Sapphire, SUS440C pin, 0.3 N

AO exposure Hyperthermal (LANL) 1.5 eV, 1024 AO/cm2

Test Ex-situ

Results Oxide:MoO2,MoO3,SO2, Crystallite independence, Intial high friction on 60 °C, but no effect on 200 °C deposited samples

Arita (Ref: 3)

RF sputtered on SUS440C, Ti pin, 10 N

Hyperthermal (PSI) 5 eV, 1020 AO/cm2

Ex-situ

Initial high fricion (20-30%) Oxide: 6 nm, Crystallite dependence,

Wei (Ref: 4)

Ion beam assisted sputter film, Al2O3 ball

Thermal (RF asher) 0.02 eV, flux unknown

In-situ

Stable friction and wear properies

Matsumoto (Ref: 5)

RF sputtered on SUS440C, SUS440C ball, 10 N

Hyperthermal (PSI) & LEO 5 eV, 1020 AO/cm2

Ex-situ

Initial low frction, Elongation of wear life for 160 °C deposited sample by AO-exposure

Tagawa (Ref: 6)

RF sputtered on SUS440C, Ti pin, 2 N

Hyperthermal (Kobe Univ.) 5 eV, 1018 AO/cm2

In-situ

Oxide: MoO3, SO, Initial high friction with AO exposure as low as 1017 AO/cm2.

exposure tests, especially in wear life. This is because further oxidation at the wear track surface after sliding does not occur in the case of ex-situ tests. In-situ tribological testing result was reported by Wei et al.[4], however, they used a thermal O-atom source; i.e., the impinging energy of 5 eV was neglected. We have studied in-situ tribological tests in 5 eV atomic oxygen beam, and found that friction coefficient became greater under AO beam exposures. In this study, we are reporting some ground-based simulation results on the AO effects on the tribological properties to the MoS2 sputter-deposited films. The AO effect on diamond like carbon (DLC) lubricants, which is a candidate of lubricants in next generation, was also evaluated. Tribological properties in small loads (micro-tribology), which related not only to a small satellite technology, but also to a fundamental aspect of macro-tribology, also are investigated.

477 2. 0 Experiments The MoS2 specimens used in this study were sputter-deposited MoS2 Films. The sputtered films were prepared by the radio frequency (RF) sputtering technique at the National Aerospace Laboratory, Japan. Sputtering conditions of the RF sputtered specimens are reported elsewhere.5 The samples prepared by the same procedure have also been aboard STS-85 and correlated with the ground-based experiment reported here. The hydrogenated diamond like carbon (DLC) samples, which contain 45 at% of hydrogen were prepared by chemical vapor deposition (CVD) on the SUS440C substrates. TABLE 2. Tribological test conditions

Pin Disk Load Sampling frequency Track length Sliding speed AO energy flux repetition rate

SUS440C ball (1/16 inch) Sputtered MoS2, DLC on SUS440C 3.6 N 5 Hz 8 mm 2.0 mm/s 4.5 eV 2.3x1014AO/cm2/s 1 Hz

The AO source used in this study was based upon the laser-induced detonation phenomenon and originally developed by the Physical Sciences Incorporation (PSI) [7]. The AO source was attached to the AO testing facility developed in our laboratory [6]. The AO beam was monitored by the time-of-flight measurement system consisting of a quadrupole mass spectrometer and a multi channel scalar. The mean energy and full width at half maximum of the hyperthermal AO was measured to be 4.7 eV and 5.5 eV, respectively. Typical translational energy spectrum of AO beam measured was reported elsewhere [8]. The AO fraction in the beam was approximately 45 % and balance molecular oxygen. The AO flux of the beam was measured by an Ag-coated quartz crystal microbalance (QCM) with an accommodation coefficient of 1. Since the reaction of AO with Ag is a non-linear reaction as reported previously [9], only the initial reaction, which gave a linear mass gain, was used to calculate AO flux. A typical AO flux at the tribological test position (47 cm from the nozzle) was 2.3 x 1014 atoms/cm2/s. The macroscopic tribological properties were evaluated in the AO source chamber. An ultrahigh vacuum (UHV)-friction tester used in this study was especially designed for this facility. This UHV-friction tester is based on the conventional pin-on-disk layout. The testing conditions are summarized in Table 2. The microscopic tribological properties were studied by the microtribometer developed in our laboratory [10]. The principle of operation is similar to that of a commercially available atomic

478 force microscope (AFM). A sharp tungsten tip, which was finished by electropolishing, was attached to the aluminum cantilever. The deflection and torsion of the cantilever were detected by the laser diode and the photo diode array. The load and the friction force were simultaneously calculated using the output of the photo diode array. The microtribometer was built on a vacuum flange so that the measurement of microtribological properties can be made without breaking the vacuum.

3. 0 Results and Discussion

3.1 MACROSCOPIC TRIBOLOGICAL PROPERTIES OF MOS2 In the previous study, we have proposed the method for the tribological testing to analyze the effect of AO [6]. One of the points is that the initial high friction of AO exposed MoS2 lubricants was due not only to the AO effect, but to the resistance of wear track formation at the beginning of friction. It has been demonstrated that the wear track formation before the AO exposure can eliminate the later effect. In this case, a pin cannot be separated form the wear track during the test. 0.15

AFTER AO EXPOSURE

Friction Force (N)

0.10 0.05 0.00 -0.05 -0.10 -0.15

0

20

40

60

80

100

Data Number

Figure 1. The initial high friction in the tribological test after AO exposure with 1.0 x 1018 2

Therefore, all of the macroscopic experiments in this study were carried out in the AO source chamber of the facility under AO beam exposures. Indeed, no initial high friction was observed without AO beam exposure. In contrast, after AO exposures, a high friction was observed at the first friction (Figure 1). Thus, the initial high fiction in Figure 1 is attributed to the AO-induced effect. From the steady-state friction during AO exposure, the friction coefficient was calculated to be 0.005 which is identical to that without AO exposure. In the continuous friction condition examined, AO does not affect the friction force. This is due to the removal of the surface oxide layer by

479 friction which was evident in XPS and AES measurements. This phenomenon was also observed when AO beam exposure was started during the friction test as indicated in Figure 2. In Figure 2 AO exposure was started at the 70th cycle of the sliding. However, no increase in friction coefficient was measured even the AO beam irradiated 0.06

0.04

Friction coefficient

Friction coefficient

0.05 0.03

0.02

0.01

0.04 0.03 0.02 0.01

0.00

0

50

100

150

200

250

0.00 0.0

300

0.5

1.0

1.5 18

AO fluence (x10

Cycle number

Figure 2. Effect of AO exposure on the continuous sliding test of MoS2 after steady-state friction was obtained.

2.0

2.5

3.0

2

atoms/cm )

Figure 3. The fluence dependence on the initial friction force of the MoS2 sputtered film.

the surface (AO flux: 3.0 x 1014atoms/cm2/s). This is not a contradictory results reporting that the friction coefficient increased by AO beam exposure [6]. The difference came form the testing conditions. The friction coefficient increased by repetition rate of 0.05 Hz with a Ti-6Al-4V pin and 2.0 N of normal load, however, it did not increase by repetition rate of 0.15 Hz with a SUS440C ball and 3.6 N of normal load. The AO fluence between every slide is calculated to be 6 x 1015 atoms/cm2 for 0.05 Hz, and 2 x 1015 atoms/cm2 for 0.15 Hz. A larger fluence of AO between frictions as well as a low contact pressure and/or chemically active Ti-pin lead to an AO-induced 0.6

Friction coefficient

0.5 0.4 0.3

DLC(H:0%) DLC(H:45%)

0.2 0.1 0.0

0

10

20

30

40

50

Cycle number

Figure 4. Friction coefficient of DLCs in vacuum.

high friction. The fluence dependence on the initial friction force of the MoS2 sputtered film is shown in Figure 3. It was obvious that the initial friction force of MoS2 film increased with increasing AO fluence on the friction surface. AO fluences

480 in the order of 1015-1016 atoms/cm2 do not influence friction coefficient of MoS2. However, an AO fluences of 2.5 x 1018 atoms/cm2 leads to the friction coefficient more than 9 times larger than the steady-state friction. It is, therefore, concluded that the AO induced high friction requires the accumulation of AO at the friction track. 3.2 MACROSCOPIC TRIBOLOGICAL PROPERTIES OF DLC

0.16

0.7

0.14

0.6

Friction coefficient

Friction coefficient

DLC is a candidate of future lubricants usable in space. Its low friction coefficient both in air and in vacuum is suitable for space applications. However, the effects of hyperthermal AO on tribological properties of DLC have not been studied. Herein, we are reporting the effects of AO on the tribological properties of DLC. Figure 4 represents the friction coefficient of DLCs in vacuum (10-4 Pa). The DLC film in which no hydrogen was included in the film showed a high friction in vacuum,

0.12 0.10 0.08 0.06 0.04

0.4 0.3 0.2 pristine DLC under AO-exposure

0.1

0.02 0.00

0.5

0.0

0

20

40

60

80

100

120

140

0

100

1.5

1.5

1.0

1.0

0.5 0.0 -0.5

-1.5

(a) BEFORE EXPOSURE NORMAL LOAD: 23 µN 20

40

60

Time (s)

400

500

0.5 0.0 -0.5 -1.0 -1.5

0

300

Figure 6. Wear life of DLC under AO beam exposure (4.6 eV, 2.4 x 1014 atoms/cm2/s)

Friction force (µN)

Friction Force (µN)

Figure 5. The effect of AO beam exposure (4.6 eV, 2.4 x 1014 atoms/cm2/s) in the sliding test of DLC.

-1.0

200

Cycle number

Cycle number

80

100

(b) AFTER AO EXPOSURE NORMAL LOAD: 23 µN 0

20

40

60

80

100

Time (s)

Figure 7. The micro-tribological properties of the AO-exposed MoS2 films. (a): before exposure, and (b): after AO exposure, AO fluence: 3.2 x 1018 atoms/cm2.

whereas, DLC with 45 at.% of hydrogen showed a friction coefficient as low as 0.01 after 10 cycle of sliding. It was, therefore, demonstrated that DLC needs a high hydrogen content to be used as a lubricant in vacuum. The effect of AO exposure on

481 the DLC with 45 at.% hydrogen was evaluated in this study. The AO beam exposure (4.6 eV, 2.4 x 1014 atoms/cm2/s) was started during the sliding test. The results are indicated in Figure 5. The AO exposures were carried out in the cycle numbers of 38-68 and 93-118 which are shown by the hatched area in Figure 5. The AO exposure obviously induced a high friction 3 to 6 times larger than the normal value. However, the wear life was elongated more than 5 times by AO exposure as shown in Figure 6. Note that AO exposure was started at the 40th cycle. The reason of this phenomenon is not clear at this moment, but the polymerization of DLC by friction, which is considered to be a key for the tribological properties of a hydrogenated DLC in vacuum, may be affected by AO due to its high reactivity. However, a wear life of the DLC tested was insufficient for space applications. A longer wear life should be realized by optimizing either sample preparation or testing conditions. 3.3 MICROSCOPIC TRIBOLOGICAL PROPERTIES OF MoS2 The micro-tribological properties of the AO-exposed MoS2 film were preliminary evaluated by the ex-situ testing. The results are shown in Figure 7. Figure 7 (a) indicated the variation of the friction force of the pristine MoS2 specimen during the tests. Different from the macroscopic testing, the normal load was kept constant by the feedback control, so that a constant load of 23 µN was kept during the test. The result of AO-exposed MoS2 showed the steady-state friction of 0.25 µN (µ=0.011) from the beginning of the sliding, i.e., no initial high friction was obvious (Figure 7(b)). The AO fluence of 3.2 x 1018 atoms/cm2 oxidized the MoS2 surface as is reported previously.6 Therefore, it was suggested that the MoS2 surface was oxidized but the microtribological properties are hardly affected in this experimental condition. Micro-tribology is important not only as a future small satellite technology, but also as a tool for the fundamental study of the macroscopic tribology. Micro-tribological studies on the effect of AO exposure to space lubricants will provide deeper understanding of the mechanism of AO-induced degradation of tribological performances of lubricants used in space.

4. 0 Conclusions The macro-scale and micro-scale tribological properties of MoS2 and DLC films were evaluated in hyperthermal AO beam exposure conditions. The high friction at the beginning of sliding was induced by an accumulation of AO at the wear track on MoS2. Appearance of the AO effect in macroscopic tribological properties depends on the sliding and exposure conditions. The DLC film that contains 45 % of hydrogen showed a high friction under the AO exposure, but a longer wear life was measured. Microtribological properties of MoS2 were hardly affected by AO exposure conditions tested (friction coefficient of 10-2).

482 5. 0 Acknowledgments A part of this study was supported by the Hyogo Science and Technology Association.

6. 0 References 1. Martin, J. A., Cross, J. B., and Pope, L. E., 1989, MoS2 Interactions with 1.5 eV Atomic Oxygen, Mat. Res. Symp. Proc., 140, 271-276. 2. Cross, J. B., Martin, J. A., Pope, L. E., Koontz, S. L., 1990, Atomic oxygen-MoS2 chemical interactions, Surface and Coating Technology, 42, 41-48. 3. Arita, M., Yasuda, Y., Kishi, K. and Ohmae, N., 1991, Investigation of Tribological Characteristics of Solid Lubricants Exposed to Atomic Oxygen, Proceedings of STLE/ASME Tribology Conference, St. Louis, paper # 91-TC-3D-1. 4. Wei, R., Wilbur, P. J., Bughholz, B. W., Kustas F. M., 1995, In situ Tribological Evaluation of Greases and Solid Lubricants in a Simulated Atomic Oxygen Environment, Tribology Trans., 38, 950-958. 5. Matsumoto, K., Suzuki, M., Imagawa, K., Okada, Y., and Tagashira, M., 1998, Evaluation of tribological characteristics of sputtered MoS2 films exposed to LEO environment, Proc. 21st Int. Symp. Space Technology and Science, Omiya, paper #98-b-30. 6. Tagawa, M., Ikeda, J., Kinoshita, H., Umeno, M., and Ohmae, N., 2001, Effect of Atomic Oxygen Exposures on the Tribological Properties of Molybdenum Disulfide Lubricants, Protection of Space Materials from the Space Environment, Kleiman, J. I. and Tennyson R. C. Eds, Kluwer Academic Publishers, Dordrecht, 73-84. 7. Caledonia, G. E., Krech, R. H., and Green, B. D., 1987, High Flux Source of Energetic Oxygen Atoms for Material Degradation Studies, AIAA Journal., 25, 59-63. 8. Tagawa, M., Yokota, K., Ohmae, N., Kinoshita H., 2002, Effects of Ambient Air Exposure on Atomic Oxygen- Exposed Kapton-H Films, Journal of Spacecraft and Rockets, 39, 447-451. 9. Tagawa, M., Kinoshita, H., Umeno, M., Ohmae, N., 1996, Recent Development of the Laser Induced Breakdown Type Atomic Oxygen Source at Osaka University, Proc. 20th Int. Symp. Space Technology and Science, Gifu, paper #96-b-41. 10. Tagawa, M., Kinoshita, H., Umeno, M., Ohmae, N., Matsumoto, K., Suzuki, M., 1999, Ground-Based Tribological Testing of the Sputter-Deposited Molybdenum Disulfide Films in hyperthermal Atomic Oxygen Exposure, Proc. 8th European Space Mechanism and Tribology Symp., Toulouse, ESA SP-438, 291-296.

ZnSe COATINGS FOR SPACECRAFT ELECTROCHROMIC THERMAL CONTROL SURFACES LI YAN JOHN A. WOOLLAM Center for Microelectronic and Optical Materials Research, Department of Electrical Engineering, University of Nebraska-Lincoln, Lincoln, NE 68588-0511 EVA FRANKE Institut für Oberflächenmodifizierung Leipzig, Leipzig, Germany

Abstract We previously reported development of multilayer infrared-operating electrochromic thermal-control surface devices for spacecraft. One problem preventing use of such devices for spacecraft in low earth orbit (LEO) is the adverse effect of atomic oxygen on the exposed surface layer. In this paper we show that zinc selenide can be used as both an infrared transparent optical coating and an AO protective surface. Zinc selenide films of different thicknesses were exposed to an electron cyclotron resonance (ECR) generated oxygen plasma to simulate LEO. To permit realistic device performance simulations, ZnSe optical constants, before and after oxygen plasma exposure, were determined using variable angle spectroscopic ellipsometry from the vacuum ultraviolet at 146 nm through the middle infrared to 40 µm. Comparing the pre- and post- oxygen plasma exposure data, few changes were observed in the middle infrared region, while drastic changes were seen in the vacuum ultraviolet through visible to near infrared. This suggests that chemical changes upon plasma exposure are found mainly in a thin layer near the surface. As the proposed application is for infrared coating applications, ZnSe will indeed be useful for space infrared applications. Performance simulations of ZnSe coated infrared-operating electrochromic thermal-control surfaces confirm this conclusion. 1.0 Introduction Compound semiconductors have been heavily researched for applications in optoelectronics. Zinc selenide (ZnSe) is an especially well-developed material, as short-wavelength optical devices can be made with it. In addition, ZnSe is of interest as an infrared coating material due to its broad spectral band of high transmission from 600 nm to about 20 microns [1]. To obtain true optical constants of ZnSe, the effects on

483

484 measurements due to overlayers need to be accounted for, including both native oxide and surface roughness 2-9. Despite differences in their physical nature, the affects of either type overlayer on apparent dielectric function of the film are similar.7 In the present study, we assumed overlayer affects to be adequately represented by a roughness layer in the regression analysis of the optical data. With the high sensitivity of ellipsometry to surface overlayers and ability to “model away” these affects, the final as-determined optical constants are representative of ZnSe with minimal overlayers. As well, the effects of overlayers, including roughness due to AO exposure, on electrochromic device performance operation can be accounted for in modeling. We recently reported an infrared (IR)-active electrochromic thin film device that operates over the spectral range from 2 microns to 35 microns.10-12 This is potentially useful for thermal control of spacecraft operating at and just above room temperature (300 to 350 K). However, the optical properties of WO3 proposed for use in these electrochromic devices change significantly after exposure to AO. To protect infrared devices from AO erosion in LEO, a ZnSe infrared anti-reflection coating provides enhanced switching performance as well as AO protection. Here, we present optical constants of ZnSe at room temperature over the spectral range 146nm to 40 microns, before and after oxygen plasma exposure, using variable angle spectroscopic ellipsometry (VASE£), and use these results to simulate the performance of electrochromic thermal-control surface devices for use in LEO.13-14 2.0 Experimental ZnSe samples of 100nm, 200nm, and 300nm nominal thicknesses were deposited on 2inch-diameter silicon wafers, using electron-beam evaporation. X-ray diffraction data indicate a preferred [111] orientation for these polycrystalline films. Spectroscopic ellipsometry (SE) is a well-known surface sensitive, non-destructive optical technique widely used to determine film thickness and optical constants.13-15 Reflection ellipsometry measures change in polarization state of light upon reflection from a sample surface. The measurement is expressed as psi (Ψ) and delta (∆), which are related to the Fresnel reflection coefficients by:15 ρ ≡ tan(ψ )e i∆ = R p R s (1) where p- and s- correspond to directions parallel and perpendicular to the plane of incidence, respectively. In this work, measurements were performed over a wide spectral range, using three ellipsometers covering the spectral ranges from 146 nm to 285 nm, 190 nm to 1700 nm, and 1.25 to 40 microns, respectively. Thus, the entire spectral range from 146nm to 40 microns was covered. For details on the ECR chamber and oxygen plasma and calibration see Ref 16. 3.0 Optical Modeling A parametric optical constant model was developed by Herzinger and Johs for analysis of semiconductors, and is of great benefit for analyzing a large variety of optical data

485

486 The ZnSe films were found to get thinner by about 20% (decreased from ~98 nm to ~78 nm) after 20hr oxygen plasma exposure. Thickness non-uniformity was ~17%, a considerable increase from before irradiation, where the non-uniformity was negligible.

Index of refraction, n

4.0

3.0 2.5 2.0 1.5 1.0

Extinction Coefficient, k

Before exposure After 20hr exposure

(a)

3.5

0

2

2.5 2.0

4

6

Photon Energy (eV)

8

10

(b)

1.5

Before exposure After 20hr exposure

1.0 0.5 0.0 0

2

4 6 Photon Ener gy (eV)

8

10

Figure 1 Optical constants before and after 20 hrs oxygen plasma exposure. (a) n. (b) k.

To study chemical changes on the ZnSe sample surfaces, energy dispersive X-ray spectroscopy (EDX) was used to examine surfaces before and after oxygen plasma exposure. A tiny oxygen peak was detected after oxygen plasma exposure. However, we should have been able to see corresponding new peaks (predicted to be at ~500 cm-1 for ZnO, and ~350 cm-1 for ZnO2, for instance) in the IR24 but did not. One explanation is that oxygen atoms may have been (physically) trapped into ZnSe lattice structures,

487 forming O2. Presumably, few chemical bonds were formed between trapped oxygen atoms and the ZnSe lattice, no oxide-associated peaks were observed in the IR spectra. 5.0 Electrochromic Device Performance Simulations

Emittance = 1 - Reflectance

Theoretical device performance was evaluated based on previous experimental device results and the optical data on degraded ZnSe layers presented in this paper.11-12 The ZnSe layer thickness was optimized for each temperature examined in order to obtain the best electrochromic emissivity modulation .25 The ZnSe layer produces two effects: (i) it serves as a protective layer for WO3, and (ii) improves the optical device performance in the IR spectral region due to optical frequency impedence matching. An example for the emissivity modulation using a 500 nm thick ZnSe layer is given in Figure 2. As is clearly seen, the emissivity modulation is considerably higher with, than without the ZnSe layer. The oxygen plasma treatment of the ZnSe top layer was found to only slightly decrease the 300K emissivity modulation and modulation ratio of the electrochromic device.

After 20hr exposure Before exposure Without ZnSe

1.00 0.75

colored

0.50 0.25 0.00 0

bleached

10

20 30 wavelength [µm] Figure 2

Simulated thermal emittance spectra for an electrochromic device, with and without a 500 nm ZnSe layer on top of the Al top electrode grid.

40

488 6.0 Summary The Herzinger-Johs parametric optical model was successfully used to obtain excellent fits over the entire spectral range available, starting from the VUV up to the middle IR (146 nm to 40 µm). Six critical points were observed, corresponding to E0, E0 + ∆E0, E1, E1 + ∆E1 E2, and E0′, at 2.79 eV±0.01, 3.54±0.06 eV, 4.2±0.3 eV, 5.0±0.5 eV, 6.29±0.07 eV, and 8.1±0.3 eV, respectively, with the latter two being reported experimentally for the first time. Drastic changes were detected upon oxygen plasma irradiation in the VUV-UVVIS-NIR region, while few were seen in the Middle IR. The ZnSe thin films got thinner by about 20% after 20hr of AO exposure, an equivalence of ~16 years in LEO orbit; meanwhile, film thickness became fairly non-uniform (~17%). These, however, have fairly small affects on infrared device performance. Theoretical infrared electrochromic device performance was evaluated based on previous experimental material and device results, along with optical data obtained from the present experiments. A ZnSe layer was found to improve device performance significantly. For 300K applications ZnSe acted sufficiently as protective layers and improved the infrared optical device performance, with little degradation due to oxygen exposure. 7.0 Acknowledgements The ZnSe film work was supported by NASA Glenn Research Center, Grant NAG32219, and the electrochromic device work by the National Science Foundation Contract NSF II-9901510-EE-UNL. We thank James Hilfiker, Corey Bungay, Tom Tiwald, and Dan Thompson, for enlightening discussions. 8.0 References 1.

Masetti, E., Montecchi,. M., and da Silva, M. P. (1993) Analysis of the oxidation of polycrystalline zinc selenide by spectroscopic ellipsometry and photothermal deflection spectroscopy, Thin Solid Films 234, 557-560.

2. 3.

Adachi, S. and Taguchi, T. (1991) Optical properties of ZnSe, Phys. Rev. B 43, 9569-9577. Kim, Y. D., Cooper, S. L., Klein, M. V., and Jonker, B. T. (1993) Optical characterization of pure ZnSe films grown on GaAs, Appl. Phys. Lett. 62, 2387-2389.

4.

Jans, J. C., Petruzzello, J., Gaines, J. M., and Olego, D. J. (1993) Determination of the optical properties of II-VI compounds by spectroscopic ellipsometry, SPIE 1985, 260-265.

5.

Dahmani, R., Salamanca-Riba, L., Nguyen, N. V., Chandler-Horowitz, D., and Jonker, B. T. (1994) Determination of the optical constants of ZnSe films by spectroscopic ellipsometry, J. Appl. Phys. 76, 514-517.

6.

Kato, K., Akinaga, F., Kamai, T., and Wada, M. J. (1994) In-situ monitoring by spectroscopic ellipsometry in ZnSe crystal growth by molecular beam epitaxy, J. Cryst. Growth 138, 373-378.

489 7.

Kim, C. C. and Sivananthan, S. (1996) Optical properties of ZnSe and its modeling, Phys. Rev. B 53,

8

Lee, J., Collins, R. W., Heyd, A. R., Flack, F., and Samarth, N. (1996) Spectroellipsometry for

1475-1484. characterization of Zn1-xCdxSe multilayered structures on GaAs, Appl. Phys. Lett. 69, 2273-2275. 9.

Koo, M. S., Kim, T. J., Lee, M. S., Oh, M. S., Kim, Y. D., Yoo, S. D., Aspnes, D. E., and Jonker, B. T. (2000) Dielectric function of epitaxial ZnSe films, Appl. Phys. Lett. 77, 3364-3366.

10.

Hale, J. S., DeVries, M. D., Dworak, B., and Woollam, J. A. (1998) Visible and infrared optical constants of electrochromic materials for emissivity modulation applications, Thin Solid Films 313-314, 205-209.

11.

Franke, E. B., Trimble, C. L., Schubert, M., Woollam, J. A., and Hale, J. S. (2000) All-solid-state electrochromic reflectance device for emittance modulation in the far-infrared spectral region, Appl. Phys. Lett. 77, 930-932.

12.

Franke, E. B., Trimble, C. L., Hale, J. S., Schubert, M., and Woollam, J. A. (2000) Infrared switching electrochromic devices based on tungsten oxide, J. Appl. Phys. 88, 5777-5784.

13.

Woollam, J. A., Johs, B., Herzinger,C. M., Hilfiker, J., Synowicki, R., and Bungay, C. L. (1999) Overview of Variable Angle Spectroscopic Ellipsometry (VASE), Part I: Basic Theory and Typical Applications, SPIE CR 72, 3-28.

14.

Johs, B., Woollam, J. A., Herzinger, C. M., Hilfiker, J., Synowicki, R., and Bungay, C. L. (1999) Overview of Variable Angle Spectroscopic Ellipsometry (VASE), Part II: Advanced Applications, SPIE CR 72, 29-58.

15.

Azzam, R. M. A. and Bashara, N. M. (1977) Ellipsometry and Polarized Light, North-Holland, New York.

16.

Yan, L, Gao, X, Bungay, C., and Woollam, J.A. (2001) Study of surface chemical changes and erosion rates for CV-1144-O silicone under electron cyclotron resonance oxygen plasma exposure, J. Vac. Sci. Technol. A 19, 447-454.

17

Herzinger, C. M. and Johs, B. (1998) Dielectric function parametric model, and method of use, US Patents #5,796,983.

18.

Johs, B., Herzinger, C. M., Dinan, J. H., Cornfeld, A., and Benson, J. D. (1998) Development of a parametric optical constant model for Hg1-xCdxTe for control of composition by spectroscopic ellipsometry during MBE growth, Thin Solid Films 313-314, 137-142.

19.

Cardona, M. (1961) Fundamental reflectivity spectrum of semiconductors with zinc-blende structure, J. Appl. Phys. (Suppl.) 32, 2151-2155.

20.

Aven, M. et al. (1961) Some electrical and optical properties of ZnSe, J. Appl. Phys. (Suppl.) 32, 22612265.

21.

Pollak, F. H. (1967), in D. G. Thomas (eds.), II-VI Semiconducting Compounds, Benjamin, New York, pp.552.

22.

Greenaway, D. L. and Harbeke, G. (1968) Optical Properties and Band Structure of Semiconductors, Pergamon, Oxford.

490 23.

Jellison, G. E., Jr. and McCamy, J. W. (1992) Sample depolarization effects from thin films of ZnS on GaAs as measured by spectroscopic ellipsometry, Appl. Phys. Lett. 61, 512-514.

24.

(1997) R. A. Nyquist and R. O. Kagel (eds.), Handbook of Infrared and Raman Spectra of Inorganic Compounds and Organic Salts, Academic Press, San Diego, CA.

25.

Franke, E., Neumann, H., Schubert, M., Trimble, C. L., Yan, L., and Woollam, J. A., (2002) Low-orbitenvironment protective coating for all-solid-state electrochromic surface heat radiation control devices, Surface and Coating Techn. 151-152, 285-288.

PERFLUOROCYCLOBUTANE (PFCB) POLYARYL ETHERS FOR SPACE-BASED APPLICATIONS ARTHUR GAVRIN, JON NEBO, NORM RICE, AND LAWINO KAGUMBA Triton Systems, Inc. Chelmsford, MA 01824 DENNIS W. SMITH, JR., JACK JIN, AND CHRIS M. TOPPING Clemson University Clemson ,SC 29634 Abstract Perfluorocyclobutanes represent a relatively new class of partially fluorinated polymers that offer the many of the process advantages of engineering thermoplastics combined with the optical, electrical, thermal, and extreme environment resistant properties of traditional fluoropolymers. Our research has focused on modifications, blends and composites of PFCB’s to enhance their space durability and conductivity to make them suitable for a variety of space based applications. Enhancement of these thermally stable, low α, UV resistant polymers makes them suitable for a variety of space-based applications. Preliminary results indicate that modification, blending, and composite approaches can be used to produce materials with a balance of electrical and thermal conductivity, solar absorbtivity, and AO resistance suitable for a range of space-based applications.

1. 0 Introduction Perfluorocyclobutyl (PFCB) polymers are prepared by the free-radical mediated thermal cyclodimerization of aryl trifluorovinylether monomers (see Scheme 1). Step growth polymerization of these monomers takes place neat or in solution at temperatures ≥150ºC to yield materials with precisely controlled viscosities, molecular weights, and polydispersities.

491

492

F

F

F

F

∆ F

O Ar

O

F

F F ArO

F

F

..

F F OAr

F F O

F

F

F F O Ar

n

Scheme 1. Cyclodimerization of aryl trifluorovinyl ethers to form PFCB polymers

A variety of thermoplastic and thermosetting materials have been developed taking advantage of the thermal polymerization of this class of materials [1-5]. The facile polymerization of these materials and the properties of the hexafluorocyclobutane linkage (e.g. excellent thermal stability, enhanced ignition resistance, and excellent solubility as thermoplastics or pre-gel thermosets) make PFCB polymers an interesting class of materials. This combination of properties provides excellent processibility compared to many perfluorinated materials. The PFCB family has emerged as a candidate for coatings, polymer waveguides, polymer matrix composites, and other applications where tunable optical and mechanical properties are desirable. Polymers containing phenyl phosphine oxide (PPO) have been developed for a variety of applications. Radiation and atomic oxygen (AO) resistant PPO-containing poly (arylene ether benzimidizoles) were developed by researchers at NASA Langley for applications on spacecraft operating in low earth orbit (LEO) [6]. Triton Systems has developed proprietary processing techniques to make this polymer available in powder, solution, adhesive tape, fabric, thread, fiber and film forms and markets these polymers under the trade name TOR™ [7] under and exclusive license agreement with NASA. Triton Systems has developed process technology for a family of PPOcontaining space durable polymer products including products including polyarylene ethers (COR™) and polyimides (TOR-NC™) [8,9]. We have sought to incorporate the PPO group into the PFCB backbone to develop this class of polymer for future space applications. The synthesis of 1,1,1-tris-(4-trifluorovinlyoxy)phenylphosphine oxide has been reported previously [5]. However, poly(1,1,1-tris-(4trifluorovinlyoxy)phenylphosphine oxide) is a thermoset, which greatly limits its use in many space applications. In this paper, we describe two new thermoplastic PFCB polymers containing the PPO unit, initial evaluations of their space durability, and our efforts to scale-up and further develop these materials. Monomers used are shown in Figure 1.

493 F

F

F

C

F

F

F

CH 3

O

BPVE

TVE

O

F

F

F

F

F

O

F

O

O

F

F

Bis(trifluorovinyloxy) biphenyl (BPVE)

F

Tris(trifluorovinyloxyphenyl) ethane (TVE) F

F O

PPO-1

F

O

P

F

F

O

F

PPO2 F 2C

O

FCO

P

OCF

CF2

Bis(4-trifluorovinyloxy phenyl) phenyl phosphine oxide (PPO1) Bis(4-trifluorovinyloxybiphenyl) phenyl phosphine oxide (PPO2)

Figure 1. Four PFCB monomers used in this work

2. 0 Experimental 2.1 MATERIALS AND INSTRUMENTATION Dibromotetrafluoroethane was generously donated by the Dow Chemical Company and was distilled over CaH2 prior to use. Phenylphosphonic dichloride was distilled at reduced pressure prior to use. Other chemicals and reagents were purchased from Aldrich, Acros, or Fisher Scientific and used as received. NMR spectra were obtained on an Eclipse+ 500 spectrometer in CDCl3 or DMSO-d6 solvents. Thermal analyses were performed using a Mettler Toledo 851TGA/SDTA system and a DSC820. Gas Chromatography/Mass spectra (GC/MS) were obtained from a Varian Saturn GC/MS. AO and VUV tests were carried out at the NASA Marshall Space Flight Center Atomic Oxygen Beam Facility. 2.2 PREPARATION OF LITHIUM REAGENTS [(3), (4)] 4-trifluorovinlyoxybromobenzene (1) and 4-bromo-4’-(trifluorovinlyoxy)biphenyl (2) were prepared from 4-bromophenol and 4-(4bromophenyl)phenol respectively [1,10]. Typically, a 100 mL, three-neck, round bottom flask, equipped with a magnetic stirring bar, a N2 inlet, a rubber septum, and an additional funnel, was purged with N2 and charged with 4-(Trifluorovinyloxy)bromobenezene (1) (7.6g, 30 m mol)and dry diethyl ether (45ml). The mixture was cooled down to -78 °C with a dry-ice acetone bath. tBuLi (17.7 ml, 31 m mol) was added dropwise. The mixture was then stirred for 45 min to 1 h at -78 oC before addition of the substrate. This process uses a lithium (Li) reagent that is effective and produces the monomer in high yield.

494 2.3 PREPARATION OF GRIGNARD REAGENTS [(3*), (4*)] 4-trifluorovinlyoxybromobenzene (1) and 4-bromo-4’-(trifluorovinlyoxy)biphenyl (2) were prepared from 4-bromophenol and 4-(4-bromophenyl)phenol respectively [1,10]. Typically, a 250 mL, three-neck, round bottom flask, equipped with a magnetic stirring bar, a N2 inlet, a reflux condenser, and an additional funnel, was purged with N2 and charged with 1.91 g (78 m mol) of Mg turnings and 100 ml of dry THF. 18.0 g (71 m mol) of 4-(Trifluorovinyloxy)bromobenezene (1) was added dropwise to the reaction flask. The temperature of the mixture was maintained at <30°C with an ice water bath during the addition. The reaction is continued for 3-6 hrs at room temperature. 2.4 SYNTHESIS OF PPO-1 MONOMER (5) PHENYL(BIS(4TRIFLUOROVINYLOXY)PHENYL) PHOSPHINE OXIDE: Typically, phenylphosphonic dichloride (3.9g, 20mmol) was added dropwise via an additional funnel into an ether solution of lithium reagent 3 (40mmol) at -78oC. The mixture was warmed to room temperature and stirred for 2h. The organic was washed with to remove the salt by-product. The organic phase was dried with anhydrous MgSO4 and filtered. Similarly, phenylphosphonic dichloride (3.9g, 20mmol) was added dropwise via an additional funnel into a THF solution of Grignard reagent 3* (40mmol) and the temperature of the mixture was maintained at <30°C with an ice water bath. The reaction is continued for 15-18 hrs at room temperature. The reaction is quenched with a weak (0.5-2M) HCl solution and ether is added to the organic phase to improve the solubility of the product. In either case, removal of solvents under reduced pressure afforded viscous crude product. The crude product then is purified by extracting starting materials and side products with hexane and removing the residual solvent at reduced pressure and/or using column chromatograph on silica gel using hexane and ethyl acetate (9:1) as an eluent. The compound 5 is liquid with slightly yellowish color (Rf=0.6), yield ~80%. GC/MS (M+ calcd. as C22H13O3PF6 469) m/z: 469, 372, 281, 200, 152, 125, 77, 50; 1H NMR (500MHz, DMSO-d6) δ: 7.44-7.51ppm(4H,mm), 7.51-7.57ppm(2H,mm), 7.5713 7.63ppm(1H,mm), 7.63-7.71ppm (2H,mm), 7.73-7.83ppm(4H,mm); C NMR(125MHz, DMSO-d6) δ: 116.4ppm, 122.0ppm, 129.3ppm, 130.3ppm, 132.1ppm, 132.7ppm, 134.4ppm(ddd, CF=CF2), 134.7ppm, 146.9(ddd, CF=CF2), 157.5ppm; 19F δ: NMR(188MHz, DMSO-d6) -117.7ppm(1F,dd), -125.5ppm(1F,dd), 135.0ppm(1F,dd); 31P NMR: 28.3 ppm. 2.5 SYNTHESIS OF PHENYL(BIS(4-TRIFLUOROVINYLOXY)BIPHENYL) PHOSPHINE OXIDE (6): Typically, phenylphosphonic dichloride (3.9g, 20mmol) was added dropwise via an additional funnel into an ether solution of lithium reagent 4 (40mmol) at -78oC. The mixture was warmed to room temperature and stirred for 2h. The organic was washed with to remove the salt by-product. The organic phase was dried with anhydrous MgSO4 and filtered. Removal of solvents under reduced pressure afforded viscous

495 yellow crude product. The crude product then is purified by column chromatography on silica gel using hexane and ethyl acetate (9:1) as an eluent. The compound 6 is yellow solid (Rf=0.58), yield rate: 70%. 1H NMR(500MHz, DMSO-d6) δ: 7.36-7.45ppm(mm), 7.45-7.51ppm(mm), 7.55-7.62ppm(mm), 7.62-7.66ppm(mm), 7.66-7.78ppm (mm), 7.78-7.90ppm (mm); 13C NMR(125MHz, DMSO-d6) δ: 116.8ppm, 120.9ppm, 125.5ppm, 127.5ppm, 129.4ppm, 129.7ppm, 131.5ppm, 132.1ppm, 132.8ppm, 134.6(ddd, CF=CF2), 136.7ppm, 143.0ppm, 147.2(ddd, CF=CF2), 154.9ppm; 19F δ: -118.0ppm(1F,dd), -126.0ppm(1F,dd), NMR(188MHz, DMSO-d6) 134.3ppm(1F,dd); 31P NMR: 28.3 ppm. 2.6 POLYMERIZATION OF MONOMER 5 AND 6:

Monomers were thermally polymerized to form PPO containing PFCB polymers. The bulk polymerization was carried out at 160oC for 24h followed by additional polymerization at 200oC for 8-24h. In some cases, diphenyl ether or NMP was added to reduce the viscosity during the 200ºC polymerization phase. 3. 0 Results and Discussion 3.1 MONOMER AND POLYMER SYNTHESIS New phosphine oxide monomers (5) and (6) were prepared using the established intermediate strategy of delivering the trifluorovinyl ether group via Grignard and aryl lithium chemistry. Intermediates (1) and (2) were converted to the desired monomers using these organometallic reagents in good to excellent yield (see Scheme 2-3). The successful conversion of (1) to the PPO-containing monomer via Grignard reagents is a significant advance compared to previous efforts [10]. Lithium reagents require reaction temperatures of –78ºC and these Li reagents decompose exothermally above – 20ºC [11]. Grignard reagents can be reacted at or near room temperature and their use in the formation of monomers will facilitate the scale-up and commercialization of this class of polymers. In the later discussion, we designate the compound 5 as PPO1 monomer, the Poly (5) as polyPPO1, the compound 6 as PPO2 monomer and the Poly(6) as polyPPO2.

496 O F

F

Cl

F

P

Cl

F

O

F

O F

F O

F

tert-BuLi/Et2O

F

Br

P

O

F

F

O

F

Li

Heat

3

1

F F

F F F

F

O P

O

5

O

Scheme 2. Synthesis of poly PPO1 via Lithium Reagent OCF=CF2

OCF=CF2

O

1) Mg 3

F2C=CFO

P

OCF=C

THF RT/ ~18hr

THF RT/ ~6hr

Br

PhP(O)Cl2

Li MgBr

5

3 3 monomer via Grignard Reagent Scheme 13. Synthesis of PPO1 * Scheme 3. Synthesis of PPO1 monomer via Grignard Reagent

Thermal polymerizations were carried out at 200ºC for bifunctional PPO-1 and PPO-2 polymers. In addition, a number of co-polymers with non-phosphine oxide containing aryl trifluorovinyl ether monomers such as, Bis(trifluorovinyloxy) biphenyl (BPVE) and Tris(trifluorovinyloxyphenyl) ethane (TVE) (see Figure 1) were made in order to assess the potential development of a family of polymers with variable Tg, strength, modulus, and elongation by co-polymerization and blending. Initial polymerization experiments indicate that the reactivity of these new PPO-containg monomers should allow the formation of a wide range of homopolymers and co-polymers.

497 3.2 THERMAL CHARACTERIZATION The thermal behavior of PPO1 monomer is shown in Figure 2. The onset of thermal cyclodimerization of aryl trifluorovinyl ether in PPO1 monomer is around 140oC and the exothermal peak reaches to maximum at 243oC. This suggests the PPO group has negligable influence on the PFCB polymerization kinetics based on the behavior of their exotherms. In terms of the enthalpy of thermal cyclopolymerization, however, both the PPO1 and PPO2 monomer are much lower than the polyTVE thermoset, which are 12.7 kcal/mole vinyl group, 13.1 kcal/mole, and 18.1 kcal/mole, respectively. Due to noncoplanar structure of triphenyl phosphine oxide, both polyPPO1 and polyPPO2 were found to be amorphous with no crystalline transition temperatures being observed. The glass transition temperature Tgs of PPO1 and PPO2 are 122oC and 188oC respectively. The increase in Tg of polyPPO2 is probably due to bulk size of the biphenyl group in the backbone. 1.5

1.0

PPO1 monomer

w/g

Exo

0.5

0.0

-0.5

-1.0 -100

0

100

200

300

400

Temp/oC

Figure 2 . Polymerization of PPO1 monomer by DSC

A series of PPO1/TVE co-polymers were prepared. The glass transition temperature of the new TVE-PPO1 co-polymer can be varied from 122 oC to 240 oC (see Table 1). The trend in Tg follows the Fox equation of mixtures (see Figure 3), indicating the copolymerization of PPO1 and TVE is random. The fact that the copolymer prepared from PPO1 monomer and TVE is random is a valuable asset since most properties can then be predicted based on monomer feed ratio. Figure 4 shows the results of TGA measurement of poly-(PPO1-co-TVE). The thermo-oxidative stability of PPO1 as judged by the 5% weight loss temperature was found to be at 404oC. In case of the TVE 40wt% copolymer the 5% weight loss occurred at 416oC. Co-polymerization with TVE led to the increase in thermal stability and is an effective to tailor the thermal properties of these materials.

498 TABLE 1 Control of Tg by Co-Polymers of TVE-PPO-1

Tg of CoPolymer*

Wgt %

121.95 oC 142.44 159.75 195.74 231.76 -

0 / 100 19.75 / 80.25 38.63 / 61.37 60.24 / 39.76 79.59 / 20.41 100/0

Fox Equation Fit

TVE / PPO-1

121.9 138.5 160.0 195.0 240.0 300 (DSC), 350 DMA

* Tg taken as midpoint of glass transition curve, heat rate 10oC/min

poly(PPO1-co-TVE) Tg measured Fox eq fit

300

Tg/ oC

250

200

150

100

0

20

40

60

80

TVE wt% Figure 3. Tg (DSC) vs. TVE content for PPO-1 random copolymer.

100

499 Thermal stability of poly(TVE) Poly(PPO1-co-TVE) 60:40 poly(PPO1)

100

90

wt %

80

70

60

50

40

0

200

400

600

800

Temp/ oC Figure 4 . TGA Thermogram of PPO1 in nitrogen

3.3 AO EXPOSURE TESTING Ground based AO exposure was performed on thin film samples at the NASA Marshall Space Flight Center. The films were exposed to an average AO fluence of 2-4 x 1020 atom/cm2 and the effects on α, ε and mass loss determined. Exposure was roughly equivalent to four months exposure to AO in a LEO environment. We prepared BPVE – PPO-1 copolymers with 87.5 % BPVE / 12.5 % PPO and to a BPVE control, and to a BPVE-CNT conductive control for AO testing. The BPVE (no PPO) control is a partially fluorinated reference that might simulate other fluoropolymer films such as Teflon (Fluoropolymers). Table 2 shows that we did obtain a 10-fold (order of magnitude) increase in AO-erosion resistance (by bulk weight loss) and significant improvements on an area basis using the PBVE-PPO co-polymer, compared to a BPVE control, and to a BPVE-CNT conductive control. Table 2 also shows that there is no appreciable change in absorptance or emittance for any of the fluoro-samples under the UV irradiation experienced in the test.

500 TABLE 2. 10-Fold Improvement in AO Resistance of BPVE-PPO-1* Compared to BPVE [no PPO] Control Polymer

AO Fluence AO/cm2 ¶

Mass Loss (Bulk Wt) [AO Erosion]

Mass Loss/ Unit Area (g/cm2) [AO Erosion]

Solar Absorptance (α)

Thermal Emittance (ε)

Pre Exp

Post Exp

Pre Exp

Post Exp

BPVE-PPO-1 AO Resistant (87.5%/12.5%)

2.6x 020

2.2 %

0.0008

.163

.193

.680

.680

BPVE Control [No PPO]

3.9x 020

33.2 %

0.0012

.144

.164

.565

.411

4.9x 020

31.9 %

0.0016

.834

.818

.668

.628

BPVE-CNT Conductive [no PPO] Control

* 87.5 % BPVE / 12.5 % PPO-1

4. 0 Conclusions and Future Work We present a unique class of phenyl phosphine oxide-containing, partially fluorinated polymers, based on the thermal polymerization cyclodimerization of aryl trifluorovinylether monomers. The PPO-Perfluorocyclobutyl (PFCB) polymers will possess many of the advantages of traditional fluoropolymers such as excellent optical properties, thermal and thermal-oxidative stability along with atomic oxygen resistance afforded by PPO. Initial DSC experiments show that the incorporation of the PPO moiety into aryl trifluorovinylether monomers does not significantly affect their thermal polymerization as homopolymers or co-polymers. We have synthesized a variety of thermally stable thermoplastic and thermosetting materials from PPO1 and PPO2 and their co-polymers with other PFCB monomers. This new class of PPO polymers has excellent potential in a variety of space-based film and composite applications. Significant enhancement of AO resistance at relatively modest levels of PPO content in the polymers was observed. Typical TOR™ family polymers contain one PPO group per repeat unit (~4-8 wt% Phosphorous). However, the PPO-1 copolymer blend tested in this work contains one PPO group in approximately every four repeat units (~1.1 wt% Phosphorous). Despite the sub-stoichiometric use of PPO containing monomers in this polymer, it still exhibited significantly improved AO resistance compared to phenyl phosphine oxide free poly-BPVE films tested in parallel. This indicates that adequate AO resistance for a variety of space-based applications may be possible at modest levels of PPO incorporation. This offers the possibility of balancing the cost and mechanical properties of our entire family of TOR™ polymers against AO

501 resistance based on PPO content in the polymer backbone. We are actively pursuing these types of optimizations for TOR-NC™ [9] and other polymers. Further evaluation of this unique class of polymers and identification of promising space applications require scale-up of these monomers and polymers. The synthesis of PPO-1 monomers is an important step in the scale–up of these materials. Future work will focus on scale-up PPO-containing PFCB monomers, synthesis of PPO1 containing homopolymers and co-polymers and characterization of their mechanical properties and AO resistance to identify appropriate space-based applications. 5. 0 Acknowledgements The authors would like to thank Dr. Charles Lee of the Air Force Research Laboratory. Pin Go and Alan Shepp of Triton Systems Inc., and Dr. Dave Carroll of Clemson University for their intellectual contributions. We would also like to thank Dow Chemical for supplying many of the starting materials. This work funded under Air Force contract F49620-02-C-0018. 6. 0 References 1.

Smith, D. W.; Babb, A. Macromolecules 1996,29, 852.

2.

Babb, D.A.; Boone, H.; Smith, D.W. Jr.; Rudolf, P. J. Appl. Polym. Sci. 1998, 69, 2005.

3.

Smith, D.W. Jr.; Babb, D.A.; Shah, H.V.; Hoeglund, A.; Traiphol, R.; Perahia, D.; Bone, H.W.; Langhoff, C.; Radler, M. J. Fluorine Chem. 2001, 104, 109.

4.

Traiphol, R.; Shah, H.V.; Smith, D.W. Jr.; Perahia, D. Macromolecules 2001, 34, 3954

5.

Smith, D.W. Jr.;Babb, D.A.; Boone, H.; Snelgrove, R.V.; Latham, L.E., Polymer Preprints, 1997, 38(2), 361.

6.

Hergenrother, P. M; Smith Jr., J. G.;. Connell, J. W.; Polymer 1993, 34, 856.Connell, J. W.; Smith Jr., J. G.; Hergenrother, P. M. Polymer 1995, 36, 5.

7.

Schuler, P.; Haghighat, R. “Space Durable Polymeric Films: Advanced Materials for Inflatable and Thermal Applications” SAMPE J. 1999, 35(5), 37-44.

8.

Connel, J.W.; Watson, K.A. In Gossamer Spacecraft: Membrane and Inflatable Structures Technology for Space Applications; C.H.M. Jenkins, Ed.; AIAA, Reston, VA, 2001, p. 243.

9.

Arthur J. Gavrin, Siu Wa Au-Yeung, Bob Mojazza, Kent A. Watson, Joseph G. Smith, Jr., and John W. Connell; “Atomic Oxygen resistant, low alpha, anti-static polyimides for potential space applications”, this Proceedings.

11.

Jin, J.; Kumar, M.S.;Fougler, S. H.; Smith, D.W. Jr Lui, H.; Mojazza, B.; Go, P.; Shepp, A.; Poly Preprints, 2002, 43 (1), 610.

12.

Narayan Ji, J., Neilson, R.; Oxley, J.; Babb, D.; Rondan, N.; Smith, D. W., Jr.; Organometallics. 1998, 17, 783.

This page intentionally left blank

FAST THREE-DIMENSIONAL METHOD OF MODELING ATOMIC OXYGEN UNDERCUTTING OF PROTECTED POLYMERS AARON SNYDER AND BRUCE A. BANKS National Aeronautics and Space Administration Glenn Research Center Cleveland, Ohio 44135 Abstract A method is presented to model atomic oxygen erosion of protected polymers in low Earth orbit (LEO). Undercutting of protected polymers by atomic oxygen occurs in LEO due to the presence of scratch, crack or pin-window defects in the protective coatings. As a means of providing a better understanding of undercutting processes, a fast method of modeling atomic-oxygen undercutting of protected polymers has been developed. Current simulation methods often rely on computationally expensive raytracing procedures to track the surface-to-surface movement of individual “atoms”. The method introduced in this paper replaces slow individual particle approaches by substituting a model that utilizes both a geometric configuration-factor technique, which governs the diffuse transport of atoms between surfaces, and an efficient telescoping series algorithm, which rapidly integrates the cumulative effects stemming from the numerous atomic oxygen events occurring at the surfaces of an undercut cavity. This new method facilitates the systematic study of three-dimensional undercutting by allowing rapid simulations to be made over a wide range of erosion parameters. 1. 0 Introduction Undercutting of protected polymers by atomic oxygen occurs in LEO due to the presence of scratch, crack or pin-window defects in the protective coatings. Methods to accurately model characteristics of undercutting beneath such defects exist, but due to the extremely complicated nature of the problem fall short of providing a complete representation of the undercutting process. For example, Monte-Carlo methods [1-2], when used in conjunction with ray-tracing procedures, produce remarkably faithful representations, yet are computationally exhaustive for small cell sizes, and thus often restricted to two spatial dimensions. Another deficiency for any method is the lack of knowledge as to what effective values should apply for atomic oxygen reaction and recombination rates at polymer and protective coating surfaces. Such lack of knowledge complicates correlation between in-space and ground based results. Proper correlation between ground-based systems and flight results are required for meaningful durability studies. A means of establishing such a correlation is

503

504 the use of computer codes to simulate ground-based and in-space environments. To advance modeling of three-dimensional undercutting erosion yet minimize computational requirements, a procedure has been developed to provide fast and accurate simulations for a wide variety of conditions. This method shortcuts many computational bottlenecks and avoids lengthy ray tracing techniques by applying geometric configuration factors to determine the exchange of atoms between cavity surfaces, which in turn are approximated by simple geometric shapes. The primary goal was to construct a procedure that provides reasonably accurate predictions and serves as a guide for the design of more complex codes. Simple geometries that are easily parameterized were selected. To limit the scope, individual geometries were chosen that exhibit qualities peculiar to directed beam undercutting resulting from a fixed ram direction. Therefore the geometric models used in this paper do not apply to simulation of arbitrary three-dimensional cavities, but are restricted to modeling classes of cavities that exhibit symmetry about an axis or a plane. 2. 0 Geometric Model As mentioned above, the code models three-dimensional undercut cavities by using a few prescribed geometries to serve as representative shapes for some of the more simple cavities observed from in-space undercutting below pin-window defects. In the model, undercutting occurs directly below a defect present in the upper protective coating of the film. With a double-coated film, a lower protective coating also exists. The protective coating is assumed to be infinitely thin. It is convenient and reasonable to use circular disks to represent the pin-window defects. Consistent with this simple defect shape, all model-geometry cavity cross sections in planes parallel to the polymer film are circular. In the present code, four specific geometric classifications exist. These four geometries are: (1) normal cylinder, (2) normal-truncated cone, (3) oblique-cylinder, and (4) oblique-truncated cone. Each of these individual models was chosen to capture a specific set of features characteristic of the posthole-type erosion pattern arising from undercutting due to directed beam atomic oxygen. The degree of obliqueness of model (3) or model (4) is determined by the angle that the ram direction deviates from the film’s surface-normal direction. The conical geometries allow for modeling of converging and diverging undercut cavities. An equation for volume that is valid for the above geometric models is given by the following equation:

V=

π 3

h ( r12 + r1r2 + r22 ) ,

(1)

where r1 and r2 are the upper and lower circular disk radii (see Figure 1), which are equal in the case of a cylindrical model. In order to provide a reasonable

505 characterization, the radial and vertical dimensions of the model cavity must evolve rdef

2

r1

1

h <= hk

3

r2

4 Figure 1.

synchronously with cavity volume in a suitable fashion. To automate the process and provide a degree of control in specifying model geometries, the radial coordinate, r1, is chosen to be a simple function of the cavity height, h, for h < hk, where hk is the polymer-film thickness. This function is given as follows:

r1 = rdef + a h ,

h < hk .

(2)

For cylindrical geometries the radius is initially equal to the defect radius, rdef, and increases linearly with cavity height for h < hk. In the case of conical geometries an additional parameter, b, is introduced to relate the radii of the circular disks. Equation (2) is again used to specify r1, and then r2 is put in terms of r1 as given by the following relation:

r2 = r1 + b h ,

h ≤ hk .

(3)

As the value of the parameter b is negative, zero, or positive, the conical section is converging, cylindrical, or diverging, respectively. Since the volumes of these models are known functions of the radius and height, the cavity height can be solved in terms of the volume. The cavity’s height, h, cannot exceed the polymer-film thickness, hk. The protective-coating thickness (typically 0.13 micron) is ignored. In this manner the cavity geometry is specified once the cavity volume is known. The parameters a and b control

506 the cavity aspect ratio and cavity wall slope, respectively. As mentioned above, the degree of variation in the ram direction from normal determines the obliqueness of the model. 3. 0 Physical Model The present model assumes that non-reacting atomic-oxygen atoms reflect off surfaces diffusely, and consequently their movements between surfaces can be treated mathematically like diffuse radiation exchange between surfaces. The fraction of uniform diffuse radiation leaving one surface that reaches another is equivalent to the configuration factor between two surfaces, because it depends solely on the geometric orientation of the surfaces with respect to one another [3]. The geometric dependence of configuration factors can be used to derive algebraic relationships between factors. One pair of configuration factors exists for each pair of finite surfaces. A reciprocity relationship exists for the factors between two finite surfaces. It can be expressed by the following formula:

Ai Fi − j = Aj F j −i ,

(4)

where

F j −i =

1 Aj

cosθ i cosθ j

³ ³ Ai

Aj

π S2

dAj dAi

(5)

is the configuration factor from area Aj to area Ai. The angles θi and θ j are the angles between the line segment S, connecting the differential areas dAi and dAj, and surface normals ni and nj, respectively. To complete the general relations among configuration factors needed here, a conservation relation stating that the sum of fractions of emission leaving a surface and arriving at other surfaces of an enclosure (including the emitting surface) must sum to unity is expressed by equation (6). With the emitting surface denoted by the subscript index j and the N receiving surfaces of the enclosure identified by the running subscript index i, then the sum of configuration factors can be written compactly as: N

¦F i

j −i

= 1.

(6)

Given the above set of relationships, a complete set of configuration factors can be derived for an enclosed system starting from a relatively small initial subset, which for the geometries used here can be obtained from the existing literature on configuration factors [3]. An oblique truncated cone (converging downward from the defect surface) is shown in Figure 1. This geometry is one of the simple cavity

507 geometries used here for thin film modeling. The upper surface of the enclosure is comprised of a central disk that represents the defect, and a circular annulus that represents that portion of the coating from which the substrate has been removed by undercutting. These two areas are denoted as area 1 and area 2, respectively. Connecting the upper and lower disks is the lateral wall area denoted as area 3. The lower disk is denoted as area 4. Assuming F4-1 and F4-(1+2) are either known by catalogued formula or calculated directly using equation (5), where the summed subscripts in the last factor indicate that the corresponding surface is a composite surface, then the complete set of factors for the enclosure is found by simple algebra. A list of this set of configuration factors and their origin are given in Table 1. TABLE 1. List of Configuration Factors Associated with the Geometry Illustrated in Figure 1 and Their Origin Configuration Factor F4-1 F4-(1+2) F4-2 = F4-(1+2)-F4-1 F1-1 = F2-2 = F4-4 = 0 F1-2 = F2-1 = 0 F1-4 = A4 F4-1/A1 F2-4 = A4 F4-2/A2 F1-3 = 1- F1-4 F2-3 = 1- F2-4, F4-3 = 1- F4-1- F4-2 F3-1 = A1 F1-3/A3 F3-2 = A2 F2-3/A3 F3-4 = A4 F4-3/A3 F3-3 = 1- F3-1- F3-2- F3-4

Origin of Configuration Factor Known Known From definition of F4-(1+2) = F4-1 + F4-2 Planar Areas, Equation (5) Area 1 and area 2 in same plane, equation (5) Reciprocity, equation (4) Reciprocity, equation (4) Conservation of atoms from area 1, equation (6) Conservation of atoms from area 2, equation (6) Conservation of atoms from area 4, equation (6) Reciprocity, equation (4) Reciprocity, equation (4) Reciprocity, equation (4) Conservation of atoms from area 3, equation (6)

4. 0 Computational Model The first step in arriving at a computational model is to provide an algorithm to determine the fraction of atoms arriving at a given surface after surviving an arbitrary number of bounces within the enclosure. Let f i 0 be the fraction of atoms arriving directly through the defect that reach the ith surface without reflecting off a surface, and let qin be the probability that an atom arriving at the ith surface after n previous cavity bounces will not react, recombine, or exit during the next surface collision, but reflect again. The fraction of atoms surviving one reflection is qi0 f i 0 . Thus, by summing over all reflecting surfaces j=1,2,3…N and employing the configuration factor, the fraction of atoms reaching the ith surface after one bounce is given by the following sum: N

f i1 = ¦ F j −i q 0j f j0 . j

(7)

508 Although the configuration factors must be recalculated as the cavity grows, they are held stationary over a given time-step interval. By telescoping from the first bounce, it can be shown that the fraction, f i n , of initial atoms entering the cavity during a given time-step interval that reach the ith surface after n bounces is given by the following summation over the N surfaces: N

f i n = ¦ F j −i q nj −1 f jn −1 ,

n = 1, 2, 3,...

(8)

j

The fraction of atoms recombining at a polymer surface, f p , is found by summing (over the total number of impacts in a time step) the product of the fraction of atoms arriving at the surface times the probability Prn of reaction, which in general changes between surface impacts. On average, energy is lost following a collision, and this loss is generally reflected in the code by selecting a lower reaction probability for the next impact. For convenience, consider the set of all polymer surfaces as being one surface denoted by the subscript p. Then the net fraction of atomic-oxygen atoms reacting at the polymer to cause erosion during a time step consisting of NB bounces is given by the following expression: NB

N

n =1

j

f netNB = ¦ Prn −1 ¦ F j − p q nj −1 f jn −1 .

(9)

The total number of atoms reacting is given by the product of the net fraction of atoms reacting and the total atoms entered, which itself is given by multiplying the atom fluence, ∆F , during a time-step interval by the defect area, Adef. Continuing in this fashion, the product of the total number of atoms reacting and the erosion yield, E, (volume/atom) equals the amount of volume erosion, ∆V , obtained during a time step, which is given by the following:

∆V = f netNB ∆F Adef E .

(10)

The final undercut volume, V, is obtained by computing a series of ∆V volumes. In preparation for discussing results, it is convenient to define the volume erosion gain, G, as the ratio of the volume of undercutting erosion below a defect site to the volume of erosion obtained on the surface of an unprotected “smooth-witness” sample, which had been exposed to the same number of atoms that entered the defect. In terms of simulation parameters, the above gain can be determined by dividing the NB covering Ns time steps by the initial reaction probability, average of the incremental f net

Pr0 , and is given by

509

¦f G=

NB net

Pr0 N s

.

(11)

The “instantaneous” gain, G′ , corresponding to the gain for a single time-step interval is given by

G′ = f netNB / Pr0 ,

(12)

and provides a normalized erosion rate useful for comparison purposes. When G′ equals one, the rate of undercutting erosion equals the rate of erosion of an unprotected smooth sample of area Adef exposed to the same atom flux. 5. 0 Results and Discussion Values of the erosion parameters representative of the LEO environment were used in simulations. These nominal values along with other code parameters are listed in Table 2. Any variation from these nominal values is noted in the text as the case occurs. Similar to ground-based experiments, simulated fluence is an effective fluence, equivalent to the atoms per unit area in LEO that would produce the same volume erosion. The in-space erosion yield E for Kapton® representative of LEO is 3 x 10-24 cm3/atom based on atomic oxygen energy of 4.5 eV. It is emphasized that the tabulated entry for atomic-oxygen fluence is scaled by polymer-film thickness. Thus, normalized fluence-per-time-step-interval is listed as 2 × 1022 atoms/cm2 per cm of film thickness, giving a non-scaled cumulative fluence over 100 time steps, for example, of 5 × 1021 atoms/cm2 for a 0.0025 cm thickness film. In this fashion, having normalized the erosion rate by film thickness, the results in this paper hold for any film thickness. TABLE 2. Nominal Computational Model Parameters for LEO Atomic Oxygen Interaction with Kapton® Atomic oxygen initial impact at polymer reaction probability Atomic oxygen asymptotic impact at polymer reaction probability Atomic oxygen impact at coating recombination probability Atomic oxygen impact at polymer recombination probability Fractional decrease in polymer reaction probability per impact Number of atom bounces per time step Atomic oxygen fluence (atoms/cm2) per cm of film thickness per time step Number of erosion time steps Atomic oxygen fluence (atoms/cm2) for 0.0025 cm thickness film Erosion yield (cm3/AO-atom) Value of parameter a in equation [2] Values of parameter b for h = hk in equation [3] for conical geometry

0.11 0.001 0.13 0.24 0.368 500 2 × 1022 100 5 × 1021 3 × 10-24 0.0759

Protective coating defect size

0.1 hk

± rdef

510 To give an example of effects obtained by changing cavity shape, results are given comparing data for the normal-cylinder cavity model with data for the converging and diverging normal-cone cavity models. For the conical geometries, the final wall angle converges or diverges (downward from the defect) by 7.1 degrees from axial. For these three geometric models, the variation of normalized erosion rate G′ versus fractional fluence is presented in Figure 2. It is readily seen in Figure 2 that whereas each curve is distinguished by the location of a steep dip, the characteristic shape of the curves remains the same for the different geometries. The dip in a curve coincides with the cavity reaching the lower protective coating. The reason that the dip occurs at a lower fluence for a converging cavity is that its volume grows more slowly with cavity height than the other cases, causing it to reach the lower coating first. For the same reason, the cavity of the cylindrical model reaches the lower coating before the cavity of the diverging-cone model. 1.8

1.6

Normalized Rate of Ersoion

Diverging 1.4

1.2

Converging

1

Cylindrical

0.8

0.6

0.4

0.2

0 0

0.2

0.4

0.6

0.8

1

Fractional Fluence Figure 2.

Two of the most important parameters influencing undercutting erosion are the defect size rdef and reaction probability Pr0 . To identify any gross effects observed with changes in these two parameters, the variation of final erosion gain, G, as a function of rdef is shown in Figure 3 for various values of Pr0 using the normal-cylinder geometry. For this data, rdef ranges from 0.2 hk to 0.001 hk. For a given Pr0 , G increases rapidly with decreasing rdef until, at a value of rdef approximately equal to 0.1 hk, little increase in G is observed. Gains for the largest defect size, rdef = 0.2 hk, are close to one, while gains for the smallest defect size, rdef = 0.001 hk, are about 75% greater because of

511 enhanced trapping of the atomic oxygen. It is seen that variations of plus or minus 20% from the nominal reaction probability ( Pr0 = 0.11) produce respective changes in G of approximately -2.1% and 2.2% for the smallest defects and -0.6% and 1.5% for the largest defects. As a final example, a comparison of erosion rates is presented for single-coated and double-coated Kapton® films. To simulate the effect of having no lower protective coating, the recombination probability is set equal to one at the lower film surface. In Figure 4 curves of the erosion rate, G′ , as a function of fractional fluence are shown for single-coated and double-coated Kapton® film using the normal-cylinder model. As expected, the curves are identical until the bottom surface is reached. In the double-coated polymer case, the presence of the lower coating results in additional erosion, beyond that obtained in the single-coated case, due to reflected atoms. 2.0

0.088

0.11

1.8

1.6

0.132

Erosion Gain

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0 0.001

0.01

0.1

Normalized Defect Radius, (radius / hk) Figure 3.

1

512 1.8 1.6

Normalized Erosion Rate

1.4 1.2

Double coated

1 0.8 0.6 0.4

Single coated 0.2 0 0

0.2

0.4

0.6

0.8

1

Fractional Fluence Figure 4.

6. 0 Summary A new method to simulate three-dimensional undercutting below pin-window defects in protected polymer thin films has been constructed. This method is very fast. To minimize computational requirements, it uses a geometric configuration-factor technique to govern the exchange of atoms between surfaces. The model assumes that non-reacting, non-recombining atoms reflect diffusely off surfaces. The basic equations to derive the view factors are given. In addition, a table listing the sequence of steps to derive a full set of configuration factors is provided for a representative case. A set of simple geometric shapes is used to model undercut cavities. A given geometric shape is selected based on its suitability to faithfully model particular qualities. These qualities belong to a small set of cavity characteristics, which include the cavity aspect ratio, cavity obliqueness to the film surface due to ram angle, and cavity-wall divergence. The equations relating the amount of volume erosion and the rate of volume erosion to the atomic-oxygen fluence are presented. Using a suitable set of simple geometries as models, rapid simulations can be obtained over a wide range of parameter space. Examples are given that illustrate typical results that may be obtained using the associated computer code.

513 7. 0 Acknowledgements The authors would like to thank Edward Sechkar for his contributions to the figures. 8. 0 References 1. B. Banks, T. Stueber, S. Snyder, S. Rutledge, and M. Norris, Atomic Oxygen Erosion Phenomena, American Institute of Aeronautics Defense and Space Conference, Huntsville, Alabama, September 2325, 1997. 2. B. Banks, T. Stueber, and M. Norris, Monte-Carlo Computational Modeling of the Energy Dependence of Atomic Oxygen Undercutting of Protected Polymers, NASA TM-1998-207423, Fourth International Space Conference, ICPMSE-4, Toronto, Canada, April 23-24, 1998. 3. R. Siegel and J. Howell, “Thermal Radiation Heat Transfer,” Second edition, Hemisphere Publishing Corporation, Washington D.C., 1981.

This page intentionally left blank

DEVELOPMENT AND VERIFICATION OF A PREDICTIVE MODEL AND ENGINEERING SOFTWARE GUIDE FOR DURABILITY EVALUATION OF POLYMER-BASED MATERIALS IN LEO J. I. KLEIMAN, Z. ISKANDEROVA, D. TALAS Integrity Testing Laboratory Inc. 80 Esna Park Drive, Markham, Ontario, L3R 2R7, Canada M. VAN EESBEEK European Space Agency, ESTEC, P.O. Box 299, Keplerlan 1, 2200 AG Noordwijk ZH, The Netherlands R. C. TENNYSON University of Toronto Institute for Aerospace Studies 4925 Dufferin Street, Toronto, Ontario, Canada, M3H 5T6

Abstract A “stand-alone” Predictive Erosion Resistance Software (PERSTM) package was developed, based on recently established predictive models of interaction of polymerbased materials with space environment and erosion resistance in LEO. The PERSTM package includes user-friendly on-line software and a database of more then 40 polymeric materials. The PERSTM program provides predictions for materials that are included or not included in the database as well allows the user to check how changes to certain properties of a polymer material will affect its erosion resistance. Experimental ground-based LEO simulating accelerated testing and evaluation of erosion resistance of more than 40 polymeric materials was also conducted. The results are presented in PERS and will be also used for a follow-up comparison study with data to be received on the same set of materials after a 1.5 year LEO space flight in “Materials on International Space Station Experiment” (MISSE). 1.0 Introduction The external surfaces of spacecraft in the low Earth orbit (LEO) environment (200 to 700 km) are exposed to a number of natural environmental factors capable of degrading sensitive spacecraft materials and affecting spacecraft major systems performance. Among these one can mention atomic oxygen (AO), solar ultraviolet (UV) and vacuum ultraviolet (VUV) radiation, thermal cycling, ionizing radiation, etc. [1-3]. Performance

515

516 degradation to varying degree depends on spacecraft orientation, orbital inclination, orbital altitude, state of the solar cycle, and mission duration. AO is the major constituent of the natural space environment in LEO and has been shown to be one of the most important factors in the environmental degradation of several important classes of spacecraft materials [1-4], with more or less influence and synergism of other factors [5-8]. Polymers and carbon-based composite materials, such as carbon-fiber reinforces plastic (CFRP) composites and carbon-carbon composites undergo various types of accelerated erosion and degradation in LEO, the degree of which depends on their positioning, extend of exposure, as well as on the nature of the chemical and physical processes involved. The consequences of materials degradation in space can be catastrophic, and considerable effort is put toward selection and development of the most stable materials or various protection schemes [4,9,10]. The description of the degradation and erosion reactions under various environmental factors is generally a very complex problem, and significant effort is currently made to understand and describe the mechanisms of highly accelerated erosion and etching of polymer-based materials in LEO. A project was initiated a few years ago to develop a basis for a user-friendly software based on a number of developed predictive models of materials erosion in LEO environment [11-13]. A reasonable agreement was found between the predictive model results and results of ground-based testing experiments in a space simulator for a number of selected materials [14]. The work reported in the present paper is a continuation of this effort. The software model was developed further and the database was widened to include the results from testing of over 40 polymer materials that were chosen also for space flight experiments and are flown presently in space on the MISSE experiment [15, 16]. The project was based on a concept of great interest to space applications, and lead to development of a new space-related high-technology product – Predictive Erosion Resistance Software package. Initial computer modeling, durability predictions, and a ground-based testing program were conducted on a limited number of selected samples [12, 14]. The results from MISSE flight missions will be used later, in the follow-up work, for confirmation and refining of the predictive model and software package. In the MISSE experiment [15], the same 40 materials as in this project are exposed to LEO environment for 1.5 year and are scheduled to be brought back to the investigators for analysis shortly. When combined with LEO environmental models, developed, for instance, by NASA and ESA [17, 18], PERSTM can be used for the development of a comprehensive Engineering Guide for quantitative prediction and evaluation of materials durability and performance in space environment, that will be used as a basis for Materials Selection for Space Application, first of it’s kind. No such models and/or interactive software Engineering Guide exist presently in the space community worldwide. 2.0 The Model of Hydrocarbon Polymers Erosion by FAO A unique predictive model of interaction of LEO space environment with polymers has been recently developed in which very useful quantitative correlations have been found between the erosion yields ReLEO in LEO and the chemical composition and structure of

517 many hydrocarbon and fluorinated polymeric materials, as well as their flammability and surface energy [11-14]. Using fundamental, well-established approaches to calculate and predict various polymer properties based on their composition [19,20], a general approach was found for predictive evaluation of erosion resistance and LEO space durability of copolymers, terpolymers, various polymer compositions, polymer blends, alloys, etc. These results were used as a starting point in the creation of the PERSTM software. In its essence, the predictive model first established a correlation between the observed erosion yields of major hydrocarbon polymers in LEO and their repeating unit structure, through an introduction of a dimensionless structural parameter, called γ or γ', and later a correlation was found between ReLEO and materials flammability [11-13]. All these models were used in the present work to describe the found correlations between the erosion yield and composition/properties of the materials. The predictive capabilities of the developed correlations were confirmed by results from many LEO flight experiments and by ground-based testing results for a number of common and new or specially synthesized polymeric materials. 3. 0 Experimental As a part of this project, a series of 40 polymer sample materials and Kapton 500HN witness samples were ground-based tested by expose to a directional beam of fast atomic oxygen (FAO) species. The fast atomic oxygen testing was performed in the Atomic Oxygen Beam (AOB) Facility, or the LEO Space Simulator for accelerated ground-based testing, design of which combines several unique components to produce a system which simultaneously simulates the low Earth orbit effects of atomic oxygen with fluxes up to 1017atomsxcm-2s-1, vacuum ultraviolet radiation with intensities of up to 10 equivalent suns (ES) in the 120-400-nm wavelength range, and cryogenic thermal cycling from -100°C to +125°C (for a detailed description of AOB facility, see [21] ). The facility was recently upgraded and modernized, using a stainless steel chamber. A fixture was developed to hold seven samples in an axi-symmetric array so that each 1.25-cm-diameter disk sample would receive as uniform flux of AO as possible. Six samples were placed symmetrically about a central Kapton witness sample. The samples were oriented at 90° angle to the AO bean to maximize the AO dose. The samples were out-gassed in a vacuum of 10-5 Torr in the test chamber for a period of 24 hr. A similar set of samples was put in the chamber but out of the AO exposure, i.e. was exposed only to high vacuum, to account precisely for the outgassing processes. The initial mass of each sample was then promptly measured on a Mettler 240 balance that has a resolution of 10µg. The samples were then exposed to AO for a period of 6-12 hours. Mass measurements were made for every sample at the end of every test. The Kapton 500 HN witness sample mass loss indicated that during FAO tests the average AO flux was ~1x1016 atoms cm-2 s-1 for the test array, so the testing was performed up to the fluence of~(1-2)x1020 atomsxcm-2. Qualitative observations of the changes in the optical appearance of the samples were also made, that indicated “matte” appearance of the eroded polymers.

518 4. 0 Predictive Erosion Resistance Software (PERS¥) Based on all the correlations discussed above a computer software was designed to automate the task of predicting erosion yields in LEO for polymeric materials. The software is designed to choose the optimum method of calculating ReLEO(pred) depending on what kind of data is available for the polymer-based material. For example, if only the chemical composition of the material is known, the software will predict ReLEO , using the so called γ-correlation, while if the density of the material is also known it will use what was called γ'-correlation to predict ReLEO [11,13]. The software can also give predictions for a variety of materials, including proprietary materials or polymer mixtures, based on the established correlation of hydrocarbon polymers erosion yield with the characteristic parameter of a polymer that defines the materials flame resistance and is called Oxygen Index [12,19]. Oxygen Index may be available from various sources [19,22-24], can be measured experimentally, using ASTM Standard [25], or, if the Oxygen Index value is unavailable, the heat of combustion [19,22,24] can be used in the software to derive the Oxygen Index value. In addition to providing predictions for materials that are included in its database, the software can make predictions for a random material as long as some of its key properties are known, i.e. structural composition, composition and density, OI value, or heat of combustion. The program also allows the user to make changes to the properties of the materials stored in the database before making a prediction to see how changes in the key properties would affect the erosion yield. Linear regression line based on gamma-correlation of ReLEO, as well as linear regression lines for ReLEO versus ReLEO(pred), and ReLEO vs 1/(OI) are graphed and provided as a basis of the final erosion yield in LEO predictions evaluation. The user has a choice as to which materials are included in the regression and which are not. The user can also alter any entry in the database as well as add new entries. Besides the key properties (chemical composition, density, OI, heat of combustion), the user can add an image of the structural formula of the polymer to the database provided the image is of *.gif or *.jpg format. Information on trade name and manufacturer of the polymer as well as erosion yield data from ground-based testing and flight experiments can be also added. To increase the functionality of the software, the user has the option of entering ranged data for erosion yield results as well as single-valued data. Information regarding the Software Development and Runtime Environment, the Algorithm, and Calculation Decision Tree is presented in next section. 4.1 DEVELOPMENT AND RUNTIME ENVIRONMENTS Predictive Erosion Resistance Software (PERSTM) package tool was developed using Java language, utilizing JDK 1.1.8, JSP technology and IBM Visual Age development environment. It is currently running on Tomcat Java Server (http://www.apache.org) and Internet Information Server (IIS 5, http://www.microsoft.com/iis), both on Windows 2000 Server platform. MySQL (http://www.mysql.com) is used as the data store for the application.

519 4.2 ALGORITHM In general, the calculation algorithm is derived from statistical information collected from flight data and applying certain formulae according to material information. On the lower level, all flight data is stored in the database, including such information as material trade name, detailed chemical information, flight erosion yield data and ground-based testing erosion yield data. Data store allows for multiple flight data entries, as well as erosion yield ranges. Figure 1 depicts schematically the application data model:

R-LEO Data Model

materials PK

material_id

LONG

chemical_name code trade_name N_count S_count H_count O_count C_count F_count carbonil_count carboxil_count sulfone_count cf_count Cl_count carbonate_count hydroxyl_count cf2_count density heat_combustion oxygen_index oxygen_index_reference

VARCHAR(255) TEXT(50) TEXT(255) SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT SHORT LONG LONG LONG TEXT(255)

trade_names PK

trade_name_id

LONG

FK1

material_id trade_name manufacturer

LONG TEXT(255) TEXT(255)

material_data PK

material_data_id

LONG

FK1

material_id value unit_id reference is_ground is_regression_member is_range_regression_member value_min value_max

LONG LONG LONG LONGTEXT CHAR(10) CHAR(10) CHAR(10) LONG LONG

FK2

units PK

unit_id

LONG

name

TEXT(50)

Figure 1. Schematic depiction of the application of the R (LEO) data model

4.3 COMPUTATIONAL METHODS The algorithm utilizes numerous methods when predicting the erosion yield as described in section 3. It looks at the property of a material to decide which method is to be used. Figure 2 shows the decision tree the application uses to decide which method to apply to the calculation:

520 RLeo Calculation Decision Tree

OR

Use Chemical Formula?

Use Oxygen Index?

Yes

Enter Number of Carbonils, Carboxils and Sulfone (Step 2)

Is Density of the Material Available? (Step 4)

OR

Use Correlation with Oxygen Index? (Step 3)

Yes

Calculate OI

No

Calculate with Method 4

Yes

Calculate with Method 1

Is it Fluorinated? (Step 2)

No

Yes

Calculate with Method 3

Calculate with Method 2

Figure 2. The schematic presentation of the decision tree for RLEO calculation

521 Figures 3 and Figure 4 show examples of windows in the PERSTM software that depict the table with properties of a chosen polymer (Fig.3) and the final table with erosion rates calculated thorough the use of different models as described above (Fig. 4). Figures 5 and 6 show the windows from the PERSTM program that are displaying in graphic form the results of the calculated data point for Kapton (represented by an open square) and the correlations that are available for erosion rate prediction using the various models (represented by the solid line) as well as the flight data (solid squares), that represent a variety of materials. It should be noted that all graphs are interactive. By hovering over any of the data points, information on that particular data can be displayed.

Name

Code

Trade Names

Composition

Kapton H polyimide

PI

Kapton HN (DuPont); Kapton H (DuPont);

C22H10O5N2

Actions

Figure 3. Table of Material Properties as developed in the present version of the PERS TM software. Properties of Kapton HN are shown.

Calculated Erosion Yield (10-24cm3/at)

Measured Erosion Yield (10-24cm3/at)

Gamma-correlation

OI-correlation

R (LEO)

Ground-based

2.76

2.88

3.0;

<standard>

Figure 4. The Table of calculated erosion yield using two models, γ-correlation and OI-correlation, and measured erosion yield (flight data in LEO and ground-based testing data).

522

Figure 5. Results of a calculation of the erosion yield using one of the predictive models for erosion. The γcorrelation model was chosen for this case. The calculated data for Kapton is presented by the open square. The solid squares present the available flight data, every point for a particular material. The solid line is the result of a regression analysis using the available flight data and the γ-correlation model.

Figure 6. Results of a calculation of the erosion yield using the Oxygen Index (OI)- correlation model The calculated data for Kapton is presented by the open square. The solid squares present the available flight data every point for a particular material. The solid line is the result of a regression analysis using the available flight data and the OI-correlation model.

5.0 The Ground-based Testing Program of the 40 Polymer Materials The 40 polymers that were chosen to build the database of the PERSTM software were also extensively tested in the Fast Atomic Oxygen (FAO) Beam Facility, or Space Simulator for accelerated ground-based testing at the University of Toronto Institute for

523 Aerospace Studies [26]. It was shown, that for the majority of tested materials the experimental results are in reasonable agreement with flight data and predicted values. However, for two classes of materials, namely, fluorinated polymers (Kynar, Halar, Tedlar, Teflon and Teflon FEP) and polyolefins (polyethylene, polypropylene and polymethyl methacrylate) results are lower that flight data and predicted values (see Table I). TABLE I. Results of the performed ground-based testing for some of the selected materials, together with the predicted erosion yields and with flight data

Material

Sample Testing Date

Mass loss (µg/cm2)

RePred by Gamma Parameter

Flight Data (10-24) cm3/a [26]

Kapton 500HN

28/04/2001 29/04/2001

360 370

2.76

3

06/05/2001

40

Fluorinated ethylene propylene

16/06/2001

70

0.3

0.03-0.3

16/06/2001

110

06/05/2001 16/06/2001 06/05/2001 24/06/01 11/08/01 01/09/01 (12 hours)

70 160 390 180 310

Ultra High Molecular Weight Polyethylene Kevlar Polypropylene

Polymethyl methacrylate

Polyetherimide Polyetherimide

Polyethylene

Ground-based Experimental Results (10-24) cm3/a Standard Standard 0.09 0.07 0.04

3

3.7

2.3

2.1-4.1

3

4.4

1.71 2.01 4.27 2.35 1.91

510

2.3

24/06/01

240

1.96

10/08/01

400

01/09/01 (12 hours) 10/08/01 11/08/01 01/0902/09 (12 hours) 10/08/01 11/08/01 01/0902/09 (12 hours)

3.75

3.9-4.8

1.8

830

2.81

420 430

1.77 1.96 2.2

2.1

650

1.97

260 280

1.43 1.42 3.1-3.7; 4.8

420

3

1.66

For fluorinated polymers the differences might appear because of the absence of UV/VUV exposure during this series of ground-based testing. The range of erosion yield data at various space flight experiments and synergistic effects due to combined effect of FAO and UV/VUV exposure affecting the erosion processes have been shown in many previous experiments (see ref. in [5,26]). Therefore, combined FAO and

524 UV/VUV tests should be conducted to resolve this ambiguity. For poly-olefins the cause of the significantly lower erosion yield is still unclear. The results of the MISSE experiment samples once they have returned from the International Space Station will be used to compare to our data in an attempt to understand the observed discrepancies. The results of the ground-based testing have been plotted vs. the gamma (compositional) parameter, vs. the Re(calc.0), i.e. using the gamma-modified approach, and vs. the inverse Oxygen Index. In all three cases, linear trends have been confirmed as predicted by our models. These trends will be refined once the ground-based tests combining FAO with UV/VUV are performed, as well as when the MISSE experiment in LEO space flight environment will be finished and the results of the erosion yield values for the space exposed 40 polymers will be obtained. 6.0 Conclusions An extensive analysis of LEO space flight data was conducted using all available literature sources worldwide, including publications, personal communications and discussions with the Principal Investigators of space flight materials’ experiments. The obtained data was critically assessed against the results predicted by recently developed models, and data was collected from a series of experiments on a comprehensive set of 40 selected polymers tested in ground-based Space Environment Simulator for Accelerated Testing at UTIAS/ITL as part of the program. The same set of materials, representing various polymer families, among which approximately half have not been tested in space or in ground-based testing facilities before, is also flown now in MISSE flight experiment. The PERSTM software package and a computer Database including 40 materials for predictive evaluation of polymer-based materials durability in LEO were developed. This prototype Software package can be also used for LEO erosion resistance prediction of untested, proprietary, or new materials. It may be later combined with one of the existing LEO Environmental Models. The conducted research verified and extended the predictive models and certain relationships, which were found between the composition, structure, and some properties of space-related polymers, and their erosion in LEO. It was confirmed that the developed models may be used for LEO durability predictive evaluation of a variety of polymeric materials. 7.0 Acknowledgements This work was supported in part by European Space Research and Technology Center, of the European Space Agency under the ESTEC contract No. 14.861/00/NL/PB. 8.0 References 1. 2. 3.

Hastings, D., Garret, H., “Spacecraft-Environment Interactions”, Cambridge University Press, 1996. Tribble, A.C., “The Space Environment. Implication for Spacecrafts Design”, Princeton University Press, 1995. “Natural Orbital Environment Guidelines for Use in Aerospace Vehicle Development”, Jeffrey B., (Editor), Smith R., (Compiler), NASA TM 4527, June 1994.

525 4. 5. 6. 7. 8. 9. 10.

11. 12.

13. 14.

15. 16.

17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Banks, A.B., “Atomic Oxygen”, LDEF Materials Data Analysis Workshop, NASA CP 10046 (1990), 191-217. Tennyson, R.C., “Atomic Oxygen and Its Effects on Materials”, in “The Behavior of Systems in the Space Environment” Ed. R.N. De Witt, Kluwer Academic Publishers (1993), 233-257. Kleiman, J.I., Iskanderova, Z.A., Gudimenko, Y..I., Morison, W.D. and Tennyson, R.C. “Polymers and Composites in the Low Earth Orbit Space Environment: Interaction and Protection”, Canadian Aeronautics and Space Journal, Vol. 45, No. 2, (1999),148-160. Proceedings of the 7th International Symposium “Materials in the LEO Space Environment”, Toulouse, France, 16-20 June 1997. Proceedings of the 8th International Symposium on “Materials in the Space Environment/ 5 International Conference “Protection of Materials and Structures from the LEO Space Environment”,” Arcachon, France, 5-9 June, 2000. Reddy, M.R. “Review: Effect of Low Earth Orbit Atomic Oxygen on Spacecraft Materials”, Journal of Material Science, 30 (1995), 281-307. Kleiman J., Iskanderova Z., “Technological Aspects of Protection of Polymers and Carbon-Based Materials in Space”, in Proceedings of the 8th International Symposium on “Materials in the Space Environment”/ 5 International Conference “Protection of Materials and Structures from the LEO Space Environment”,” Arcachon, France, 5-9 June, 2000. Iskanderova, Z. A, Kleiman, J. I., Gudimenko, Yu., and Tennyson, R. C., “Influence of Content and Structure of Hydrocarbon Polymers on Erosion by Atomic Oxygen,” J. of Spacecraft and Rockets, Vol. 32, No. 5, Sept-Oct. 1995, pp. 878-884. Kleiman, J.I., Iskanderova, Z. A., Gudimenko, Yu., Lemberg, V., Talas, D., and Tennyson, R.C., “Software Engineering Guide Development for Prediction of Erosion of Polymer Material in Space Environment,” NATO Advanced Research Workshop on Computer Modeling of Electronic and Atomic Processes in Solids, Wroclaw, Poland, May 20-23, Kluwer Academic Publishers, 1997, 277-288. Iskanderova, Z., Kleiman, J.I., Gudimenko, Y., Cool, G., and Tennyson, R.C. “Development of a Model of Polymer Interaction with Fast Atomic Oxygen.” Proceedings of the 7th International Symposium on “Materials in the Space Environment”, Toulouse, France, 16-20 June 1997. Cool, G.R., Iskanderova, Z., Gudimenko, Yu., Kleiman, J., Tennyson R.C., “Ground-Based Experimental Verification of the Predictive Model of Polymer-Based Materials Erosion by Atomic Oxygen”, in Proceedings of the Third International Conference “Protection of Materials and Structures from the LEO Space Environment”, , Kluwer Academic Publishers, 1999, 101-106. 15. De Groh, Kim, Banks, B. A., at al., “MISSE PEACE Polymers: an International Space Station Environmental Exposure Experiment”, AIAA-2001-4923, Nov. 2001. 16.Woll, S.L.B., Pippin, H.G., Stropki, M.A. and Clifton, S. “Materials on International Space Station Experiment (MISSE)”., Proceedings of the 8th International Symposium on “Materials in the Space Environment”/ 5 International Conference “Protection of Materials and Structures from the LEO Space Environment”,” Arcachon, France, 5-9 June, 2000. NASA ENVIRONET, Internet Database Van Eesbeek, M., Levadou F., Berthoud, L., “Interaction Between Environmental Models and Materials Evaluation”, in Proceedings of ESA Environmental Modeling for Space-based Applications Symposium, Noordwijk, Netherland, 1996; ESA, 1996, 23-28. Van Krevelin, D. W, Properties of Polymers, Elsevier, 1990. Biserano, J., Predictions of Polymer Properties, Marcel Dekker Inc., 1993. Tennyson, R. C., “Atomic Oxygen Effects on Polymer-Based Materials,” Canadian Journal of Physics, 69, 8 and 9 (1991), 1190-1208. Hilado, C., Flammability Handbook for Plastics, Technomic, 1990. Zaikov, G.E. (ed.), Flammability of Polymeric Materials, Nova Science Publishers, 1990. Landrock A., Handbook of Plastics Flammability and Bombustion Toxicology, Noyes Publishers, 1983. Book of Annual ASTM Standards, “Standard Test Method for Measuring the Minimum Oxygen Concentration to Support Candle-like Combustion of Plastics (Oxygen Index)” , D 2863-91, 1991. Report “ Prediction of Erosion of Polymer-based Materials by Atomic Oxygen in LEO ”, Integrity Testing Laboratory Inc., Toronto, Canada; GRC Contract C-72917-G, 1998.

This page intentionally left blank

COMPARATIVE STUDY OF LOW ENERGY C AND O ATOMS IMPACT IN A HYDROCARBON SURFACE

M. MEDVEDEVA AND B. J. GARRISON Department of Chemistry, The Pennsylvania State University, University Park, PA 16802

Abstract Molecular dynamics simulations of the interaction of low energy (1-10 eV) carbon atoms with a hydrogenated diamond (110) surface are performed. The goal is to examine similarities and differences in interaction of carbon and oxygen atoms with hydrocarbon surfaces. Comparison of the results obtained with the experimental data appears that scattering of C atoms from the hydrogenated diamond surface is qualitatively similar to O+ scattering from a saturated hydrocarbon liquid, squalane. 1. 0 Introduction Various structural materials with carbon-carbon and carbon-hydrogen bonds are used on spacecrafts functioning at Low Earth Orbit (LEO). During a space mission, complex chemical reactions occur on the material surface in result of an atmospheric atomic oxygen attack. These reactions cause an oxygen atomic-induced degradation of materials on the spacecrafts [1]. An abstraction of a hydrogen atom to form a hydroxyl radical is a major initial reaction for the interaction of O(3P) with hydrocarbon molecules [2,3]. Another reactive event occurs when oxygen atoms attack aromatic hydrocarbon molecules. An addition of an oxygen atom to a benzene ring with subsequent elimination of a hydrogen has been observed in crossed beam experiments [4] as well as in the computational study of the reaction C6H6 + O(3P) [5]. A scission of carbon-carbon bonds because of the O(3P) attack of ethane and several higher alkanes has been predicted by ab initio techniques [6]. The threshold calculated for the breaking of the C-C bonds due to the reaction of O(3P) is 1.7 – 2 eV. This value is significantly less then the initial energies of oxygen in LEO (~ 5 eV). Reaction events similar to gasphase reactions have been presented for interactions of oxygen atoms with a hydrocarbon surface in [7,8]. Both direct inelastic oxygen scattering and a H-atom abstraction to produce volatile OH and H2O have been manifested as the main initial pathway at O+ scattering from a saturated hydrocarbon liquid, squalene. More complex reactions arise on the surface under the continuous oxygen atom attack. Certainly, radical sites created by the initial atom are open to further reactions with incoming O

527

528 atoms. Every reaction event is influenced by the incident oxygen flux and angle, initial oxygen energy, and surface temperature [1,9]. The entire picture of an oxygen atom influence the surfaces with C-C and C-H bonds is far from being completed, in spite of a considerable number of studies. Development of computational models that simulate the oxygen atom interactions with various materials can potentially provide predictions of degradation pathways of materials under O-atom attack in space. Molecular dynamics (MD) simulations have been demonstrated to be a powerful tool for studying chemical reactions at hydrocarbon surfaces [10]. The advantage of MD simulations is that many different events can be traced for large systems relatively quickly. Detailed MD studies, however, are complex because no empirical or semi-empirical potential energy surface exists that is capable of describing chemical processes resulting from oxygen attack on the surface. The Brenner potential and its modification [11,12] do successfully model hydrocarbon reactions but without oxygen atoms. Recently, the comparative studies of interactions of atomic Cl and O with the squalene [13] and the reactivity of C and O atoms with benzene molecules [14] have been carried out. The interactions of Cl and O atoms with the continuously refreshed liquid squalene surface exhibit very similar dynamics. The reactivity of C and O atoms with benzene, however, is remarkably different. Until now, no studies have been done to compare the effects of C and O atoms on surfaces with the C-C and C-H bonds. To explore influence the carbon atom impact on a hydrocarbon surface as well as to compare it with the known data for the O-atom attack, MD simulations of the interaction of carbon atoms with a hydrogenated diamond (110) surface are initiated. This structure, in spite of its simplicity, exposes C-C and C-H bonds. At the first step, we explore what kinds of events occur under impact of a C atom with different initial energies and angles on the hydrocarbon surface maintained at two temperatures. Dependence of reactive events on the initial energy is determined. Then, the energy distributions of scattered C atoms are calculated at different initial energies and final angles. The dependence of average final energies of scattered C-atoms on the final angles corresponding to different initial angles is obtained. The results are compared with the experimental data by [8] for O-atom scattering. 2. 0 Method MD simulations of scattering of carbon atoms with initial energies Ei = 1 - 10 eV from a hydrogenated diamond (110) surface are performed. The semi-empirical many-body Brenner potential [11] is used to describe the interaction between C-C, H-H and C-H atoms. The computational cell is composed of carbon atoms arranged in a diamond lattice of 8 layers thick, with 32 atoms per layer, and 32 hydrogen atoms attached to the top-layer carbon atoms. The bottom layer is held fixed to simulate an infinite substrate. The motions of the top three layers of C atoms and all H atoms are determined from the forces of the appropriate potential. Friction and stochastic forces via the generalized Langevin equation method for maintaining a substrate temperature at a constant magnitude of 0 K and 300 K are added to the remaining four layers [10]. Periodic lateral boundary conditions are applied. The trajectories of atoms in the system last for

529 1 ps, at which time the configurations appeared to be stable. For each impact parameter set (initial energy, incidence angle, surface temperature), 2700 trajectories are calculated. 3. 0 Results and Discussion 3.1 REACTION EVENTS To study possible reactions on the hydrocarbon surface due to the C atom attack, the events occurring under impact of C atom at Ei = 5 eV and incident angles θi = 0°, 30º, 45º, and 60º to the surface normal and surface temperatures, Ts, of 0 K and 300 K are studied. The raw numbers of possible events are shown in Table 1. The events are sorted according to the outgoing species. Inelastic scattering of C atoms and reactions of abstraction with the outgoing species of CH and CH2 are observed. TABLE 1. Events for 5 eV carbon atom impact in the hydrogenated diamond (110) surface.

Surface temperature

Initial angle

Scattering

CH abstraction

CH2 abstraction

H elimination

0K

0° 45° 0° 30° 45° 60°

996 270 951 1498 2165

31 102 269 25

3 5 -

9 -

300 K

Sticking to the surface 2700 1704 2399 1635 928 510

The same events are observed in the experiments for oxygen and chlorine ions scattering from the squalene [8, 13]. Moreover, a reaction of H atom elimination occurs at θi = 30º and Ts = 300 K in the simulations. This reaction is observed for the O(3P) atom interaction with C6H6 in Refs. [4,5]. Table 1 displays that a large number of trajectories where the initial C atoms are accommodated completely by the surface without any outgoing particles. To identify the situations of these incoming carbon atoms the distribution of all primary C atoms vs Z-coordinate, normal to the surface, in 1 ps after impact are calculated for following initial parameters: Ei = 5 eV, θi = 45º, and Ts = 300 K. The results are plotted in Figure 1.

530

Figure 1. Distribution of primary carbon atoms vs the normal to the surface in 1 ps after impact.

The distribution consists of two parts. One of them comprises C atoms escaped from the surface at distances more than 10 Å. These atoms no longer interact with the surface. A fraction of them has been inelastically scattered from the surface. Others are involved in the formation of CH and CH2 compounds. Atoms staying on the surface form another part of the distribution. The Z-coordinate distribution of these atoms is shown in the inset in Figure 1. The dashed straight lines schematically point out the hydrogen atoms and the first layer of surface carbon atoms. It is seen that the particles do not penetrate deeper than the first carbon layer. Various reaction events can occur on the surface with these carbon atoms including insertion in C-C and C-H bonds, adsorption of the impinging atom with one or two H atoms abstraction, and creation of the radical sites. These events affect the surface, changing its chemical properties and, as consequence, influence the interaction of subsequent hitting atoms with the surface. To determine the influence of the C atom initial energy on the reaction events mentioned in Table 1 the probability of inelastic scattering, abstraction, elimination and sticking to the surface are calculated at θi = 0º and Ts = 300 K. The results are displayed in Figure 2.

531

Figure 2. Probability of reaction events vs the initial energy for Ts = 300 K.

Inelastic scattering and C atom sticking to the surface are the major events at Ei < 6 eV. These processes interplay with each other. There are no sticking events for an initial energy less than 2 eV. In other words, there an energy threshold exists for sticking of C atoms to the surface. Only at Ei > 4 eV does this type of event becomes dominate. A small contribution of abstraction events occurs at Ei >3 eV and elimination ones for energy of more than 8 eV are observed. Thus, the initial energy influences strongly the material reactivity for C atom impacts. 3.2 ENERGY AND ANGLE DISTRIBUTION OF SCATTERED CARBON ATOMS The energy distributions of scattered C atoms as a function of both the initial and final angles, as well as the dependence of these distributions on the initial energy are calculated and compared with the experimentally obtained distributions of O+ scattered from the squalene surface in the experiments from the Minton’s group [8]. The initial conditions of scattering used are approximately the same as in the experiment [8], Ei = 3 eV, and 5 eV, and θi = 60º. The scattered atoms are detected at θf = 45º and 70º in the calculations. The surface temperature is maintained at 300 K. The normalized energy distributions of scattered C atoms are presented in Figure 3 (straight lines). The dashed lines represent the experimental measured distributions of hyperthermal scattered O

532 atoms provided by Minton’s group. It can be seen that the scattering of atomic oxygen from the squalene and C atoms from the hydrogenated diamond (110) surface is very similar. In the both cases, the shapes of distributions depend on the initial energy and the final angles. The large final energy of scattered atoms is detected for larger Ei and more glancing scattering (θi = 60º, θf = 70º). The squalene surface has a larger diversity of sites than hydrogenated diamond surface thus it is not surprising that the experimental distributions are broader.

Figure 3. Energy distributions of C atoms scattered from the hydrogenated diamond (110) (straight lines) and O+ ions scattered from the squalane (dashed lines). Experimental data are from [8].

Another argument of the similarity of the O and C atoms scattering from the hydrocarbon surfaces is presented in Figure 4.

533

Figure 4. Average energies of scattered C and O atoms as a function of the final angles. Experimental data are from [8].

The average final energies of scattered C atoms at each final angle, are calculated at θi = 30º, 45º, and 60º and plotted as a function of θf in the lower panel of Figure 4. The initial energy is 5 eV. The experimental data from Ref. [8] for scattered O-atoms are shown in the upper panel. One can see that the average energy of both scattered carbon atoms and oxygen ones increases with increasing a final angle for all the incident angles. The maximum of energy deposited into the surface is observed at θi and θf closer to the surface normal for both impinging atoms.

534 4. 0 Conclusions Results of the MD simulations indicate that the interaction of C atoms with the hydrogenated diamond (110) surface reproduces qualitatively the O atom interactions with hydrocarbon surfaces. All reactive events occurring at the O atom attack on the hydrocarbon molecules and surfaces are detected in our simulations with C atoms. Inelastic scattering, reactions of the abstraction to form CH (OH) and CH2 (H2O), and H-atom elimination with an insertion of incident C (O) atom to the surface are observed. The probability of C atom reactions and, as consequence, the material reactivity, depends on the initial energy, incident angle, and surface temperature. Comparison of the calculated energy distributions of scattered C atoms with the experimental data at O atom scattering concludes that scattering dynamics of both atoms is qualitatively similar. 5. 0 Acknowledgements The financial support of the Multidisciplinary University Research Initiative Program of the Air Force Office of Scientific Research is gratefully acknowledged. Additional computational resources were provided in part by the IBM Selected University Resource Program and the Center of Academic Computing of Penn State University. The authors thank Timothy K. Minton and Jianming Zhang for providing experimental data. 6.0 References 1.

2. 3.

4. 5.

6. 7.

Dooling, D. and Finckenor, M.M.: Material selection guidelines to limit oxygen effects on spacecraft surface, NASA TP-209260, National Aeronautics and Space Administration, Washington, DC, (1999). See also references therein. Lin, M.C.: Dynamics of oxygen atom reactions, in Lawley, K. P. (ed.), Advances in Chemical Physics: Potential Energy Surfaces, John Wiley & Sons, Ltd., Chichester, Great Britain, (1980). Minton, T.K. and Garton, D.J.: Dynamics of atom-oxygen-induced polymer degradation in low Earth orbit, in Dressler, R. A. (ed.), Advanced Series in Physical Chemistry, World Scientific, Singapore, (2000). Sibener, S.J., Buss, R.J., Casavecchia, P., Hirooka, T., and Lee, Y.T.: A crossed molecular beams investigation of the reactions O(3P) + C6H6, C6D6, J. Chem. Phys. 72 (1980) 4341-4349. Hodgson, D., Zhang, H.-Y., Nimolos, M.R., and McKinnon, J.T.: Quantum chemical and RRKM investigation of the elementary channels of the reaction C6H6 + O(3P), J. Phys. Chem. A 17 (2001) 4316-4327. Gindulytè, A., Massa, L., Banks, B. A., and Rutledge, S. K.: Can hydrocarbon chains be disrupted by fast O(3P) atoms?, J. Phys. Chem. A 104 (2000) 9976-9982. Garton, D.J., Zhang, J. and Minton, T.K.: Atomic oxygen interaction with saturated hydrocarbon surfaces: probing polymer degradation mechanisms, Proceedings of the 8th International Symposium on Material in a Space Environment and 5th International Conference on Protection of Materials and Structure from the LEO Space Environment, France, (2000).

535 8.

Zhang, J., Garton, D.J., and Minton, T.K.: Reactive and inelastic scattering dynamics of hyperthermal oxygen atoms on a saturated hydrocarbon surface, J. Chem. Phys. (2002), in press. 9. Koontz, S.L., Albyn, K., and Leger, L.J.: Atomic oxygen testing with thermal atom systems: a critical evaluation, J. of Spacecraft and Rockets 28 (1991) 315-323. 10. Garrison, B.J., Kodali, P.B.S., and Srivastava, D.: Modeling of surface processes as exemplified by hydrocarbon reactions, Chem. Rev. 96 (1996) 1327-1341. 11. Brenner, D.W.: Empirical potential for hydrocarbons for use in simulating the chemical vapor deposition of diamond films, Phys. Rev. B 42 (1990) 9458-9471. 12. Stuart, S.J, Tutein, A.B., and Harrison, J.A.: A reactive potential for hydrocarbons with intermolecular interactions, J. Chem. Phys. 112 (2000) 6472-6486. 13. Garton, D.J., Minton, T.K., Alagia, M., Balucani, N., Casavecchia, P., and Volpi, G.G.: Comparative dynamics of Cl(2P) and O(3P) interactions with a hydrocarbon surface, J. Chem. Phys. 112 (2000) 5975-5984. 14. Hahndorf, I., Lee, Y.T., Kaiser, R.I., Vereecken, L., Peeters, J., Bettinger, H.F., Schreiner, P.R., Schleyer, P.v.R., Allen, W.D., and Schaefer III H. F.: A combined crossed-beam, ab initio, and Rice-ramsperger-Kassel-Marcus investigation of the reaction of carbon atoms C(3Pj) with benzene, C6H6(X1A1g) and d6-benzene C6D6(X1A1g), J. Chem. Phys. 116 (2002), 3248-3262.

This page intentionally left blank

A DIRECT TRAJECTORY DYNAMICS INVESTIGATION OF FAST O + ALKANE REACTIONS RONALD Z. PASCUAL AND GEORGE C. SCHATZ Northwestern University, Evanston, Illinois, USA 60208-3113 DONNA J. GARTON Montana State University Bozeman, MT 59717

Abstract Direct trajectory calculations of collisions of fast O(3P) with methane, ethane and propane molecules are presented. We find that OH+alkyl and H+alkoxy formation are the major reactive channels for both systems, with OH formation being more important are low energy (2-3 eV) and H formation being more important at high energy (4-6 eV). For ethane and propane, C-C bond cleavage and other processes such as direct water formation, are found, but the cross sections are smaller than for OH+alkyl and H+ alkoxy.

1. 0 Introduction The mechanism of erosion of polymers in low Earth orbit (LEO) conditions has long been the subject of speculation [1]. Even for simple hydrocarbon polymers, the initial products of reaction with 5 eV O(3P) atoms are not known, and there has been uncertainty as to whether intersystem crossing is essential to the formation of oxy radicals or to carbon-carbon bond cleavage. In order to provide new insight about the reaction of high velocity O(3P) atoms with hydrocarbon polymers, we have performed direct dynamics simulation (i.e., density functional theory calculations on the fly in trajectory simulations) of O(3P) collisions with methane, ethane and propane. There have been several past experimental and theoretical studies of atomic oxygen atom reactions with aliphatic hydrocarbons, both in the triplet ground state [3, 4], and the singlet excited state [5, 6]. All but one of these considered only low energies, where OH is the only product. The recent work of Massa and coworkers [2] demonstrated that the barriers for carbon-carbon bond cleavage are low enough to make this process feasible under LEO conditions, but no dynamical calculations were

537

538 presented. The importance of other possible products, such as oxy radical formation, is unknown. 2. 0 Computational details Direct trajectory density functional theory calculations were done at two relative translational energies for each reaction. The translational energies Erel were taken to be 2.36 and 3.92 eV for O + methane, 3.01 eV for O + ethane, and 3.46 and 5.75 eV for O + propane. The initial separation of the O atom from the hydrocarbon was adjusted such that the closest distance between the O and any atom in the hydrocarbon was 4.0 bohr. The maximum impact parameter used was 7.0 bohr for methane and ethane, 8.0 bohr for propane. To save computer time, collisions were prescreened by assuming straight line motion so as to remove any trajectory where the closest O/alkane distance was greater than 2.0 bohr. This procedure eliminated about half the trajectories. To test this prescreening, we ran 200 trajectories for O + methane at 2.36 eV in which only trajectories outside the 2.0 bohr cutoff were calculated. We found only four reactive trajectories, all of them corresponding to OH formation, corresponding to a 10% effect on the cross section. In studying the O + methane reaction, 501 trajectories were initiated at Erel = 2.36 eV, 289 of which were explicitly calculated. At Erel = 3.92 eV, these numbers were 403 and 250, respectively. For O + ethane, these numbers were 300 and 205 for Erel =3.01 eV, and for O + propane, these numbers were 316 and 214, respectively for Erel = 3.46 eV, and 400 and 243 for Erel = 5.75 eV. We used density functional theory (B3LYP/6-31G*) for the direct trajectories with the GAMESS [7] electronic structure package. The alkane molecules were started with zero point energy in all cases, and the usual Monte Carlo sampling of molecular coordinates and velocities was performed. In addition, the Q-Chem electronic structure package [8] was used to determine equilibrium structures and energies. 2.1 REACTION PATHS AND PROFILE In the present direct trajectory calculations, we observed three reaction paths for the O + methane (Eq. 1-3), five for ethane (Eq. 4-8) and twelve for O + propane (9-20) as follows: O(3P) + Methane Æ OH + CH3

(1)

Æ H + OCH3

(2)

Æ 2H + OCH2

(3)

3

O( P) + Ethane Æ OH + C2H5

(4)

Æ H + OC2H5

(5)

Æ H2O + CH-CH3

(6)

Æ OCH3 + CH3

(7)

539 Æ 2H + OCHCH3 O(3P) + Propane Æ OH + C3H7

(8) (9)

Æ H + OC3H7

(10)

Æ H2O + CH3-C-CH3

(11)

Æ OCH3 + CH2-CH3

(12)

Æ OCH2-CH3 + CH3

(13)

Æ H + CH3-COH-CH3

(14)

Æ 2H + CH3-CO-CH3

(15)

Æ H + CH3-CH2-O-CH2

(16)

Æ H + OH + CH2=CH-CH3

(17)

Æ H + CH2-CH2-CH2OH

(18)

Æ OCH2 + 2CH3

(19)

Æ O + CH2CH3 + CH3

(20)

Note that in processes (3), (8), (15) and (19), a singlet aldehyde or ketone is produced, while (6) and (11) lead to a triplet methylene containing radical. For O + methane, channel (1) is endoergic by 0.27 eV (based on B3LYP, and including zero point corrections), while channels (2) and (3) are 0.53 and 1.62 eV uphill, respectively. These thermochemistry predictions are typically within 0.2 eV of the experimental values. Barriers for channels (1) and (2) are 0.30 and 1.83 eV. For O + ethane, channels (4) and (5) are 0.08 and 0.27 eV uphill, with barriers of 0.10 and 1.83 eV, respectively. For O + propane, the endothermicities of channels (9) and (10) are 0.09 and 0.41 eV, for reaction at the primary carbon, and the corresponding barriers are 0.09 and 1.80 eV, respectively. The corresponding secondary carbon atom reactions, which are also lumped into channels (9) and (10), are about 0.2 eV downhill from their primary carbon atom counterparts. The carbon-carbon bond cleavage channels, (12) and (13) are exothermic by 0.34 and 0.40 eV, with barriers of 1.48 and 2.17 eV, respectively. These estimates of carbon-carbon bond cleavage barriers are similar to those noted previously.[2] 2.2 REACTION CROSS SECTIONS The reactive cross sections for O + methane are presented in Table 1. Here we see that at 2.36 eV, the dominant path involves OH + alkyl production (channel 1). At 3.92 eV, the OH + alkyl cross section has decreased, while H + methoxy production is now dominant. The corresponding results for O + ethane, presented in Table 2, show cross sections for OH + ethyl (path 4) and H + ethoxy (path 5) that are each more than double their methane counterparts at a similar energy. In addition, we see small but nonzero cross sections for direct production of water (path 6), for breaking carboncarbon bonds (path 7) and for producing acetaldehyde (path 8). The O + propane results in Table 3 show that OH + propyl (path 9) is dominant at 3.46 eV, with a cross section

540 that is about triple its methane counterpart. At 5.75 eV, the OH production cross section has dropped while H + propoxy has risen significantly. Indeed the total H atom production cross section, including processes 9 and 14-18 accounts for 57% of the total at this energy. The cross sections for direct water production (process 11), and for carbon-carbon bond cleavage (processes 12 and 13) are smaller, but still significant, and small cross sections for the production of three fragments in a single collision are also found. TABLE 1. Reaction cross sections (in bohr2) for O(3P) + methane.

Reaction Path 1 2 3

Erel = 2.36 eV Cross section 5.5 ± 0.9 0.7 ± 0.2 0.4 ± 0.4

Erel = 3.92 eV Cross section 4.6 ± 0.9 5.5 ± 0.5 0.3 ± 0.1

TABLE 2. Reaction cross sections (in bohr2) for O(3P) + ethane.

Reaction Path 4 5 6 7 8

Erel = 3.01 eV Cross section 12.2 ± 2.3 2.1 ± 0.7 0.5 ± 0.5 0.2 ± 0.2 0.1 ± 0.1

TABLE 3. Reaction cross sections (in bohr2) for O(3P) + propane.

Reaction Path 9 10 11 12 13 14 15 16-20

Erel = 3.46 eV Cross section 18.6 ± 2.2 3.0 ± 0.8 0.4 ± 0.4 0.3 ± 0.3 0.2 ± 0.2

Erel = 5.75 eV Cross section 12.0 ± 1.8 10.7 ± 1.3 0.9 ± 0.6 0.9 ± 0.4 0.3 ± 0.2 0.4 ± 0.2 0.2 ± 0.1 0.1 ± 0.1 – 0.2 ± 0.2

3. 0 Conclusion The present calculations show that fast oxygen atoms can participate in many reactive processes with small alkanes. OH + alkyl radical is the most important product at low energies, with a cross section that decreases with increasing energy for the energies that we considered. H + alkoxy radical formation is especially important at high energies, and carbon-carbon bond cleavage is also possible. These results provide important

541 benchmarks for the development of LEO erosion mechanisms. In addition we will use them to interpret crossed molecular beam experiments currently underway at Montana State. 4. 0 Acknowledgements This research was supported by AFOSR MURI Grant F49620-01-1-0335. We thank Tim Minton (Montana State) for valuable discussions. 5. 0 References 1.

Cohen, L. K. (1993) A Lower Bound on the Loss of Graphite by Atomic Oxygen Attack at Asymptotic Energy, J. Chem. Phys. 99, 9652-9663.

2.

Gindulytơ, A., Massa, L., Banks, B. A., and Rutledge, S. K. (2000) Can Hydrocarbon Chains Be Disrupted by Fast O(3P) Atoms?, J. Phys. Chem. A 104, 9976-9982.

3.

Liu, X., Gross, R. L., Suits, A. G. (2002) Differential Cross Sections for O(3P) + Alkane reactions by

4.

Sweeney, G. M., Watson, A., McKendrick, K. G. (1997) Rotational and Spin-Orbit Effects in the

Direct Imaging, J. Chem. Phys. 116, 5341-5344. Dynamics of O(3Pj) + hydrocarbon Reactions 106, 9172-9189. 5.

Lin, J. J., Shu, J., Lee, Y. T., Yang, X. (2000) Multiple Dynamical Pathways in the O(1D) + CH4 Reaction: A Comprehensive Crossed Beam Study, J. Chem. Phys. 113, 5287-5301.

6.

Shu, J., Lin, J. J., Lee, Y. T., Yang, X. (2001) A Crossed Beam Study of the O(1D) + C3H8 Reaction:

7.

Schmidt, M. W., Baldridge, K. K., Boatz, J. A., Elbert, S. T., Gordon, M. S., Jensen, J. H., Koseki, S.,

Multiple Reaction Pathways, J. Am. Chem. Soc. 123, 322-330. Matsunaga, N., Nguyen, K. A., Su, S. J., Windus, T. L., Dupuis, M., Montgomery, J.A. (1993) GAMESS version = 25 June 2001 from Iowa State University, J. Comput. Chem. 14, 1347-1363. 8.

J. Kong, C. A. White, A. I. Krylov, C. D. Sherrill, R. D. Adamson, T. R. Furlani, M. S. Lee, A. M. Lee, S. R. Gwaltney, T. R. Adams, C. Ochsenfeld, A. T. B. Gilbert, G. S. Kedziora, V. A. Rassolov, D. R. Maurice, N. Nair, Y. Shao, N. A. Besley, P. E. Maslen, J. P. Dombroski, H. Dachsel, W. M. Zhang, P. P. Korambath, J. Baker, E. F. C. Byrd, T. Van Voorhis, M. Oumi, S. Hirata, C. P. Hsu, N. Ishikawa, J. Florian, A. Warshel, B. G. Johnson, P. M. W. Gill, M. Head-Gordon, J. A. Pople, (2000) Q-Chem, Version 2.0, Q-Chem, Inc., Export, PA.

This page intentionally left blank

MATHEMATICAL SIMULATION METHODS TO PREDICT CHANGES OF INTEGRAL AND SPECTRAL OPTICAL SURFACE CHARACTERISTICS OF EXTERNAL SPACECRAFT MATERIALS AND COATINGS V.N. VASILIEW, A.V. GRIGOREVSKIY, Y.P. GORDEEV Closed joint-stock company “Institute “Kompozit-Test” 4, Pionerskay street, 141070 Korolev, Moscow region, Russia Phone: + 7 095 513 20 20 Fax: + 7 095 513 20 75 E-mail: [email protected] [email protected]

Abstract The methods of statistical prediction of the change of optical properties of materials and coatings applied on spacecraft surfaces with regard to data of laboratory and in-flight experiments are presented. The prediction methods are based on mathematical simulations that describe the change of optical characteristics of spacecraft materials under the influence of environmental factors, namely, solar UV-radiation, atomic oxygen, protons and electrons of the radiation belts, solar wind and flares, affecting materials properties both separately or simultaneously. The methods of analysis of experimental data by means of structural and parametric identification of mathematical models of materials damage, the methods of predicting the changes of materials optical properties in space environment as well as results of prediction for specific spacecraft orbits are presented. 1. 0 Introduction The optical characteristics of materials and coatings on external surfaces of spacecraft are integral coefficient of absorption of solar radiation As, and the coefficient of thermal radiation, or thermal emission ε. These characteristics are some of the main factors, which determine the absorption of external radiative thermal fluxes, and, therefore, the required areas of refrigerating radiators and, finally, temperature of spacecraft structural elements. The spectral reflectance of materials determine the spectrum and intensity of the reflected solar energy, and the thermal radiation of spacecraft. This value is used for the analysis of reflective characteristics of spacecraft, which are determined by groundbased telescopes. This is associated with the problem of monitoring of the spacecraft thermal conditions during its orbital flight [1].

543

544 Therefore the thermo-optical characteristics of materials and coatings of spacecraft need to be taken into account both in the designing of spacecraft thermal control system and the analysis of its reflective characteristics. The numerous results of orbital experiments and laboratory tests testify that under influence of the damaging factors of environmental space the optical characteristics of materials and coatings of spacecraft have undergone changed (degraded). The value of change of coefficients As and ρλ depends on the type and intensity of influencing factors of space environment which are determined by height and inclination of a spacecraft orbit. The UV-radiation of the Sun influences the materials properties on all types of orbits, atomic oxygen is important at height not more than ~900 km, protons and electrons flows-in radiation belts of the Earth at height 100060000 km, solar wind is significant at height more than 60000 km. Besides the damaging factors of the space environment there is influence of temperature T, contamination under outgassing of polymeric materials of spacecraft external surfaces and jets of spacecraft engines. Contamination is one of the main factors which determined the change of optical characteristics of specular and metal coatings of type OSR, vacuum-deposited Al, photo-cell of solar arrays. The value of change of coefficient As and ρλ is determined in orbital or laboratory experiments. The orbital experiments have the large cost and long duration and are determined the change of optical characteristics of materials and coatings for a particular orbit. The laboratory experiments have the limited opportunities of imitation of some factors of environmental space-power spectra of protons and electrons of radiation belts, solar wind and burst, cyclic change of temperature, irradiation by UVradiation of the Sun and others. By these reasons the mathematical simulation of effects of the environmental factors on optical characteristics external materials and coatings of spacecraft were developed. 2. 0 Methods of Mathematical Simulation of Radiation Damage of Optical Characteristics of Spacecraft Materials and Coatings. The system for prediction of the change of coefficient As of materials and coatings of spacecraft in space environment is represented as follows:

& & As = F (a , x , t ) & & x = x (t )

(1)

& &

where F ( a , x , t ) -mathematical model of damage (MMD);

& a = {a i }-vector of parameters of MMD; & x = {x i }-vector of intensity of the environmental space and condition of operation

(for example, the temperature or the intensity of UV-radiation of Sun) which can vary during the orbital flight; t – time of operation, hours.

545 The solution of the equation (1) is conducted by a numerical method with the use of method of linear summation of damage (change of coefficient As). The type of MMD is determined by space environment namely height of spacecraft orbit [2,3]. 2.1 THE HEIGHT OF ORBIT IS 200-600 KM. In this space environment the materials and coatings of spacecraft have undergone influence of UV-radiation of Sun and atomic oxygen. The MMD is:

{

[

∆As = a 1 − exp − b0 E s′k1 exp(− k 2 / T )t sβ

]}

,

(2)

where a , b0 , k 1 , k 2 , β - parameters;

E s′ = 0...1 -irradiation of materials (coatings) by UV-radiation of Sun; T = T(t)-temperature, K; t-time of irradiation by Sun during orbit, hours.

[

K bl = ∆As bl ∆As 0 = 1 − exp − bbl ( h, H s′ )(t sh cos α ) β bl

]

,

(3)

where K bl - coefficient of bleaching of damage by atomic oxygen;

bbl , β bl - parameters; h – orbit altitude, km; Hs – equivalent exposure, ESH t sh - time of shadow during orbit, hours; α - angle of attack of atomic oxygen flow. The equation (2) is MMD by UV-radiation of Sun, equation (3)-mathematical model of bleaching of damage by atomic oxygen. 2.2 THE HEIGHT OF ORBIT IS 400-1000 KM (EXCEPT THE POLAR ORBIT) In this condition the materials and coatings of spacecraft have undergo influence only UV-radiation of Sun and MMD is type (2). 2.3 THE HEIGHT OF ORBIT IS 1000-60000 KM AND POLAR ORBIT H>600 KM It is the most complicated case for prediction of change of integral optical characteristics of materials and coatings of spacecraft in space environments. In the laboratory experiments we can not to imitate the power spectra of protons and electrons of radiation belts. In this connection, a method of simulation of influence of spectra of particles by mono-energetic radiation has been devised. This method is based on the concept of the effective particle density.

546 In case when the damage by protons and electrons is described by equation:

[

∆As p ,e = a{1 − exp − b0 E γ exp(− k / T )Ɏ β

]}

,

where a , b0 , k , γ , β - parameters;

Ɏ - flax of particles, sm-2;

t – time, sec.; T – temperature; the effective particle density is:

∞

ϕ p ,e eff =

³E 0

γ

β

dϕ dE dE γ

E0

,

β

where E 0 - the energy of mono-energetic particles;

γ,β dϕ dE

- parameters; - energy spectrum, particles ×sm-2×s-1×Mev-1×sterad-1.

and the damage under influence of particles with energy spectrum

dϕ

dE

is:

β º½ ­° ª ° § · γ ∆A = a ®1− exp « − b0 E exp ( − k / T )¨ ϕ p , e t ¸ » ¾ s p ,e eff ¹ » ° © °¯ «¬ ¼¿

(4)

In this case the materials and coatings of spacecraft in space environments have undergone influence of combined radiation UV+p+e and the MMD is: K ­ º½ ª · 2 K §ϕeeff ° ′ E s ¸ ϕ p 3 exp§¨ − K4 ·¸t β »°¾ ∆As = a®1 − exp«− b0K1 ¨ eff ¨ ϕ p,eeff ¸ T¹ » « © °¯ ¹ © ¼°¿ ¬

where a , b0 , γ ,

β , K1 , K 2 , K 3 , K 4

– parameters;

ϕp,e eff – effective particle density; T – temperature; t – time, hours. ; E s′ – irradiation of materials by UV-radiation of Sun;

,

(5)

547 The parameters of MMD types (2)-(5) are evaluated by results of laboratory studies using of multi-dimentional optimization methods [3]. Criterion of efficiency of a developed technique of prediction is comparison of the values, received with its help, with the results of orbital experiments. Fig.1 presents the comparison of values received on base of laboratory results with data of orbital tests for geostationary orbit on spacecraft "Ecran", "Gorizont". Similar results for the space station "Salut-7"[1] are presented in Fig.2.

Figure 1. Comparison of orbital tests of enamel KO-5191 with prediction of variation of coefficient As on base of laboratory experiments for geostationary orbit 1 - orbital tests of spacecraft "Ecran", "Gorizont"; 2 - estimation of ∆As (UV +p +e ).

Figure 2. Comparison of orbital tests of enamel KO-5191 with prediction of variation of coefficient As on base of laboratory experiments for low orbit: 1 – orbital tests of space station "Salut-7"; 2 - estimation of ∆A (UV + atomic oxygen).

548 It is clear, that the results of prediction on base the laboratory tests are in good agreement with the results of orbital experiments. 3. 0 Methods of Prediction the Change of Spectral Reflectance of Spacecraft Materials and Coatings The methods of prediction of the changes of spectral reflectance is based on results of estimating the change of integral coefficient ∆As in space environments. If the spectrum of Sun to divide on the region with same energy we have: n

∆ρ =

¦ ∆ρ λ

i

i =1

n

,

(6)

where n - number of the region. ρλ

1.0 0.8 0.6 0.4 0.2 0

UV 0.2 – 0.4 Mkm - Basic value

Visible 0.4 – 0.8 IR 0.8 – 2.5 Mkm mkm After 5 year operation

Figure 3. Estimation of change of spectral reflectance of enamel AK-573 on geostationary orbit

The prediction of the value of change ρλ for the enamel AK-573 (ZnO + resin KO-08) on the geostationary orbit is presented on Fig.3. From this figure follows, that the maximum change of spectral reflectance occurs in UV- and visible ranges of spectrum, whereas in IR range of spectrum the change is insignificant.

549 If we know the change of coefficient As in space environments ∆Assp and the change of spectral reflection and coefficient ∆As in laboratory or in orbital tests ∆ρ λl ,

∆Asl

then:

k= and:

∆Assp ∆Asl

∆ρ λl =

=

∆ρ λsp ∆ρ λl

∆ρ λsp k

(7) (8)

4. 0 Conclusion It is proposed the mathematical models for the description of the change of optical characteristics of materials and coatings of spacecraft exposed to the environmental space factors during long operation. The results of the prediction received on base of laboratory tests and proposed mathematical models are in good agreement with results of orbital experiments of thermal control coatings. 5. 0 References 1. 2. 3.

Chaburov P.N., Vasiliev V.N. et al. (1997), Spacecraft temperature monitoring by ground based telescopes. Proceeding of the Sixth European Symposium on Space Environmental Control System, Noordwijk, The Netherlands, (ESA SP-400, August 1997). Vasilyev V.N et al. (1992), Manual for designers on providing thermal regimes. vol.4. Central Institute of Machinebuilding, Korolev, Russia. RD 92-0284-91 (1991) The methodical instruction. “Methods of tests of polymeric materials and the thermal control coatings”. Central Institute of Machine-building, Korolev, Russia.

This page intentionally left blank

SUBJECT INDEX atomic oxygen- degradation of materials 62, 388, 419, 420 atomic oxygen- effective fluence 396 atomic oxygen- erosion 65, 66, 197, 237, 240, 241, 317, 347, 419, 420, 433, 434, 449, 484, 503

A absorptance 16, 53, 73, 81, 82, 87-94, 98, 114-119, 140, 149, 153, 177, 178, 193, 195, 197, 199-201, 213, 246, 249, 313, 324, 347, 392, 401, 402, 404-406, 454, 499, 543 acrylic 60, 217, 218, 221, 223-227, 231, 232, 236, 238, 462, 467-471, 473 activation energy 110, 304, 351, 356, 384, 388, 397 adhesion 208, 213, 231, 307, 308, 311, 313, 316, 454, 456, 459 adhesive 142, 195, 196, 199, 205, 206, 232, 462, 467, 472, 473, 474, 492 algorithm 185, 503, 507, 519 alkanes 302, 537, 538 Al-Li alloy 414-416 Aluminum 58, 64, 69, 70, 231 Aluminum foil 231 angle dependence 391, 393, 394, 395 angular distribution 285-287 anodization 60, 64, 152, 326 antenna 6, 49, 60, 62, 63 anti-static 435, 501 array 32, 33, 34, 35, 36, 37, 38, 40, 57, 60, 63, 66, 67, 237, 243, 467, 472, 478, 517 atmosphere 2, 3, 10, 12, 14-18, 24, 25, 26, 115, 141, 198, 257, 300, 312, 319, 322-324, 341, 387, 389, 436, 437, 462, 463, 464 atmospheric plasma spraying 452, 456 Atomic Force Microscopy (AFM) 177, 216, 308-314, 316, 379, 381, 383, 385, 445, 447, 478 atomic oxygen (AO) 5 ,15, 17, 31, 32, 53, 54, 56, 62, 63, 64, 65, 66, 67, 73, 156, 160, 163, 183, 184, 191, 193, 194, 195, 197, 198, 200, 201, 203, 204, 205, 211-216, 235-243, 260, 272, 283, 284, 289, 291, 292, 295, 296, 299, 300, 305, 317, 319, 324, 326, 335, 336, 340, 345, 346, 349, 351-358, 359-366, 379-385, 387390, 391-393, 395, 396, 419-421, 423-427, 429, 431-434, 435, 436, 439, 440, 441, 443, 444, 448, 449, 451, 454, 455, 456, 458-460, 462, 467, 474-481, 483, 484, 485, 488, 491-493, 499-501, 503, 504, 509, 511, 515, 517, 527, 532, 537, 543545, 547 atomic oxygen- attack 54, 236, 237, 239, 241, 380, 391, 396, 443, 449, 450, 458, 527 atomic oxygen- bombardment 359

B backbone 293, 300, 439, 460, 492, 497, 501 baffle 422 barrier to reaction 283 binder 132, 144, 197, 211, 214, 292, 293, 422, 429, 431 blends 181, 491, 517 bond cleavage 537, 539, 540 BRDF 140, 152, 153 C Canadarm 2, 5, 291-293, 295-298 Canadian Space Agency 1, 4, 5, 6, 150, 359, 360, 366 carbon-carbon composite 516 charged particle(s) 7, 11, 19, 21, 26, 27, 52, 53, 82, 86, 91, 96, 110, 130, 132, 183, 322, 325, 451 charging 7, 21, 22, 32, 33, 34, 35, 36, 37, 38, 39, 40, 53, 54, 55, 73, 81, 86, 89, 206, 213, 425 clean room 257, 258 cleanliness 257, 258, 260, 262 coatings (also see protective coatings) 34, 63, 65, 66, 74, 92, 139, 147, 172, 183, 184, 193, 194, 197, 200, 201, 211, 217, 218, 221, 228, 232, 233, 235, 238, 239, 240, 242, 294, 310, 316, 319, 322, 327, 330, 333, 359, 360, 390, 443-449, 451, 452, 454458, 460, 492, 503, 543-546, 549 collision energy 283, 285, 306, 391 composite materials (see also CFRO composites, glass epoxy, carbonepoxy) 232, 249, 263, 264, 268-270, 271, 272, 276, 279-281, 420, 451, 456, 457, 459, 516 contaminants 21, 66, 260, 313, 367, 368, 406 conversion 495 coronal mass equation (CME) 12 cosmic rays 12, 23, 26, 27, 28, 29, 31, 32, 44, 45, 49, 50, 51, 52 coverglass 462-467 crack(s) 67, 109, 134, 156, 159, 206, 207, 208, 209, 212, 213, 217, 221, 223, 228, 232, 239, 240, 241, 245-

551

552 249, 251, 253, 271, 273, 279, 352, 503 creep 148, 272, 453 cross-linking 103, 106, 108, 109, 131, 132, 133, 135, 136, 137, 213 D defect(s) 92, 93, 97, 235, 238-242 database 47, 113, 114, 122, 139-147, 150, 152-154, 515, 516, 518, 519, 522 debris 5, 7, 8, 13, 14, 31, 32, 34, 35, 41, 64, 183, 195, 272, 280, 292, 296, 451, 460, 462, 463 degradation 7, 14, 16, 17, 22, 23, 26, 45, 53, 57, 60-64, 71, 82, 83, 90, 91, 92, 94, 105, 106, 109, 114, 120, 121, 131, 132, 135, 137, 140, 155, 156, 157, 158, 162, 163, 165, 166, 171, 180, 183, 193, 197, 198, 200, 201, 211, 214, 215, 216, 217, 221, 232, 271, 272, 279, 280, 294, 299, 300, 304, 305, 319, 322, 324 , 336, 340, 345, 347, 349, 359, 367, 368, 379, 387-390, 407, 414, 415, 418, 429, 431, 443, 447, 449, 462, 474, 475, 481, 488, 516, 527, 528, 534 dehydration 437 detector(s) 26, 53, 69, 124, 127, 146, 147, 152, 205, 210, 218, 265, 337341, 403, 404, 405, 463 diffusion coefficient 28, 247, 273, 278, 279 dose rate 32, 53, 99-110, 113, 115, 140, 141, 147, 148, 149, 160, 163 E Earth 1-5 Earth radius 19 EDS 203, 205, 210, 213, 215 elasticity 148, 232, 264, 266, 274, 473 electrochromic 483, 484, 485, 487, 488, 489 electron(s)10, 14, 21, 22, 26, 27, 31-33, 35-37, 40, 41, 44, 53-56, 78, 80-83, 86-89, 91-95, 98-100, 102, 103, 105, 110, 122, 123, 124, 126, 127, 130, 132, 140, 147, 148, 156, 157, 160, 191, 198, 203, 205, 212, 213, 214, 216, 217, 218, 221, 223, 278, 296, 303, 319, 322, 336, 380, 387, 388, 389, 390, 415, 425, 439, 454, 461, 483, 484, 489, 543-546 electron beam 53, 54, 160 electron cyclotron resonance (ECR) 191, 439, 483-485, 489 electron flux 40, 124, 126, 127, 216, 390

electron radiation 53, 81, 83, 86, 87, 89, 91, 94, 100, 102, 103 ellipsometry 183, 184, 185, 191, 307309, 312, 316, 483, 484, 488, 489 enamel 462, 467-471, 473, 547, 548 Equivalent Sun Hours (ESH) 160, 199, 341, 343, 345-347, 545 erosion effect(s) 420, 470 erosion rate 54, 183, 191, 205, 213, 300, 360, 380, 384, 386, 388, 389, 391-393, 395, 397, 398, 439, 489, 509, 510, 511, 521 erosion yield 203, 206, 208, 209, 213215, 238, 300, 364, 379, 381, 384, 385, 389, 419, 420, 425, 433, 508, 509, 516-519, 521-524 European Space Agency (ESA) 5, 40, 48, 60, 64, 68, 69, 100, 111, 113, 122, 168, 171, 176, 181, 191, 216, 298, 334, 358, 390, 482, 515, 516, 524, 525, 549 Extra Vehicle Activities (EVA) 32, 34, 35, 39-42, 61, 64, 66, 483 F fabric 60, 64, 65, 146, 492 fairings 257 fast atomic oxygen (FAO) 279, 425, 427, 431, 433, 451, 456, 458, 516, 517, 522, 523, 524 FEP/Teflon 53-56, 65, 66, 71, 91-99, 155-168, 196, 203, 204, 205, 208, 294, 296, 300, 335, 340, 345, 346, 349, 385, 387, 394, 398, 499, 523 FEP/Teflon- aluminized (Al-FEP) 91, 92, 93, 94, 95, 97, 98, 155-158, 161, 162 FEP/Teflon- silvered (Ag-FEP) 156 fiber(s) 142, 144, 146, 152, 217, 253, 260, 263-270, 272, 279, 401, 403, 419, 422, 424, 434, 445, 451, 492, 516 fiberglass 65, 66, 257, 258 flexibility 156 fluence 28, 49, 50, 81-83, 85-91, 93, 94, 98, 131, 133-137, 160, 163, 194, 195, 203, 205-213, 236, 241, 300, 320, 321, 346, 363, 364, 389, 393, 395, 426, 429, 455, 475, 479, 481, 499, 508-512, 517 fluorinated polymer(s) 491, 500, 517, 523 fluorine 299, 300, 303, 426 friction coefficient 475, 476, 478, 479, 480, 481

553 G galactic (cosmic rays) 12, 23, 26, 27, 28, 29, 31, 44, 45, 49, 50 GAMESS 538, 541 gears 292, 293, 296, 297 gearbox(es) 292, 297 geomagnetic 15, 19, 21, 29, 31, 32, 36, 37, 40, 41, 44, 45, 49, 123, 124, 125, 128, 129, 130 glass transition temperature 135, 269, 273, 276, 280, 437, 497 glass- optical 113, 114, 121, 122 glass- protective 322, 324, 461, 463, 464, 473 glass fiber 65, 253, 254, 257, 462 glue 472, 473 graphite 238, 294, 396, 421, 451, 453 grapple fixture 291, 419 ground-based facility 360, 366 ground-based testing 360, 516-519, 521-524 ground laboratory facility (see also AO laboratory simulation) 155

I inclination 14, 21, 23, 29, 31, 55, 516, 544 Indium Tin Oxide (ITO) 73, 74, 75, 76, 77, 78, 79, 80, 81 inelastic scattering 530, 535 Information Technology (IT) 188 inorganic 65, 66, 139, 140, 144, 147, 217, 229, 230, 232, 292, 293 insulator 53, 76 International Space Station (ISS) 2, 3, 31-41, 43, 44-57, 59-72, 121, 235240, 242, 243, 436, 441, 515, 524, 525 interplanetary 8, 9, 10, 12, 25, 26, 27, 28, 29 ionospheric 1, 15, 17, 20, 31-37, 40, 41, 320, 326 ionization 26, 29, 106 ionizing radiation 17, 31, 32, 44, 49, 52, 54, 111, 141, 149, 158, 204, 334, 515 K

H Halar 523 halocarbon 235 hardness 122, 131, 133-135, 137, 148, 459 hazard 5, 22, 23, 26, 34, 40, 42, 204, 387 heating 16, 92, 100, 115, 133, 155, 158, 159, 161, 162, 164, 166, 173, 174, 183, 184, 186, 187, 189, 190, 246, 249, 264, 265, 267, 273, 276, 312, 424, 455 hermetic 145, 328 hexamethyldisiloxane (HMDSO) 443, 444, 445, 447, 448, 449 high temperature 10, 174, 251, 252, 255, 294, 453 high-energy 7, 9, 12, 20, 22, 26, 31 Hubble Space Telescope (HST) 71, 99, 155-168 hydrocarbon compounds 283 hydrocarbon polymers 239, 517, 518, 537 hydrocarbon- surface 527, 528, 529, 532, 534, 535 hyperthermal atomic oxygen beam 391, 392, 475, 481 hyperthermal- atomic oxygen (AO)284, 351, 391, 392, 398, 475, 477, 480, 481 hyperthermal- beam 284 hyperthermal- reactions 282, 283

Kapton 57, 58, 63, 66, 100, 133, 174, 181, 184, 193, 194, 195, 199, 200, 203-213, 215, 237-243, 319, 335, 340, 345, 346, 349, 362, 363, 364, 365, 379-390, 393, 394, 398, 420, 426, 439, 443, 445, 448-450, 451, 456, 457, 459, 482, 485, 509, 511, 517, 521, 522, 523 Kapton erosion yield in LEO 387, 389, 456 Kapton H 205, 346, 363, 364, 365, 380, 383, 521 Kapton HN 205, 346, 380, 383, 521 Kapton polyimide 382, 388 L lacing tape 419, 422, 424, 426-428, 434 LEO environment 62, 64, 66, 156, 166, 185, 204, 212, 263, 270, 280, 292, 351, 388, 389, 420, 433, 443, 462, 475, 482, 499, 509, 516 linear energy transfer 49, 106 Long Duration Exposure Facility (LDEF) 62, 72, 156, 232, 239, 243, 359, 366, 394, 398, 434, 524 Low Earth Orbit 13, 14, 17, 32, 35, 44, 52, 62, 64, 66, 68, 69, 72, 99, 156, 158, 166, 183, 184, 185, 188, 191, 193, 194, 198, 201, 203-209, 211213, 215, 233, 235, 236, 237, 239, 260, 262, 263, 269, 270, 271, 272, 273, 275, 276, 277, 279, 280, 281, 291, 292, 294, 295, 296, 298, 299,

554 300, 304-306, 324, 335, 343, 351, 359, 379, 388-390, 397-399, 419, 420, 433, 434, 435, 441, 443, 450, 451, 456, 460, 462, 474, 475, 476, 482-484, 485, 488, 492, 499, 503, 509, 515, 516, 517, 518, 519, 521, 524, 525, 527, 534, 537, 541 low temperature 110, 132, 251, 253, 279, 293, 319, 320 lubricant(s) 260, 291-298, 475, 476, 478, 480, 481 lubrication 291, 292, 293, 294, 295, 297, 298, 475 M magnetic field 9, 10, 11, 12, 15, 19, 21, 23, 24, 25, 26, 29 magnetosphere 19, 21, 26, 27, 29 magnetron (sputtering) 17, 21, 75, 184, 294, 307-309, 311, 316, 323, 325, 477 mass loss 148, 150, 171, 173, 174, 175, 176, 177, 180, 205, 208, 209, 272, 273, 346, 379-381, 387, 389, 391, 392, 393, 396, 397, 398, 425, 426, 439, 440, 441, 448, 456, 499, 517 materials- outgassing 31, 59, 60 mechanical properties 55, 56, 99, 132, 155, 172, 272, 277, 279, 280, 296, 336, 345, 347, 379, 418, 421, 435, 436, 492, 500, 501 MEDET (satellite) 368-370, 373-375, 377 MgF2 73-77, 79, 81, 337, 340, 341, 349, 380, 382 microcrack(s) 222, 223, 230, 232, 248, 276, 278, 280, 418 microhardness 272 microparticles 462, 467, 474 miniature spectrometer 367, 368, 377 MISSE 436, 515, 516, 524, 525 model(s) 7, 8, 26, 27, 29, 31, 35, 37-40, 42, 44, 52, 57, 59-61, 68, 70, 110, 114, 115, 123, 125, 127, 128, 129, 151, 173, 185, 188, 190, 214, 239, 241, 245, 248, 249, 252, 253, 255, 299, 300, 304, 307, 313, 315, 316, 327, 328, 330-334, 337, 366, 407411, 425, 484, 485, 488, 489, 503507, 510-512, 516, 517, 519, 521, 522, 524, 528, 543-545, 549 modeling 11, 40, 126, 151, 152, 240, 241, 243, 253, 314, 316, 334, 359, 366, 484, 485, 488, 503, 504, 507, 516 model- AE-8 53, 130 model- AP-8 123, 125, 129, 130

molecular oxygen 211, 320, 336, 380, 388, 423, 456, 477 molybdenum disulphide (MoS2) 292 monolithic miniature spectrometer (MMS) 371, 372, 374 morphology 133, 206, 209, 214, 272, 279, 307, 322, 383, 388 multi-layer insulation (MLI) blankets 65, 156, 158, 159 Mylar 335, 340, 345, 347, 349, 394, 420 O O-atom 211, 284, 285, 287, 289, 300, 476, 528, 533 on-orbit 54, 57, 60, 71, 121, 155, 156, 158, 164, 167, 292 optical characteristics 218, 250, 402, 543-545, 549 optical constants 183, 185, 187-190, 307-309, 312-317, 483-485, 488, 489 optical density 87, 88, 100, 103, 149 optical properties 14, 17, 64, 65, 86, 90, 92, 121, 141, 148, 155, 156, 171, 174, 176, 178, 180, 183, 184, 187, 190, 194-197, 199, 200, 249, 262, 402, 421, 435, 440, 443, 449, 456, 457, 484, 488, 489, 500, 543 optical materials 367, 377 optical microscopy 316, 454 organic 60, 108, 139, 141, 144, 193, 214, 218, 227, 231, 232, 257, 293, 295, 300, 307, 317, 434, 435, 436, 494 oscillator 185, 186, 187, 189, 485 outgassing 59, 60, 61, 62, 66, 97, 174, 177, 211, 245, 257, 258, 277, 294, 295, 327-334, 419, 544 outgassing/contamination tests 211 oxidation 239, 242, 268, 351-358, 419, 447, 476, 488 oxide layer 214, 215, 352, 478 oxide(s) 81, 203, 211, 214, 215, 230, 235, 250, 252, 295, 296, 307, 312, 320, 323, 351-358, 435, 436, 439, 478, 484, 485, 487, 489, 492, 494497, 500 oxygen atom(s) 88, 194, 204, 210, 211, 212, 215, 293, 301-304, 306, 351353, 364, 379, 382, 388, 389, 395, 445, 455, 486, 506, 508, 527, 528, 534, 535, 537, 540 oxygen plasma(s) 191, 203, 206-215, 319, 320, 322, 323, 324, 325, 326, 384, 390, 426, 427, 429, 447, 456, 483, 484, 485, 486, 487, 488, 489

555 P paint- black 433 paint- gray 419, 428 paint- polyurethane 66, 193, 194, 200, 419 paint- silicone 193, 194, 200, 203, 204, 215 paint- thermal control 60, 194 paint- white 433 pathways (reaction) 283, 285, 288, 289 peroxide 351 photo-electric converters (PEC) 319, 320, 321, 322, 323, 324, 325 planetary 3, 9, 12, 18, 24, 25, 171, 173 plasma- afterglow 214, 380 plasma- asher 203, 204, 335, 336, 340, 342, 348, 349, 419, 426, 431, 433, 455 phenyl phosphine oxide (PPO) 436, 437, 492, 494-501 plasma- polymerization 443, 444, 449 polar orbit 545 Polymer-based composites 217, 293, 294 Polymer materials- CV-1144-O silicone183, 184, 185, 187, 188, 190, 191, 489 Polymer materials- Kapton 57, 58, 63, 66, 100, 133, 174, 181, 184, 193195, 199, 200, 203-213, 215, 237243, 319, 335, 340, 345, 346, 349, 362-365, 379-390, 393, 394, 398, 420, 426, 439, 443, 445, 448-450, 451, 456, 457, 459, 482, 485, 509, 511, 517, 521, 522, 523 Polymer materials- Kynar 66, 523 Polymer materials- polycarbonate 69, 100 Polymer materials- Mylar 335, 340, 345, 347, 349, 394, 420 Polymer materials- polyethylene (PE) 52, 99, 105, 146, 299, 300, 304, 315, 316, 335, 340, 345, 349, 391, 392, 393, 396, 398, 420, 421, 523 Polymer materials- polyimide 66, 176, 181, 205, 237, 294, 340, 380, 383, 390, 391, 392, 393, 395, 397, 398, 435, 436, 437, 438, 439, 440, 521 Polymer materials- Tedlar 523 Polymer materials- Tefzel 335, 340, 345, 347, 349 polysiloxane 193, 195, 200, 201 propane 302, 537-540 Porosity 204, 214, 215, 253, 255, 457 porous material 247, 248, 256 Power 71, 153, 450 Predictive Erosion Resistance Software (PERS) 515, 516, 518

predictive model 515, 516, 517, 522, 524 product- volatile 379, 380, 381, 385, 387, 389 protective coating- defects 238 proton flux(es) 47, 92, 100, 124, 128, 130 proton (ir)radiation 81, 82, 83, 84, 85, 86, 87, 88, 89, 91, 94, 96-98, 131137, 156, 157, 160 Q Quartz crystal microbalance (QCM) 379, 380, 382, 385-388, 391-393, 396, 477 R radiation 9, 21, 22, 32, 44, 46, 52, 53, 55, 59, 67, 69, 70, 71, 89, 90, 91, 96, 99, 101, 102, 104, 110, 111, 114, 120, 122, 123, 128, 130, 141, 147, 149, 296, 492, 513, 544 radio-frequency (RF) plasma 203, 204, 205, 206, 208, 209, 211, 212, 213, 214, 215, 216, 242, 336-339, 342344, 347, 349, 364, 380, 384, 388390 radiolysis 106, 108 reaction pathways 283, 285, 288, 289 reaction products 107 reflectance 91, 93, 95, 140, 153, 195, 199, 236, 401, 402, 404-406, 448, 449 reflectance- spectral 91, 92, 93, 94, 98, 149, 195, 199, 489, 543, 548 reliability 366 resin-epoxy 459 retro-reflector 236, 238, 242 roughness 214, 228, 235, 242, 307-309, 311, 313-316, 381, 383, 385, 388, 391, 396, 398, 445, 454, 456, 457, 459, 484, 485 S satellite(s) 1-6, 13, 17, 21, 23, 26, 31, 40, 47, 55, 125, 126, 130, 193, 194, 198, 211, 300, 421, 475, 476, 481 Scanning Electron Microscopy (SEM) 203, 205-209, 215-218, 232, 271, 272, 425-428, 431-433, 461, 462, 463, 464, 465, 468, 469, 471, 472 Second Surface Mirrors (SSM) 193, 194, 195, 196, 197, 201 sensors 52, 151, 260, 297 shielding 22, 26, 32, 44, 45, 46, 49, 50, 51, 52, 54, 63, 124, 145, 221, 296, 337, 338, 341

556 short-circuit current 320, 322, 324, 325, 326 shrinkage 264 silylation 420, 423, 424 silica 66, 75, 90, 183, 307-310, 313, 316, 422, 445, 494, 495 silicone 53, 60, 62, 65, 66, 100, 104, 131-137, 146, 183, 184, 191, 193, 194, 196, 197, 200, 203, 204, 206, 211, 214, 215, 227, 228, 229, 232, 295, 317, 462, 467, 472, 473, 474, 489 silver oxide 352, 356 simulated LEO environment 193, 194, 276 simulation tests 92 software 49, 50, 60, 61, 140, 151, 153, 306, 315, 515-518, 521, 522, 524 solar- absorptance 156, 172, 179, 454, 457 solar- array 17, 60, 63, 66, 73, 155, 156, 158, 235, 242, 257, 262, 319, 390, 444, 461, 474, 544 solar- cell 23, 26, 60, 132, 319, 320, 322, 325, 326, 461, 472 solar- cycle9, 11, 12, 14, 17, 28, 29, 30, 31, 516 solar- flares 11, 13, 29, 156, 158 solar- maximum 12, 28, 29, 41, 44, 45, 53, 438 solar- minimum 12, 27, 28, 30, 40, 44, 53, 160 solar- particles 25, 29 solar- protons 29, 44 solar- radiation 27, 149, 171, 300, 320, 321, 322, 324, 325, 543 solar- x-ray 44 South Atlantic Anomaly (SAA) 21, 22, 23, 24, 44, 45, 46, 52, 53 space- applications 181, 272, 293, 316, 414, 419, 420, 435, 475, 480, 481, 492, 501, 516 space- components 419, 421, 423, 433 space- durability 491, 492, 517 space- environment 4, 7-9, 12, 31, 81, 90, 92, 99, 132, 140, 155, 156, 158, 159, 161, 163, 165, 167, 181, 217, 232, 262, 269, 271, 272, 281, 291, 307, 319, 324, 327, 333, 334, 359, 366, 379, 387, 390, 391, 398, 401, 402, 435, 436, 460, 462, 467, 475, 504, 515, 516, 543-546, 548, 549 space- exposure 157, 217, 218, 223, 227, 230, 231, 232, 269, 276, 343, 463 space- factors 218, 232, 281, 402, 549 space- mission 140, 174, 527

Space Station- International2, 3, 14, 32, 31, 68, 69, 70, 71, 121, 235, 436, 441, 515, 524, 525 spacecraft materials 52, 183, 204, 280, 299, 335, 390, 406, 434, 460, 474, 515, 543 spectral reflectance 91, 92, 93, 94, 98, 149, 195, 199, 543, 548 spectrometer52, 92, 150, 195, 199, 425, 454, 477, 493 spectrum 50, 51, 53, 54, 74, 91, 94, 95, 97, 98, 103, 116, 117, 124, 125, 129, 131-133, 135, 147, 185, 198, 199, 211, 260, 322, 383, 384, 388, 406, 477, 489, 543, 546, 548 SPRUT-6 124, 125, 127, 128 stability 122, 123, 139, 141, 148, 149, 154, 171-173, 177, 178, 180, 193, 194, 223, 270, 294, 307, 402, 408, 415, 431, 451, 497, 500 strength 15, 24, 25, 86, 105, 109, 131, 135-137, 148, 149, 190, 222, 271, 272, 274, 279, 291, 294, 296, 345, 349, 410, 414, 416-418, 420, 453, 459, 474, 496 stress- bending 271, 273, 274, 275, 276, 277, 278, 279, 280 stress- residual 276, 278, 316 sun 10-12, 14, 21, 26-29, 36-38, 160, 171, 184, 341, 343, 345 surface modification by Photosil technology 419-421, 423-429, 431, 433, 434 surface morphology 134, 177, 203, 206, 209, 212, 213, 214, 215, 310, 381, 426, 427 surfaces- black 151 surfaces- front 463-466, 470-473 surfaces- transmissive 150 synergistic effects 32, 190, 204, 205, 398, 523 T Teflon 53-56, 65, 66, 71, 91-95, 98, 99, 155, 156, 158, 161, 162, 164-166, 203, 204, 205, 208, 294, 296, 300, 335, 340, 345, 346, 349, 385, 387, 394, 398, 499, 523 Tefzel 335, 340, 345, 347, 349 telescope 11, 53, 156 tensile properties 155, 158, 162, 163, 166, 407, 437, 438 tensile strength 105, 131, 134, 135, 137, 148, 161, 342, 345, 349, 438 textile 140, 146 thermal- analysis 173, 361, 362

557 thermal- conductivity 149, 245, 246, 248, 250, 252, 255, 256, 351, 454, 491 thermal control- coatings 32, 92, 193, 194, 198, 199, 200, 201, 204, 211, 217, 280, 406, 549 thermal control- materials 99, 156, 198 thermal- cycling 53, 65, 81, 82, 83, 84, 86, 89, 156, 157, 158, 160, 163, 184, 193, 264, 269, 270, 272, 273, 275, 276, 277, 278, 279, 280, 407-409, 411, 414-417, 419, 451, 456, 459, 515, 517 thermal- deformation 264, 266, 268, 270 thermal- diffusivity 149, 245 thermal emittance 92, 156, 172, 176, 179, 454, 457 thermal- expansion 86, 149, 249, 253, 264, 266-270, 273, 279, 407, 411, 414, 454 thermal- stability 171, 172, 173, 175, 176, 180, 273, 492, 497 thermal- stress(es) 249, 407, 408, 410, 412, 417, 418, 420, 459 thermo-gravimetric analysis (TGA) 171, 173, 174, 175, 176, 177, 180, 437, 497, 499 thermophysical properties 245, 255 thermosphere 8, 15, 17 thread(s) 146, 294, 492 time-of-flight distribution 284-287 total mass loss (TML) 211, 295, 439, 440 trajectory 29, 425, 537, 538 transition state 300, 301, 302, 303, 304 transmission 113, 114, 116, 119, 120, 122, 149, 171, 176, 177, 178, 180, 205, 404, 405, 438, 440, 483 transmittance 74, 82, 86, 87, 90, 140, 153, 339, 341, 381, 401, 402, 405, 406 transparency 438 tribological properties 475, 476, 477, 480, 481 tribological testing 476, 478 U ultimate tensile strength 342, 345, 346, 348 undercutting 206, 212, 239-241, 243, 503, 504, 507-510, 512 UV exposure 17, 163, 184, 186, 187, 189, 190, 198 UV irradiation 190, 193, 199, 200, 201, 499

UV radiation 14, 157, 188, 189, 190, 213, 216, 379, 380, 387, 389, 423, 435, 436, 460 V Van Allen (belts) 20, 21, 156 Variable Angel Spectroscopic Ellipsometry (VASE) 183, 184, 308, 313, 484, 489 vibration cleaning 258 VII. The Ground-Based Testing Program of the 40 polymers 522 viscoelastic 273, 276 void(s) 313 volatile products 379-381, 385, 387, 389 VUV- degradation 67, 335 VUV- durability 344, 349 VUV- exposure 209, 213, 335, 340, 341, 345, 346, 348, 349, 523 VUV- radiation 201, 203, 204, 209, 212, 213, 335, 339, 360, 379, 380, 382, 388 VUV- source(s) 335, 336, 340, 349 VUV- test(s) 493, 524 W wavelength 14, 75, 87, 88, 92, 93, 120, 147, 152, 153, 156, 176, 178, 187, 314, 322, 335-337, 339, 341-345, 347, 448, 454, 483, 485, 517 witness sample 161, 208, 209, 346, 426, 517 X X-ray- diffraction 309, 316, 352, 353, 356, 484 X-ray- Photo electron Spectroscopy (XPS) 91, 93, 95, 98, 160, 203, 205, 210, 211, 213-216, 379, 381, 384, 388, 390, 393, 425, 426, 428, 429, 445, 449, 454, 458, 479 X-ray- spectra 217, 461 X-ray- spectrum 218, 220, 230, 232 Z Zenit- 2 257 zinc selenide 483, 484, 485, 486, 487, 488, 489

This page intentionally left blank

AUTHOR INDEX G

A Abraimov, V.V. Alred, J. Altstatt, R. L. Antoniazzi, J. Atabaev, B. G. Au-Yeung, S. W.

Garneau, M. 1 Garrison, B. J. 527 Garton, D. J. 283, 537 Gatsenko, L. S. 461 Gavrin, A. J. 435, 491 Gendron, S. 359 Geng, H. 81, 91, 131, 407, 414 Gindulyte, A. 299 Golden, J. 31 Gordeev, Y. P. 327, 543 Gouzman, I. 203, 379 Grachov, Y. A. 123 Grigorevskiy, A. V. 327, 543 Grigoryan, O. R. 123 Grossman, E. 203, 379 Gudimenko, Y. 419 Guillaumon, J. C. 193 Gusakov, A. G. 351 Gusarov, A. 113

81 31 31 291 319 435 B

Bakina, E. A. Banks, B. A. Barsamian, H. Barth, J. L. Boeder, P. Briskman, B. A. Bruckner, E.

461 235, 299, 503 31 7 31 99, 139 335 C

Cashman, T. Clark, A.A. Clark, A.J. Christiansen, E. Connell, J. W.

73 139 139 31 435

H Haffke, J. E. 183 Hambourger, P.D. 73 He, S. 81, 91, 131, 407, 414 Heiman, R. B. 451 Heltzel, S. 171 Hughes, P. C. 419

D de Groh, K. K. Deev, I. S. Demko, R. Demont, P. Dever, J. Dinguirard, M. Doyle, D.

155 213, 461 235 271 335 203, 271, 367 113

I Iskanderova, Z. Issoupov, V. V.

139, 419, 451, 515 263, 271 J

E Edwards, D. Eesbeek, M. V. Eremenko, V. V.

Jiang, X-X. Jin, J. Jin, Z.

31 171, 367, 515 401

359 491 443 K

F Fenchenko, V. N. Francke, M. Franke, E. Fritsche, B. Fruit, M.

Kablov, E. N. 213 Kagumba, L. 491 Kaur, J. 73 Kern, J. 31 Khassanchine, R. H. 327 Khristoforov, D. A. 263 Kleiman, J. 139, 245, 419, 451, 515

401 451 483 451 113

559

560 Kida, T. Kinoshita, H. Klinshpont, E. R. Koontz, S. L. Kozyrski, E. N.

391 283, 475 99, 139 31 351 L

Lacabanne, C. Lempert, G. Letin, V. A. Li, C. Lifanova, L. F. Lifshitz, Y. Litovsky, E. Liu, H. Lorenz, M. J. Lucier, L.

274 203, 379 461 91 319 203, 379 245 81 31 359

O Ohmae, N. P Pascual, R. Z. Pechnikova, L. V. Pedley, M. Polsak, A. Potapovych, L. P. R 319 351 193 491 263 S

461 319 155 299 475 31 139 335 527 245 31 235, 299 291, 419 213 31 283 451 435 475 73

N Nabarra, P. Naumenko, V. M. Nebo, J. Ng, R. Nikanpour, D. Nikishin, E. F. Novikov, L. S. Noter, Y.

537 319 31 171 257

Rakhimova, F. Raspopov, S. A. Remaury, S. Rice, N. Rumyantsev, A. F.

M Malenkov, A. V. Markov, A. V. Martin, M. Massa, L. Matsumoto, K. Mayeaux, B. McCall, S. H. C. P. McCracken, C. Medvedeva, M. Menn, N. Mikatarian, R. R. Miller, S. K. Milligan, D. Minakov, V. T. Minow, J. I. Minton, T. K. Moc, A. Mojazza, B. Muromoto, M. Muhieddine, L. K.

391, 475

193 401 491 419 359 213, 263, 271, 461 123 203, 379

Schatz, G. C. Schneider, T. Semprimoschnig, C. O. A. Shanbhag, M. Shavarin, Y. Sitalo, V. G. Smith, D. W. Smith, Jr., J. G. Snyder, A. Soares, C. Startsev, O. V. Stepanov, V. F. Suzuki, M.

537 31 171 73 139 257 491 435 235, 503 31 263, 271 99 475

T Tagawa, M. Talas, D. Tchourilo, I. V. Tennyson, R. C. Tighe, A. P. Topping, C. M. Tykhyy, V. G.

391, 475 515 123, 319 419, 515 367 491 257, 401 U

Ubaid, S.H.

73 V

Vasiliew, V. N. Vecher, A. A.

543 351

561 Vemulapalli, J. Viel-Inguimbert, V. Voropaev, A. G.

73 203, 271 351

W Wang, H. Wang, J. Wang, Y. Watson, K. A. Wei, Q. Welch, B. Woollam, J. A.

81 443 443 435 131 73 183, 307, 483 Y

Yang, D. Yan, L. Yang, S. Yu, Z. Yokota, K.

81, 91, 407, 414 307, 483 81, 91, 131 443 391 Z

Zhang, L. Zheludekevich, M. I. Zhou, X.

131 351 443