Radioactive Aerosols

RADIOACTIVE AEROSOLS RADIOACTIVITY IN THE ENVIRONMENT A companion series to the Journal of Environmental Radioactivit...

Author: Papastefanou C.

131 downloads 1351 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

RADIOACTIVE AEROSOLS

RADIOACTIVITY IN THE ENVIRONMENT A companion series to the Journal of Environmental Radioactivity Series Editor M.S. Baxter Ampfield House Clachan Seil Argyll, Scotland, UK Volume 1: Plutonium in the Environment (A. Kudo, Editor) Volume 2: Interactions of Microorganisms with Radionuclides (F.R. Livens and M. Keith-Roach, Editors) Volume 3: Radioactive Fallout after Nuclear Explosions and Accidents (Yu.A. Izrael, Author) Volume 4: Modelling Radioactivity in the Environment (E.M. Scott, Editor) Volume 5: Sedimentary Processes: Quantification Using Radionuclides (J. Carroll and I. Lerche, Authors) Volume 6: Marine Radioactivity (H.D. Livingston, Editor) Volume 7: The Natural Radiation Environment VII (J.P. McLaughlin, S.E. Simopoulos and F. Steinhäusler, Editors) Volume 8: Radionuclides in the Environment (P.P. Povinec and J.A. Sanchez-Cabeza, Editors) Volume 9: Deep Geological Disposal of Radioactive Waste (R. Alexander and L. McKinley, Editors) Volume 10: Radioactivity in the Terrestrial Environment (G. Shaw, Editor) Volume 11: Analysis of Environmental Radionuclides (P.P. Povinec, Editor) Volume 12: Radioactive Aerosols (C. Papastefanou, Author)

RADIOACTIVE AEROSOLS

Constantin Papastefanou Physics Department, Aristotle University of Thessaloniki Thessaloniki, Greece

AMSTERDAM – BOSTON – HEIDELBERG – LONDON – NEW YORK – PARIS SAN DIEGO – SAN FRANCISCO – SINGAPORE – SYDNEY – TOKYO

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK First edition 2008 Copyright © 2008 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-08-044075-0 ISSN: 1569-4860

For information on all Elsevier publications visit our website at books.elsevier.com

Printed and bound in Hungary 08 09 10 11 12

10 9 8 7 6 5 4 3 2 1

This book is dedicated to all whose purpose in searching for knowledge is to share it unselfishly with others and to use it with understanding for the benefit of their fellow men

The wilderness and the parched land shall be glad, and the desert shall rejoice, and blossom as the rose. It shall blossom abundantly, and rejoice even with joy and singing. . . And the parched land shall become a pool, and the thirsty ground springs of water. Isaiah 35: 1,2,7

This page intentionally left blank

vii

Contributions

Andersen, A.A. (1976). ACFM Ambient Particle Sizing Samplers. Operating Manual TR 76900016, pp. 1–21. Andersen, A.A. (1979). Low-Pressure Impactor. Operating Manual TR 76-900016, pp. 1–29. Berner, A., Lurzer, C. (1980). Mass size distributions of traffic aerosols at Vienna. J. Phys. Chem. 84, 2079–2083. Bondietti, E.A., Papastefanou, C. (1989). Large particle nitrate artifacts in the aerodynamic size distributions of ambient aerosols. J. Aerosol Sci. 20 (6), 667–670. Bondietti, E.A., Papastefanou, C. (1993). Estimates of residence times of sulfate aerosols in ambient air. Sci. Total Environ. 136, 25–31. Bondietti, E.A., Papastefanou, C., Rangarajan, C. (1987). Aerodynamic size associations of natural radioactivity with ambient aerosols. In: Hopke, P.K. (Ed.), Radon and Its Decay Products: Occurrence, Properties, and Health Effects, In: ACS Symposium Series, vol. 331, pp. 377–397. Cember, H. (1987). Inhaled radioactivity. In: Introduction to Health Physics, second ed., revised and enlarged. Pergamon Press, New York. Chamberlain, A.C. (1991). Radioactive Aerosols. Cambridge University Press, Cambridge, UK. Cheng, Y.S., Guilmette, R.A., Zhou, Y., Gao, J., LaBone, T., Whicker, J.J., Hoover, M.D. (2004). Characterization of plutonium aerosol collected during an accident. Health Phys. 87 (6), 596–605. Demokritou, P., Lee, S.J., Ferguson, S.T., Koutrakis, P. (2004). A compact multistage (cascade) impactor for the characterization of atmospheric aerosols. J. Aerosol Sci. 35, 281– 299. Grundel, M., Porstedorfer, J. (2003). Characterisation of an electronic radon gas personal dosemeter. Radiat. Prot. Dosim. 107 (4), 287–292. Hinds, W.C. (1999). Aerosol Technology: Properties, and Measurement of Airborne Particles. John Wiley & Sons, New York. Jost, D.T., Gaggeler, H.W., Baltensperger, U. (1986). Chernobyl fallout in size-fractionated aerosol. Nature 324, 22–23. Kauppinen, E.I., Hillamo, R.E., Aaltonen, S.H., Sinkko, K.T.S. (1986). Radioactivity size distributions of ambient aerosols in Helsinki, Finland, during May 1986 after the Chernobyl accident. Environ. Sci. Technol. 20, 1257–1259.

viii

Contributions

Kesten, J., Butterweck, G., Porstendorfer, J., Reineking, A., Heymel, H.-J. (1993). An online α-impactor for short-lived radon daughters. Aerosol Sci. Technol. 18, 156–164. Kim, D.S., Lee, K.W., Kim, Y.J. (2006). Characterization of a particle trap impactor. J. Aerosol Sci. 37, 1016–1023. Kondo, K., Muramatsu, H., Kanda, Y., Takahara, S. (1984). Particle size distribution of 7 Beaerosols formed in high energy accelerator tunnels. Int. J. Appl. Radiat. Isotopes 35 (10), 939–944. Kulmala, M., Mordas, G., Petaja, T., Gronholm, T., Aalto, P.P., Vehkamaki, H., Hienola, A.I., Herrmann, E., Sipila, M., Riipinen, I., Manninen, H.E., Hameri, K., Stratmann, F., Bilde, M., Winkler, P.M., Birmili, W., Wagner, P.E. (2007). The condensation particle counter battery (CPCB): A new tool to investigate the activation properties of nanoparticles. J. Aerosol Sci. 38 (3), 289–304. Kwon, S.B., Lim, K.S., Jung, J.S., Bae, G.N., Lee, K.W. (2003). Design and calibration of a 5-stage cascade impactor (K-JIST cascade impactor). J. Aerosol Sci. 34, 289–300. Marple, V.A., Willeke, K. (1979). Inertial impactors. In: Lundgren, D.A., Harris Jr., F.S., Marlow, W.H., Lippmann, M., Clark, W.E., Durham, M.D. (Eds.), Aerosol Measurement. University Presses of Florida, Gainesville, FL, pp. 90–107. Marple, V.A., Rubow, K.L., Behm, S.M. (1991). A microorifice uniform deposit impactor (MOUDI): Description, calibration, and use. Aerosol Sci. Technol. 14, 434–446. McMahon, T.A., Denison, P.J. (1979). Empirical atmospheric deposition parameters—A survey. Atmos. Environ. 19, 571–585. National Research Council (1978). Airborne Particles. University Park Press, Baltimore. Papastefanou, C. (2006). Residence time of tropospheric aerosols in association with radioactive nuclides. Appl. Radiat. Isotopes 64, 93–100. Papastefanou, C. (2006). Radioactive nuclides as tracers of environmental processes. J. Radioanal. Nucl. Chem. 267 (2), 315–320. Papastefanou, C., Bondietti, E.A. (1987). Aerodynamic size associations of 212 Pb and 214 Pb in ambient aerosols. Health Phys. 53 (5), 461–472. Papastefanou, C., Bondietti, E.A. (1991). The unattached fraction of radon progeny in ambient aerosols. J. Environ. Radioact. 13 (1), 1–11. Papastefanou, C., Bondietti, E.A. (1991). Mean residence times of atmospheric aerosols in the boundary layer as determined from 210 Bi/210 Pb activity ratios. J. Aerosol Sci. 22 (7), 927–931. Papastefanou, C., Ioannidou, A. (1995). Aerodynamic size association of 7 Be in ambient aerosols. J. Environ. Radioact. 26, 273–282. Papastefanou, C., Ioannidou, A. (1996). Influence of air pollutants in the 7 Be size distribution of atmospheric aerosols. Aerosol Sci. Technol. 21, 102–106. Porstendorfer, J. (1994). Properties and behaviour of radon and thoron and their decay products in the air. J. Aerosol Sci. 25 (2), 219–263. Porstendorfer, J. (2001). Physical parameters and dose factors of the radon and thoron decay products. Radiat. Prot. Dosim. 94 (4), 365–373. Porstendorfer, J., Pagelkopf, P., Grundel, M. (2005). Fraction of the positive 218 Po and 214 Pb clusters in indoor air. Radiat. Prot. Dosim. 113 (3), 342–351. Raabe, O.G. (1979). Design and use of the Mercer-style impactor for characterization of aerosol aerodynamic size distributions. In: Lundgren, D.A., Harris Jr., F.S., Marlow, W.H.,

Contributions

ix

Lippmann, M., Clark, W.E., Durham, M.D. (Eds.), Aerosol Measurement. University Presses of Florida, Gainesville, FL, pp. 135–140. Ruiperez, L.G., Garcia, B.A., Uria, J.M.J., Iglesias, J.M.P. (1984). A new counter of Aitken particles. Atmos. Environ. 18 (8), 1711–1714. Sinclair, D., Countess, R.J., Liu, B.Y.H., Pui, D.Y.H. (1979). Automatic analysis of submicron aerosols. In: Lundgren, D.A., Harris Jr., F.S., Marlow, W.H., Lippmann, M., Clark, W.E., Durham, M.D. (Eds.), Aerosol Measurement. University Presses of Florida, Gainesville, FL, pp. 544–563. Thomas, J.W. (1953). The Diffusion Battery Method for Aerosol Particle Size Determination. ORNL Report No. 1648, pp. 1–68. Tokonami, S., Takahashi, F., Iimoto, T., Kurosawa, R. (1997). A new device to measure the activity size distribution of radon progeny in a low level environment. Health Phys. 73 (3), 494–497. Wileke, K., Baron, P.A. (1993). Aerosol Measurement: Principles, Techniques, and Applications. Van Nostrand Reinhold, New York. Yu, C.C., Tung, C.S., Hung, I.F., Tseng, C.L. (1993). Analyses of radioactive aerosols to support accurate internal dose assessments at Chin-Shan nuclear power plant. Health Phys. 65 (2), 147–153.

This page intentionally left blank

xi

Contents

Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

Chapter 1: Atmospheric aerosol particles . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction and definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. The size distribution of atmospheric aerosol particles . . . . . . . . . . . . . . 3. Aerosols and radiation: Generation of radioactive aerosols . . . . . . . . . . . 4. Aerodynamic size distribution of radionuclide-associated aerosol particles (radioactive aerosols) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Aitken nuclei mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Accumulation mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Coarse particle mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Mechanisms of formation and growth of aerosol particles . . . . . . . . . . . 5.1. Coagulation and condensation . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Gas-to-particle conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 3 4 6

Chapter 2: Radioactive aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Radioactive aerosols associated with the cosmic-ray produced radionuclides 2.1. Beryllium-7 aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Radon and thoron decay product aerosols . . . . . . . . . . . . . . . . . . . . 3.1. Formation of radon decay product aerosols . . . . . . . . . . . . . . . . . . . . . . 3.2. Diffusivity of radon decay products . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Attachment of radon decay products to aerosol particles . . . . . . . . . . . . . . . . 3.4. Activity size distributions of the radon product decay aerosols . . . . . . . . . . . . . 3.4.1. Lead-214 and 212 Pb aerosol size distributions . . . . . . . . . . . . . . . . . 3.4.2. Lead-210 vs 214 Pb aerosol size distributions . . . . . . . . . . . . . . . . . 3.4.3. Lead-212 vs sulfate, SO2− 4 aerosol size distributions . . . . . . . . . . . . .

11 11 11 12 16 16 18 19 21 21 26 26

7 7 7 8 8 8 8 9

xii

Contents 3.4.4. 3.4.5. 3.4.6. 3.4.7.

214 Pb and 212 Pb aerosol size distriLead-210, 7 Be, 35 S, 32 P and sulfates, SO2− 4 vs butions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . α-Recoil model: An explanation for the 214 Pb shift in the aerosol size distributions Recoil redistribution of 210 Pb following 214 Po α-decay . . . . . . . . . . . .

Radioactive aerosol particle sizes relative to growth mechanisms of sulfate, SO2− 4 aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The unattached fraction of radon decay product aerosols . . . . . . . . . . . .

27 29 32

4. Mine aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Fission product radionuclide aerosols . . . . . . . . . . . . . . . . . . . . . . . 6. Radioactive aerosols associated with the operation of high-energy accelerators 7. Plutonium aerosols due to nuclear weapons testing or nuclear reactor accidents References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

32 33 38 40 49 52 54

Chapter 3: Radioactive nuclides as tracers of environmental processes . . . . . . . 1. Radioactivity in the environment . . . . . . . . . . . . . . . . . . . . . . . . . 2. Atmospheric particle deposition . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Dry deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Wet deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Resuspension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Air-to-vegetation transfer of radionuclides associated with submicron aerosols References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

59 59 61 61 63 66 67 68

3.4.8.

Chapter 4: Residence time of tropospheric aerosols in association with radioactive nuclides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Residence time of aerosol particles . . . . . . . . . . . . . . . . . . . . . . . . 2. Residence time of tropospheric aerosol particles associated with the cosmic ray produced 7 Be . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Residence time of tropospheric aerosol particles associated with the radon decay product radionuclides 210 Pb, 210 Bi, 210 Po and the fission product radionuclides 89 Sr, 90 Sr, and 140 Ba and their activity ratios . . . . . . . . . . . . . . . 4. Residence time of tropospheric aerosol particles associated with the fission product radionuclides 89 Sr, 90 Sr, 140 Ba and their activity ratios . . . . . . . . 5. Residence times of sulfate aerosols in the atmosphere . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter 5: Radioactive particles and human subjects . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Radioactive dose from inhalation of radon decay product aerosols 3. Deposition of radioactive aerosol particles in the lung . . . . . . . 4. Risk assessment due to inhalation of radon decay product aerosols References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

71 71 73

75 79 80 82

. . . . . .

85 85 85 102 106 110

Chapter 6: Aerosol sampling and measurement techniques . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Cascade impactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

113 113 116

Contents

3. 4.

Inertial impactors . . . . . . . . . . Andersen cascade impactors . . . . 4.1. 1 ACFM ambient cascade impactor . 4.2. Low-pressure cascade impactor . . .

. . . .

. . . .

. . . . 4.3. Four-stage low-pressure cascade impactor . 4.4. High volume cascade impactor . . . . . .

. . . . . .

. . . . . .

xiii

. . . . . .

. . . . . .

. . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Activity size distribution of radioactive aerosol particles . . . . . . . . . . . . 5. Online α-impactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. A micro-orifice uniform deposit cascade impactor (MOUDI impactor) 7. Particle trap impactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. The diffusion battery method for aerosol particle size determination . 9. Condensation particle counter battery . . . . . . . . . . . . . . . . . . . 10. The Mercer-style impactor . . . . . . . . . . . . . . . . . . . . . . . . . 11. Berner-type impactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12. Dichotomous sampler . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. Compact multistage cascade impactor . . . . . . . . . . . . . . . . . . 14. K-JIST 5-stage cascade impactor . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

117 121 123 124 127 128 129 130 135 139 141 146 148 150 151 153 154 157

Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

161

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

169

This page intentionally left blank

1

Foreword

The radioactive nuclides, such as radio-iodine and other fission products, that remain airborne for more than a few hours after the explosion of atomic bombs (Hiroshima and Nagasaki, August 1945) or nuclear weapons testing (Nevada, April 1953; Bikini Atolls, 1953; Mururoa and Fangatoufa, 1996) or nuclear reactor accidents (particularly at Chernobyl, April 26, 1986) might become attached to aerosol particles producing radioactive aerosols. Radioactive aerosols can also contain radionuclides of terrestrial origin, such as radon (222 Rn) and thoron (220 Rn) decay products and those of cosmogenic origin, such as 7 Be, which are continuously present in ambient air. The pathways of exposure that must be considered primarily concern the radioactivity which is inhaled through the respiratory tract and the actual cases of exposure to radioactive aerosols considered include the exposure of workers to radon and its decay products in uranium and other mines, such as coal mines. The subject of this particular volume relates to aerosol particle physics including aerosol characterisation, the formation mechanism, the aerodynamic size distribution of the activity and aerosol residence time, instrumentation techniques, aerosol collection and sampling, various kinds of environmental (atmospheric aerosols), particularly radioactive aerosols and the special case of radon decay product aerosols (indoors and outdoors) and the unattached fraction, thoron decay product aerosols, the deposition patterns of aerosol particles in the lung and the subsequent uptake into human subjects and risk assessment. The objective of this special volume is to provide today’s readers of aerosol science and atmospheric physics, in general, with an up-to-date summary of knowledge about radioactive aerosols and, perhaps most importantly, with a vision of future developments in this field of research. I hope that readers will find that this objective has been achieved. I would like to express my gratitude to Murdoch Baxter, the series editor and my friend, for his idea of this book series on radioactivity in the environment and for continual encouragement in writing this book and the contributors for their research on aerosol science which has appeared in the literature over more than three decades. Constantin Papastefanou Aristotle University of Thessaloniki Atomic & Nuclear Physics Laboratory Thessaloniki 54124, Greece E-mail: [email protected]

This page intentionally left blank

3

Chapter 1

Atmospheric aerosol particles

1. Introduction and definitions The aerosol particles in the atmosphere are liquid or solid particles. Their sizes range from a fraction of a micron to several hundreds of microns (µm). Various names, such as Aitken nuclei, smokes, fumes or hazes (for the smaller particles) and dusts, mists, fogs or ash (for larger ones) have been used in the nomenclature of atmospheric aerosols. All aerosol particles are formed by condensation of gases or vapours or by mechanical processes. They may be transformed by coagulation or condensation at the same time as they are transported by air movement and dilution. They might disappear from the atmosphere and settle on some surfaces which act as a sink. The residence times of aerosol particles in the atmosphere vary from some days near the earth’s surface in the troposphere to a year or more in the stratosphere. A particle is a small quantity of liquid or solid. Many particles are unstable. They can change by growing or can even disappear on contact with a surface, such as a raindrop striking a surface and coalescing or an ion losing its charge after contact with a surface or an oppositely charged particle (radioactive aerosols). The atmospheric aerosols are a mixture of gases and particles that exhibits some stability in a gravitational field. The definition of fine and coarse particles as smaller and larger than 1 µm diameter, respectively, originated from the fact that atmospheric aerosol size distributions are bimodal, i.e. one mode occurs below 1 µm diameter, the other above 1 µm diameter (Whitby, 1975) (Figure 1.1). In atmospheric aerosol particle size distributions, fine particles result almost entirely from condensation and coagulation, whereas coarse particles are produced by mechanical processes. Some particles serve as nuclei, cloud condensation nuclei and ice nuclei reflecting their effects. Small and large ions carry electrical charges. Small ions are clusters of atoms or molecules around a charge. Their stability depends on their having a charge. Large ions are merely particles that carry an electric charge. Large particles, for those 0.1 to 1.0 µm in diameter, and giant particles, for those above 1.0 µm diameter, are terms also applied in characterising the atmospheric aerosol particles. As is known, the size of an atom is of the order of 10−10 m or 10−4 µm or 0.1 nm (the Bohr radius r1 = 0.5292 × 10−10 m), while the size of an atomic nucleus is of the order of 10−14 m or 10−8 µm or 10−5 nm. Thus an aerosol particle having a diameter of some nanometres (or 10−2 µm) is a hundred times or more larger than an atom and so includes some number of atoms or clusters of atoms. RADIOACTIVITY IN THE ENVIRONMENT VOLUME 12 ISSN 1569-4860/DOI: 10.1016/S1569-4860(07)12001-5

© 2008 Elsevier B.V. All rights reserved.

4

C. Papastefanou

Fig. 1.1. Bimodal mass distributions measured with a set of special impactors and a cascade impactor. Run 14 contains many more coarse particles than the average because of construction activity upwind. Note the negligible effect of this increased concentration of coarse particles on the fine particle mode. From Whitby (1975, p. III-21) in NRC (1979).

Primary particles, such as road dust, salt (sea-) spray from the oceans and cement dust do not change form after emission, whereas a substantial fraction of mass of the secondary particles, such as photochemically produced sulfates and photochemical smog, is formed by in situ chemical reactions involving gases. Monitoring and particular aerosol characterisation studies have led to a revolution in both our knowledge and our understanding of atmospheric aerosols. Atmospheric sciences, particle technology, industrial hygiene and health effect studies of pollutants are different contributing fields in that context.

2. The size distribution of atmospheric aerosol particles Very early on, Aitken (1923) showed that most particles in the atmosphere are smaller than 0.1 µm diameter and that their concentrations vary from some hundreds per cm3 over the ocean to millions per cm3 in urban areas. Junge (1955, 1963, 1972) measured the atmospheric aerosol number size distribution and concentration in urban and non-urban areas as functions of altitude and site. He established the standard form for plotting size distribution data: log of dN/dDp versus log Dp , where N = number and Dp = particle diameter. He observed that this plot was a straight line that could be described by the equation: dN/dDp = ADp−k , where A and k were constants. He also noted that in the range from 0.1 to 10.0 µm particle diameter, k was approximately equal to 4.0. This distribution mode was widely known as the Junge distribution or the power law distribution. Friedlander (1961) later showed that by balancing aerosol source and removal rates a portion of the resultant theoretical number distribution steady state could be fitted reasonably well by the Junge distribution. Clark and Whitby (1967), by fitting the Junge distribution to 52 at-

Atmospheric aerosol particles

5

Fig. 1.2. Normalised frequency plots of number, surface, and volume (particle volume times particle density) distributions for the grand average 1969 Pasadena, California smog aerosol. Note the bimodal distribution of mass. Each weighting shows features of the distribution not shown by the other plots. From Whitby (1975, p. II-11) in NRC (1979).

mospheric distributions, found that the constant A was equal to 0.4 multiplied by the aerosol volume concentration (µm3 /cm3 ). This agrees with the value predicted by Friedlander. Whitby et al. (1972) found that the number size distribution established by Junge was not a good model for the surface and mass or volume size distribution which was normally bimodal, with one mode being around about 0.3 µm diameter and the other ranging from 5.0 to 15.0 µm diameter. In Figure 1.2 (NRC, 1979), the normalised frequency plots of number, surface and volume or mass distributions are presented. In this figure, the apparent area under the curves is proportional to the number, surface area and volume or mass in a given size range. Most particles, i.e. the number distribution, are of approximately 0.01 µm diameter. The number of particles decreases sharply with increasing size. Most of the surface area is provided by particles averaging 0.2 µm diameter. The volume or mass distribution is bimodal: one mode is around 0.3 µm diameter, the other about 10.0 µm diameter. The mass of fine particles of size smaller than 2.0 µm is almost equal to the mass of coarse particles of size larger than 2.0 µm. Atmospheric aerosol size distributions consist basically of three separate modes: (i) Aitken nuclei mode for particles smaller than 0.1 µm diameter, i.e. 0.015 µm, (ii) accumulation mode for smaller than 2.0 µm, but larger than 0.1 µm, diameter particles, i.e. 0.3 µm, and (iii) coarse mode for particles larger than 2.0 µm diameter, i.e. 10.0 µm. Depending on their source there may be from one to three distinct maxima in the surface and volume or mass distributions. The activity size distribution of a radionuclide-associated aerosol particle is a surface distribution (Papastefanou and Bondietti, 1987).

6

C. Papastefanou

Washout by rain greatly reduces the Aitken nuclei mode and the coarse particle mode but has little effect on the accumulation mode in the trimodal size distribution (Whitby, 1975). The origin of each mode of atmospheric aerosol size distribution can be associated with various aerosol formation mechanisms, such as Brownian motion of the particles smaller than 0.1 µm in diameter, which causes the particles to diffuse and by collisions to coagulate to larger sizes. Coagulation generates multimodal distributions and affects the shape and the chemical composition of the particles.

3. Aerosols and radiation: Generation of radioactive aerosols When a radioactive nuclide decays, electrons are stripped from the parent atom by its recoil and decay products are formed as positive ions. These ions can attract liquid and even solid material, thus forming clusters of atoms or particles in the submicron region ranging from 0.001 to 0.01 µm. Air is permanently ionised by radiation from the natural radioactivity of air and by cosmic radiation which consists mostly of positively charged particles, 85% protons, 10% alpha particles with a smaller percentage of positively charged stripped nuclei of heavier elements, such as Fe, Co and Ni, etc. Production of an ion pair requires 35.6 eV if ionisation is by alpha particles and 32.5 eV if by fast electrons. In the free atmosphere, the rate of production of small ions is in balance with the rate of neutralisation by recombination and the rate of attachment to condensation nuclei. Condensation nuclei are mostly the Aitken nuclei, which are submicrometre particles in the range 0.005 to 0.01 µm. In air containing water vapour, the positive ions are mostly hydrated protons, p+ or H+ (H2 O)n , where n may be any number between 1 and 8. Negative ions are probably mostly hydrated O−− or OH− . The formation of clusters of water molecules round ions is very rapid, but the clusters do not grow beyond about 0.001 µm (1 nm) diameter and remain as small ions until they become attached to condensation nuclei. They then become large ions. Large ions themselves can be classified in two size ranges. The nuclei mode (Aitken nuclei) centred on 0.01 µm (10 nm) is distinct from the accumulation mode, centred on 0.1 µm (100 nm). The number of particles in the nuclei mode is greater than the number in the accumulation mode, but their total surface area is less, and it is surface area which determines the probability of attachment of small ions to particles with diameter of the order of 0.1 µm (100 nm) or less (Papastefanou and Bondietti, 1987). Thus, the large ions are mostly in the accumulation mode. The distinction between small and large ions is well established in the study of atmospheric electricity. The existence of intermediate ions in the size range 0.001 to 0.010 µm (1–10 nm) has been confirmed. If sulfate or nitrate vapours are present in air, molecules of these will dissolve in the water clusters and these will then grow into the intermediate ion range. A possible mechanism for formation of radiolytic nuclei is radiolysis of water, leading to formation of H2 O2 , which then oxidises traces of SO2 to give H2 SO4 . Addition of O3 to the air also increased nuclei production, whereas addition of NO, a well-known radical scavenger, inhibits it. The radionuclide ions in the atmosphere exist in two forms: (1) as “unattached clusters” with a diffusion equivalent diameter size ranging from 0.5 to 5 nm and (2) as “aerosolattached clusters” with particle diameters varying between 5 and 3000 nm. Reported values of the diffusion coefficients in the literature range from 0.01 to 0.1 cm2 s−1 (Raabe, 1968;

Atmospheric aerosol particles

7

Porstendörfer and Mercer, 1979). Most of the previous studies on the diffusion coefficient did not consider the importance of the electrical charge of the radionuclide ions and of the air humidity. Neutralisation of radionuclide ions depends on their concentrations and on the humidity in air. Recently, Dankelman et al. (2001) determined the neutralisation rates of 218 Po ions in air by an electron transfer process in “normal” environmental outdoor and indoor air with natural trace gases and found that they are small and can be neglected under “normal humidity” (RH: 309–95%) and normal radionuclide concentrations.

4. Aerodynamic size distribution of radionuclide-associated aerosol particles (radioactive aerosols) The aerodynamic size distribution of radionuclide-associated aerosol particles is, as mentioned in Section 2, a surface distribution, and so it is trimodal, the first mode being the socalled Aitken nuclei mode, the second the accumulation mode and the third the coarse particle mode. Analytically, these modes can be summarised as follows: 4.1. Aitken nuclei mode The nuclei or Aitken nuclei mode accounts for most of the Aitken nuclei count and originates primarily from the condensation and coagulation of highly supersaturated vapours. There is evidence that a prominent nuclei mode in the size distribution indicates the presence of substantial amounts of fresh aerosol. Many particles in the nuclei mode raise the Aitken nuclei. Usually, they do not greatly increase the aerosol mass concentration because the nuclei mode rarely accounts for more than a few percent of the total mass. Whitby et al. (1975) found that the nuclei mode may contain over 25 µg m−3 of aerosol. Whitby et al. (1976) also observed distributions in which the nuclei mode contained more volume than the accumulation mode. Because particles may serve as nuclei for the condensation of water vapour, condensation is an important growth mechanism for submicrometre aerosol particles. Examples are fogs and hazes formed when the humidity exceeds 60%. 4.2. Accumulation mode The twin mechanisms of coagulation and heterogeneous nucleation (condensation of one material to another) tend to accumulate submicrometre aerosol particle mass in this mode (Whitby and Cantrell, 1976; Willeke and Whitby, 1975). Because of the sharp decrease in particles larger than 0.3 µm in diameter, little mass is transferred from the accumulation mode to the coarse particle size range. Sedimentation and impaction tend to increase the relative concentration of the smallest mechanically produced particles, and then accumulate in this mode. Salt from sea spray is typically present as particles in the 1–5 µm size range, outside the normal accumulation mode. The radionuclide-associated aerosols (radioactive aerosols) peak in the accumulation mode as this mode is the main one in terms of surface area distribution (Papastefanou and Bondietti, 1987; NRC, 1979) (Figure 1.2).

8

C. Papastefanou

4.3. Coarse particle mode In the coarse particle mode, practically all aerosol particles at relative humidities below 100% originate from mechanical processes. Most of the particles originate from condensation processes occurring in the atmosphere. Coarse particles are produced from natural and/or man-made (anthropogenic) mechanical processes. The origin, behaviour and removal processes of fine particles are almost entirely independent of the coarse particles. The multimodal nature of the size-mass distribution is supported by evidence from the size distributions of chemical elements, radionuclides or compounds in aerosols. These are emitted as smokes or fumes and exist mostly in the fine particle size range. These aerosols eventually coagulate and become mixed throughout the accumulation mode. In the accumulation mode, the geometric mean size is nearly equal to the mass or volume geometric mean size. Elements in the soil, e.g. silicon, are mostly in the coarse particle size range. Elements, such as sulphur, which are produced by condensation processes from anthropogenic sources, are principally in the fine particle size range. The size distribution of elements in soil is similar to the mass distribution of the coarse particle mode. 5. Mechanisms of formation and growth of aerosol particles The aerosol particles are formed either by coagulation and condensation processes or by gasto-particle conversion. Analytically: 5.1. Coagulation and condensation Aerosol particles tend to coalesce when they collide with each other. Since at normal humidities most particles are sheathed with moisture, the sticking probability is close to unity. Collisions between two particles lead to the formation of a new particle of larger size. This process, called coagulation, causes the size distribution to change in favour of large particles. Coagulation must be distinguished from condensation, which describes the deposition of vapour-phase material onto particulate matter. In the absence of pre-existing particles, condensation lead to the formation of new Aitken nuclei, provided that the vapour pressure of the condensing substance is sufficiently high. The last process is termed homogeneous nucleation or gas-to-particle conversion. 5.2. Gas-to-particle conversion Atmospheric gas-phase reactions may lead to the formation of condensible products, which subsequently associate with the atmospheric aerosol. Condensation may either cause the formation of new particles in the Aitken size range (homogeneous nucleation) or deposit material onto pre-existing particles (heterogeneous condensation). The gas-to-particle conversion usually starts with air free from particles. The development of the particle size range goes through three successive stages, dominated by nucleation, coagulation and heterogeneous condensation. In the atmosphere, all three processes take place concurrently. The generation of new particles requires conditions that allow the growth of molecular clusters by condensation in the

Atmospheric aerosol particles

9

phase of competition from heterogeneous condensation. Molecular clusters are formed due to weakly attractive forces between molecules, the van der Waals’ forces. Except under conditions of low temperature, it is difficult to observe clusters containing more than a few molecules.

References Aitken, J. (1923). In: Knott, C.G. (Ed.), Collected Scientific Papers of John Aitken. Cambridge University Press, London, 591 pp. Clark, W.E., Whitby, K.T. (1967). Concentration and size distribution measurements of atmospheric aerosols and a test of the theory of self-preserving size distributions. J. Atmos. Sci. 24, 677–687. Dankelman, V., Reineking, A., Porstendörfer, J. (2001). Determination of neutralisation rates of 218 Po ions in air. Radiat. Prot. Dosim. 94, 353–357. Friedlander, S.K. (1961). Theoretical considerations for the particle size spectrum of the stratospheric aerosol. Am. Meteor. Soc. J. Meteor. 18, 753–759. Junge, C.E. (1955). The size distribution and ageing of natural aerosols as determined from electric and optical data on the atmosphere. Am. Meteor. Soc. J. Meteor. 12, 13–25. Junge, C.E. (1963). Air Chemistry and Radioactivity. Academic Press, New York, 382 pp. Junge, C.E. (1972). Our knowledge of the physico-chemistry of aerosols in the undisturbed marine environment. J. Geophys. Res. 77, 5183–5200. National Research Council, NRC (1979). Airborne Particles. University Park Press, Baltimore. Papastefanou, C., Bondietti, E.A. (1987). Aerodynamic size associations of 212 Pb and 214 Pb in ambient aerosols. Health Phys. 53, 461–472. Porstendörfer, J., Mercer, T.T. (1979). Influence of electric charge and humidity upon the diffusion coefficient of radon decay products. Health Phys. 37, 191–199. Raabe, O.G. (1968). Measurement of the diffusion coefficient of radium, A. Nature 217, 1143–1145. Whitby, K.T. (1975). Modeling of Atmospheric Aerosol Particle Size Distributions, A Progress Report on EPA Research Grant No. R800971. Sampling and Analysis of Atmospheric Aerosols, Particle Technology Laboratory Report No. 253. Environmental Division, Mechanical Engineering Department, University of Minnesota. Whitby, K.T., Cantrell, B.K. (1976). Atmospheric aerosols—characteristics and measurements. In: International Conference on Environmental Sensing and Assessment, Las Vegas, Nevada, September 14–17, 1975. Institute of Electrical and Electronics Engineers, New York. Whitby, K.T., Husar, R.B., Liu, B.Y.H. (1972). The aerosol size distribution of Los Angeles smog. J. Colloid Interface Sci. 39, 177–204. Whitby, K.T., Clark, W.E., Marple, V.A., Sverdrup, G.M., Sem, G.J., Willeke, K., Liu, B.Y.H., Pui, D.Y.H. (1975). Characterization of California aerosols—1. Size distribution of freeway aerosol. Atmos. Environ. 9, 463–482. Whitby, K.T., Kittelson, D.B., Cantrell, B.K., Barsic, N.J., Dolan, D.F., Tervestad, L.D., Nieken, D.J., Wolf, J.L., Wood, J.R. (1976). Aerosol size distributions and concentrations measured during the General Motor Proving Grounds sulfate study. In: The General Motors/Environmental Protection Agency Sulfate Dispersion Experiment, Rep. EP A-600/3-76-035. Environmental Protection, Research Triangle Park, NC, pp. 29–80. Willeke, K., Whitby, K. (1975). Atmospheric aerosols: size distribution interpretation. J. Air Pollut. Control Assoc. 25, 529–534.

This page intentionally left blank

11

Chapter 2

Radioactive aerosols

1. Introduction Radioactive aerosols can be classified in the following categories: (a) Radioactive aerosols associated with radioactive nuclides of cosmogenic origin, such as 7 Be, 22 Na, 32 P and 35 S, (b) radon and thoron decay product aerosols associated with 218 Po 214 Pb, 212 Pb, 210 Pb, 210 Bi and 210 Po, (c) fission product radionuclide aerosols associated with 89 Sr, 90 Sr, 137 Cs, 103 Ru, 131 I, 132 Te and 140 Ba, (d) radioactive aerosols associated with the operation of high-energy accelerators, such as 7 Be, 22 Na, 24 Na and 52 Mn, (e) plutonium aerosols due to nuclear weapons testing or nuclear reactor accidents, and (f) mine aerosols.

2. Radioactive aerosols associated with the cosmic-ray produced radionuclides Relatively short-lived radionuclides with half-life of about a few days, of cosmogenic origin, such as 7 Be (T1/2 = 53.3 d), 32 P (14.3 d), 33 P (25.3 d), and 35 S (87.4 d) occur permanently in the atmosphere. There is another category of long-lived radionuclides with half-life larger than a year, such as 3 H (T1/2 = 12.5 y), 10 Be (1.5 × 106 y), 14 C (5730 y), 22 Na (2.6 y), 26 Al (7.3×105 y), 32 Si (280 y), 36 Cl (3.01×105 y), 39 Ar (269 y) and 81 Kr (2.1×105 y), which also occur in the atmosphere. All the above-mentioned radionuclides are formed continuously by the interaction of cosmic-ray particles with matter (atmosphere). Most of them are formed by spallation processes of light atmospheric nuclei, such as nitrogen (Z = 7), oxygen (Z = 8) and even carbon (Z = 6), or heavier atmospheric nuclei, such as sodium (Z = 11), phosphorus (Z = 15), sulphur (Z = 16), potassium (Z = 19) and calcium (Z = 20) (NCRP, 1987), when they absorb protons and even neutrons of cosmic origin. Exceptions are the production of 24 Na and 38 Cl by neutron activation of the stable isotopes 23 Na and 37 Cl, respectively. Global average production rates and concentrations of cosmogenic radionuclides in the atmosphere are summarised in Table 2.1 (Lal and Suess, 1968; UNSCEAR, 2000). Of the above radionuclides, 7 Be has a high isotope production rate in the atmosphere (8.1 × 10−2 atoms cm−2 s−1 ) and as it is a gamma-emitter (477.6 keV gammas, 11% yield), it can easily be detected and measured in the atmospheric air, in precipitation and in vegetation as well. Its average concentration in the tropospheric air is about 12.5 mBq m−3 (UNSCEAR, 2000) and 700 Bq m−3 in rainwater (UNSCEAR, 1982). Sodium-22 is also a RADIOACTIVITY IN THE ENVIRONMENT VOLUME 12 ISSN 1569-4860/DOI: 10.1016/S1569-4860(07)12002-7

© 2008 Elsevier B.V. All rights reserved.

12

C. Papastefanou

Table 2.1 Production rates and concentrations of cosmogenic radionuclides in the atmosphere (UNSCEAR, 2000) Radionuclide

3H 7 Be 10 Be 14 C 22 Na 26 Al 32 Si 32 P 33 P 35 S 36 Cl 37 Ar 39 Ar 81 Kr

Production rate Per unit area (atoms m−2 s−1 )

Annual amount (PBq a−1 )

2500 810 450 25,000 0.86 1.4 1.6 8.1 6.8 14 11 8.3 56 0.01

72 1960 0.000064 1.54 0.12 0.000001 0.00087 73 35 21 0.000013 31 0.074 1.7 × 10−8

Global inventory (PBq)

Fractional amount in troposphere

Concentration in troposphere (mBq m3 )

1275 413 230 12,750 0.44 0.71 0.82 4.1 3.5 7.1 5.6 4.2 28.6 0.005

0.004 0.11 0.0023 0.016 0.017 7.7 × 10−8 0.00011 0.24 0.16 0.08 6 × 10−8 0.37 0.83 0.82

1.4 12.5 0.15 56.3 0.0021 1.5 × 10−8 0.000025 0.27 0.15 0.16 9.3 × 10−8 0.43 6.5 0.0012

positron- and gamma-emitter (511 keV gammas, annihilation peak), but its production rate in the atmosphere is too small (8.6 × 10−5 atoms cm−2 s−1 ) and its concentration in the tropospheric air is 0.0021 mBq m−3 (UNSCEAR, 2000). All the other radionuclides referred to above are mostly beta-emitters with low production rates in the atmosphere and very low concentrations in the troposphere. 2.1. Beryllium-7 aerosols Beryllium-7 is a relatively short-lived (T1/2 = 53.3 d, τ = 1/λ = 77 d) naturally occurring radionuclide of cosmogenic origin formed by spallation processes of light atmospheric nuclei, such as carbon (Z = 6), nitrogen (Z = 7) and oxygen (Z = 8), when they absorb protons and even neutrons of the primary component of cosmic radiation (Bruninx, 1961, 1964; Rindi and Charalambous, 1967; Silberberg and Tsao, 1973), according to the following reactions: 12 C + p

→ 7 Be + 6 Li, 14 N + p → 7 Be + 24 He, 16 O + p → 7 Be + 10 B, 16 O + p → 7 Be + 7 Li + 3 He, 7 Be

12 C + n

→ 7 Be + 6 He, 14 N + n → 7 Be + 8 Li, 16 O + n → 7 Be + 10 Be, 16 O + n → 7 Be + 6 He + 4 He.

(2.1)

Once is formed in the troposphere, it rapidly associates primarily with submicronsized aerosol particles (Bondietti et al., 1987). Beryllium-7 in these fine aerosols may subsequently enter the marine as well as the terrestrial environment and vegetation via wet or dry depositional events. Following deposition, 7 Be will tend to associate with particulate material (a particle-reactive element). Beryllium-7 has come to be recognised as a potential tool in studying the description of environmental processes such as aerosol transit and residence times in the troposphere (Martell,

Radioactive aerosols

13

1970), aerosol deposition velocities (Young and Silker, 1980) and aerosol trapping by aboveground vegetation (Bondietti et al., 1984). Beryllium-7, as well as other natural radionuclides like 212+210 Pb, 22 Na, 35 S and 32+33 P, participates in the formation and growth of the accumulation mode aerosols (0.07 to 2 µm diameter), which are a major reservoir of pollutants in the atmosphere. Following its production by gas-phase nuclear transformation, this isotope condenses on the aerosol population, growing by condensation of non-radioactive species (e.g., H2 SO4 or organics) and the fate of 7 Be will become the fate of the carrier aerosols (Bondietti et al., 1988). On the behaviour of 7 Be with atmospheric aerosols it was concluded from early aerosol studies that considerable coagulation occurred during migration from the stratosphere and upper troposphere to ground level (Lockhart et al., 1965b; Gaziev et al., 1966). A histogram of the activity size distribution of 7 Be versus aerodynamic diameter (Dp ) is presented in Figure 2.1. This distribution was selected by Papastefanou and Ioannidou (1995) from 11 atmospheric aerosol sampling measurements made over an almost 2-year period at Thessaloniki, Greece (40◦ 38 N, 22◦ 58 E) by using Andersen 1 ACFM cascade impactors at a flow rate of 1.7 m3 h−1 (28.3 l min−1 or 1 ft3 min−1 ). Atmospheric aerosol size appeared to follow a trimodal distribution expected for condensation-derived aerosols. This trimodal distribution of atmospheric aerosols showed the following mode ranges: the Aitken nuclei mode ranges from 0.003 to 0.07 µm (average 0.015 µm); the accumulation mode ranges from 0.07 to 2 µm (average 0.3 µm); and the coarse mode ranges from 2 to 36 µm (average larger that 10 µm) (NRC, 1979). Young et al. (1975) reported that 7 Be is attached primarily to submicron-sized particles in the atmosphere. About 88% of 7 Be was found to be present as particles smaller than 1.1 µm in diameter, and less than 1% was on particles larger than 7 µm in diameter. This means that 7 Be aerosols are accumulation mode aerosols. This is also evident from the plot of Figure 2.1a. From the 11 measurements carried out over a 2-year period including all seasons, Papastefanou and Ioannidou (1995) reported that the activity median aerodynamic diameter (AMAD) varied from 0.76 to 1.18 µm (average 0.90 µm) and the geometric standard deviation (σg ) varied from 1.86 to 2.77 (average 2.24). The AMAD and (σg ) calculations were made by plotting the cumulative distributions on log-normal probability paper. They also showed that 60% of the 7 Be activity was associated with particles with diameter smaller than 1.1 µm. Beryllium-7 aerosol measurements carried out by Papastefanou and Ioannidou (1995) at sea level in a coastal area, in a hilly area, at 250 m height and on the top of a mountain, at 1000 m altitude, showed that the 7 Be activity size distribution dominated a smaller size range of aerosol particles with an AMAD of 0.68 µm (σg = 2.18) at a height of 250 m and an AMAD of 0.68 µm (σg = 2.24) at a height of 1000 m, showing a dependency on altitude. In marine environments at sea-level, the 7 Be activity size distribution dominated a higher size range of aerosol particles with an AMAD of 0.82 µm (σg = 1.88). Bondietti et al. (1987) in 13 measurements in an almost one-year period (June 1985–March 1986) at Oak Ridge, Tennessee at temperate latitude (35◦ 58 N, 84◦ 30 W) and with a wet climate showed that the activity median aerodynamic diameter, AMAD, varied from 0.29 to 0.50 µm (average 0.35 µm) and that the fraction of 7 Be-associated aerosols above 1.4 µm was usually between 5 and 10%, i.e. analytically 4.5% was found in the 1.4 to 2.1 µm size range, 1.1% in the 2.1 to 4.2 µm size range, and only 0.2% in sizes greater than 4.2 µm.

14

C. Papastefanou

(a)

(b) Fig. 2.1. (a) Aerodynamic size distribution of 7 Be ambient aerosols. (b) Relative activity size distribution of 7 Be in outdoor air.

They also concluded that cosmogenic radionuclides such as 7 Be and 35 S were associated with smaller aerosols than the longer-lived radionuclides such as 210 Pb, a decay product of 222 Rn of terrestrial origin, which were associated with larger aerosols. Röbig et al. (1980) reported that the distribution of the long-lived radionuclide 7 Be was shifted to large particle sizes due to long residence times of 7 Be in the atmosphere. An equivalent aerodynamic diameter of about 0.65 µm for 7 Be might have resulted from the plot of the activity size distribution of the ambient air obtained by a high volume cascade impactor

Radioactive aerosols

15

(flow rate 68 m3 h−1 ) at Göttingen, Germany (51◦ 32 N, 9◦ 55 E). Shifts to large particle sizes were also observed when the relative humidity increased during rainfall. Much later, Grundel and Porstendörfer (2004) observed that the results of the 7 Be measurements for a period of 4 weeks, carried out in outdoor air in a suburb of the town of Göttingen, showed no activity fraction in the nucleation (Aitken nuclei) mode, but a small amount of the activity (5%) in the coarse mode size range (Figure 2.1b). The accumulation mode of 7 Be-aerosols with an activity fraction of 95% has an activity median aerodynamic diameter AMAD value of 702 nm. The activity size distribution of 7 Be-aerosols depends probably on their location of formation. Most of the 7 Be atoms and the 7 Be-aerosols are generated in the upper region of the atmosphere, where different aerosol conditions exist than those in the lower atmosphere. In the activity size distribution of ambient aerosols, 7 Be is shifted to large particle sizes in the presence of pollutants. Removal of small aerosol particles in the submicron size range of the activity size distribution either by scavenging or by deposition of particles on any surface may result in a depletion of small particles in the activity size distribution. Subsequent 7 Be condensation on all aerosols effectively enriches large particles in the activity size distribution. The radioactive aerosol is only generated by attachment and there is no nucleation, in contrast to the sulfate aerosol (nucleation + attachment) (Hopke, 1991). Freshly produced 7 Be should attach to an existing aerosol or coagulate with other nuclei during its lifetime, as mean attachment half-lives are of the order of a minute or less (Porstendörfer and Mercer, 1980). Papastefanou and Ioannidou (1996) reported that sites where the AMAD of 7 Be aerosol varied between 0.62 and 0.74 µm were located on lines of direction of wind blow (local wind). These streamlines of air masses transfer pollutants from an industrial zone to urban air and the marine environment over the sea surface. In the area influenced by the industrial zone, the AMAD of the 7 Be aerosol varied between 0.82 and 1.00 µm which is higher than that observed outside the industrial zone. Winkler et al. (1998) in 46 measurements in a period of 1 1/3 years (December 1994– March 1996) at Munich-Neuherberg, Germany (48◦ 13 N, 11◦ 36 E) showed that the activity median aerodynamic diameter, AMAD, of 7 Be-aerosols ranged from 0.44 to 0.74 µm (average 0.57 µm) and that because of seasonal effects during the period of high 7 Be air concentrations, i.e. in the summer period, relatively low values of the AMAD (0.45–0.52 µm) occur. They also concluded that the activity median aerodynamic diameters, AMADs, ranged between the mass median aerodynamic diameters, MMAD, and the surface median aerodynamic diameters, SMAD, of the ambient aerosols, indicating that this radionuclide is involved in the transformation process of the tropospheric aerosols after their formation in the stratosphere and upper troposphere. Yu and Lee (2002) in 14 measurements in Hong Kong (22◦ 18 N, 114◦ 10 E) for a 3 1/2 month period (26 November 2001–8 March 2002) demonstrated that the activity median aerodynamic diameter, AMAD, of 7 Be-aerosols varied from 0.33 to 1.15 µm (average 0.67 µm). They concluded that the AMADs of 7 Be-aerosols are anticorrelated with 7 Be concentrations in air, are correlated with relative humidity, RH and mean cloud cover, while temperature does not affect the AMADs of the 7 Be-aerosols. Apart from these, Lujaniene et al. (2001) reported larger 7 Be AMAD values varying from 1.12 to 2.06 µm (average 1.45 µm) at a northern latitude in Vilnius, Lithuania (54◦ 41 N,

16

C. Papastefanou

25◦ 19 E). The data rather indicate that the activity median aerodynamic diameter, AMAD, of 7 Be-aerosols increases with increasing latitude (latitudinal effect). As cosmic radiation increases with latitude, the numbers of 7 Be atoms and ions formed also increase with latitude and so there are more 7 Be atoms and ions available either to form small aerosol particles in the nucleation (Aitken nuclei) mode and then growing or to be attached directly to the existing large particles in the accumulation and in the coarse particle modes thereby increasing the AMAD of the 7 Be-aerosols.

3. Radon and thoron decay product aerosols The gases radon (222 Rn) and thoron (220 Rn) are formed as decay products of uranium and thorium in the uranium and thorium series in soil and rocks. They are emitted from the ground into the atmosphere, where they decay and form decay products, radioisotopes of polonium (Z = 84), bismuth (Z = 83) and lead (Z = 82), which either remain airborne till they decay or are deposited in rain and by diffusion to the ground. The decay of 222 Rn (radon) and 220 Rn (thoron) in the atmosphere produces low vapour pressure decay products which coagulate with other freshly produced nuclei or condense on existing accumulation-mode aerosols. These radioisotopes include 218 Po (T1/2 = 3.05 min), 214 Pb (26.8 min), and 212 Pb (10.64 h). A longlived radioisotope in the 222 Rn decay chain, 210 Pb (22.3 y), is produced about an hour after attachment of 218 Po. There is considerable information about radon decay product aerosols in ambient air (outdoors) and even in residences (indoors) (Mercer and Stowe, 1971; Kruger and Andrews, 1976; Porstendörfer and Mercer, 1978; Kruger and Nöthling, 1979; Porstendörfer and Mercer, 1979; Porstendörfer et al., 1979; George and Breslin, 1980; Porstendörfer and Mercer, 1980; Becker et al., 1984; George et al., 1984; Bondietti et al., 1987; Papastefanou and Bondietti, 1987; Porstendörfer, 1994; Grundel and Porstendörfer, 2004). The concentration of radon (222 Rn) in ambient air is about 10 and 40 Bq m−3 in indoor air, while the concentration of thoron (220 Rn) in outdoor air is about 10 Bq m−3 and approximately the same indoors (UNSCEAR, 1993, 2000). The concentrations of the radon (222 Rn) decay products in outdoor air are 19.05 Bq m−3 for 218 Po, 3.88 Bq m−3 for 214 Pb, 5.26 Bq m−3 for 214 Bi, 0.5 mBq m−3 for 210 Pb and 0.050 mBq m−3 for 210 Po, while for thoron (220 Rn) decay products in outdoor air, they are 0.055 Bq m−3 for 212 Pb and 0.575 Bq m−3 for 212 Bi (UNSCEAR, 2000). 3.1. Formation of radon decay product aerosols Alpha decay carries away positive charge and electrons are stripped from the parent atom by its recoil. Therefore, the decay products are formed as positive ions. Air is ionised by radiation from the naturally occurring radionuclides in the air and on the ground and by cosmic rays. Production of one ion pair requires 32.5 eV if ionisation is caused by fast electrons, 35.6 eV if by alpha particles. The total energy dissipated in air per decay of 222 Rn depends on the equilibrium ratio of the radon decay products. In the free atmosphere, the rate of production of small ions is in balance with the rate of neutralisation by recombination and the rate of attachment to condensation nuclei. Condensation nuclei, otherwise called Aitken nuclei, are submicrometre particles mainly pro-

Radioactive aerosols

17

duced by combustion processes. In air containing water vapour, positive ions are mostly hydrated protons, H+ (H2 O)n , where n may be any number between 1 and 8. Negative ions are mostly hydrated O−− or OH− . The formation of clusters of water molecules round ions is very rapid, but in unpolluted air the clusters do not grow beyond 1 nm diameter and remain as small ions until they become attached to condensation nuclei. They then become large ions. Large ions themselves can be classified in two size ranges, the nuclei mode centred on 0.01 µm which is distinct from the accumulation mode, centred on 0.1 µm (Whitby, 1978). In urban air, the number of particles in the nuclei mode is greater than the number in surface area which determines the probability of attachment of small ions to particles with diameters of order 0.1 µm or less. Thus the large ions are mostly in the accumulation mode. The distinction between small and large ions is well established in the science of atmospheric electricity. Intermediate ions exist in the size range 0.001–0.01 µm. If sulfates and nitrates are formed photochemically in air, molecules of acid will dissolve in the water clusters and these will then grow into the intermediate ion range (Raes, 1985). Reaction products can be created in air by radiolytic as well as photolytic processes. The radiolytic nuclei are initially uncharged, indicating that ions are not required for their formation. A possible mechanism for formation of radiolytic nuclei is radiolysis of water vapour, leading to formation of H2 O2 , which then oxidises traces of SO2 to give H2 SO4 . Increased formation of radiolytic nuclei occurs in air when SO2 is added. Addition of O3 to the air also increases nucleus production, whereas addition of NO, a well-known radical scavenger, inhibits it. The radon decay product aerosols in the atmosphere are generated in two steps. After formation from the radon isotope (222 Rn), the freshly generated decay product radionuclides react very fast (<1 s) with trace gases and air vapours, and become small particles, called clusters or unattached radionuclides with diameters ranging from 0.5 to 500 nm (Figures 2.2a and 2.2b) (Porstendörfer, 1994; Porstendörfer et al., 2005). Besides the cluster formation, these radionuclides attach to the existing aerosol particles in the atmosphere within 1–100 s, forming the radon decay product aerosols (radioactive aerosols).

(a) Fig. 2.2. (a) Basic processes of radon decay product behaviour in air defining “unattached” and “aerosol-attached” particle activities. (b) Processes of 218 Po and 214 Pb in air.

18

C. Papastefanou

(b) Fig. 2.2. (continued)

3.2. Diffusivity of radon decay products Most of the newly formed decay product clusters are positively charged and have a high mobility. The mobility is characterised by the diffusion coefficient that chiefly controls the formation of the radioactive aerosols by attachment and the deposition on surfaces. The measured values for the diffusion coefficient, D, range from 0.03 to 0.085 cm2 s−1 , depending on

Radioactive aerosols

19

the atmospheric conditions with regard to air humidity and trace gases in the air. With very low humidity in air (<2%) the diffusion coefficient of the positively charged decay products is about 0.024 cm2 s−1 (Porstendörfer and Mercer, 1979). For higher water vapour concentrations (>30%), the diffusion coefficient is about 0.068 cm2 s−1 . Water vapours and trace gases (NOx , NH3 , SO2 ) influence the neutralisation and the cluster formation of the radioactive polonium ions. 3.3. Attachment of radon decay products to aerosol particles Attachment is mainly by diffusion. If the decay products are ions, electrostatic attraction to charged nuclei of opposite sign has a small additional effect. The rate constant for attachment, λa , is given by an equation originally applied to evaporation from small droplets, λa =

4πrND , D(rvm α)−1 + r(r + lo )−1

(2.2)

where N is the number of condensation nuclei per unit volume of air, r is their radius, lo is the mean free path of decay product molecules, D is the diffusion coefficient, α is the accommodation coefficient or sticking probability of decay products on nuclei, and vm = (RT /2πM)1/2 is the component of mean kinetic velocity of vapour molecules 2 perpendicular to a surface, as RT = 2πMvm , where R = 8.61738 × 10−5 eV/K is the gas constant (Boltzmann’s constant), T is the absolute temperature, and M is the molecular weight of decay products. The mean free path of decay product molecules lo is about 0.015 µm and if the particle radius, r, is much less than this, the second term in the denominator of Equation (2.2) is small and λa = 4πr 2 Nvm α,

(2.3)

which is the rate of collision between molecules and particles multiplied by the sticking probability, α. The attachment rate, λa , is then proportional to the surface area of the condensation nuclei. For particles of about 1 µm radius, r(r + lo )−1 is near unity. Either term in the denominator of Equation (2.2) may then be dominant, depending on whether D(rvm α)−1 is greater or less than unity. If the sticking probability, α, is less than 0.1, its value determines the attachment rate, λa , which then depends on r 2 up to r = 1 µm. As the particle size increases further, the sticking probability, α, becomes unimportant, unless it is very small and Equation (2.2) reduces to Smoluchowski’s equation λa = 4πrND.

(2.4)

20

C. Papastefanou

Fig. 2.3. Attachment coefficients of radon decay products ions. (×) Kruger and Andrews (1976); (+) Kruger and Nöthling (1979); (!) Porstendörfer and Mercer (1978); (P) Porstendörfer et al. (1979). Line is Equation (2.2) with D = 7 × 10−6 m2 s−1 .

The rate of attachment is then determined by the rate of diffusion through the boundary layer round the particle. The attachment coefficient β is λa /N , where N is the number of condensation nuclei per unit volume of air. Then, the attachment rate, λa , is λa = βN.

(2.5)

A plot of the attachment coefficient, β, of radon decay product ions is shown in Figure 2.3 (Chamberlain, 1991). The line is Equation (2.2) with the diffusion coefficient value, D = 7 × 10−6 m2 s−1 , vm = 44 m s−1 and α = 1. In the natural aerosol size distribution, typical of well-populated country districts, Junge’s (1963) natural aerosol size distribution includes particles such as sea salt and resuspended dust which extend the distribution at the largediameter end, the rate constant for attachment λa = 2.1 × 10−2 s−1 , and since N = 1.7 × 1010 m−3 for the Junge’s aerosol, the corresponding value of the attachment coefficient is β = 1.2×10−12 m3 s−1 . Measured values for the attachment coefficient β for outdoor aerosols

Radioactive aerosols

21

Fig. 2.4. Decay rates of aerosol radioactivity found on low-pressure cascade impactor, LPI stages, illustrating the relative α-counting rates, decay slopes and background.

average 1.4×10−12 m3 s−1 (Porstendörfer, 1984). This is also the normal value for attachment of ordinary atmospheric small ions to nuclei. 3.4. Activity size distributions of the radon product decay aerosols 3.4.1. Lead-214 and 212 Pb aerosol size distributions The α-activity of an atmospheric aerosol sample initially decays with the 26.8 min half-life of 214 Pb; after 3 h, the rate approaches that of 10.64 h 212 Pb (Papastefanou and Bondietti, 1987). The measured alphas in the sample are actually derived from the decay products of these

22

C. Papastefanou

Fig. 2.5. Representative plots from 46 low-pressure cascade impactor (LPI) measurements illustrating aerodynamic size (Dp ) distributions of 212 Pb and 214 Pb (R = radioactivity). (a) Type results occurred 46% of the time, (b) 39% of the time, (c) 8.7% of the time, and (d) 6.5% of the time. Lower Dp limits are arbitrary.

nuclides. Polonium-218 (T1/2 = 3.05 min), due to its significantly short half-life, does not directly contribute alpha particles in a 3-h aerosol sample, whereas alphas from 210 Po (T1/2 = 138.38 d), due to its significantly longer half-life, are negligible as might be confirmed easily by repeated examination of the count rate after the normal 24-h recording period. An example of the α-decay curves from 214 Po and 212 Bi+ 212 Po, measured with the low pressure impactor, LPI, is presented in Figure 2.4 (Papastefanou and Bondietti, 1987). The <0.52-µm stages contain most of the radioactivity, while the two stages above 0.52 µm have very low levels. In much the same way that a total air-filter sample of ambient air behaves, the measured count rates decline rapidly during the first hours due to decay of 214 Pb and 214 Bi. After 180 min (3 h), the count rates begin to follow a slope similar to that for 10.64-h 212 Pb decay. The 1σ counting uncertainties were 5% or less for the stages corresponding to aerosols below 0.52 µm and <15% for those stages collecting aerosols above 0.52 µm. Extended measurements of 24 to 66 h showed that the longer-lived radioisotope has a half-life of about 10 h. The activity size distributions of 214 Pb and 212 Pb versus aerodynamic diameter (Dp ) are represented by four subplots (Figure 2.5). These distributions were selected from 46 measurements made over a 10-month period at Oak Ridge, Tennessee (35◦ 58 N, 84◦ 17 W) by

Radioactive aerosols

23

Table 2.2 Mean observed 214 Pb and 212 Pb activity size distributions and frequency of dominance (44 measurements) Size range (µm)

214 Pb (%)

212 Pb (%)

Equal (%)

<0.08 0.08–0.11 0.11–0.23 0.23–0.52 0.52–0.90 0.90–1.4 1.4–2.0 >1.4

28.2 52.2 65.2 58.7 73.9 71.7 90.9 66.7

69.6 47.8 34.8 41.3 23.9 23.9 9.1 33.3

2.2

2.2 4.4

Papastefanou and Bondietti (1987). The radioactivity distribution appears to follow the surface area distribution expected for condensation-derived aerosols (Whitby et al., 1972) as predicted by Junge’s model of radioactivity associations with natural aerosols (Junge, 1955). About 46% of the measurements showed a radioactivity peak in the 0.11- to 0.23-µm region (subplot a), while about 39% showed a peak in the 0.23- to 0.52-µm region (subplot b). The remaining 15% of the measurements resulted in distributions similar to those in subplot (c) (8.7%) or in subplot (d) (6.5%), where 214 Pb and 212 Pb activities were highest in different size ranges in the same spectrum. On average, about 76% of the 214 Pb activity and 67% of the 212 Pb activity were found to be associated with aerosols in the 0.08- to 1.4-µm size range. These particles originate and grow by condensation and droplet-phase processes and dominate the surface area and usually the volume (mass) of the atmospheric aerosol. The activity associated with aerosols smaller than 0.08 µm (Aitken nuclei) can also be substantial, as indicated in Figure 2.5. Table 2.2 presents the average aerodynamic distributions of 212 Pb and 214 Pb as well as the frequency with which 214 Pb or 212 Pb was the dominant isotope in each size range. The Aitken nuclei fraction (below 0.08 µm) contained a higher percentage of 212 Pb activity compared to 214 Pb in 69.6% of the measurements. The predominance of 212 Pb in this fraction is also illustrated by the distributions reported in Figure 2.5. In the remaining measurements, where 214 Pb was fractionally more abundant below 0.08 µm, the disparity between the relative amounts of each isotope was not nearly as dramatic. Conversely, Figure 2.5 and Table 2.2 illustrate that 214 Pb is generally enriched in the accumulation mode aerosol, particularly between 0.11 and 0.52 µm, where most of the surface area and mass occurs (Papastefanou and Bondietti, 1987). The shift of 214 Pb to a slightly higher size distribution compared to 212 Pb was found using normal (1-ACFM) and high volume (HVI) cascade impactors (Figure 2.6). The higher flow rates of these cascade impactors, as well as the ability to measure the HVI-activity by gamma spectroscopy, give confidence that this shift is real and not a data analysis artifact. The 214 Pb activity median aerodynamic diameters, AMADs, determined with the low pressure impactors (LPI) varied from 0.10 to 0.37 µm (mean value 0.16 µm). For 212 Pb, the AMADs varied from 0.07 to 0.245 µm (mean value, 0.13 µm). These AMAD calculations were made assuming log-normal distributions. An abbreviated version of these results is presented in Table 2.3 (Papastefanou and Bondietti, 1987).

24

C. Papastefanou

Fig. 2.6. Aerodynamic size (Dp ) distributions of 212 Pb and 214 Pb activity (R) found with 1-ACFM and highvolume, HVI cascade impactors illustrating the large particle shift of 214 Pb. Lower Dp limits are arbitrary. Table 2.3 Summary of mean monthly activity median aerodynamic diameters (AMADs) and geometric standard deviations (σg ) of radon (222 Rn) and thoron (220 Rn) decay products size distributions in ambient aerosols Period (month)

Number of samples

AMAD (µm)

σg

AMAD (µm)

σg

December 1984 January 1985 March April May June 7 July August September October November December

2 2 5 5 3 4 6 6 10 2 1 2

0.16 0.17 0.17 0.18 0.20 0.21 0.16 0.12 0.13 0.09 0.07 NRa

2.35 2.46 2.33 2.08 2.28 2.98 3.27 32.51 3.68 3.13 3.20 NRa

0.15 0.09 0.12 0.17 0.20 0.17 0.23 0.10 0.11 0.09 0.12 0.13

2.80 2.96 2.46 2.25 2.35 2.38 3.52 3.11 3.44 2.85 6.70 4.35

a NR = not recorded.

Pb-214

Pb-212

Radioactive aerosols

25

Grundel and Porstendörfer (2004) more recently carried out measurements in outdoor air using an on-line alpha cascade impactor (OLACI) for a period of about one month at a northern latitude in Göttingen, Germany (51◦ 32 N, 9◦ 55 E) and showed that the activity size distributions of the short-lived radon (222 Rn) and thoron (220 Rn) decay products are about 12–19% in the nucleation mode (Aitken nuclei, 0.05–60 nm), 81–88% in the accumulation size range (60–1000 nm), and with no coarse mode (particle diameter >1000 nm). The activity median aerodynamic diameter (AMAD) of the accumulation mode varied between 332 nm for 218 Po and 347 nm for 214 Po for the short-lived radon (222 Rn) decay products and between 382 nm for 212 Po and 421 nm for 212 Pb for the thoron (220 Rn) decay products. In comparison to the short-lived radon (222 Rn) decay products, the AMADs of the thoron (220 Rn) decay products were shifted significantly to larger values. In El-Minia, Egypt, El-Hussein and Ahmed (1995) showed that the activity size distributions of the 214 Pb- and 214 Bi-attached aerosols were nearly identical and that most of the activities were associated with aerosol particles of the accumulation mode. The mean activity median aerodynamic diameter, AMAD, of 214 Pb (range 261–458 nm) and 214 Bi (range 190–620 nm) had the same value of 380 nm, but the relative geometric standard deviation, σg , of the log-normal distribution of 214 Pb (σg : 1.67–2.81, average 2.20) shows a broader activity distribution than for 214 Bi (σg : 1.67–2.25, average 2.05). Later, El-Hussein et al. (1998) from measurements in El-Minia, Egypt, found that the activity median aerodynamic diameter, AMAD, varied from 175 to 485 nm (average 330 nm) and the geometric standard deviation, σg , varied from 1.62 to 3.21 (average 2.45) for 214 Pb and from 170 to 477 nm (average 316 nm) and σg from 1.79 to 3.12 (average 2.35) for 214 Bi. Apart from this, Mohammed (1999) reported that the AMAD for the very short-lived radon decay product 218 Po-aerosols varied from 280 to 386 nm (average 340 nm) and the standard deviation, σg , varied from 2.5 to 2.9 (average 2.7), while for its decay product 214 Pb-aerosols the AMAD varied from 296 to 360 nm (average 320 nm) and the σg varied from 2.1 to 2.88 (average 2.7), which means that they are quite similar. For the relatively longer-lived thoron decay product 212 Pb-aerosols, the AMAD of the accumulation mode varied from 240 to 390 nm (average 360 nm) and the σg varied from 2.1 to 3.2 (average 2.7) (Mohammed et al., 2000). The long-lived radon (222 Rn) decay products 210 Pb and 210 Po are almost all, i.e. 93–96%, adsorbed on aerosol particles in the accumulation size range and only 4–7% of their activities are attached on nuclei with diameters smaller than 60 nm. A coarse mode of the long-lived radon (222 Rn) decay products was not measured. An AMAD-value of 558 nm for 210 Pb and 545 nm for 210 Po was measured. These are significantly larger values than those of the shortlived radon (222 Rn) decay products. Winkler et al. (1998) reported that the activity size distribution of 210 Pb in ambient aerosols was unimodal (log-normal) and associated with submicron aerosols of about 0.5 to 0.6 µm. On average, the activity median aerodynamic diameter, AMAD, of 210 Pb-aerosols (0.53 µm) has been found to be significantly lower than the average mass median aerodynamic diameter, MMAD (0.675 µm), and higher than or at most equal to the respective surface median aerodynamic diameter, SMAD, (0.465 µm) of the aerosols: SMAD < AMAD < MMAD. Variation of the atmospheric processes resulted in a variability of the activity median aerodynamic diameter, AMAD, between 0.28 and 0.74 µm for 210 Pb. While in the winter period (October to April) the AMAD of 210 Pb averaged 0.595 µm, in the summer period 210 Pb was associated with significantly smaller aerosols (AMAD: 0.43 µm).

26

C. Papastefanou

Table 2.4 Median aerodynamic diameters (MADs) and geometric standard deviations (σg ) of 210 Pb and 7 Be based on radioactivity and SO2− 4 measured with high-volume cascade impactors (HVI) Month

June 1985 July #1 July #2 July #3 August #1 August #2 September November December January #1, 1986 January #2 February #1 February #2 March #1 March #2

MAD, µm (±σg ) Pb-210

Be-7

SO2− 4

– 0.49 (1.8) – – 0.40 (2.0) 0.40 (1.9) – – – – 0.32 (2.1) 0.32 (2.0) 0.36 (1.8) 0.28 (1.6) –

0.50 (2.3) 0.30 (3.5) 0.31 (2.7) 0.48 (2.1) 0.36 (2.5) 0.30 (2.2) 0.34 (2.2) 0.34 (2.2) 0.32 (2.5) 0.32 (2.6) – – 0.34 (2.3) 0.32 (2.5) 0.29 (2.2)

0.50 0.49 0.38 0.48 0.41 0.40 0.45 0.45 0.58 0.42 – – 0.43 – 0.41

(1.80) (1.8) (2.0) (2.1) (2.2) (1.9) (2.3) (2.3) (2.5) (2.3)

(2.1) (2.1)

Suzuki et al. (1999) reported that 77% of 210 Pb and 70% of 210 Po activities were measured in size-fractionated aerosols with a diameter smaller than 0.70 µm from the coast of the Japan Sea. 3.4.2. Lead-210 vs 214 Pb aerosol size distributions Lead-210 is produced from the α-decay of 214 Po, the event used to quantify 214 Pb activity size distributions on the low-pressure (LPI) cascade impactors. While the relationship between the aerodynamic sizes of 214 Pb and 210 Pb is complicated because of the large differences in their atmospheric lifetimes, 210 Pb has always been found associated with aerosol particles larger than 214 Pb, as indicated by the differences in AMADs reported in Tables 2.3 and 2.4. Grundel and Porstendörfer (2004) showed that the long-lived radon decay products 210 Pb and 210 Po are almost all (93–96%) adsorbed on aerosol particles in the accumulation size range and only 4–7% of their activities are attached on nuclei with diameters smaller than 60 nm. AMAD-values of 558 nm for 210 Pb and 545 nm for 210 Po were measured, i.e. significantly larger values than those of the short-lived radon and thoron decay products. 3.4.3. Lead-212 vs sulfate, SO2− 4 aerosol size distributions The activity median aerodynamic diameters (AMADs) of 212 Pb (Table 2.3) and mass median aerodynamic diameters (MMADs) of SO2− 4 (Table 2.4) determined from a series of low- pressure (LPI) cascade impactor measurements made during the period January to October (1985) by Papastefanou and Bondietti (1987) are illustrated in Figure 2.7. The 212 Pb data were derived from measurements made at the same time as SO2− 4 and from measurements made to 212 214 compare Pb versus Pb. The mean aerodynamic diameter of 212 Pb was about three times smaller than that of SO2− 4 . Much less sulfate was found in the aerosol fraction below 0.08 µm

Radioactive aerosols

27

Fig. 2.7. A comparison of mean 212 Pb activity (R), and SO2− 4 mass (M) aerodynamic size (Dp ) distributions from low-pressure cascade impactor, LPI measurements. Lower Dp limits are arbitrary.

(Aitken nuclei), compared to 212 Pb. While 212 Pb was largely absent above 0.52 µm, about 20% of the SO2− 4 occurred above this size. The aerodynamic size distributions of 212 Pb and SO2− 4 were quite different, reflecting the different dependencies of surface area and volume on aerosol diameter (Friedlander, 1977). Lead-212, like the other radionuclides, becomes associated with atmospheric aerosol particles by condensation or coagulation processes which are surface-area related. Sulfate, on the other hand, is the main solute in the accumulation mode aerosol so that its steady-state distribution is proportional to volume, even though condensation of H2 SO4 may dominate its initial aerosol association. The difference also reflects residence times of 16.7 h for 212 Pb + 212 Bi and of about a week for sulfate. 214 Pb and 212 Pb aerosol size distribu3.4.4. Lead-210, 7 Be, 35 S, 32 P and sulfates, SO2− 4 vs tions The longer-lived radionuclides are associated with larger aerosols than 214 Pb or 212 Pb. An example of these differences is presented in Figure 2.8, which compares the activity size distributions of 212 Pb, 210 Pb, 7 Be, 35 S and 32 P as well as the mass size distribution of sulfate SO2− 4 found on two six-stage high volume (HVI) cascade impactors operated continuously for one week in the winter period. Figure 2.8 illustrates that 210 Pb of terrestrial origin, the cosmogenic radionuclides 7 Be, 35 S and 32 P, and sulfates SO2− 4 , were associated with larger 212 7 aerosols than Pb. In most of the analyses, the fraction of Be associated with aerosol particles above 1.4 µm was usually between 5 and 10%. In this measurement, 4.5% of 7 Be was

28

C. Papastefanou

Fig. 2.8. Aerodynamic size (Dp ) distributions of radionuclides (R) and SO2− 4 (M) derived from a 7-day aerosol sampling made using two high-volume cascade impactors, HVI. Pb-212, 35 S, 32 P, and 210 Pb were only measured on three stages, <1.4 µm. Lower Dp limits are arbitrary.

found in the 1.4- to 2.1-µm size range, 1.1% in the 2.1- to 4.2-µm size range, and only 0.2% in sizes greater than 4.2 µm. Figure 2.8 also shows that cosmogenic 35 S, measured as sulfate 2− SO2− 4 , does not have the same aerosol size distribution as stable SO4 . This same result oc35 curred in two other measurements. The abundance of S above 1.4 µm is unknown. The 35 S and 32 P activity size distributions presented in Figure 2.8b (as well as the 210 Pb activity size distribution in Figure 2.8a) represent only the three collection stages below 1.4 µm, due to the current limitation in the detection of low concentrations of these beta-emitters. Table 2.4 summarises the activity median aerodynamic diameter (AMAD) of 210 Pb and 7 Be and the mass median aerodynamic diameter (MMAD) of SO2− found in measurements 4 made in the spring period. Beryllium-7 activity size distributions are substantially smaller 210 Pb data included in Table 2.4, while limited, than SO2− 4 , regardless of the time of year. The suggest that summer aerosol particle sizes are larger than winter aerosol particle sizes. Lujaniene et al. (1997) reported that the activity median aerodynamic diameter, AMAD, of the soluble aerosols of 7 Be varied in the range 3.03 to 0.46 µm; 32+33 P in the range 0.13 to 0.30 µm; and 35 S in the range 0.11 to 1.94 µm at Vilnius, Lithuania, in northern latitudes (54◦ 41 N, 25◦ 19 E). In most cases, at ground level, insoluble aerosols carrying the cosmogenic radionuclides were larger (1 to 5 µm) than soluble ones. No coarse insoluble carriers of cosmogenic radionuclides were observed in the samples taken in the middle troposphere.

Radioactive aerosols

29

3.4.5. α-Recoil model: An explanation for the 214 Pb shift in the aerosol size distributions The longer half-life of 212 Pb compared to 214 Pb might favour the presence of larger aerosol particle associations of 210 Pb if coagulation rates are fast relative to radioactive decay rates, although impactor measurements may not be sensitive enough to record this effect. Instead, the measurements indicated that the shorter-lived chain was more often associated with larger aerosol particle sizes than the longer-lived chain. The high volume impactor measurements reported by Röbig et al. (1980) also indicated a large shift of 210 Pb relative to 212 Pb, although they did not offer an explanation for the observation. The 214 Pb shift might be explained by the fact that a significant fraction of the 3.05 min 218 Po parent of 214 Pb should attach to an existing aerosol particle or coagulate with other nuclei during its lifetime, as mean attachment half-lives are on the order of a minute or less (Porstendörfer and Mercer, 1980). When this attached 218 Po α-decays, the recoiling 214 Pb decay product can escape the aerosol particle. Complete recoil loss would occur if the diameter of the aerosol particle was smaller than the range of the recoiling nucleus. In water this recoil range is 0.13 µm (Mercer, 1976) and should be somewhat less in the atmospheric aerosol particle which has a density closer to 1.5 g cm−3 (Hering and Friedlander, 1982). By contrast, very little of the 0.46 µs 216 Po would attach before decaying to 212 Pb because of its short half-life. A considerable fraction of the 214 Pb should undergo recoil detachment, particularly from aerosol particles with diameters smaller than 0.1 µm (diameters approximating the recoil range). The probability of loss would decrease with increasing radius (Mercer, 1976). If the recondensing 214 Pb behaves like the original 214 Po, the net effect would be a shift of 214 Pb to a larger size distribution. A recoil model established by Papastefanou and Bondietti (1987) which accounts for the 214 Pb shift is as follows: First, the radon decay products are defined in terms of their atmospheric distributions at some time during their life: A is the fraction of the total 218 Po that attaches to any aerosol particle before it decays to 214 Pb; 1 − A is the fraction of the total 218 Po that decays before aerosol attachment; Ri is the fraction of the 214 Pb that, through recoil, is lost from aerosol particle size interval i following 218 Po decay; 1 − Ri is the fraction of 214 Pb that remains with aerosol particle size interval i following 218 Po decay; fi is the fraction of the total unattached atoms (218 Po or 214 Pb) that condense on particle size interval i. Then, it is assumed that fi , the fractional distribution of condensing isotopes on particle size interval i, can be derived from the measured 212 Pb size distributions (i.e., the half-life of 218 Po is short enough that the size distribution of 212 Pb represents the initial fate of condensing species). For A, the fraction of the total 218 Po atoms that attaches before decay, the recoiling 214 Pb atoms produced in any particle size interval can fractionate as follows: Fei = Afi Ri ,

(2.6)

30

C. Papastefanou

where Fei is the fraction of the total escape particle size interval i, and

218 Po

atoms that result in recoiling

Fni = Afi (1 − Ri ),

214 Pb

atoms that (2.7)

where Fni is the fraction of the total attached 218 Po atoms that result in recoiling 214 Pb atoms that do not escape particle size interval i. Summing all particle size intervals in terms of Equation (2.6) gives Fes = (2.8) Afi Ri = A fi , Ri , where Fes represents the sum of all 218 Po atoms that condense and then escape (as 214 Pb) the atmospheric aerosol particles after decay. Condensing 214 Pb, while assumed to follow the same particle size distribution as observed for 212 Pb, is derived from two sources which must be condensed separately. First, Fui = fi (1 − A),

(2.9)

is the fraction of the 214 Pb in particle size interval i

that originated by condensation where Fui of 214 Pb that originated from the decay of 218 Po before condensation; and second, Fdi = fi Fes ,

(2.10)

where Fdi is the fraction of the total 214 Pb atoms associated with particle size interval i that originated from the deposition of the 214 Pb that had escaped from all aerosol particle fractions (Fes ) following decay of attached 218 Po. These 214 Pb atoms, which originated by recoil detachment from the general aerosol population, are condensing for the second time. Because the probability of recoil detachment decreases with increasing aerosol particle diameter, the greater depletion of the smaller aerosol particles effectively results in a shift of the total 214 Pb atoms to a higher size distribution. From the above definitions the total 214 Pb atoms can be accounted for as follows: Fti = Fui + Fdi + Fni ,

(2.11)

where Fti is the fraction of the total measured 214 Pb atoms that occur in particle size interval i. Summing Equation (2.11) for all particle size intervals gives Fts = (2.12) Fti . Equations (2.11) and (2.12) were solved for values of Ri and A that gave the best fit to observed and calculated 214 Pb size distributions. The various impactor stages are used for deriving the particle size intervals. An example of the results of this model is presented in Figure 2.9. In this calculation, A was assumed to be 0.963 (i.e. 96.3% of the 218 Po was assumed to attach to ambient aerosol particles before decay to 214 Pb). The best agreement between calculated and measured 214 Pb size distributions was found for the case when calculated recoil losses (Ri ) of 100, 70, 65, and 35% occurred from the <0.11-, 0.11- to 0.23-, 0.23- to 0.52-, and 0.52- to 0.90-µm size ranges, respectively. In this example, the percentage of the total 214 Pb that underwent recoil detachment was calculated to be 79.6%, very similar to values predicted in a theoretical analysis (Mercer, 1976) and calculated from experimental data

Radioactive aerosols

31

Fig. 2.9. Results of an empirical model that calculated, for each low-pressure cascade impactor, LPI size range, the percentage recoil losses of 214 Pb (subplot a) necessary to produce a calculated 214 Pb distribution in best agreement with the observed distribution (subplot b). Lower Dp limits are arbitrary.

obtained when atmospheric aerosols were exposed to an enriched radon atmosphere (Mercer and Stowe, 1971). In solving for the best fit between calculated and observed 214 Pb size distributions, the less than 0.08-µm and 0.08- to 0.11-µm measurements were combined to improve the fit. This simplification was made because of: (1) the low fractional contribution of the 0.08- to 0.11-µm region to the total activity, (2) the likelihood that recoil losses will approach 100% in this particle size range (Mercer, 1976), and (3) coagulative growth of 212 Pb out of the <0.08-µm region might be significant during its lifetime. This model does not explicitly consider that a fraction of the measured 214 Pb actually deposits in the impactor stages as particle-associated 218 Po. The 214 Pb atoms produced under this condition would either not recoil off the plate or, if they did, they might end up associated with a smaller fraction on a lower stage. In terms of both the model and the measurements, this fraction of the total 218 Po is not operationally different from the fraction which decays

32

C. Papastefanou

before attachment (1 − A) or is not lost following recoil; both represent 218 Po which does not undergo recoil redistribution. 3.4.6. Recoil redistribution of 210 Pb following 214 Po α-decay In addition to 214 Pb, 210 Pb would also undergo α-recoil following the decay of 214 Po. Since the 214 Po is separated from 214 Pb only by β-decays, the model parameters derived above can be used to calculate the aerodynamic size distribution of 210 Pb which would result if condensation processes (and recoil) alone affect radioactivity sizes. For the values of Ri derived from Figure 2.9, the amount of recoiling 210 Pb can be calculated using Equation (2.8), substituting the measured 214 Pb size distribution (Table 2.2) for fi . The size distribution of condensing 210 Pb can be derived from Equation (2.10) where Fdi now represents the fraction of the total 210 Pb associated with particle size interval i that originated from 210 Pb which had escaped from all aerosol fractions following 214 Po decay. The calculated 210 Pb activity size distributions, using this model, are: <0.11 µm sizes, 31.7% (vs 32.7% for 214 Pb); 0.11 to 0.23 µm size range, 26.2% (vs 27.4% for 214 Pb); 0.23 to 0.52 µm size range, 30.8% (vs 30.5% for 214 Pb); 0.53 to 0.9 µm size range, 5.1% (vs 4.1% for 214 Pb); and 0.9 to 1.4 µm size range, 3.5% (vs 2.6% for 214 Pb). The activity median aerodynamic diameter (AMAD) derived from this new 210 Pb activity size distribution is 0.18 µm, substantially lower than actually measured, indicating that the presence of 210 Pb in larger aerosol particle sizes must result from post-condensation growth. 3.4.7. Radioactive aerosol particle sizes relative to growth mechanisms of sulfate, SO2− 4 aerosols Hering and Friedlander (1982), in a study of Los Angeles Basin sulfate aerosols, observed the occurrence of two categories of sulfate mass size distributions: one with a mass median diameter of 0.54 ± 0.07 µm and one with a mass median diameter of 0.20 ± 0.02 µm. Based on a growth model, they concluded that condensational processes dominated the formation of the smaller particle size distributions. At much higher sulfate loadings than found in the measurements, i.e. 11 µg cm−3 (Figure 2.8), they concluded that coagulation times of several weeks are required to increase aerosol volume distributions from 0.25 to 0.50 µm. Because air masses have much shorter residence times in the Basin, they concluded that particle size distributions near 0.5 µm were the result of rapid growth in the liquid phase. McMurry and Wilson (1982, 1983) examined aerosol particle growth in urban plumes and remote locations and concluded that both condensation and droplet-phase reactions contributed to aerosol particle growth, with humidity and sunlight being important variables. For example, an analysis of measurements in the St. Louis, Missouri, urban plume indicated that 75% of secondary aerosol volume formation was attributed to condensation and 25% to droplet-phase reactions. They further concluded that, in humid climates, droplet-phase growth was responsible for the presence of submicron volume distributions peaking in the 0.3 to 0.5 µm size range. In arid climates, the presence of submicron volume distributions, which peaked near 0.2 µm, was considered to be due to condensation-dominated growth. Measurements on radioactive aerosol particle sizes can be evaluated in terms of these mechanisms, although the significance of droplet-phase growth remains uncertain. Lead-212 size distributions reflect, of course, condensational growth. The rate at which mass (as 212 Pb atoms) deposits onto the aerosol particles decreases with increasing diameter, as predicted

Radioactive aerosols

33

by the growth law (McMurry and Wilson, 1982). The 214 Pb size distributions also reflect condensational growth but with the added complication of recoil. The oxidation rate of SO2 is at a minimum during winter, affecting both the rate of new particle production and the growth of aerosol particles by chemical reactions in the liquid phase. Therefore, the growth of 210 Pb from less than 0.2 µm to slightly over 0.3 µm during the winter may establish a lower limit for growth by coagulation during the approximately one week 210 Pb resides in the atmosphere. McMurry and Wilson (1982) calculated aerosol particle growth rates by determining the change in diameter, Dp , with respect to time. By analogy, the growth of 210 Pb in the atmosphere can be calculated by dividing the about 0.2 µm change in diameter, Dp , by the residence time of 210 Pb. The resulting growth rate is approximately 0.001 µm h−1 , indicating growth by condensation or coagulation (McMurry and Wilson, 1982, 1983). It is also concluded by the measurements that the upper limit of aerosol particle growth is 2 to 4 µm, as indicated by the 7 Be results. Additional insights into aerosol particle growth rates may be gained by evaluating the ages of 210 Pb on the 0.73 to 2.1 µm size range (stages 5 and 4 of the Sierra 236 high volume HVI cascade impactor) using the 210 Bi in-growth method (Poet et al., 1972; Moore et al., 1980). Very little 214 Pb condenses in this size range, so that the age of those fractions should reflect post-condensation ageing. This approach may be especially useful during the summer when growth rates are fast. Measurements made in climates where soil resuspension is minimal, however, mean that the fractional distribution of 7 Be can be used as a guide for evaluating if 210 Pb size distributions can be safely attributed to a condensation origin. The rates and mechanisms of aerosol particle growth in the atmosphere are obviously quite complicated. Natural radionuclides are unique tracers in that they can show: (1) where condensational growth is occurring (212 Pb), (2) the average growth which occurs during the life of an aerosol particle (210 Pb), and (3) how growth rates differ with respect to season and entry point into the mixed layer (7 Be and 210 Pb). The presence of 7 Be as a global tracer of aerosol mixing into the boundary layer also suggests that the influences of humidity, aerosol precursor concentrations, and other variables can be compared geographically. For example, it was observed that much larger 210 Pb aerosol particle sizes than reported by Moore et al. (1980) occur in Boulder, Colorado (40◦ 01 N, 105◦ 17 W). Is this a humidity or pollution effect or a methodology difference? Sanak et al. (1981) found larger 210 Pb aerosol particles in the marine boundary layer. Why? Was it because they used glass-fibre substrates on each impactor stage or do aerosols attain different particle sizes over the oceans? More systematic comparisons are obviously needed. 3.4.8. The unattached fraction of radon decay product aerosols The size distribution of 222 Rn decay product aerosols in ambient air exerts a marked effect on the calculated dose to lung tissue following inhalation of radon-contaminated air because of the alpha radiation from its decay products (218 Po, 214 Po). The maximum permissible concentration in air (MPC)a for exposure to radon and radon decay products associated with aerosol particles was found to be a function of the unattached fraction, f (ICRP, 1981): (MPC)a =

3000 1 + 1000f

pCi per litre of air,

where 1 pCi/l of air = 37 Bq/m3 of air.

(2.13)

34

C. Papastefanou

Table 2.5 The unattached fraction of radon decay product aerosols Test aerosol

Methodology

Unattached fraction

References

(1)

Open air

0.03

Duggan and Howell (1969)

(2)

Open air Open air Tunnel air Laboratory air (indoors) Laboratory air (indoors) House air (indoors) House air (indoors)

0.04–0.332 (med. 0.052) 0.005–0.01 0.49 0.07–0.40

Raghavayya and Jones (1974)

(3) (4) (5)

two filter method (parallel plate) two filter method (wire screen prefilter) screen diffusional sampler ionisation chamber two filter method impaction method

0.02–0.44

Mercer and Stowe (1971)

wire screen method

George (1972b)

(unknown)

0.01–0.06 (average 0.04) 0.05–0.12 (average 0.09) 0.02 (212 Pb) 0.015–0.16

Philips and Pai (1976)

electrodeposition

0.05–0.15

Reineking et al. (1985)

electrodeposition

0.06–0.15

Porstendörfer et al. (1987)

(6) (7) (8)

(12) (13) (14) (15) (16) (17) (18)

House air (indoors) House air (indoors) House air (indoors) Uranium mine Uranium mine Uranium mine Uranium mine Uranium mine Uranium mine Uranium mine

(19) (20)

Uranium mine Thorium mine

(9) (10) (11)

(unknown)

diffusion chamber electrodeposition diffusion tube diffusion tube electrostatic precipitation electrostatic sampler impaction method (Mercer) (unknown) impaction method (Mercer) (radial flow device)

0.10 0.01 0.00–0.73 0.10–0.20 0.05–0.20 0.05 0.002–0.12 (average 0.03) 0.028–0.088 0.00027 (212 Pb)

Van der Vooren et al. (1982) Shimo and Ikebe (1984) Duggan and Howell (1969)

Fisenne and Harley (1973)

Chamberlain and Dyson (1956) Billard et al. (1964) Craft et al. (1966) Fusamura and Kurosawa (1967) Pradel et al. (1970) Chapuis et al. (1970) George and Hinkliffe (1972) Cooper et al. (1973) Kotrappa et al. (1975)

Since a wide range of values for the unattached fraction of radon decay product aerosols exists in the literature (Table 2.5), there is considerable uncertainty in the calculation of doses to lung tissue and in determining the MPC for radon and radon decay products in air. In outdoor air, Duggan and Howell (1969), using the two filter method, found the unattached fraction to contain about 3% of the ambient radioactivity. Raghavayya and Jones (1974), using the wire screen method (screens of 120 mesh/inch), found f to range from 4.0 to 33.2%, with a median value of 5.2%. The ICRP (1981) recommends a value of between 0.00 and 0.05 for the unattached fraction in most occupational exposure conditions. In experimental work on the aerodynamic size associations of radon decay products in ambient aerosols, Papastefanou and Bondietti (1987) suggested that the abundance of unattached 222 Rn decay product aerosols in outdoor air was very small, probably less than 1%. Van der Vooren et al. (1982) calculated from literature data that an apparent unattached fraction for

Radioactive aerosols

35

Table 2.6 Lead-214 and 212 Pb activities deposited on stainless steel wire screens using a low air face velocity (u = 10 cm s−1 ) Date

20 Aug. 1987 21 Aug. 1987 24 Aug. 1987 26 Aug. 1987 27 Aug. 1987 31 Aug. 1987 4 Sept. 1987a 8 Sept. 1987a 10 Sept. 1987a 16 Sept. 1987a 17 Sept. 1987 21 Sept. 1987 22 Sept. 1987a 23 Sept. 1987a 24 Sept. 1987a

Concentration in air

Mi = 60 mesh/inch

Mi = 200 mesh/inch

214 Pb

212 Pb

214 Pb

212 Pb

214 Pb–212 Pb

(Bq m−3 )

(%)

(%)

unattached fraction (%)

214 Pb (%)

212 Pb (%)

214 Pb–212 Pb

(Bq m−3 ) − 24.28 15.33 31.49 22.02 28.91 26.55 20.10 15.65 12.41 10.98 29.10 28.48 20.50 9.21

– 0.18 0.18 0.23 0.17 0.18 0.23 0.16 0.11 0.09 0.08 0.15 0.20 0.14 0.12

0.50 2.88 2.64 1.20 – 2.03 0.74 1.18 1.08 – 1.24 1.44 0.74 – –

0.40 1.90 1.94 1.10 – 1.85 0.47 0.54 0.30 – 0.42 0.88 0.46 – –

0.10 0.98 0.70 0.10 – 0.18 0.27 0.64 0.78 – 0.82 0.56 0.28 – –

– 3.66 5.07 – 1.94 – 2.01 2.14 1.70 2.63 1.39 2.50 1.72 2.10 1.46

– 2.89 4.02 – 1.49 – 1.33 1.20 1.00 1.57 1.20 2.28 1.10 1.40 0.93

– 0.77 1.05 – 0.45 – 0.68 0.94 0.70 1.06 0.19 0.22 0.62 0.70 0.53

unattached fraction (%)

a Air passed through the screen and the filter after impaction on the stages of 1 ACFM normal cascade impactor (flow

rate: 28.3 l min−1 ).

ambient air might be 0.5–1%. Other work under laboratory conditions, using radon sources or artificial aerosols in houses or uranium mines with high particle concentrations, usually leads to higher estimates of the unattached fraction. Literature values range from 0.015 (Philips and Pai, 1976) to 0.73 (Craft et al., 1966) (Table 2.6). Papastefanou and Bondietti (1991a, 1991b) performed experiments on the diffusive deposition of aerosol particles on wire screens and, in particular, used 212 Pb deposition as a measure of the collection efficiency of the screens for aerosol-associated attached radionuclides in outdoor air, at Oak Ridge National Laboratory, Oak Ridge, Tennessee (35◦ 58 N, 84◦ 17 W) during the summer period. Stainless steel wire screens (60, 200, as well as 40 and 100 mesh/inch) were used in the experiments to collect the unattached species of radon decay products in ambient aerosols. Glass fibre filters were used as back-up to collect the radon decay products which passed the wire screens. The screens were separated from the back-up filter by a spacer screen (4 mesh/inch) to prevent contamination by the filter deposit (e.g., 214 Pb atoms) via α-recoil. They performed two series of experiments at different linear velocities of air drawn through the screen and filter at a range of face velocities. The length of each collection period was about 18 h stopping at 08:30 h. The samples were collected 1 m above ground level in the open air (although under shelter). The screens were leached with a 3 ml solution of 1 M HNO3 and the leachate was evaporated on 5.08 cm stainless steel plates using a hot plate. The activities of 214 Pb and 212 Pb were then measured by alpha counting. Less than 15 min were needed to prepare all the samples for counting after termination of sampling. It was experimentally

36

C. Papastefanou

proved that leaching is 50% more efficient than direct activity measurement of the screens. Six ZnS, silver-activated alpha-scintillation counters, each with an effective counting diameter of 5.26 cm, were used to determine the alpha activities of the samples. The counter backgrounds averaged less than 0.025 cpm. Cascade impactors were also used in the measurements. An Andersen SA 236, six-stage (with ECDs, effective cutoff diameters, of 0.41, 0.73, 1.4, 2.1, 4.2 and 10.2 µm) high-volume (20 CFM) impactor (HVI) was combined with a 200 mesh/inch stainless steel, wire screen prefilter (fitted ahead of the back-up filter). Hence the air passing through the screen and the back-up filter was relatively free of aerosol particles >0.4 µm. The face velocity of the air in these experiments was about 70 cm s−1 . Similar experiments were performed using 1 ACFM (28.3 l min−1 ) normal cascade impactors with the lowest stage having ECD 0.4 µm; almost the same as that of the 20 CFM (>0.5 m3 min−1 ) high-volume impactor, but with a face velocity of about 10 cm s−1 . The unattached fraction of radon decay product aerosols in ambient air should be contributed to by the short-lived decay products of 222 Rn; namely 218 Po (T1/2 = 3.05 min), 214 Pb (26.8 min) and 214 Bi (19.8 min), but not the relatively long-lived decay product of 220 Rn; namely 212 Pb (T 1/2 = 10.64 h), which has such a long residence time that it will be almost entirely associated with aerosols. If it is accepted that the screen collection efficiency for the unattached fraction (which has a high diffusion coefficient of 0.06 cm2 s−1 ) is 99.2% (almost 100%) for the finest screens such as 200 mesh/inch and for a low air face velocity, i.e. 10 cm s−1 , as the screen diffusion theory predicts (Thomas and Hinchliffe, 1972), then 214 Pb should dominate the screen activity. For aerosol particles >0.5 µm, the screen efficiency predicted by the diffusion theory is practically zero. However, the diffusion coefficient of the unattached particles (atoms, ions or molecules and even small clusters of atoms) is not expressed by a certain value, such as 0.06 cm2 s−1 , as considered by George (1972a) and Raghavayya and Jones (1974), or 0.054 cm2 s−1 , as considered by Chamberlain and Dyson (1956), Barry (1968), Thomas and Hinchliffe (1972), Porstendörfer et al. (1987), Reineking et al. (1985) and Mercer and Stowe (1969). Screens do not collect unattached species only. Van der Vooren et al. (1982) believed that screens always collect attached species, probably with an efficiency of 1–2%. From the screen diffusion theory (Thomas and Hinchliffe, 1972), the screen collection efficiency, η, is an exponential function of the diffusion coefficient of the collected aerosol particles, which is a function of the particle diameter, the mesh size of the screen used (either expressed in mesh/inch or mesh/cm), the wire diameter and the linear air velocity (face velocity): η = 1 − P = 1 − (0.82e−0.233μ + 0.18e−16.7μ ), where P is the aerosol penetration, 15.5Mi2 dD 100Mc2 dD = , a constant (dimensionless), u u Mc is the mesh size of the screen (openings per cm), Mi is the mesh size of the screen (openings per inch), d is the wire diameter (cm), μ=

(2.14)

Radioactive aerosols

37

Table 2.7 Lead-214 and 212 Pb activities deposited on stainless steel wire screens using a high air face velocity (u = 70 cm s−1 ) Date

Concentration in air

9 May 1985 16 May 1985 19 Aug. 1987 20 Aug. 1987 16 July 1987 16 July 1987 17 July 1987 21 July 1987 21 July 1987 22 July 1987 23 July 1987 28 July 1987 28 July 1987 29 July 1987 29 July 1987 30 July 1987 31 July 1987 5 Aug. 1987 6 Aug. 1987 7 Aug. 1987 11 Aug. 1987 12 Aug. 1987

214 Pb (Bq m−3 )

212 Pb (Bq m−3 )

16.86 11.26 16.70 35.06 19.54 19.54 19.56 27.89 27.89 26.74 34.27 8.34 8.34 18.74 18.74 21.68 23.32 5.42 8.23 11.50 17.09 24.22

0.17 0.14 0.13 0.33 0.18 0.18 0.19 0.23 0.23 0.25 0.23 0.13 0.13 0.21 0.21 0.27 0.19 0.06 0.10 0.10 0.13 0.19

u (cm s−1 )

Number of screens

Mi (mesh/inch)

214 Pb

212 Pb

214 Pb–212 Pb

(%)

(%)

unattached fraction (%)

70 70 70 70 70 70 70 70 70 70 70 70 70 70 70 60 60 60 60 60 60 60

1 1 1 1 1 2 2 1 2 3 1 1 3 1 3 1 1 1 1 1 1 1

60 60 60 60 40 40 60 100 100 100 200 200 200 200 200 200 200 200 200 200 200 200

1.19 0.56 2.00 0.50 0.40 0.80 1.70 1.12 1.90 1.94 2.00 2.93 4.57 1.77 4.34 0.84 0.58 3.48 3.54 1.43 1.26 1.21

0.56 0.27 1.60 0.40 0.40 0.80 0.90 0.90 1.60 1.90 1.10 2.44 3.66 1.63 3.35 0.72 0.60 2.47 2.97 1.10 1.03 1.10

0.63 0.29 0.40 0.10 0.00 0.00 0.80 0.22 0.30 0.04 0.90 0.49 0.91 0.14 0.99 0.12 −0.02 1.01 0.57 0.33 0.23 0.11

Measurements with u = 60 cm s−1 were performed by a 200 mesh/inch wire screen fitted ahead of a 8 × 10 inch glass fibre filter in a 20 CFM high-volume impactor.

u is the face velocity (air linear velocity) (cm s−1 ), and D is the diffusion coefficient of aerosol particles (cm2 s−1 ). The theory of diffusion through the screens in diffusion batteries, as described by a semiempirical equation derived by Cheng and Yeh (1980) correlating the screen fractional efficiency with experimental parameters, does not strongly differ from that formulated by Thomas and Hinchliffe (1972) as the penetration of particles through the screen (or the screen collection efficiency) is also an exponential function of the above parameters. One can conclude from the above summary that the attached species collected by screens might be estimated using 212 Pb to correct the 214 Pb + 218 Po for the attached component. Table 2.6 shows the results for an air face velocity of u = 10 cm s−1 and it can be seen that the screen 212 Pb activity ranged from 0.3 to 1.9% for a mesh size of 60 mesh/inch and from 0.9 to 4.0% for a 200 mesh/inch screen. For a high air face velocity (u = 70 cm s−1 ) (Table 2.7), the 212 Pb activity ranged as follows: from 0.4% (single) to 0.8% (multiple) for a mesh size of 40 mesh/inch; from 0.3 to 1.6% (single) and 0.9% (multiple) for a mesh size of 60 mesh/inch; from 0.9% (single) and 1.6 to 1.9% (multiple) for a mesh size of 100 mesh/inch and finally from 1.1 to 2.4% (single)

38

C. Papastefanou

and 3.4 to 3.7% (multiple) for a screen mesh size of 200 mesh/inch. The highest values for the finest screens are predicted by the theory (Thomas and Hinchliffe, 1972; Cheng and Yeh, 1980). According to the semi-empirical equation for diffusion of aerosols through the screens, as established by Thomas and Hinchliffe (1972), the 212 Pb activity is associated with aerosols of particle diameters ranging from 0.03 to 0.08 µm or with diffusion coefficients from 1×10−5 to 7 × 10−5 cm2 s−1 , in the case of the 200 mesh/inch, and from 4 × 10−5 to 2.5 × 10−4 cm2 s−1 for 60 mesh/inch. These 212 Pb particles are in the activity size distribution of ambient aerosols as determined by Papastefanou and Bondietti (1987) and are present in the back-up filter of the impactor used, i.e. either low-pressure impactors where the back-up filter collected particles with an aerodynamic diameter of <0.08 µm or 1 ACFM normal impactors (the filter retained particles of a diameter <0.4 µm) and 20 CFM high-volume impactors (the filter retained particles of a diameter <0.41 µm). In fact, in the experiments using high-volume impactors, HVI (u = 60 cm s−1 ), the 212 Pb screen activity, relative to that of the filter, ranged from 0.6 to 3.9% (Table 2.6). The 214 Pb activity deposited on the screen was found to be from about 0.1 to 1% higher than the 212 Pb activity, for a low air face velocity (u = 10 cm s−1 ) (Table 2.6); roughly the same result was derived at u = 70 cm s−1 (i.e., from about 0.0 to 1.0%) (Table 2.7). When a 200 mesh/inch stainless steel wire screen was used, fitted ahead of the back-up filter, with a high-volume impactor, HVI (u = 60 cm s−1 ), the 214 Pb screen activity was higher than that of 212 Pb, ranging from 0.1 to 1% (Table 2.7). The data show that the unattached fraction of 214 Pb atoms is independent of the 214 Pb concentration in air; in agreement with theoretical predictions based on the interactions occurring between radon decay products and aerosols (Raabe, 1969). It is now evident that the deposition of 214 Pb 3% in the upper stages of a 1 ACFM normal cascade impactor is not due to unattached species, as Mercer and Stowe (1971) stated, but is from other particle loadings (e.g., resuspended material) (Papastefanou and Bondietti, 1987).

4. Mine aerosols Aerosols from mine sources have particle size distributions that are determined by both the mining method and the source of the aerosol (Cantrell et al., 1987). A size distribution summarising the physical characteristics of mine aerosols is shown in Figure 2.10 (Willeke and Baron, 1993). The shape of the aerosol size distribution is influenced by the different sources contributing to the aerosol. This figure displays some of these sources and the physical mechanisms, such as condensation and coagulation, that transfer aerosol mass from one size to another. It should be noted that these mechanisms and the general shape of the distribution are not unique to mine aerosols. There are three distinct aerosol size ranges identifiable by features in measured mine aerosol size distributions: (1) the smallest of these from 0.001 to 0.08 µm is the Aitken nuclei range, which contains primary aerosol from combustion sources, such as diesel engines and secondary aerosols or chain aggregates formed by coagulation of primary aerosols,

Radioactive aerosols

39

Fig. 2.10. Size distribution summarising the general physical characteristics of a mine aerosol (from Cantrell et al., 1987).

(2) the next size range from 0.08 to approximately 1.0 µm is termed the accumulation range. It contains emissions in this size range plus aerosol accumulated by mass transfer through coagulation and condensation processes from the Aitken nuclei range, and (3) the last range from 1.0 to approximately 40 µm is termed the coarse particle range. Aerosols within this size fraction generally result from mechanical processes such as rock fracture and bulk material handling. Mineral dust aerosol re-entrained by mine haulage vehicles during the load–haul–dump cycle is an example of an in-mine emission that will contribute aerosol to this size range. For convenience, the Aitken nuclei and the accumulation ranges are combined in a single fine particle range. A division is usually made between this range and the coarse particle range at 1.0 µm. This distinction is possible because sources of aerosol in the two ranges are usually different and the coarse particle range contains very little mass transferred from the accumulation range by coagulation. In each of the ranges mentioned, the size distribution of mine aerosol can exhibit a maximum or mode which takes its name from the size range in which it occurs. Hence, the maximum in the accumulation range is termed the accumulation mode. Figure 2.11 presents a typical size distribution of aerosol mass concentration measured in a haulage entry of a diesel-equipped coal mine (Cantrell and Rubow, 1990). Here, the modal character of the size distribution is discernible even though the nuclei mode is suppressed compared to the accumulation mode. In contrast, Figure 2.12 shows a mass size distribution measured in the haulage way of an all-electric coal mine (Willeke and Baron, 1993). Here, the accumulation mode is much smaller than the coarse particle mode. Taken together, the figures indicate that diesel aerosols can make a strong contribution to accumulation mode aerosol in a diesel-equipped mine. Radioactive nuclides, such as 238 U, 218 Po, 214 Pb–214 Bi and 210 Pb–210 Po are characteristic of the mine aerosols.

40

C. Papastefanou

Fig. 2.11. Mass size distribution of mine aerosol in a diesel-equipped mine (from Cantrell and Rubow, 1990).

Fig. 2.12. Mass size distribution of mine aerosol in an all-electric equipped mine (from Cantrell and Rubow, 1990).

5. Fission product radionuclide aerosols Following an accident at a nuclear power station, great amounts of radioactive aerosols are emitted into the troposphere, because airborne fission product radionuclides interact with the environment and a carrier is responsible for their long-range transport and atmospheric diffusion. Radioactive nuclides, such as 103 Ru, 131 I, 132 Te and 137 Cs, characterise these aerosols.

Radioactive aerosols

41

The particle size of a fission product aerosol and the distribution of fission products between particulate and vapour phases depend on the mechanism of release to the atmosphere. In a nuclear explosion, some physicochemical fractionation of radionuclides may occur, particularly if the explosion is near the ground. Everything in the vicinity is vapourised by the heat of the explosion, but within less than a minute the fireball cools to a temperature in the range 1000–2000 ◦ C, and refractory materials, such as metal oxides and silicates, condense to form aerosol particles (Glasstone and Dolan, 1977). Refractory fission products and plutonium are incorporated in these aerosol particles. Less refractory fission products condense later onto the surfaces of the aerosol particles. Those with gaseous precursors, e.g. 90 Sr and 137 Cs, condense as they are formed by decay of their parent nuclides. The refractory fission products are distributed between particles according to their mass or volume (diameter cubed). The more volatile fission products and those with volatile parents are distributed according to a power of dp , between dp and dp2 . Per unit mass, there is then more activity in the smaller particles. The particle size depends on the volumetric concentration of condensing material and decreases as the fireball expands. The earlier a fission product condenses, the larger the particles in which it is incorporated and the quicker it is lost by fallout to the ground. At distances of about 100 km from ground zero of the explosion site the aerosol particle size of fallout ranges from a few µm to a few 100 µm. Fission products with volatile precursors are enhanced by about a factor of 2 compared with refractory fission products. The fractionation is greater in the smaller particles (Hicks, 1982). The fireballs created in an explosion at high altitude are very large, but no ground-based material is incorporated. So, the condensation aerosol particles are very small, with 90% of the activity in particles less than 0.3 µm in diameter (Drevinsky and Pecci, 1965). Coagulation takes place with the natural stratospheric aerosol which is thus labelled with fission products. In the troposphere, there is further coagulation and near ground most activity is in the 0.3–1µm size range (Lockhart et al., 1965a). Any fractionation of fission products in the explosions is evened out by the process of coagulation and there is no evidence of fractionation in particles sampled near ground. The Chernobyl reactor accident (April 26, 1986) produced elevated radioactivity in ambient air. Jost et al. (1986) sampled aerosols with an Andersen impactor in Spiez (46◦ 41 N, 7◦ 39 E), Switzerland from April 30 to May 13, 1986 and in Zurich (47◦ 23 N, 8◦ 32 E), Switzerland from May 2 to 8, 1986 with a Berner impactor, so soon after the accident. In Figure 2.13 (data from Spiez), it is shown that 131 I had a rather different activity size distribution from other fission product radionuclides such as 103 Ru and 137 Cs. Most 131 I was found in the size fraction on the back-up filter (<0.47 µm), whereas 103 Ru and 137 Cs showed a pronounced maximum at 0.93 µm (geometric mean diameter). In Figure 2.14 (data from Zurich), a similar activity size distribution is shown with the maximum for 131 I at 0.35 µm and at 0.71 µm for 103 Ru, 132 Te and 137 Cs. Different activity size distributions were due to the different cutoff diameters of the aerosol sampling devices. The activity size distributions for 103 Ru, 132 Te and 137 Cs are 2− − very similar to the concentration patterns of prominent ions, such as NH+ 4 or SO4 and NO3 as determined by ion chromatography. As iodine (131 I) was mainly (about 80%) present in the gas or vapour phase, the adsorption of gaseous species onto aerosol particles was strongly surface correlated (Papastefanou and Bondietti, 1987). These surface distributions for environmental aerosols usually peak at size fractions with particle diameters 0.4 µm (Seinfeld, 1986; Whitby et al., 1972). Therefore,

42

C. Papastefanou

Fig. 2.13. Activities of 103 Ru, 137 Cs and 131 I in size-fractionated aerosol samples in Bq m−3 sampled at Spiez, Switzerland from 09:00 h April 30 to 09:00 h May 13, 1986 using an Andersen aerosol cascade impactor. Note that the low activities are due to averaging over the 14-day sampling period.

it is concluded that the activity size distribution observed for 131 I does not reflect the primary particles released at Chernobyl, but rather the surface distribution of the aerosols at the sampling sites. By contrast, 103 Ru, 131 Te and 137 Cs were presumably ejected as particles or attached very early to aerosol particles and grew by coagulation with other particles during transport. It is therefore reasonable that their activity size distribution resembles that of the main inorganic ions (sulfate, nitrate and ammonium) which may also be transported over long distances. Bondietti and Brantley (1986) evaluated the aerodynamic sizes of the aerosol-associated fission product radionuclides from the Chernobyl accident in four aerosol samplings made during the period from May 7 to June 13, 1986 at Oak Ridge, Tennessee (35◦ 58 N, 84◦ 17 W), using Sierra high-volume cascade impactors and low-background intrinsic Ge coaxial and well type detectors. The Chernobyl activity was shown to be mostly sub-micrometre in size, and the particle size increased significantly over the measurement period. They also observed that only a small fraction of the aerosol-associated 131 I was soluble in CHCl3 , whereas both 103 Ru

Radioactive aerosols

43

Fig. 2.14. Activities of 131 I, 132 Te, 137 Cs, and 103 Ru in size-fractionated aerosol samples at Zurich-Hoenggerberg, Switzerland from 09:15 h May 2 to 23:15 h May 8, 1986 using a Berner cascade impactor.

and 134+137 Cs were less soluble than natural radionuclides, indicating some association with aerosols produced by the accident. They measured the aerosol-associated 131 I and its chemical composition using one cascade impactor operated from May 7 to 16, 1986 with a charcoalimpregnated final filter instead of the normal glass-fibre filter. Based on the 131 I activity size distributions on both charcoal-equipped and normal cascade impactors, they estimated 40% of the airborne 131 I was aerosol-associated and that very little gas-to-aerosol transformation of 131 I occurred during its transit over long distances. The <0.41-µm aerosol 131 I fraction consisted of about 20% I2 or I− (or other CHCl3 -soluble species). No IO− 3 was detected. The 0.41- to 0.73-µm fractions from the measurements showed similar CHCl3 -soluble fractions. Table 2.8 lists the activity mean aerodynamic diameter, AMAD, in µm, of fission product radionuclide aerosols of Chernobyl origin together with the 7 Be-associated aerosols of cosmogenic origin at Oak Ridge, Tennessee (35◦ 58 N, 84◦ 17 W). The particle sizes increased considerably during the measurement period. By contrast, cosmogenic 7 Be, which is initially associated with very small aerosol particles, was continually being added to the aerosol population. In the first two measurements, 2% of the fission product radionuclide aerosol particles were found to be larger than 2.1 µm, strongly suggesting the absence of any significant mechanical suspension of reactor material into the free troposphere. The behaviour and sizes of the fission product radionuclides are consistent with the long-range transport of isotopes re-

44

C. Papastefanou

Table 2.8 Activity median aerodynamic diameters (AMADs, in µm) of Chernobyl fission products and cosmogenic 7 Be measured at Oak Ridge, Tennessee Date

131 I

137 Cs

134 Cs

103 Ru

106 Ru

140 Ba

7 Ba

7–16 May 20–23 May 30 May–3 June 6–13 June

0.37 0.32 ND ND

0.40 0.48 0.67 0.72

0.40 0.43 0.66 0.71

0.37 0.44 0.60 0.62

ND 0.44 ND ND

ND 0.45 ND ND

0.38 0.39 0.48 0.36

ND, not detected or measured with >25% uncertainty.

leased by volatilisation, which condensed on or with ambient or accident-produced aerosols. Zircon-95 or other rare-earth fission product radionuclides were not found in the aerosol samples. About 27% of the 134+137 Cs and 46% of the 103+106 Ru were not rapidly dissolved in cold 2 M HCl. This behaviour was very different from that observed for natural 7 Be, 210 Pb or 212 Pb, which become associated with natural aerosols by surface condensation and are quite soluble. That poor solubility may indicate that some of the aerosol-associated fission product radionuclides were incorporated into reactor-derived materials which condensed in the hot plume leaving the reactor core. In contrast to the findings of Bondietti and Brantley (1986), Jost et al. (1986) found no increase of the activity median aerodynamic diameter, AMAD, during the measurement period. They believed that this increase might not reflect the original release at Chernobyl, but was due to transport effects. Apart from this, the activity size distribution of 137 Cs from the Chernobyl fallout was very similar to the activity size distribution in the fallout from nuclear weapons tests (Lockhart et al., 1965b), in contrast to the activity size distribution found in the stratosphere, which is shifted towards smaller particles (Persson and Sisefsky, 1971). Lujaniene et al. (1997) reported that the activity median aerodynamic diameter, AMAD, of the soluble aerosols of 137 Cs of Chernobyl origin varied in the range 0.10 to 0.86 µm at Vilnius, Lithuania (54◦ 41 N, 25◦ 19 E), and the size of caesium radioisotopes stuck to insoluble aerosol particles in all the samples was similar (about 1 µm). Kauppinen et al. (1986) also sampled ambient radioactive aerosols of Chernobyl origin at Helsinki, Finland (60◦ 10 N, 24◦ 58 E), during May 7–14, 1986, for evaluation of the activity size distribution. They used 11-stage multijet compressible flow low-pressure Berner impactors (modified HAUKE 25/0.015 LPI) covering 0.03–0.16 µm aerodynamic diameter size range, AMAD. A Liu-type aerosol inlet was connected to the LPI inlet to minimise the effect of wind on the aspiration efficiency. Polycarbonate films of 10-µm thickness were used as impaction substrates. To prevent coarse-particle bounce, the films on the stages collecting particles greater than 1 µm were greased thinly with a uniform layer of Apiezon L-grease. The amount of grease on the film was 30–300 µg. The mass of collected particles was determined by weighing substrate films carefully before and after sampling with a microbalance. Before gravimetric analysis, the films were exposed to an ion stream generated by a polonium α-active source, in order to reduce the effect of electrical charge on the weighing results. Before the radioactivity analysis, the films (doughnut shaped, in the middle of which are the particle deposition spots) were cut into four equal pieces and laid above each other, in order

Radioactive aerosols

45

Fig. 2.15. Mass size distribution curves of ambient aerosols in Helsinki, Finland during May 1986 after the Chernobyl accident.

Fig. 2.16. Activity size distribution curves of Helsinki, Finland atmospheric aerosols.

to achieve better γ -counting geometry. As a result, the samples could be approximated by a diskette geometry. The amount of radioactive isotopes in the aerosol samples was determined by measuring its γ -spectrum with a cylindrical Ge(Li) detector. During the measurements, the crystal was ventilated with aged pressurised air to reduce the iodine background. In Figure 2.15, the log-normal fits of mass size distribution of three samples of Helsinki atmospheric aerosols are shown. They are clearly bimodal, exhibiting accumulation and coarseparticle modes and a gap between these at a size range of 1–2 µm a.e.d. In Figure 2.16, the log-normal fits of 103 Ru, 131 I, 132 Te and 137 Cs activity size distributions of four samples of Helsinki atmospheric aerosols are shown. All the activity size distributions are unimodal. The modal parameters, such as the activity mean aerodynamic diameter, AMAD, the geometric standard deviation, σg , and the modal concentration for the accumulation-mode mass and surface area size distributions (calculated from the mass size distribution assuming unit density spherical particles) and for the activity size distributions are given in Table 2.9. Iodine-131

46

C. Papastefanou

Table 2.9 Geometric mean aerodynamic diameters, DGae , in µm, geometric standard deviations, σg , noted as SG, and modal concentrations, C, in mBq m−3 for activity size distributions and for the accumulation mode mass and surface area size distributions Sample 2

Mass Surface area 103 Ru 131 I 132 Te 137 Cs

Sample 3

Sample 4

DGae , µm

SG

C, mBq/m3

DGae , µm

SG

C, mBq/m3

DGae , µm

SG

C, mBq/m3

0.42 0.32 0.83 0.33 0.93

1.7 1.7 1.8 1.7 ∼1.5

20a 321b 14 98 13

0.44 0.31 0.63 0.36

1.8 1.8 1.9 2.3

15a 251b 46 39

0.57 0.38 0.65 0.57

1.9 1.9 1.7 2.0

25a 318b 12 17

0.63

1.8

9

a Accumulation mode mass concentration, µg/m3 . b Accumulation mode surface area concentration, µm2 /cm3 .

activity size distributions differ clearly from the activity size distributions of 103 Ru, 132 Te and 137 Cs. The activity mean aerodynamic diameter, AMAD, of 131 I was smaller than that of 103 Ru, 132 Te and 137 Cs. In two samples (samples 2 and 3), the AMAD was almost equal to the surface area mean aerodynamic diameter, SAMAD, of the accumulation mode, and in one sample (sample 4), the AMAD was equal to the mass mean aerodynamic diameter, MMAD, of the accumulation mode. The AMAD, of 131 I seems to grow, whereas the AMAD, of 103 Ru decreases, as the aerosol particles age. Another case of radioactive aerosols associated with fission product radionuclides is that formed in the nuclear power plant environment. Environments in nuclear power plants are quite complex. Vapourisation, condensation, fragmentation, and gas-to-particle conversion all evolve in reactor vessels and containments. Practices such as sandblasting, grinding and thermal arc cutting further complicate the situation of aerosol production and dispersion. These aerosols, however, may be grouped into several categories, e.g., the coarse mode and the accumulation mode. Coarse mode particles with a diameter greater than about 1 µm are generally created by mechanical processes such as fragmentation. Finer particles, with a diameter <1 µm, are mostly derived from processes such as combustion and gas-to-particle conversion. Analyses of such radioactive aerosols were performed by Yu et al. (1993) at Chin-Shan Nuclear Power Plant, NPP in Taiwan, comprising a 636-MWe boiling water reactor (BWR) that has been in operation since 1977. They used Andersen ambient cascade impactors (1 ACFM normal cascade impactors at an airflow rate of 28.3 l min−1 ) with a preseparator to discriminate large particles with an aerodynamic diameter greater than 10 µm. For the segregation of smaller size aerosol particles in the accumulation mode, they applied a low-pressure adaptor to the Andersen cascade impactor. At an airflow rate of 3 l min−1 , the effective cutoff aerodynamic diameter of sampled aerosols by the low-pressure Andersen cascade impactor (LPI) was between 0.08 and 35 µm. Fission products including 131 I, 140 Ba and 140 La and the activation product 60 Co were the main radionuclides of the aerosols generated by sandblasting the steam turbine components. Long-lived activation products 60 Co, 54 Mn and 134 Cs and the

Radioactive aerosols

47

(a)

(b) Fig. 2.17. Activity size distribution of 60 Co-bearing aerosols collected during the thermal-arc-cutting process, shown in terms of the following: (a) cumulative distribution plotted on log-probability axes, and (b) differential distribution. Solid curves are the fitting results using a log-normal function.

fission product 137 Cs constituted the major radionuclides in aerosols produced by the thermalarc-cutting process on the waste concentration vapour body (WCVB). Grinding activities on the heat-exchanger pipes produced aerosols containing mostly activated radionuclides, such as 60 Co and 54 Mn. Aerosols produced at the off-gas sump room mainly contained short-lived noble gas decay products, such as 88 Rb, 138 Cs and 139 Ba. A plot of the activity size distribution of 60 Co which is a unimodal log-normal distribution is shown in Figure 2.17, where C is the aerosol concentration and dae is the aerodynamic diameter. The AMAD is 3.2 µm and the geometric standard deviation, σg , is 1.9. A similar plot of the activity size distribution which is a bimodal log-normal distribution for 137 Cs-bearing aerosols is shown in Figure 2.18. The AMAD of the accumulation mode is 0.29 µm and the geometric standard deviation, σg , is 2.2, while the AMAD of the coarse mode is 3.2 µm and

48

C. Papastefanou

(a)

(b)

(c) Fig. 2.18. Activity size distribution of 137 Cs-bearing aerosols collected during the thermal-arc-cutting process, shown in terms of the following: (a) cumulative distribution with one mode fitting, (b) cumulative distribution with two modes fitting, and (c) differential distribution with two modes fitting. Solid curves are the fitting results using a log-normal function.

the geometric standard deviation, σg , is 1.9. Caesium-137 could be produced directly from fission or as the decay product of 137 Xe, a gaseous fission product radionuclide. The direct 137 Cs (fission product radionuclide) gives rise to aerosols in the coarse mode, whereas the coagulation of 137 Xe decay product (137 Cs) gives rise to aerosols in the accumulation mode. A plot of the activity size distribution for some major radionuclides associated with aerosols is shown in Figure 2.19. Corresponding values of AMAD and geometric standard deviation, σg , are listed in Table 2.10.

Radioactive aerosols

49

Fig. 2.19. Fitting results of the differential activity size distribution of radioactive aerosols collected during different practices and in various environments.

6. Radioactive aerosols associated with the operation of high-energy accelerators During accelerator operation radioactive nuclides are produced by the interaction between the primary and secondary particles (E 30 MeV) from the machine and the atmospheric air in the accelerator halls. Spallation reactions in solid machine parts can also lead to the formation of radioactive nuclides. If the air is confined in the accelerator hall there will be no release of radioactive nuclides into the outside zones during operation of the machine. When the machine is stopped an unexpected concentration of radioactive nuclides may be present in the air. Table 2.11 lists all the radioactive nuclides with a half-life >1 s (in decreasing half-life order) which can be produced by irradiation of the atmospheric air in a proton ac-

50

C. Papastefanou

Table 2.10 Activity median aerodynamic diameters, AMADs, in µm and geometric standard deviations, σg , noted as GSD, of the activity size distributions for aerosol collected in different work environments Work environment

Sandblasting

Coarse mode Fraction (%)

AMAD (µm)

GSD

60 Co

100 100 100 100 100 100

3.4 2.9 3.0 3.1 3.1 3.1

2.3 2.4 2.3 2.3 2.3 2.4

97 97 51 50

3.2 3.1 3.2 3.2

100 100 100 100 100

131 I 137 Cs 140 Ba 140 La 141 Ce

Thermal arc cutting

60 Co 54 Mn 134 Cs 137 Cs

Grinding

Accumulation mode

Nuclide

51 Cr 54 Mn 58 Co 59 Fe 60 Co

Noble gas progeny (off-gas sump room)

88 Rb

Noble gas progeny (SJAE room)

Gross beta

138 Cs

Fraction (%)

AMAD (µm)

GSD

− − − − − −

– – – – – –

– – – – – –

1.9 1.9 1.9 1.9

3 3 49 50

– – 0.29 0.28

– – 2.2 2.2

4.1 4.2 3.9 4.2 3.7

2.0 2.0 1.9 2.0 2.0

− − − − −

– – – – –

– – – – –

− 4

– 7.8

– 2.4

100 96

0.41 0.43

1.8 1.9

4

6.2

1.7

96

0.44

1.9

Table 2.11 High energy hadron fluxes around the aerosol sampling site (particles cm−2 s−1 ) Position

30 MeVa

50 MeV

100 MeV

400 MeV

A B C

1.2 × 107 4.9 × 106 7.6 × 106

1.9 × 106 6.1 × 105 9.5 × 105

4.4 × 105 1.1 × 105 3.7 × 105

1.8 × 105 1.0 × 105 1.5 × 105

27 Al(p, spal.)24 Na, 27 Al(p, spal.)22 Na, 27 Al(p, spal.)7 Be, 63,65 Cu(p, spal.)52 Mn,

Eth : ∼30 MeV, Eth : ∼50 MeV, Eth : ∼400 MeV, Eth : ∼100 MeV,

σ σ σ σ

= 7.92 mb(6) , = 10.4 mb(7) , = 8.3 mb(7) , = 7.6 mb(8) .

a The 27 Al(n, x)24 Na reaction contributes somewhat to the 24 Na production (E : ∼6 MeV), however the calculation th was made by assuming that all 24 Na was produced by the spallation reaction.

celerator hall, from spallation, (γ , n) reaction and thermal neutron capture. The contribution from other possible reactions with high energy neutrons is negligible (Rindi and Charalambous, 1967).

Radioactive aerosols

51

Fig. 2.20. The log-normal distribution of 7 Be aerosol. (—) Experimental; (- - -) calculated [β(r)N (r)].

Radioactive dust is produced from activated machine parts by normal erosion or mechanical wear. Dust in the accelerator halls is also activated when it is in suspension in the air or when it is deposited on the machine components. The principal radioisotopes found in dust are those produced from the constituents of the machine parts. Radioactive aerosols are produced from activated air in the accelerator hall (Charalambous and Rindi, 1967). High-energy accelerators are in operation at CERN, Geneva, Switzerland (46◦ 12 N, 6◦ 09 E), at Fermilab, Battavia, Chicago, IL, USA (41◦ 53 N, 87◦ 38 W), at JINR, Dubna, Moscow Region, Russia (55◦ 45 N, 37◦ 35 E) and at DESY, Hamburg, Germany (53◦ 33 N, 9◦ 59 E). Thus far, little is known of the size distribution and the mechanisms of aerosol formation. Radioactive atoms are formed with high kinetic energy and/or charges, and can interact with their surroundings as hot atoms. Subsequent reactions with ambient non-radioactive aerosols are supposed to result in the formation of radioactive aerosols. It seems likely that the 7 Beaerosols are formed by the attachment of 7 Be to non-radioactive aerosols. Kondo et al. (1984) used a parallel plate diffusion battery and a condensation nuclei counter, CNC, to collect aerosols at the KEK 12-GeV proton synchrotron in the National Laboratory for High Energy Physics at Tsukumba-gun, Ibaraki-ken, Japan (35◦ 42 N, 139◦ 46 E). They examined the aerosol size distribution of 7 Be produced by the spallation reaction of nitrogen, N and/or oxygen, O atoms in air in an integrated high energy hadron flux around the sampling inlet of about 5 × 106 particles/cm2 s for particles with E 30 MeV, which decreased with increasing energy. A plot of the log-normal size distribution of 7 Be-aerosols is shown in Figure 2.20. The mean geometric radius was rg = 0.027 ± 0.005 µm and for the geometric standard deviation, log σg = 0.36 ± 0.02 or σg = 2.3. They observed that naturally occurring radioactive aerosols of radon and/or thoron decay products also have a similar size distribution, but their mean geometric radii were slightly greater than the values obtained for the 7 Be-aerosols. They also reported that the size distributions of radioactive aerosols of fission product radionuclides due to atomic bomb tests differ in the stratosphere and troposphere depending on the characteristics of ambient non-radioactive aerosols. For the non-radioactive aerosols, the size distribution was of a log-normal shape with a mean geomet-

52

C. Papastefanou

Fig. 2.21. Aerodynamic size distribution and the fitted log-normal distribution from the cascade impactor measurement of 238 Pu particles resuspended from the fixed air filter samplers (FAS) filter.

ric radius rg = 0.01 µm and for the geometric standard deviation log σg = 0.3 or σg = 2.3, whereas the aerosol concentration was (1.5–4.3) × 105 cm−3 . 7. Plutonium aerosols due to nuclear weapons testing or nuclear reactor accidents Plutonium (Z = 94) is a man-made transuranic element and only infinitesimal traces occur naturally. It melts at 641 ◦ C and boils at 3330 ◦ C. Plutonium-239 is formed in nuclear reactors by neutron capture in 238 U, followed by two successive beta decays: 238 U + 1 n

→ 239 U

β− β− 239 Np 239 Pu 23.5 m 2.35 d

α 2.411×104 y

.

(2.15)

Further neutron captures lead to 240 Pu and 241 Pu. Plutonium-238 is formed from 239 Pu by (n, 2n) reactions, or from 235 U by three successive neutron captures and two beta decays. Plutonium aerosols can be formed in various ways, including: (a) oxidation or volatilisation of Pu metal, (b) oxidation or volatilisation of irradiated U or UO2 , (c) droplet dispersion from aqueous solutions or suspensions of Pu, and (d) resuspension of soil or dust which has become contaminated with Pu. The particle size of Pu-aerosols is very variable, depending on the mode of formation. Plutonium is so toxic that processing and fabrication are always done in sealed cells or glove boxes, but accidental dispersion of aerosol occurs from time to time. Cheng et al. (2004) used a Lovelace Multi-jet cascade impactor for collecting aerosols during the release of an undetermined amount of 238 PuO2 from a glovebox system, in a room within a Plutonium Facility (PF-4) at Los Alamos National Laboratory (LANL), Los Alamos, NM. They examined the plutonium particle activity size distribution. A plot of the aerodynamic size distribution and

Radioactive aerosols

53

Fig. 2.22. Particle size distributions of Pu oxidising in air at 20 ◦ C (A), 123 ◦ C (B), 450 ◦ C (C) (Carter and Stewart, 1971). Pu fume from exploded wire (D) (Brightwell and Carter, 1977). Pu fume from fire (E) (Mann and Kirchner, 1967).

the fitted log-normal distribution of 238 Pu-particles resuspended from the FAS (fixed air filter sampler) filter is shown in Figure 2.21. The activity median aerodynamic diameter, AMAD, was 4.7 µm and the geometric standard deviation, σg , was 1.4. In Figure 2.22, the curves A, B and C show particle size distributions obtained by Carter and Stewart (1971) in laboratory experiments on the oxidation of Pu metal in air. In controlled oxidation at temperatures below the ignition point (about 500 ◦ C), scaly, friable, oxide particles were produced, with median diameter increasing with temperature. Few particles less than 1 µm in diameter were found. When the delta alloy of Pu was used, the oxide was more adherent and the particle size was larger. Quite different results were obtained when Pu metal was heated in argon above its melting point and droplets of molten metal were allowed to fall down a column in air. The vigorous oxidation raised the temperature of the drops sufficiently to generate Pu vapour which condensed as a fume, comprising particles of about 0.1 µm in diameter aggregated in chains. Fume was also generated when an electrical current was passed through a plutonium wire, giving the particle size distribution shown in curve D of Figure 2.22 (Brightwell and Carter, 1977). Curve E in Figure 2.22 shows the particle size

54

C. Papastefanou

Table 2.12 Plutonium aerosols in ventilation ducts of USAEC plants (Elder et al., 1974) Location

Major operations

Mean airborne activity (Bq m−3 )

AMAD (µm)

σg

A B C D E

R&D R&D Fabrication Recovery Fabrication

2 × 103 2 × 104 1 × 104 1 × 105 2 × 104

1 to 3 1 to 4 3 to 5 0.1 to 1 2 to 4

1.5 to 3 1.5 to 3.5 1.5 to 2.5 1.5 to 4 1.5 to 3

distribution of Pu fume from fire, following combustion of Pu metal chips in a production area at Rocky Flats, Colorado, in 1964 and airborne contamination was widespread (Mann and Kirchner, 1967). The activity median aerodynamic diameter, AMAD, was 0.3 µm. Moss et al. (1961) reported during normal operations in a Pu production area at Los Alamos, NM, activity median aerodynamic diameters, AMADs, in the range 0.15 to 0.65 µm. Elder et al. (1974) obtained a statistical distribution of aerosol particle sizes by using multi-stage Andersen cascade impactors in the ventilation ducts at three plutonium processing plants. When the size distribution was plotted on log-probability paper, a reasonable approximation to a straight line was obtained with about 80% of the samples, but some showed bimodal distributions. Table 2.12 shows the activity median aerodynamic diameters, AMADs, and the geometric standard deviation, σg , of plutonium aerosols in ventilation ducts of USAEC (United States Atomic Energy Commission) plants. The grinding and machining processes in the fabrication areas produced relatively large aerosols, with more than 50% of the Pu activity in the 1–5 µm range. In the plutonium recovery plant, 70% of the activity was in sub-micrometre particles, and this area presented the worst problems in designing filtration plant to reduce the effluent to the desired level of 2 × 10−3 Bq m−3 . References Barry, P.J. (1968). Sampling for airborne radioiodine by copper screens. Health Phys. 15, 234–250. Becker, K.H., Reineking, A., Scheibel, H.G., Porstendörfer, J. (1984). Radon daughter activity size distributions. Radiat. Prot. Dosim. 7, 147–150. Billard, F., Miribel, J., Madeleine, G., Pradel, J. (1964). Methodes de mesure du radon et de dosage dans les mines d’uranium. In: Radiological Health and Safety in Mining and Milling at Nuclear Materials, vol. I. IAEA, Vienna, STI/PUB/78, pp. 411–423. Bondietti, E.A., Brantley, J.N. (1986). Characteristics of Chernobyl radioactivity in Tennessee. Nature 322, 313–314. Bondietti, E.A., Hoffmann, F.O., Larsen, L.L. (1984). Air-to-vegetation transfer rates of natural submicron aerosols. J. Environ. Radioact. 1, 5–27. Bondietti, E.A., Papastefanou, C., Rangarajan, C. (1987). Aerodynamic size associations of natural radioactivity with ambient aerosols. In: Radon and Its Decay Products: Occurrence, Properties, and Health Effects. In: ACS Symposium Series, vol. 331. American Chemical Society, Washington, DC, pp. 377–397. Bondietti, E.A., Brantley, J.N., Rangarajan, C. (1988). Size distributions and growth of natural and Chernobyl-derived submicron aerosols in Tennessee. J. Environ. Radioact. 6, 99–120. Brightwell, J., Carter, R.F. (1977). Comparative measurements of the short term lung clearance and translocation of PuO2 and mixed Na2 O and PuO2 aerosols in mice. In: Walton, W.H. (Ed.), Inhaled Particles IV, Part 1. Pergamon, Oxford, pp. 285–301.

Radioactive aerosols

55

Bruninx, E. (1961). High-energy nuclear reaction cross-sections. I. CERN Report 61-1, Geneva, Switzerland. Bruninx, E. (1964). High-energy nuclear reaction cross-sections. III. CERN Report 64-17, Geneva, Switzerland. Cantrell, B.K., Rubow, K.L. (1990). Mineral dust and diesel exhaust aerosol measurements in underground metal and nonmetal mines. In: Proceedings of VIIth International Pneumonoconioses Conference. NIOSH Publication No. 90-108, pp. 651–655. Cantrell, B.K., Zeller, H.W., Williams, K.L., Cocalis, J. (1987). Monitoring and measurement of in-mine aerosol: Diesel emissions. BuMines IC 9141, Washington, DC, pp. 18–40. Carter, R.F., Stewart, K. (1971). On the oxide formed by the combustion of plutonium and uranium. In: Walton, W.H. (Ed.), Inhaled Particles III, vol. 2. Pergamon, Oxford, pp. 819–838. Chamberlain, A.C. (1991). Radioactive Aerosols. Cambridge University Press, Cambridge, UK. Chamberlain, A.C., Dyson, E.D. (1956). The dose to the trachea and bronchi from the decay products of radon and thoron. Br. J. Radiol. 29, 317–325. Chapuis, A., Lopez, A., Fontan, J., Billard, F., Madeleine, G. (1970). Spectre granulometrique des aerosols radioactif dans une mine d’uranium. J. Aerosol Sci. 1, 243–253. Charalambous, S., Rindi, A. (1967). Aerosol and dust radioactivity in the halls of high-energy accelerators. Nucl. Instrum. Methods 56, 125–135. Cheng, Y., Yeh, H.C. (1980). Theory of screen type diffusion battery. J. Aerosol Sci. 11, 313–319. Cheng, Y.S., Guilmette, R.A., Zhou, Y., Gao, J., LaBone, T., Whicker, J.J., Hoover, M.D. (2004). Characterization of plutonium aerosol collected during an accident. Health Phys. 87, 596–605. Cooper, J.A., Jackson, P.O., Langford, J.C., Peterson, M.R., Stuart, B.O. (1973). Characteristics of attached radon222 daughters under both laboratory and field conditions with particular emphasis upon underground uranium mine environments. Battele Pacific Northwest Laboratories, Richland, WA (as cited in Van der Vooren et al., 1982). Craft, B.F., Oser, J.L., Norris, F.W. (1966). A method for determining relative amounts of combined and uncombined radon daughter activity in underground uranium mines. Am. Ind. Hyg. Assoc. 27, 154–159. Drevinsky, P.J., Pecci, J. (1965). Size and vertical distributions of stratospheric radioactive aerosols. In: Klement Jr., A.W. (Ed.), Radioactive Fallout from Nuclear Weapons Tests. U.S. Department of Commerce, Springfield, VA, CONF 765. Duggan, M.J., Howell, D.M. (1969). The measurement of the unattached fraction of airborne RaA. Health Phys. 17, 423–427. Elder, J.C., Conzales, M., Ettinger, H.J. (1974). Plutonium aerosol size characteristics. Health Phys. 27, 45–53. El-Hussein, A., Ahmed, A.A. (1995). Unattached fraction and size distribution of aerosol-attached radon progeny in the open air. Appl. Radiat. Isotopes 46 (12), 1393–1399. El-Hussein, A., Mohammed, A., Ahmed, A.A. (1998). A study on radon and radon progeny in surface air of El-Minia, Egypt. Radiat. Prot. Dosim. 78 (2), 139–145. Fisenne, I.M., Harley, N.H. (1973). Lung dose estimates from natural radioactivity measured in urban air. USAEC Report HASL-TM-74-7, New York. Friedlander, S.K. (1977). Smoke, Dust, and Haze. John Wiley and Sons, New York. Fusamura, N., Kurosawa, R. (1967). Determination of f -value in uranium mine air. In: Assessment of Airborne Radioactivity. IAEA, Vienna, SM-95/26, STI/PUB/159, pp. 213–227. Gaziev, Y.I., Malakhov, S.G., Nazarov, L.E., Silantiev, A.N. (1966). The size distribution of radioactive particles from nuclear weapons tests and their transport in the atmosphere. Tellus 18, 474–485. George, A.C. (1972a). Measurement of the uncombined fraction of radon daughters with wire screens. Health Phys. 23, 390–392. George, A.C. (1972b). Indoor and outdoor measurements of natural radon daughter decay products in New York air. In: Adams, J.A.S., Lowder, W.M., Gessel, T.F. (Eds.), Natural Radiation Environment II. U.S. Energy Research and Development Administration Report, CONF-720805-P2, pp. 741–750. George, A.C., Breslin, A.J. (1980). The distribution of ambient radon and radon daughters in residential buildings in the New Jersey–New York area. In: Natural Radiation Environment III, vol. 2. National Technical Information Service, Springfield, VA, CONF-780422, pp. 1272–1292. George, A.C., Hinkliffe, L. (1972). Measurements of uncombined radon daughters in uranium mines. Health Phys. 23, 791–803. George, A.C., Knutson, E.O., Sinclair, D., Wilkening, M.H., Andrews, L. (1984). Measurement of radon daughter aerosols in Socorro, New Mexico. Aerosol Sci. Technol. 3, 277–281.

56

C. Papastefanou

Glasstone, S., Dolan, P.J. (1977). The Effect of Nuclear Weapons. U.S. Department of Defense, Springfield, VA. Grundel, M., Porstendörfer, J. (2004). Differences between the activity size distributions of the different natural radionuclide aerosols in outdoor air. Atmos. Environ. 38, 3723–37287. Hering, S.V., Friedlander, S.K. (1982). Origins of aerosol sulfur size distributions in the Los Angeles basin. Atmos. Environ. 11, 2647–2656. Hicks, H.G. (1982). Calculation of the concentration of any radionuclide deposited on the ground by offsite fallout from a nuclear detonation. Health Phys. 46, 585–610. Hopke, P.K. (1991). The initial atmospheric behaviour of radon decay products. In: Kay, J.K., Keller, G.E., Miller, J.F. (Eds.), Indoor Air Pollution. Lewis Publishers, Chelsea, UK, pp. 141–169. International Commission on Radiological Protection, ICRP (1981). Limits for Inhalation of Radon Daughters by Workers, ICRP Publication No. 32. Ann. ICRP 6 (10). Jost, D.T., Gaggeler, H.W., Baltensperger, U., Zinder, B., Haller, P. (1986). Chernobyl fallout in size-fractionated aerosol. Nature 324, 22–23. Junge, C. (1955). The size distribution and aging of natural aerosols as determined from electric and optical data on the atmosphere. J. Meteorol. 12, 13–25. Junge, C. (1963). Air Chemistry and Radioactivity. Academic Press, New York. Kauppinen, E.I., Hillamo, R.E., Aaltonen, S.H., Sinkko, K.T.S. (1986). Radioactivity size distributions of ambient aerosols in Helsinki, Finland, during May 1986 after the Chernobyl accident: Preliminary report. Environ.Sci. Technol. 20, 1257–1259. Kondo, K., Muramatsu, H., Kanda, Y., Takahara, S. (1984). Particle size distribution of 7 Be-aerosols formed in high energy accelerator tunnels. J. Appl. Radiat. Isotopes 35, 939–944. Kotrappa, P., Bhantin, D.P., Dhandayutham, R. (1975). Diffusion sampler useful for measuring diffusion coefficients and unattached fraction of radon and thoron decay products. Health Phys. 29, 155–162. Kruger, J., Andrews, M. (1976). Measurement of the attachment coefficient of radon-220 decay products to monodispersed aerosols. J. Aerosol Sci. 7, 21–36. Kruger, J., Nöthling, J.F. (1979). A comparison of the attachment of the decay products of radon-220 and radon-222 to monodispersed aerosols. J. Aerosol Sci. 10, 571–579. Lal, D., Suess, H.E. (1968). The radioactivity of the atmosphere and the hydrosphere. Ann. Rev. Nucl. Sci. 18, 407– 434. Lockhart, L.B., Patterson, R.L., Saunders, A.W. (1965a). Distribution of airborne activity with particle size. In: Klement Jr., A.W. (Ed.), Radioactive Fallout from Nuclear Weapons Tests. U.S. Department of Commerce, Springfield, VA, CONF 765. Lockhart, L.B., Patterson, R.L., Saunders, A.W. (1965b). The size distribution of radioactive atmospheric aerosols. J. Geophys. Res. 7, 6033–6041. Lujaniene, G., Ogorodnikov, B.I., Budyka, A.K., Skitovich, V.I., Lujanas, V. (1997). An investigation of changes in radionuclide carrier properties. J. Environ. Radioact. 35 (1), 71–90. Lujaniene, G., Ogorodnikov, A., Budyka, A.K. (2001). Size distribution and chemical associations of cosmogenic and artificial radionuclides in ambient aerosols. J. Aerosol Sci. 32 (S1), 535–536. Mann, J.R., Kirchner, R.A. (1967). Evaluation of lung burdens following acute inhalation exposure to highly insoluble PuO2 . Health Phys. 13, 877–882. Martell, E.A. (1970). Transport patterns and residence times for atmospheric trace constituents vs. altitude. In: Radionuclides in the Environment. In: Adv. Chem. Ser., vol. 93. American Chemical Society, Washington, DC, pp. 138–157. McMurry, P.H., Wilson, J.C. (1982). Droplet phase, heterogeneous and gas phase, homogeneous contributions to secondary ambient aerosol formation as function of relative humidity. Atmos. Environ. 16, 121–134. McMurry, P.H., Wilson, J.C. (1983). Growth laws for the formation of secondary ambient aerosols: Implications for chemical conversion mechanisms. J. Geophys. Res. 88, 5101–5108. Mercer, T. (1976). The effect of particle size on the escape of recoiling RaB atoms from particulate surfaces. Health Phys. 31, 173–175. Mercer, T.T., Stowe, W.A. (1969). Deposition of unattached radon decay products in an impactor stage. Health Phys. 17, 259–264. Mercer, T.T., Stowe, W.A. (1971). Radioactive aerosols produced by radon in room air. In: Inhaled Particles III, vol. II. Unwin Brothers Limited/The Gresham Press, Surrey, UK, pp. 839–851.

Radioactive aerosols

57

Mohammed, A. (1999). Activity size distributions of short-lived radon progeny in indoor air. Radiat. Prot. Dosim. 86 (2), 139–145. Mohammed, A., El-Hussein, A., Ali, A.E. (2000). Measurement of thorium-B, 212 Pb in the outdoor environment and evaluation of equivalent dose. J. Environ. Radioact. 49, 181–193. Moore, H.E., Poet, S.E., Martell, E.A. (1980). Size distributions and origin of 210 Pb, 210 Bi, and 210 Po on airborne particles in the troposphere. In: Gesell, T.F., Lowder, W.M. (Eds.), Natural Radiation Environment III, vol. 1. National Technical Information Service, Springfield, VA, CONF-780422, pp. 415–429. Moss, W.D., Hyatt, E.C., Schulte, H.F. (1961). Particle size studies on plutonium aerosols. Health Phys. 5, 212–218. National Council on Radiation Protection and Measurements, NCRP (1987). Exposure of the Population in the United States and Canada from Natural Background Radiation. Report No. 94. National Research Council, NRC (1979). Airborne Particles. University Park Press, Baltimore. Papastefanou, C., Bondietti, E.A. (1987). Aerodynamic size associations of 212 Pb and 214 Pb in ambient aerosols. Health Phys. 53, 461–472. Papastefanou, C., Bondietti, E.A. (1991a). Mean residence times of atmospheric aerosols in the boundary layer as determined from 210 Bi/210 Pb activity ratios. J. Aerosol Sci. 22, 927–931. Papastefanou, C., Bondietti, E.A. (1991b). The unattached fraction of radon progeny in ambient aerosols. J. Environ. Radioact. 13, 1–11. Papastefanou, C., Ioannidou, A. (1995). Aerodynamic size association of 7 Be in ambient aerosols. J. Environ. Radioact. 26, 273–282. Papastefanou, C., Ioannidou, A. (1996). Influence of air pollutants in the 7 Be size distribution of atmospheric aerosols. Aerosol Sci. Technol. 24, 102–106. Persson, G., Sisefsky, J. (1971). Health Phys. 21, 421–442 (as cited in Jost et al., 1986). Philips, C.R., Pai, H.L. (1976). Environmental radon and radon daughter measurements in Port Hope. Report to the Atomic Energy Control Board of Canada (as cited in Van der Vooren et al., 1982). Poet, S.E., Moore, H.E., Martell, E.A. (1972). Lead-210, 210 Bi, and 210 Po in the atmosphere: Accurate ratio and application to aerosol residence time determination. J. Geophys. Res. 77, 6515–6527. Porstendörfer, J. (1984). Behaviour of radon daughter products in indoor air. Radiat. Prot. Dosim. 7, 104–114. Porstendörfer, J. (1994). Properties and behaviour of radon and thoron and their decay products in the air. J. Aerosol Sci. 25, 219–263. Porstendörfer, J., Mercer, T.T. (1978). Influence of nuclei concentration and humidity upon the attachment rate of atoms in the atmosphere. Atmos. Environ. 12, 2223–2228. Porstendörfer, J., Mercer, T.T. (1979). Influence of electric charge and humidity upon the diffusion coefficient of radon decay products. Health Phys. 37, 191–199. Porstendörfer, C., Mercer, T. (1980). Diffusion coefficient of radon decay products and their attachment rate to the atmospheric aerosol. In: Gesell, T.F., Lowder, W.M. (Eds.), Natural Radiation Environment III, vol. 1. National Technical Information Service, Springfield, VA, CONF-780422, pp. 281–293. Porstendörfer, J., Röbig, G., Ahmed, A. (1979). Experimental determination of the attachment coefficients of atoms and ions on monodispersed aerosols. J. Aerosol Sci. 10, 21–28. Porstendörfer, J., Reineking, A., Becker, K.H. (1987). Free fractions, attachment rates, and plate-out rates of radon daughters in houses. In: Hopke, P.K. (Ed.), Proceedings of International Symposium on Radon and Its Decay Products: Occurrence, Properties and Health Effects. In: ACS Symposium Series, vol. 331. American Chemical Society, Washington, DC, pp. 281–293. Porstendörfer, J., Pagelkopf, P., Grundel, M. (2005). Fraction of the positive 218 Po and 214 Pb clusters in indoor air. Radiat. Prot. Dosim. 113, 342–351. Pradel, J., Chapuis, A., Lopez, A., Cabrol, C. (1970). Sur les caracteristiques des aerosols radioactifs presents dans les mines Francaises d’uranium. Radioprotection 5, 263–270. Raabe, O.G. (1969). Concerning the interactions that occur between radon decay products and aerosols. Health Phys. 17, 177–185. Raes, F. (1985). Description of the properties of unattached 218 Po and 212 Pb particles by means of comparison of the classical theory of cluster formation. Health Phys. 49, 1177–1187. Raghavayya, M., Jones, J.H. (1974). A wire-screen filter paper combination for the measurements of fractions of unattached radon daughters in uranium mines. Health Phys. 26, 417–429. Reineking, A., Becker, K.H., Porstendörfer, J. (1985). Measurements of the unattached fractions of radon daughters in houses. Sci. Total Environ. 45, 261–270.

58

C. Papastefanou

Rindi, A., Charalambous, S. (1967). Airborne radioactivity produced at high-energy accelerators. Nucl. Instrum. Methods 47, 227–232. Röbig, G., Becker, K.H., Hessin, A., Porstendörfer, J., Scheibel, H.G. (1980). A cascade impactor calibration for measurement of activity size distributions in the atmosphere. In: Proceedings of 8th Conference in Aerosol Science. Georg-August-University, Göttingen, Germany, pp. 96–102. Sanak, J., Gaudry, A., Lambert, G. (1981). Size distributions of 210 Pb aerosols over oceans. Geophys. Res. Lett. 8, 1067–1070. Seinfeld, J.H. (1986). Atmospheric Chemistry and Physics of Air Pollution. Wiley, New York. Shimo, M., Ikebe, Y. (1984). Measurements of radon and its short-lived decay products and unattached fraction in air. Radiat. Prot. Dosim. 8, 209–214. Silberberg, R., Tsao, C.H. (1973). Cross sections of proton-nucleus interactions at high energies. Report NRL 7593, Naval Research Laboratory, Washington, DC. Suzuki, T., Maruyama, Y., Nakayama, N., Yamada, K., Ohta, K. (1999). Measurement if the 210 Po/210 Pb activity ratios in size fractionated aerosol from the coast of the Japan sea. Atmos. Environ. 33, 2285–2288. Thomas, J.W., Hinchliffe, L.E. (1972). Filtration of 0.001 µm particles by wire screens. J. Aerosol Sci. 3, 387–393. United Nations Scientific Committee on the Effects of Atomic Radiation, UNSCEAR (1982). Ionizing Radiation: Sources and Biological Effects. United Nations, New York. United Nations Scientific Committee of the Effects of Ionizing Radiation, UNSCEAR (1993). Sources and Effects of Ionizing Radiation. United Nations, New York. United Nations Scientific Committee on the Effects of Ionizing Radiation, UNSCEAR (2000). Sources and Effects of Ionizing Radiation. United Nations, New York. Van der Vooren, A.W., Busigin, A., Philips, C.R. (1982). An evaluation of unattached radon, and thoron daughter measurement techniques. Health Phys. 42, 801–808. Whitby, K.T. (1978). The physical characteristics of sulphur aerosols. Atmos. Environ. 12, 135–159. Whitby, K.T., Husar, R.B., Liu, B.Y.H. (1972). The aerosol size distribution of Los Angeles smog. J. Colloid Interface Sci. 39, 177–204. Willeke, K., Baron, P.A. (1993). Aerosol Measurement: Principles, Techniques, and Applications. Van Nostrand Reinhold, New York. Winkler, R., Dietl, F., Franck, G., Tschierch, J. (1998). Temporal variation of 7 Be and 210 Pb size distribution in ambient aerosol. Atmos. Environ. 32 (6), 983–991. Young, J.A., Silker, W.B. (1980). Aerosol deposition velocity on the Pacific and Atlantic oceans, calculated from 7 Be measurements. Earth Planet. Sci. Lett. 50, 92–104. Young, J.A., Tanner, T.M., Thomas, C.W., Wogman, A., Petersen, M.R. (1975). Concentrations and rates of removal of contaminants from the atmosphere in an downwind of St. Louis. In: Pacific Northwest Laboratory Annual Report for 1974 to the USAEC Division of Biomedical and Environmental Research, Part 3: Atmospheric Sciences, pp. 70–76. Yu, K.N., Lee, L.Y.L. (2002). Measurements of atmospheric 7 Be properties using high-efficiency gamma spectroscopy. Appl. Radiat. Isotopes 57, 941–946. Yu, C.C., Tung, C.J., Hung, I.F., Tseng, C.L. (1993). Analyses of radioactive aerosols to support accurate internal dose assessments at Chin-Shan Nuclear Power Plant. Health Phys. 65.

59

Chapter 3

Radioactive nuclides as tracers of environmental processes

1. Radioactivity in the environment Natural radionuclides formed in the atmosphere by gas-to-particle conversion reactions include the relatively short-lived radionuclides with half-lives of a number of days, of cosmogenic origin, such as 7 Be (T1/2 = 53.3 d), 32 P (14.3 d), 33 P (25.3 d) and 35 S (87.4 d) and the relatively long-lived 22 Na (2.6 y), which occur permanently in the atmosphere. All the above-mentioned radionuclides are forming continuously by the interaction of cosmic-ray particles with matter (the atmosphere). Most of them are formed by spallation processes of light atmospheric nuclei, such as nitrogen, oxygen and even carbon or heavier atmospheric nuclei, such as sodium, phosphorus, sulphur, potassium and calcium (Lal and Suess, 1968; NCRP, 1987) when they absorb protons and even neutrons of cosmic origin. Like sulfates produced from gas-phase sulphur dioxide, these solid-element radionuclides rapidly attach to or coagulate with other nuclei. The small particles with which these radionuclides are associated are a significant reservoir of atmospheric pollutant when regional or continental air loadings are considered. Their size distributions appear to group into two modes: Aitken nuclei mode (<0.1 µm diameter) and accumulation mode (between 0.1 and 2.0 µm), with the latter contributing most of the mass. The accumulation mode represents a selfpreserving distribution of aerosol sizes that results from gains due to rapid aggregation of nuclei and losses due to atmospheric removal processes acting at various efficiencies on all sizes. Particles larger than about 0.3 µm do not tend to form in large amounts because accumulation mode atmospheric residence times (one week or less; Rodhe and Grandell, 1972; Moore et al., 1980) are shorter than the coagulation periods necessary for their formation (NRC, 1979). Stratosphere-to-troposphere exchange and the 11-year solar cycle influence the concentrations of these radionuclides in the atmosphere. These radionuclides have come to be recognised as tracers in various atmospheric processes, such as precipitation, washout (precipitation scavenging), resuspension, atmospheric particle deposition and deposition patterns of airborne contaminants, aerosol deposition and aerosol trapping by above-ground vegetation (air-to-vegetation transfer). The potential usefulness of the atmospheric particle deposition parameters in understanding the dynamics of aerosol particle deposition to ecosystems can be clearly seen. RADIOACTIVITY IN THE ENVIRONMENT VOLUME 12 ISSN 1569-4860/DOI: 10.1016/S1569-4860(07)12003-9

© 2008 Elsevier B.V. All rights reserved.

60

C. Papastefanou

Of the cosmogenic radionuclides, beryllium-7 has a high production rate in the atmosphere (8.1 × 10−2 atoms cm−2 s−1 ) and, as it is a gamma-emitter (477 keV gammas, 11% yield), it can easily be detected and measured in atmospheric air, in both precipitation and vegetation. Its average concentration in the troposphere is about 12.5 mBq m−3 (UNSCEAR, 2000) or 3 mBq m−3 in surface air in the temperate zone and 700 Bq m−3 in rainwater (UNSCEAR, 1982). Sodium-22 is also a gamma-emitter (511 keV gammas, annihilation peak), but its production rate in the atmosphere is too small (8.6 × 10−5 atoms cm−2 s−1 ) and its concentration in the troposphere is only 0.0021 mBq m−3 (UNSCEAR, 2000). The other radionuclides referred to above are mostly beta-emitters with low production rates in the atmosphere and very low concentrations in the troposphere. Other radionuclides occurring naturally in the atmosphere are the radon decay products, such as 218 Po (T1/2 = 3.05 min), 214 Pb (26.8 min), 214 Bi (19.9 min) of 222 Rn, 212 Pb (10.64 h) of 220 Rn, and particularly 210 Pb (22.3 y), a decay product of 222 Rn emanating from the ground. Beryllium-7 and 210 Pb have common associations with atmospheric aerosol particles (Dibb and Jaffrezo, 1993). Apart from the radionuclides of cosmogenic or terrestrial origin, 137 Cs (30.17 y), a fission product radionuclide of Chernobyl origin, was embedded into the environment and remained for a long time after the nuclear reactor accident (April 26, 1986). At the time of the accident, the concentration of 137 Cs in air reached 2 Bq m−3 (5–6 May 1986) and its deposition on the ground 24 kBq m−2 in the Thessaloniki area, Northern Greece (Papastefanou et al., 1988). In previous studies, 137 Cs from fallout due primarily to pre-1963 atmospheric nuclear explosions by the major nuclear weapons countries was taken into account (Peirson and Keane, 1962; Peirson and Cambray, 1965; Cambray et al., 1970). This radionuclide is also a tool for tracing environmental processes, such as deposition and resuspension and the long-term behaviour of airborne radioactivity. The radioactivity in air is typically measured by air sampling carried out with TFIA-2 Staplex type high-volume air samplers with type TFAGF 810 Staplex glass-fibre filters which are highly retentive of particulate material, 20.32 cm × 25.40 cm (8 × 10 ) in dimensions and with 99% collection efficiency for submicron particles as small as 0.3 µm and larger. This design involves a regulated air-flow rate of 1.7–1.92 m3 min−1 (60–68 cfm or ft3 min−1 ) (average 1.84 m3 min−1 or 65 cfm or ft3 min−1 ). The length of each collection period is 24 h. Air samplings are carried out from the ground level up to about 20 m height. The specific activity of each radionuclide in air is determined by gamma-spectrometry using a high purity low background Ge detector with resolution about 1.9 keV at 1.33 MeV of 60 Co and efficiency usually better than 42%. Figure 3.1 shows a plot of a gamma-ray spectrum of an atmospheric aerosol sample (air filter) obtained by a Ge detector, in which the gamma peaks of 7 Be, 210 Pb and 137 Cs are clearly shown (Papastefanou, 2006). Precipitation samples are collected in regular fallout funnels made of Teflon and having 300 mm diameter and total collection area about 0.30 m2 . The funnels are exposed to the atmosphere continuously, thus collecting both dry and wet precipitation. The samples are filtered through Whatman-41 filters to remove solid insoluble materials. Large volumes of rainwater are then concentrated by reducing to 1 litre by evaporation and acidified with 1 cm3 nitric acid (65%) per litre of rainwater before evaporation, thus resulting in a 0.015 N solution to prevent any loss of radionuclide atoms by absorption to the surface of the vessel used. Each concentrated sample is set in a Marinelli beaker of 1 litre volume and then the radioactivity

Radioactive nuclides as tracers of environmental processes

61

Fig. 3.1. Plot of a gamma-ray spectrum of an atmospheric aerosol sample (air filter) obtained by a Ge detector.

is measured by gamma spectrometry. The radioactivity of the solid insoluble material, due to dry deposition during the non-rainy days, is also measured by gamma-spectrometry. Grass samples, gramineae or poaceae or other species such as fescue, alfalfa, etc., are usually collected from an area of 20 × 20 = 400 m2 in the outside environment, appropriately fenced to prevent interference by animals and therefore undisturbed. Each sample is cut from an area of about 3 m2 (0.125 kg of grass per m2 ) and then the radioactivity is measured by gamma-spectrometry.

2. Atmospheric particle deposition The deposition of atmospheric particles occurs as dry deposition due to gravity and inertia with impaction and gravitational settling and wet deposition by rainfall events. 2.1. Dry deposition The deposition velocity, Vd , concept has been applied in evaluations of aerosol transfer to the earth’s surface. The deposition velocity represents the effective thickness of the atmosphere losing aerosols per unit time or the effective rate at which small aerosols are sedimenting. The deposition velocity, Vd , in m s−1 , is defined as Vd =

F (Bq m−2 s−1 ) , Cair (Bq m−3 )

(3.1)

where F is the deposition flux of the aerosol particles or any kind of atmospheric particles, e.g. soil, dust or ash associated with a radionuclide, in Bq m−2 s−1 , that is the amount of

62

C. Papastefanou

Fig. 3.2. Relationship between deposition velocities of aerosol particles to grass and particle diameter.

airborne radioactivity which is removed per unit area per unit time, and Cair is the concentration of radionuclide in air, in Bq m−3 . This expression has found use in assessment models because measured radionuclide concentrations in air and generic deposition velocities allow pollutant flux to be estimated. For vegetation, however, the deposition velocity concept is difficult to apply because leaves, for example, are not net traps for aerosol constituents. Growth dilution, precipitation and other natural processes cause additions and losses of atmospheric constituents to and from the vegetation canopy, while root uptake can mask their origins. If ground area rather than leaf area Vd is desired, the density of the canopy needs to be evaluated. The deposition velocity of atmospheric particles is dependent on the particle diameter with large particles having higher deposition velocities according to the equation Vd =

dp2 18η

(ρp − ρair )g,

(3.2)

where dp is the particle diameter, in cm, ρp is the particle density in g cm−3 (ρp = 1 g cm−3 for aerosol particles, ρp = 2.65 g cm−3 for resuspended soil particles), ρair is the air density, in g cm−3 (ρair = 0.001293 g cm−3 ), η is the viscosity of air (η = 181 × 10−6 g cm−1 s−1 at 20 ◦ C) and g is the acceleration due to gravity (g = 981 cm s−2 ). Then, the deposition velocity, Vd is expressed in cm s−1 . Figure 3.2 illustrates the relationship between the deposition velocity, Vd , and the particle diameter, dp , derived from field and laboratory measurements of deposition velocities of

Radioactive nuclides as tracers of environmental processes

63

Table 3.1 Deposition velocity of atmospheric particles, Vd (cm s−1 ) 7 Be

210 Pb

137 Cs

Investigation

Country

0.5 (0.3–0.8) 1.2 (0.5–2.1) 0.5 (0.2–3.4) 0.1–0.6 0.80 1.0 2.8 1.66 1.3 1.5 1.6

– – – – – – 0.95 – 0.7 – 1.1

3.4 (1.3–6.3) – – – – – – – – 1.46 –

Papastefanou et al. (1995) Chamberlain (1953) Small (1960) Peirson et al. (1973) Young and Silker (1980) Crecelius (1981) Turekian et al. (1983) Mahoney (1984) Todd et al. (1989) Rosner et al. (1996) McNeary and Baskaran (2003)

Greece UK Norway UK USA USA USA USA USA Germany USA

Typical range is given within brackets.

particles to grass (McMahon and Denison, 1979). This figure shows that (i) the variability of deposition velocity for large particle diameters (greater than 10 µm) results from variable wind speeds, (ii) the minimum particulate deposition velocity occurs in the range 0.1–1.0 µm, (iii) the deposition velocity is approximately a linear function of wind speed, (iv) the deposition of particulate matter beneath trees varies significantly, and with values ranging from 2 to 16 times that in adjacent open terrain, and (v) considerable care is needed in choosing a “typical” deposition velocity, as it is a function of many factors which can vary by 2 orders of magnitude. Table 3.1 shows data for the deposition velocity, Vd , of atmospheric aerosol particles associated with 7 Be, 210 Pb and 137 Cs. Regarding the 7 Be-associated particles, the deposition velocity varied from 0.1 to 3.4 cm s−1 (Chamberlain, 1953; Small, 1960; Peirson et al., 1973; Young and Silker, 1980; Crecelius, 1981; Turekian et al., 1983; Mahoney, 1984; Todd et al., 1989; Papastefanou et al., 1995; Rosner et al., 1996; McNeary and Baskaran, 2003). For the 210 Pb-associated particles, the deposition velocity varied from 0.7 to 1.1 cm s−1 (Turekian et al., 1983; Todd et al., 1989; McNeary and Baskaran, 2003). For the 137 Csassociated particles, the deposition velocity varied from 1.3 to 6.3 cm s−1 (Rosner et al., 1996; Papastefanou et al., 1995). The data of Table 3.1 refer mostly to temperate latitudes of the Northern Hemisphere, e.g. at Thessaloniki, Greece 40◦ N, Oak Ridge, TN 36◦ N (Mahoney, 1984), Norfolk, VA 37◦ N (Todd et al., 1989), New Haven, CT 41◦ N (Turekian et al., 1983), Detroit, MI 42◦ N (McNeary and Baskaran, 2003), Qulllauyte, WA 49◦ N (Crecelius, 1981), Munich, Germany 49◦ N (Rosner et al., 1996). 2.2. Wet deposition Wet deposition which consists of rainout (within cloud scavenging) and washout (below cloud scavenging) may be considered as an exponential decay process. Thus: χt = χ0 e−Λpt ,

(3.3)

64

C. Papastefanou

Fig. 3.3. Relationship between rain scavenging rates and aerosol particle diameter.

where χt is the atmospheric concentration of particles at time t, χ0 is the atmospheric concentration of particles at time zero, and Λp is the scavenging coefficient for particles, in units of time−1 (s−1 ). Figure 3.3 illustrates the relationship between rain scavenging rates expressed by the scavenging coefficient, Λ, and particle diameter, dp (McMahon and Denison, 1979). In the five cases shown by circles in this figure, the values of the scavenging coefficient are a function of rainfall intensity. A common value of 5 mm h−1 for the rainfall intensity was adopted. This figure also shows the importance of particle size as a major factor in determining the wet scavenging coefficients. The wet deposition due to precipitation is determined by the fractional rate of removal of radionuclide by rain, λ in s−1 , according to the following equation λ=

Wp , ρH

(3.4)

were p is the precipitation rate, in kg m−2 s−1 or in kg m−2 d−1 , H is the vertical extent of the plume, assumed to be well mixed, in m, ρ is the air density (ρ = 0.001293 g cm−3 = 1.29 kg m−3 ) and W is the washout ratio or precipitation scavenging ratio. The dimensionless washout ratio, W , is determined as the ratio of radioactivity per unit mass (Bq kg−1 ) or unit volume of rainwater (Bq m−3 ) and the radioactivity per unit mass (Bq kg−1 ) or unit volume of air (Bq m−3 ): W =

Crain (Bq kg−1 ) Cair (Bq kg−1 )

(3.5)

or −3 = Crain (Bq m ) . W Cair (Bq m−3 )

(3.6)

Radioactive nuclides as tracers of environmental processes

65

Table 3.2 Washout ratio of atmospheric particles, W 7 Be

210 Pb

137 Cs

Investigation

Country

144 (103–175) – – – 370 (370–375) 948 – – – –

– – – – 215 (203–228) 637 – 430 530 –

1295 (284–3810) 560 680 (600–800) 730 – – 230–6600 230 – 700

Papastefanou and Ioannidou (1991) Peirson and Keane (1962) Peirson and Cambray (1965) Cambray et al. (1970) Todd et al. (1989) McNeary and Baskaran (2003) Garland and Playford (1991) Chamberlain (1960) Peirson et al. (1966) Clark and Smith (1988)

Greece UK UK UK USA USA UK UK UK UK

Typical range is given within brackets.

By taking into consideration the density of air, ρair = 0.001293 g cm−3 = 1.29 kg m−3 , and the density of water, ρwater = 1 g cm−3 = 1000 kg m−3 , then W =

−3 ρair = 1.29 × 10−3 Crain (Bq m ) . W = 1.29 × 10−3 W ρwater Cair (Bq m−3 )

(3.7)

If τ is the mean residence time of a radioactive nuclide associated with aerosol particles, in s or in d, that is the inverse of the fractional rate of removal of the radionuclide, λ in s−1 or d−1 , then Equation (3.4) becomes for the washout ratio W =

ρH , τρ

(3.8)

where τ = 1/λ. By taking into consideration the mass of tropospheric air per m2 of earth’s surface, ρH = 9000 kg m−2 , as ρ = 1.2 kg m−3 the air density and H = 7500 m the vertical extent of the plume, the precipitation rate p = 3.6 × 10−5 kg m−2 s−1 = 3.6 kg m−2 d−1 over the northern hemisphere (Garland and Playford, 1991), the mean residence time of tropospheric aerosol particles associated with a radionuclide τ = 8 d = 6.9 × 105 s (Papastefanou and Bondietti, 1991; Papastefanou and Ioannidou, 1995), and the fractional rate of removal of a radionuclide by precipitation λ = τ1 = 0.125 d−1 = 1.4 × 10−6 s−1 , then Equation (3.8) gives W = 363 for this case. ApSimon et al. (1989) used for the fractional rate of removal of a radionuclide by rain, λ, the following formula: λ = 5 × 10−5 q 0.8 ,

(3.9)

where q is the precipitation rate in mm h−1 , equivalent to a mean effective value of about 8 × 10−5 s−1 . Table 3.2 shows data for the washout ratio, W , of atmospheric aerosol particles associated with 7 Be, 210 Pb and 137 Cs. Regarding the 7 Be-associated particles, the washout ratio varied from 103 to 948 (Todd et al., 1989; Papastefanou and Ioannidou, 1991; McNeary and

66

C. Papastefanou

Baskaran, 2003). For the 210 Pb-associated particles, the washout ratio varied from 203 to 637 (Chamberlain, 1960; Peirson et al., 1966; Todd et al., 1989; McNeary and Baskaran, 2003). For the 137 Cs-associated particles, the washout ratio varied from 230 to 6600 (Chamberlain, 1960; Peirson and Keane, 1962; Peirson and Cambray, 1965; Cambray et al., 1970; Clark and Smith, 1988; Garland and Playford, 1991; Papastefanou and Ioannidou, 1991). Washout ratios are mostly higher for the 137 Cs-associated particles than those corresponding to the 7 Be- and 210 Pb-associated particles because high values of the washout ratio W are associated with larger particle sizes in the coarse mode (>1.0 µm in diameter) in the activity size distribution of atmospheric aerosols (Chamberlain, 1991). Washout ratios are also enhanced if the airborne radionuclide concentration increases with height (Chamberlain, 1991). Like wet scavenging coefficients, Λ, washout ratios, W , decrease with precipitation amount but increase with particle size (Gatz, 1975). Gatz (1975) also observed that W increases with distance from the emission source. All these aspects indicate that, when using washout ratios, localised measurements are desirable.

3. Resuspension Resuspension is an environmental process which enhances the air concentration of any particulate material deposited onto the ground and or onto vegetation. Most radioactive particles and vapours, once deposited, are held rather firmly on surfaces, but resuspension does occur. A radioactive particle may be blown off the surface, or, more probably, the fragment of soil or vegetation to which it is attached may become airborne. This occurs most readily where soils and vegetation are dry and friable. So, resuspension has the potential to cause persistent concentrations in air and to redistribute material on the ground. Some of the sustained airborne concentrations could result from (i) return of some activity injected into the lower stratosphere, (ii) resuspension and dispersal on a continental scale of deposited material, and (iii) local resuspension. There is also the possibility of resuspension from agricultural and urban land. Inhalation of resuspended activity may be the most important route for human uptake of natural or man-made radionuclides or entry into food chains. The resuspension of particles deposited on any surface is described by the resuspension factor, kr in m−1 , which is determined as the ratio of the radioactivity per unit volume of air (Bq m−3 ) and the radioactivity deposited on the ground (Bq m−2 ): kr =

Cair (Bq m−3 ) . Ds (Bq m−2 )

(3.10)

This definition of the resuspension factor, kr , implies an equilibrium relationship between the two quantities which may be achieved only over an extensive area of uniform deposition. In principle, when the deposition varies spatially, resuspension would be better predicted by using the resuspension rate, Λ in s−1 , which is defined as the ratio of vertical flux (Bq m−2 s−1 ) and the radioactivity deposited on the ground (Bq m−2 ): Λ=

R (Bq m−2 s−1 ) . Ds (Bq m−2 )

(3.11)

Radioactive nuclides as tracers of environmental processes

67

Table 3.3 Resuspension factor of atmospheric particles, kr (×10−4 m−1 ) 7 Be

137 Cs

Investigation

Country

2.3 (1.6–4.2) – –

0.6 (0.1–1.2) 10−3 –1 10−4

Papastefanou et al. (1995) Stewart (1966) Garland and Cambray (1988)

Greece UK UK

Typical range is given within brackets.

Thus, the resuspension rate Λ is the fraction removed per second by resuspension process. The use of this quantity with a suitable dispersion and deposition model would enable the movement of a radioactive nuclide from place to place to be predicted. Such an approach is necessary for estimating the radionuclide concentration in air due to resuspension downwind of an area heavily affected by the deposition process. Whether kr or Λ is used, it is clear that the value of the parameter must be expected to vary with many environmental variables. The most important of these environmental variables will be time after deposition, surface structure, nature of the radioactivity, wind speed, surface moisture and rate of mechanical disturbance of the surface. A simple exponential function for the resuspension factor, kr , gives a reasonable description of the decrease with time (Garland and Pattenden, 1989) as follows: kr (t) = Ae−Bt ,

(3.12) m−1

10−9

months−1

10−9

× and B in varied from 3.6 × to 49 × where the parameters A in 10−9 m−1 for the parameter A and from 0.026 to 0.124 month−1 for the parameter B. Table 3.3 shows data for the resuspension factor, kr , of atmospheric aerosol particles associated with 7 Be and 137 Cs. Regarding the 7 Be-associated particles, the resuspension factor varied from 1.6×10−4 to 4.2×10−4 m−1 (average 2.3×10−4 m−1 ; Papastefanou et al., 1995). For the 137 Cs-associated particles, the resuspension factor varied from 10−8 to 1.2×10−4 m−1 (Stewart, 1966; Garland and Cambray, 1988; Papastefanou et al., 1995). Garland and Cambray (1988) stated that higher resuspension factors at some locations appeared to be linked with traffic or intense industrial or agricultural activity. For the resuspension rate Λ, relative values ranging from 0.4 × 10−5 to 300 × 10−5 s−1 for atmospheric particles were reported by McMahon and Denison (1979).

4. Air-to-vegetation transfer of radionuclides associated with submicron aerosols The removal of submicron aerosols from the atmosphere occurs by precipitation scavenging and by direct capture of the aerosols by landscape. On a regional scale, the former process (wet deposition) clearly dominates even though the magnitude of the latter process (dry deposition) has been difficult to evaluate because of the mechanisms by which submicron aerosols deposit onto surfaces. It is believed that diffusion controls the deposition of aerosol <0.1 µm in diameter, with impaction and gravitational settling becoming increasingly important as the aerosol size increases (Chamberlain and Little, 1981).

68

C. Papastefanou

Table 3.4 Air-to-vegetation transfer rates of radionuclides, CRv,a (m3 kg−1 ) 7 Be

137 Cs

Investigation

Country

9856 (5258–16180) 6500 (3800–9000)

181197 (39638–345279) –

Papastefanou et al. (1995) Bondietti et al. (1984)

Greece USA

Typical range is given within brackets.

Assuming that the uptake of a radionuclide by vegetation, e.g. grassland, occurred during deposition of submicron aerosols onto the foliage and that the radioactivity remained virtually unchanged for the first few days, e.g. 5 days (Bondietti et al., 1984) immediately after deposition, the transfer coefficient of a radionuclide from air to vegetation or the air-to-vegetation transfer rate, CRv,a in m3 kg−1 , is defined by the ratio CRv,a =

Cvegetation (Bq kg−1 ) , Cair (Bq m−3 )

(3.13)

where Cvegetation is the concentration of a radionuclide on vegetation (Bq kg−1 ) and Cair is the concentration of this radionuclide in air (Bq m−3 ) averaged over at least three half-lives and preferably five. A biomass-normalised deposition velocity, VD in m3 kg−1 s−1 , can thus be derived VD = λ (s−1 ) CRv,a (m3 kg−1 ),

(3.14)

where λ is the decay constant of a radionuclide (s−1 ) and CRv,a is the air-to-vegetation transfer rate (m3 kg−1 ). The resulting VD represents the effective air volume being depleted of aerosol particles by 1 kg of vegetation each second. By multiplying VD by the biomass density, Y in kg per m2 of ground area, the deposition velocity, Vd (m s−1 ), is derived: Vd = VD Y.

(3.15)

Bondietti et al. (1984) estimated that the biomass density, Y , averaged to 0.4 kg m−2 for fescue, a tufted perennial grass. Table 3.4 shows data for the air-to-vegetation transfer rate, CRv,a , of atmospheric aerosol particles associated with 7 Be and 137 Cs. Regarding the 7 Be-associated particles, the airto-vegetation transfer rate varied from 3800 to 16180 m3 kg−1 (Bondietti et al., 1984; Papastefanou et al., 1995). For the 137 Cs-associated particles, the air-to-vegetation transfer rate varied from 39638 to 347279 m3 kg−1 (average 181197 m3 kg−1 ; Bondietti et al., 1984), much higher than for the 7 Be-associated particles. Papastefanou et al. (1995) referred to grass, gramineae or poaceae, as the species for vegetation, while Bondietti et al. (1984) referred to fescue, a tufted perennial grass. Average values of VD and Vd , 4.3 × 10−3 m3 kg−1 s−1 and 0.21 cm s−1 , respectively, were reported for the case of fescue by Bondietti et al. (1984).

Radioactive nuclides as tracers of environmental processes

69

References ApSimon, H., Wilson, J.J.H., Simms, K.L. (1989). Proc. Roy. Soc. London A 425, 365–405. Bondietti, E.A., Hoffman, F.O., Larsen, I.L. (1984). Air-to-vegetation transfer rates of natural submicron aerosols. J. Environ. Radioact. 1, 5–27. Cambray, R.S., Fischer, E.M.R., Brooks, W.L., Peirson, D.H. (1970). Radioactive fallout in air and rain: Results to the middle of 1970. AERE Report R-6556, HMSO, London. Chamberlain, A.C. (1953). Aspects of travel and deposition of aerosol and vapour clouds. AERE Report HP/R-1261. Chamberlain, A.C. (1960). Aspects of the deposition of radioactive and other gases and particles. Int. J. Air Pollut. 3, 63–88. Chamberlain, A.C. (1991). Radioactive Aerosols. Cambridge University Press, Cambridge, UK. Chamberlain, A.C., Little, P. (1981). Transport and capture of particles by vegetation. In: Grace, J., Ford, E.D., Jarus, P.G. (Eds.), Plants and Their Atmospheric Environments. Blackwell Scientific Publications, Oxford, UK, pp. 147–173. Clark, M.J., Smith, F.B. (1988). Wet and dry deposition of Chernobyl releases. Nature 332, 245–249. Crecelius, E.A. (1981). Prediction of marine atmospheric deposition rates using total 7 Be deposition velocities. Atmos. Environ. 15, 579–582. Dibb, J.E., Jaffrezo, J.-L. (1993). Beryllium-7 and lead-210 in aerosol and snow in the Dye 3 Gas, Aerosol and Snow Sampling Program. Atmos. Environ. 27A, 2751–2760. Garland, J.A., Cambray, R.S. (1988). Deposition, resuspension and the long-term variation of airborne radioactivity from Chernobyl. In: Proceedings of IVth Symposium International de Radioecologie de Cadarache. Centre d’Etudes Nucleaire de Cadarache (France), 14–18 Mars 1988, tome 1, pp. B26–B31. Garland, J.A., Pattenden, N.J. (1989). Resuspension following Chernobyl. Commission of the European Communities, Seminar on Methods and Code for Assessing the Off-Site Consequences of Nuclear Accidents, Athens (Greece), May 1990. Garland, J.A., Playford, K. (1991). Deposition and resuspension of radiocesium after Chernobyl. In: Proceedings of Seminar on Comparative Assessment of the Environmental Impact of Radionuclides during Three Major Nuclear Accidents: Kysthym, Windscale, Chernobyl, Rep. EUR 13574, Luxembourg, 1–5 October 1990, vol. I, pp. 237– 253. Gatz, D.F. (1975). Estimates of wet and dry deposition of Chicago and Northwestern Indiana aerosols in southern Lake Michigan. In: Second Interagency Conference on Marine Science, Great Lakes, Argonne, IL, March 1975. Lal, D., Suess, H.E. (1968). The radioactivity of the atmosphere and hydrosphere. Ann. Rev. Nucl. Sci. 18, 407–434. Mahoney, L.A. (1984). Beryllium-7 deposition to terrestrial vegetation in Tennessee. Ph.D. thesis, Western Kentucky University, Bowling Green, KY, 75 pp. McMahon, T.A., Denison, P.J. (1979). Empirical atmospheric deposition parameters: A survey. Atmos. Environ. 13, 571–585. McNeary, D., Baskaran, M. (2003). Depositional characteristics of 7 Be and 210 Pb in Southeastern Michigan. J. Geophys. Res. 108, 4210–4224. Moore, H.E., Poet, S.E., Martell, E.A. (1980). Size distribution and origin of 210 Pb, 210 Bi and 210 Po on airborne particles in the troposphere. In: Gesell, T.F., Lowder, W.M. (Eds.), Natural Radiation Environment III. National Technical Information Service, Springfield, VA, pp. 415–429. CONF-780422. National Council on Radiation Protection and Measurements, NCRP (1987). Exposure of the Population in the United States and Canada from Natural Background Radiation, Report No. 94. National Research Council, NRC (1979). Airborne Particles. University Park Press, Baltimore, MD. Papastefanou, C. (2006). Radioactive nuclides as tracers of environmental processes. J. Radioanal. Nucl. Chem. 267, 315–320. Papastefanou, C., Bondietti, E.A. (1991). Mean residence times of atmospheric aerosols in the boundary layer as determined from 210 Bi/210 Pb activity ratios. J. Aerosol Sci. 22, 927–931. Papastefanou, C., Ioannidou, A. (1991). Depositional fluxes and other physical characteristics of atmospheric beryllium-7 in the temperate zones (40◦ N) with a dry (precipitation-free) climate. Atmos. Environ. 25A, 2335– 2343. Papastefanou, C., Ioannidou, A. (1995). Aerodynamic size association of 7 Be in ambient aerosols. J. Environ. Radioact. 26, 273–282.

70

C. Papastefanou

Papastefanou, C., Ioannidou, A., Stoulos, S., Manolopoulou, M. (1995). Atmospheric deposition of cosmogenic 7 Be and 137 Cs from fallout of the Chernobyl accident. Sci. Total Environ. 170, 151–156. Papastefanou, C., Manolopoulou, M., Charalambous, S. (1988). Radiation measurements and radioecological aspects of fallout from the Chernobyl reactor accident. J. Environ. Radioact. 7, 49–64. Peirson, D.H., Cambray, R.S. (1965). Fission product fallout from the nuclear explosions of 1961 and 1962. Nature 205, 433–440. Peirson, D.H., Keane, J.R. (1962). Characteristics of early fallout from the Russian nuclear explosions of 1961. Nature 196, 801–807. Peirson, D.H., Cambray, R.S., Spicer, G.S. (1966). Lead-210 and polonium-210 in the atmosphere. Int. J. Air Pollut. 3, 63–68. Peirson, D.H., Cawse, P.A., Salmon, L., Cambray, R.S. (1973). Trace elements in the atmospheric environment. Nature 241, 252–256. Rodhe, H., Grandell, J. (1972). On the removal time of aerosol particles from the atmosphere by precipitation scavenging. Tellus 24, 442–454. Rosner, G., Hotzl, H., Winkler, R. (1996). Continuous wet-only and dry-only deposition measurements of 137 Cs and 7 Be: An indicator of their origin. Appl. Radiat. Isotopes 47, 1135–1139. Small, S.H. (1960). Wet and dry deposition of fallout materials at Kjeller. Tellus 12, 308–314. Stewart, K. (1966). The resuspension of particulate material from surfaces. In: Fish, B.R. (Ed.), Surface Contamination. Pergamon Press, Oxford, UK, pp. 63–74. Todd, J.F., Wong, G.T.F., Olsen, C.R., Larsen, I.L. (1989). Atmospheric depositional characteristics of beryllium-7 and lead-210 along southeastern Virginia coast. J. Geophys. Res. 94, 11106–11116. Turekian, K.K., Benninger, L.K., Dion, E.P. (1983). 7 Be and 210 Pb total deposition fluxes at New Haven, Connecticut and at Bermuda. J. Geophys. Res. 88, 5411–5415. United Nations Scientific Committee on the Effects of Atomic Radiation, UNSCEAR (1982). Ionizing Radiation: Sources and Biological Effects. United Nations, New York. United Nations Scientific Committee on the Effects of Atomic Radiation, UNSCEAR (2000). Sources and Effects of Ionizing Radiation. United Nations, New York. Young, J.A., Silker, W.B. (1980). Aerosol deposition velocities on the Pacific and Atlantic Oceans calculated from 7 Be measurements. Earth Planet. Sci. Lett. 50, 92–104.

71

Chapter 4

Residence time of tropospheric aerosols in association with radioactive nuclides

1. Residence time of aerosol particles The residence time of atmospheric aerosol particles in the lower atmosphere, assuming that the troposphere is considered a well-mixed reservoir (closed system), is a function of various removal processes, the most important being: (a) dry deposition by impaction, diffusion and sedimentation, and (b) wet deposition by rain drops (precipitation scavenging) as a result of processes occurring both within and below the rain cloud. There could be variations in the removal rates at different continental locations on the globe (Lehmann and Sittkus, 1959), over the oceans (Koch et al., 1996), and at high altitudes of the atmosphere (Martell, 1970; Moore et al., 1973) due to changes in meteorological conditions. There is also a dependence of the tropospheric aerosol residence time on latitude (Ehhalt, 1973; Balkanski et al., 1993; Koch et al., 1996). The residence time of atmospheric aerosol particles can be estimated by means of radioactive nuclides as tracers, which become attached to aerosol particles and are removed with them as they are scavenged by precipitation or undergo dry fallout (Warneck, 1988). Several methods have been used for estimating the mean residence time of atmospheric aerosol particles. These include measurements of the activities and ratios of (i) cosmic-ray produced radionuclides, such as 7 Be (T1/2 = 53.3 d) (Shapiro and Forbes-Resha, 1976; Winkler et al., 1998; Yu and Lee, 2002), (ii) radioactive decay products of radon, 222 Rn, and thoron, 210 Rn, which emanate from continental surfaces into the atmosphere, such as 210 Pb (T1/2 = 22.3 y), 210 Bi (T1/2 = 5.01 d) and 210 Po (T1/2 = 138.38 d) (Francis et al., 1970; Poet et al., 1972; Lambert et al., 1982, 1983; Marley et al., 2000; Baskaran and Shaw, 2001), and (iii) fission product radionuclides from man-made explosions from nuclear weapons testing or reactor accidents at nuclear facilities, such as 89 Sr (T1/2 = 50.5 d), 90 Sr (T1/2 = 28.8 y) and 140 Ba (T1/2 = 12.79 d) (Kuroda, 1982). However, there is disagreement between the derived values of the residence times due to various processes, including the fact that they refer to different portions of the atmosphere, e.g., cosmic-ray produced radioisotopes refer to the upper troposphere or lower stratosphere, such as 7 Be, while radon decay products, such as RADIOACTIVITY IN THE ENVIRONMENT VOLUME 12 ISSN 1569-4860/DOI: 10.1016/S1569-4860(07)12004-0

© 2008 Elsevier B.V. All rights reserved.

72 210 Pb, 210 Bi

C. Papastefanou

and 210 Po, refer to the lower troposphere and also to the existence of different sources for some radioisotopes, such as 210 Po. The collection of atmospheric aerosol particles is typically carried out with high volume jet air samplers, of type Staplex TFIA-2 having glass-fibre filters of type Staplex TFAGF 810, which are highly retentive for particulate material, 20.32 cm × 25.40 cm (8 × 10 ) in dimensions and of 99.28% collection efficiency for submicron particles as small as 0.3 µm and over. The air-flow rate of these samplers is regulated from 1.7 m3 min−1 (60 cfm) to 1.92 m3 min−1 (68 cfm) (average 1.84 m3 min−1 or 65 cfm). The length of each collection period is 24 h. The size fractionation of atmospheric aerosol particles is carried out with aerosol cascade impactors, of type Andersen 2000, as follows: (i) the 1-ACFM design is operated at an air-flow rate of 28 l min−1 (1 ft3 min−1 ). Its stages have effective cut-off diameters (ECD) of 0.4, 0.7, 1.1, 2.1, 3.3, 4.7, 7.0 and 11.0 µm, (ii) the low-pressure modification which alters the impactor’s operation by increasing the resolution in the submicron region, involves a regulated air-flow rate of 3 l min−1 , five low-pressure (114 mm Hg which is the absolute pressure downstream of the critical orifice) stages for the submicron region and eight atmospheric pressure stages for separating aerosol particles above 1.4 µm. The ECDs of the low-pressure stages are 0.08, 0.11, 0.23, 0.52 and 0.90 µm, whereas for the upper stages they are 1.4, 2.0, 3.3, 6.6, 10.5, 15.7, 21.7 and 35.0 µm, (iii) the high volume cascade impactors have a regulated air-flow rate either of about 0.57 m3 min−1 (20 cfm) or 1.13 m3 min−1 (40 cfm) and the ECDs are 0.41, 0.73, 1.4, 2.1, 4.2 and 10.2 µm for the 20 cfm configuration or 0.49, 0.95, 1.5, 3.0 and 7.2 µm for the 40 cfm configuration at standard temperature and atmospheric pressure (250◦ and 760 mm Hg). The stainless steel plates supplied by the manufacturer are used for collection of the aerosol particles. Either polycarbonate or glass-fibre backup filters are used to collect all particles below the 0.08-µm collection plate for the low-pressure cascade impactors, below the 0.4-µm collection plate for the 1-ACFM impactors, and below the 0.41- or 0.49-µm collection plate for the high volume cascade impactors. The length of each collection period varies from 1 to 24 h for 7 Be, and 210 Pb, 210 Bi and 210 Po, depending on impactor type and objective. The activity of 7 Be-associated aerosol particles is measured through its gamma-ray peak of 477.6 keV using a high resolution (e.g., 1.9 keV at 1.33 MeV of 60 Co), high efficiency (e.g., 42%) low-background HP Ge detector. The uncertainty of the gamma-counting system for 7 Be measurements must be better than 8.5%. The activity of 210 Pb- and 210 Bi-associated aerosol particles is measured by a lowbackground phoswitch scintillation detector system1 having a background of about 2 cpm and efficiency usually higher than 40% for counting beta radiation. This system is comprised of a thin CaF2 (Eu) primary crystal with a decay time of 0.23 µs. The samples are counted for long enough to obtain a statistical accuracy better than 5%. It must be noted that 210 Pb and 210 Bi are chemically separated and measured for their activities. A detailed description of the analytical method is presented by Jaworowski (1963). The activity of lead-210 might also be measured by a surface barrier Ge detector through its gamma peak at 46.5 keV. The activity size distribution and the activity median aerodynamic diameter (AMAD) of the aerosol particles are determined upon measurement of the activities of the aerosol-associated radioactive nuclides. A plot of a gamma-ray spectrum of an atmospheric aerosol sample (air filter) obtained by a Ge detector is illustrated in Figure 3.1, in which the 477.6 keV-gamma peak of 7 Be and 1 Harshaw model TASC-12-A6.

Residence time of tropospheric aerosols in association with radioactive nuclides

73

Table 4.1 Activity median aerodynamic diameter (AMAD) of atmospheric aerosol particles (µm) 7 Be

210 Pb

Reference

0.76–1.18 (avg. 0.90) – 0.36, 0.38, 0.39, 0.48 0.65–1.09 (avg. 0.77) 0.44–0.74 (avg. 0.57) 0.33–1.15 (avg. 0.67) 0.70

– 0.60, 0.61, 0.63, 0.77 (avg. 0.60) – 0.56 0.28–0.74 (avg. 0.53) – 0.55

Papastefanou and Ioannidou (1995) Sanak et al. (1981) Bondietti and Brantley (1986) Reineking and Porstendörfer (1995) Winkler et al. (1998) Yu and Lee (2002) Porstendörfer and Gründel (2003)

46.5 keV-gamma peak of 210 Pb are clearly shown. The activity aerodynamic size distribution of a sample of 7 Be atmospheric aerosol obtained by a 1-ACFM cascade impactor is shown in Figure 2.1 (Papastefanou, 2006). The activity aerodynamic size distribution of a sample of 210 Pb atmospheric aerosol obtained by a high-volume (HVI) cascade impactor is shown in Figure 2.8 (Bondietti et al., 1987). 2. Residence time of tropospheric aerosol particles associated with the cosmic ray produced 7 Be A method for estimating the residence time of tropospheric aerosol particles associated with the cosmic-ray produced radionuclides, such as 7 Be, is based on the aerosol particle growth rate, which is the change of particle diameter with time, which was estimated to be 0.004 to 0.005 µm h−1 (McMurry and Wilson, 1982) and the difference between the activity median aerodynamic diameter, AMAD, of a radionuclide, e.g. 7 Be, and the size of the Aitken nuclei in the size distribution of the aerosol particles, which is 0.015 µm (NRC, 1979). The AMAD of all radionuclides is in the accumulation mode of the size distribution of atmospheric aerosol particles which ranges between 0.1 and 2.0 µm (NRC, 1979; Papastefanou and Bondietti, 1987). Table 4.1 shows data for the activity median aerodynamic diameter, AMAD, of atmospheric aerosol particles associated with 7 Be and 210 Pb atoms. Regarding the 7 Be aerosols, the AMAD values varied from 0.33 to 1.18 µm (Bondietti and Brantley, 1986; Papastefanou and Ioannidou, 1995; Reineking and Porstendörfer, 1995; Winkler et al., 1998; Yu and Lee, 2002; Porstendörfer and Gründel, 2003), while for the 210 Pb-aerosols, the AMAD values varied from 0.28 to 0.77 µm (Sanak et al., 1981; Reineking and Porstendörfer, 1995; Winkler et al., 1998; Porstendörfer and Gründel, 2003). The residence time, τR , is described by the formula τR =

AMADmean − SizeAitken nuclei . Mean particle growth rate

(4.1)

In the same way, the residence time of atmospheric aerosol particles associated with 210 Pb, a decay product of soil-emanated radon 222, can be determined, because 7 Be and 210 Pb have a common association with atmospheric aerosol particles (Dibb and Jaffrezo, 1993). Residence times of tropospheric aerosols derived from 7 Be and 210 Pb concentrations in air are presented in Table 4.2.

74

C. Papastefanou

Table 4.2 Residence times of tropospheric aerosols derived from 7 Be and 210 Pb activities in air (days) Investigation

Coordinates

7 Be

210 Pb

Reference

Thessaloniki, Greece Neuherberg, Germany Fullerton, California Hong Kong Northern Hemisphere Tropics Southern Hemisphere South to North

40◦ 38 N, 22◦ 58 E 48◦ 13 N, 11◦ 36 E 33◦ 52 N, 117◦ 55 W 22◦ 18 N, 114◦ 10 E 30◦ N–80◦ N 30◦ S–30◦ N 30◦ S–80◦ S 80◦ S–80◦ N

7.4–8.9 (avg. 8.0) 5–6 35.4 2.6–11.8 – – – 21

– 4.5 – – 5–10 10–15 5 9

Papastefanou and Ioannidou (1995) Winkler et al. (1998) Shapiro and Forbes-Resha (1976) Yu and Lee (2002) Balkanski et al. (1993) Balkanski et al. (1993) Balkanski et al. (1993) Koch et al. (1996)

Papastefanou and Ioannidou (1995) estimated residence times for 7 Be-aerosols varying between 7.4 and 8.9 days (average 8.0 days) with corresponding AMAD values of aerosol particles associated with 7 Be varying from 0.76 to 1.18 µm (average 0.90 µm) (Table 4.1) for 12 measurements of aerosol samplings carried out during a 1 1/2-year period, thus including all seasons, in the Thessaloniki region (48◦ 38 N, 22◦ 58 E), Northern Greece, with a dry (precipitation-free) climate at temperate latitude. Winkler et al. (1998) estimated residence times of 5–6 days for 7 Be-aerosols and 4–5 days for 210 Pb-aerosols, in 46 measurements of aerosol samplings carried out during a 1 1/4-year period in ground-level air at a semi-rural area in Neuherberg, Germany (48◦ 13 N, 11◦ 36 E), 490 m above sea level. Shapiro and Forbes-Resha (1976) much earlier estimated a mean tropospheric aerosol residence time for 7 Be-bearing aerosols of 35.4 days, significantly higher, i.e. more than four times higher, at Fullerton, California (33◦ 52 N, 117◦ 55 W), also at mid-latitude, over an almost 2-year period, with relatively light precipitation. Yu and Lee (2002) recently estimated mean residence times for 7 Be-associated aerosols ranging from 2.6 to 11.8 days in 14 measurements of aerosol samplings carried out during a 4-month period (November–March), 20 m above ground at Hong Kong (22◦ 18 N, 114◦ 10 E) including winter and spring measurements. Balkanski et al. (1993) following a global three-dimensional model which uses meteorological parameters, such as precipitation scavenging, computed residence times for 210 Pb aerosols in the tropospheric column of about 5 days at southern mid-latitudes up to 80◦ S and 10–15 days in the tropics (30◦ S–30◦ N). Values at northern mid-latitudes up to 80◦ N varied from about 5 days in winter to about 10 days in summer. The residence time of 210 Pb produced in the lowest 0.5 km of atmosphere is on average four times shorter than that of 210 Pb produced in the upper atmosphere. They found that the tropospheric residence time is a function of latitude according to the following equation (Ehhalt, 1973) C (4.2) , Φ where C is the tropospheric column of a radionuclide extending from the surface up to the model layer just below the tropopause, and Φ is the total depositional flux out of the column at a given latitude. τR =

Residence time of tropospheric aerosols in association with radioactive nuclides

75

Koch et al. (1996) using a three-dimensional chemical tracer model similar to that of Balkanski et al. (1993) calculated a tropospheric aerosol residence time of 9 days for 210 Pb aerosols at different latitudes from 80◦ S to 80◦ N. The respective residence time of 7 Be aerosols was 21 days reflecting the high altitude versus low altitude source regions of these two tracers. They also found that the tropospheric residence time is a function of latitude (Balkanski et al., 1993) according to Equation (4.2). The data of Table 4.2 admit residence times of tropospheric aerosols in the range 2.6 to 35.4 days, but gather into two groups of values at 2.6 to 15 days (average 8.8 days) and 21 to 35.4 days (average 28.2 days). The lower values are applicable only to the boundary layer near the Earth’s surface and the higher values are appropriate to the troposphere as a whole (Junge, 1963). Martell and Moore (1974) came to the opposite conclusion, namely, that the high values are due to the contribution of stratospheric aerosols, while the lower values represent the true tropospheric residence time essentially independent of altitude.

3. Residence time of tropospheric aerosol particles associated with the radon decay product radionuclides 210 Pb, 210 Bi, 210 Po and the fission product radionuclides 89 Sr, 90 Sr, and 140 Ba and their activity ratios A method for estimating the residence time of tropospheric aerosol particles associated with radon decay product radionuclides is based on the radioactivity of a pair of genetically related radioisobars, such as 210 Pb, 210 Bi or 210 Pb, 210 Po according to the sequential disintegrations in the beta decay scheme, as 210

β − , 22.3 y

β − , 5.01 d

α, 138.38 d

Pb −−−−−−→ 210 Bi −−−−−−→ 210 Po −−−−−−→ 206 Pb (stable).

(4.3)

The residence time, τR , is described by the formula λBi ·NBi

1 λPb ·NPb τR = · , λBi 1 − λBi ·NBi λPb ·NPb

(4.4)

where λBi · NBi is the activity of 210 Bi, λPb · NPb is the activity of 210 Pb in air and λBi = 0.138 d−1 is the decay constant of 210 Bi. Equation (4.4) was derived from equation of the production and removal of radionuclides assuming steady state equilibrium dNBi (4.5) = λPb NPb − (λBi + λR )NBi = 0, dt where λR = 1/τR is the first-order rate constant for the removal of aerosols from the atmosphere by all processes, that is the inverse of residence time, τR . The ratio of the activities λBi · NBi /λPd · Npd in Equation (4.4) varied from 0.48 to 0.68 (this work) or from 0.42 to 0.85 (Moore et al., 1972). If the activity of 210 Po, λPo NPo , in air is considered and λPo = 5.0 × 10−3 d−1 is the decay constant of 210 Po, then the activity ratio of 210 Po and 210 Pb is given by the equation λPo · NPo = λPb · NPb τR +

τR2 1 λBi · τR +

1 λPo

,

(4.6)

76

C. Papastefanou

from which the residence time, τR , is determined as τR =

−b + (b2 − 4ac)1/2 , 2a

(4.7)

where a = λPb · NPb − λPo · NPo , 1 1 + b = −λPo · NPo , λBi λPo 1 c = −λPo · NPo · , λBi · λPo λBi = 0.138 d−1 and λPo = 5 × 10−3 d−1 .

(4.8)

The ratio of the activities λPo NPo /λPb NPb in Equation (4.6) varied from 0.054 to 0.092 (Moore et al., 1972). The residence time, τR , can also be determined through the ratio of activities of radon, 222 Rn, λ · N , and 210 Pb, λ · N , in air according to the sequential disintegrations in the Rn Rn Pb Pb alpha and beta decay scheme as 222

a, 3.824 d

a, 3.05 m

β − , 26.8 m

β − , 19.7 m

Rn −−−−−−→ 218 Po −−−−−→ 214 Pb −−−−−−→ 214 Bi −−−−−−→ 214 Po a, 164 µs

−−−−−→

210

β − , 22.3 y

Pb −−−−−−→

(4.9)

and via Equation (4.5) by the formula τR =

1 · λPb

1 λRn ·NRn λPb ·NPb

−1

≈

1 λPb · NPb · , λPb λRn · NRn

(4.10)

·NRn where λPb = 8.5 × 10−5 d−1 is the decay constant of 210 Pb and λλRn 1. Pb ·NPb The ratio of the activities λRn · NRn /λPb · NPb in Equation (4.10) varied from 282 to 7700 (Moore et al., 1972). Residence times of tropospheric aerosols derived from 210 Bi/210 Pb, 210 Po/210 Pb and 222 Rn/210 Pb activity ratios in air are presented in Table 4.3. Papastefanou and Bondietti (1991) estimated tropospheric aerosol residence times ranging from 4.8 to 15.3 days (average 8.2 days) based on the 210 Bi/210 Pb activity ratios for 21 measurements of aerosol samplings carried out during an annual period at Oak Ridge, Tennessee (35◦ 58 N, 84◦ 17 W) at temperate latitude with high precipitation (wet climate). Poet et al. (1972) estimated tropospheric aerosol residence times ranging from 1.59 to 13 days (average 5.4 days) when based on the 210 Bi/210 Pb activity ratios and from 11 to 77 days (average 24 days) when based on the 210 Po/210 Pb activity ratios for 20 measurements of aerosol samplings carried out in surface air during a 4 1/2-year period at Boulder, Colorado (40◦ 01 N, 105◦ 17 W). They concluded that a mean tropospheric residence time of about 4 days could be applied for aerosol particles in the lower troposphere and about a week for aerosol particles in precipitation. They also found that the mean aerosol residence time increases with altitude within the troposphere by less than a factor of 3 (Moore et al., 1973). Francis et al. (1970) estimated a mean atmospheric residence time for 210 Pb of 9.6 days ±20% based on the 210 Po/210 Pb activity ratios from the dust of filtrate (larger than 0.22 µm in

Investigation

Coordinates

210 Bi/210 Pb

210 Po/210 Pb

222 Rn/210 Pb

89 Sr/90 Sr

140 Ba/90 Sr

Reference

Oak Ridge, Tennessee Boulder, Colorado

35◦ 58 N, 84◦ 17 W 40◦ 01 N, 105◦ 17 W

4.8–15.3 (avg. 8.2) 1.59–13.0 (avg. 5.4)

– 11–77

–

– –

– –

Argonne, Illinois Madison, Wisconsin Fayetteville, Arkansas Fayetteville, Arkansas Fayetteville, Arkansas Poker Flat, Alaska Eagle, Alaska Jungfraujoch, Switzerland Freiburg, Germany Gif-sur-Yvette, France

41.7◦ N, 88.0◦ W 43◦ 51 N, 89◦ 22 W 35◦ 03 N, 78◦ 54 W 35◦ 03 N, 78◦ 54 W 35◦ 03 N, 78◦ 54 W 65.1◦ N, 147.5◦ W 69.5◦ N, 141.2◦ W 46◦ 32 N, 07◦ 59 E

6–67 – 2.4–25.6 (avg. 8.5) 3–240 (avg. 20) 2–78 – – –

33–66 9.6 – 2–320 (avg. 40) 9–234 11.9–32 0–38.9 1–12 (avg. 6)

– – – – 5–12 – – –

– – – – 6–11 – – –

Papastefanou and Bondietti (1991) Poet et al. (1972) Moore et al. (1972, 1973) Marley et al. (2000) Francis et al. (1970) Fry and Menon (1962) Gavini et al. (1974) Kuroda (1982) Baskaran and Shaw (2001) Baskaran and Shaw (2001) Gäggeler et al. (1995)

47◦ 59 N, 7◦ 51 E 48◦ 52 N, 2◦ 20 E

20 –

– –

– –

Milford Haven, Wales Bombay, India El-Minia, Egypt

51◦ 40 N, 5◦ 02 W 18◦ 58 N, 72◦ 50 E 28◦ N, 30◦ 45 E

– 8.8–10.5 7–9 – 8 4.3–12.87 (avg. 9.83)

– – –

– – –

40 – –

2.2–3.4 – – – – – – – – – 6.5 6.77 – – –

Lehmann and Sittkus (1959) Lambert et al. (1982) Lambert et al. (1983) Peirson et al. (1966) Rangarajan (1992) Ahmed et al. (2000)

Residence time of tropospheric aerosols in association with radioactive nuclides

Table 4.3 Residence times of tropospheric aerosols derived from 210 Bi/210 Pb, 210 Po/210 Pb, 222 Rn/210 Pb, 89 Sr/90 Sr and 140 Ba/90 Sr activity ratios in air (days)

77

78

C. Papastefanou

diameter) collected during a 3-month period (May–August) at Madison, Wisconsin (43◦ 05 N, 89◦ 22 W). Marley et al. (2000) estimated residence times for seven aerosol samples collected at Argonne, Illinois (41.7◦ N, 88.0◦ W) during a 2-year period, three aerosol samples collected at Phoenix, Arizona, one sample collected at Socorro, New Mexico, and five samples collected at Mexico City, Mexico. Based on the 210 Bi/210 Pb activity ratios, they found residence times of 6–67 days, while the 210 Po/210 Pb activity ratios suggested residence times of 33–66 days for the aerosols below 2 µm in size. Much earlier, Fry and Menon (1962) in 12 measurements carried out during a 7-month period (spring and summer) at Fayetteville, Arkansas (36◦ 03 N, 78◦ 54 W), indicated apparent aerosol residence times ranging from 2.4 to 25.6 days (average 8.5 days), based on the 210 Bi/210 Pb activity ratios in Arkansas rain. Gavini et al. (1974) also estimated variable residence times from 3 to 240 days (average 29 days) from 210 Bi/210 Pb activity ratios and from 2 to 320 days (average 40 days) from 210 Po/210 Pb activity ratios in Arkansas rain. Baskaran and Shaw (2001), based on 210 Po/210 Pb activity ratios, estimated that residence times of arctic haze aerosols, from the upper atmosphere to the arctic atmosphere, varied between 12 and 32 days from two measurements carried out during a 1 1/2-year period (winter) at Poker Flat (65.1◦ N, 147.5◦ W) and between 0 and 39 days from eight measurements carried out during a 2-month period (winter) at Eagle (69.5◦ N, 141.2◦ W) in Alaska. Very early on, Lehmann and Sittkus (1959) estimated aerosol residence times of 20 days from 210 Po/210 Pb activity ratios in air at Freiburg, Germany (47◦ 59 N, 7◦ 51 E), while Peirson et al. (1966) estimated aerosol residence times of 40 days from 210 Po/210 Pb activity ratios in air at Milford Haven, Wales (51◦ 40 N, 5◦ 02 E). Lambert et al. (1982) estimated aerosol residence times varying from 8.8 to 10.5 days based on the 210 Bi/210 Pb activity ratios in air, while later in another work they estimated aerosol residence times of 7 to 9 days (average 8.58 days) based on the 210 Bi/210 Pb activity ratios in air near Paris over a 6-year period (Lambert et al., 1983). Gäggeler et al. (1995) carried out measurements over an annual cycle on the activities of 214 Pb, 210 Pb and 212 Pb in air at high altitude in Jungfraujoch, Switzerland (46◦ 32 N, 07◦ 59 E) as high as 3450 m using an epiphaniometer. They estimated aerosol residence times varying between 1 and 12 days (average 6 days) from the 214 Pb/210 Pb activity ratios. Rangarajan (1992) based on the 210 Bi/210 Pb activity ratios in air estimated a mean residence time of the natural radioactive aerosols in the planetary boundary layer at Bombay, India (18◦ 58 N, 72◦ 50 E) of around 8 days from 43 measurements carried out during the long dry period, October to May. The data of Table 4.3 admit residence times of values as low as 1.59 days and as high as 320 days. Mostly, the lower τR values resulted from the 210 Bi/210 Pb activity ratios and the higher τR values from the 210 Po/210 Pb activity ratios. Low τR values also resulted from the 222 Rn/210 Pb activity ratios. Poet et al. (1972) concluded that apparent residence time values for τR based on the 210 Po/210 Pb activity ratios are longer than for τ based on the 210 Bi/210 Pb activity ratios. R The difference may be explained by the presence of a mixture of aerosols of various apparent ages, of which the older aerosols (those of greater age) contribute most of the 210 Po. On the other hand, the solid products of radon-222 decay, that is 210 Pb, 210 Bi and 210 Po, might arise from sources other than radioactive decay within the atmosphere. Poet et al. (1972) showed

Residence time of tropospheric aerosols in association with radioactive nuclides

79

that up to 85% of 210 Po in the atmosphere is of terrestrial origin, and the vertical profile of 210 Po was found to differ appreciably from that expected from the decay of 222 Rn. Lambert et al. (1979) indicated that volcanic gases are very rich in long-lived radon decay products, especially in 210 Po relative to 210 Pb. Soil particles are the most likely contributors, since a part of the tropospheric aerosols originates at the Earth’s surface. Coal burning and forest fires presumably are additional sources of radionuclides.

4. Residence time of tropospheric aerosol particles associated with the fission product radionuclides 89 Sr, 90 Sr, 140 Ba and their activity ratios Apart from the activity ratios of the radon-222 decay product radionuclides, the residence time of tropospheric aerosols can be derived from the activity ratios of the fission product radionuclides released into the atmosphere during the explosions from nuclear weapons testing or nuclear reactor accidents, such as 89 Sr/90 Sr and 140 Ba/90 Sr. These nuclide ratios are considered as nuclear clocks. The applicability of the radionuclide ratios depends on whether steady-state conditions hold at the time and place of measurement and on the kind of sample, whether surface air or precipitation (rain or snow), used for the radioisotope activity determination. Kuroda (1982), following the 22nd, 23rd and 24th Chinese Weapons Tests at Lop Nor Desert (40◦ N, 90◦ E) on September 17, 1977, March 13, 1978 and December 14, 1978, respectively, calculated mean residence times of 90 Sr based on the 89 Sr/90 Sr and 140 Ba/90 Sr activity ratios in atmospheric precipitation in the corresponding periods at Fayetteville, Arkansas (35◦ 03 N, 78◦ 54 W), when the nuclear debris passed over there for the first and second times, ranging from 5 to 12 days and 6 to 11 days (Table 4.3). The difference in the mean residence times may be attributable to the difference in the average particle size distributions of the debris which passed over Fayetteville, Arkansas for the first and second times. It is likely that large fallout particles were removed from the atmosphere more rapidly than smaller particles, and hence the nuclear debris which arrived at Fayetteville for the second time consisted of finer particles than that which arrived earlier, and hence the mean residence time for the former was longer than that for the latter. The values of mean residence times based on 210 Bi/210 Pb activity ratios, of 2–78 days, and on 210 Po/210 Pb activity ratios, of 9–243 days (Table 4.3), are much greater than the values of the residence times obtained from the fission product radionuclide activity ratios. The fact that there is a large difference between these values may not necessarily be attributable to experimental errors. It is more likely that the differences are real and are attributable to the fact that, while the fresh fission product radionuclides were injected by the Chinese 22nd, 23rd and 24th tests almost exclusively into the troposphere, the radon decay product radionuclides found in rain are mixtures of two different components: tropospheric and stratospheric. The radon decay products in the lower stratosphere descend rapidly during the winter and mix with the tropospheric component, while such a mixing process occurs less frequently in the summer and fall periods (Kuroda et al., 1978). It is interesting to speculate on the following sequence of events: Suppose that a large quantity of nuclear debris was injected into the stratosphere during the fall period. A large concentration of fission product radionuclides with an apparent “age” of several months will

80

C. Papastefanou

be located in the lower stratosphere during the early spring months of the following year. Now, suppose that these fission product radionuclides were brought down by a violent storm which penetrated the tropopause (Dingle, 1965). If a mixture of radon decay product radionuclides, whose apparent “age” was also several months (Burton and Stewart, 1960), was located in the lower stratosphere together with the fission product radionuclides, approximately concordant residence times should be obtained from the 89 Sr/90 Sr and 210 Po/210 Pb activity ratios in rainwater at ground-level.

5. Residence times of sulfate aerosols in the atmosphere Gas-phase transformations in the atmosphere produce low-vapour pressure species, with the oxidation of SO2 and other reduced sulphur species dominating aerosol formation and growth. Oxidation of SO2 in the gas phase produces H2 SO4 , a readily condensible species that either combines with other molecules (new particle formation) or condenses on existing aerosols. The size distribution of sulfates on aerosol particles in the atmosphere is the result of a combination of various independent processes including coagulation and growth on the small Aitken particles, large particle sedimentation, precipitation scavenging, etc. (Moore et al., 1980). As an aerosol population ages, coagulation depletes the mass in the smallest size ranges, while the wet and dry removal processes preferentially deplete the mass in the largest sizes. Both trends are offset by new aerosol production (Milford and Davidson, 1987), hence size characteristics might affect the atmospheric residence time. Moore et al. (1980) reported that calculated mean residence times increase with increasing particle size interval in the troposphere. Measurements of sulfate aerosols are usually carried out with an Andersen cascade highresolution submicron region low-pressure ambient particle-sizing cascade impactor (LPI) for aerosol collection. This low-pressure impactor involves a regulated flow rate of 3 l min−1 , five low-pressure (114 mm Hg) stages for the submicron region and eight atmospheric pressure stages for collecting aerosols larger than 1.4 µm in diameter. The effective cut-off diameters (ECDs) of the low-pressure stages are as follows: 0.08, 0.11, 0.23, 0.52 and 0.90 µm, whereas for the upper stages they are 1.4, 2.0, 3.3, 6.6, 10.5, 15.7, 21.7 and 35.0 µm. The stainless steel plates supplied by the manufacturer are used for aerosol collection. A back-up polycarbonate filter (pore size 0.4 µm Nuclepore, type 111807) is used to collect particles below the 0.08-µm collecting plate. These polycarbonate membranes are usually used for SO2− 4 measurements to avoid artifacts that might result from SO2 adsorption by glass fibre filters when used as backup filters (Bondietti and Papastefanou, 1989). The length of each collection period varies from 2 to 4 days. The stainless steel collecting plates of the impactors are leached with a 2-ml solution of 0.001 M HCl. Sulfates are measured with a Dionex anion-exchange chromatograph after deionised water-leaching of the impactor plates. Procedure blanks are used to test for anion contamination from sources other than the sampled air. The samples are collected from the ground level up to 20 m height. Figure 4.1 shows the mass aerodynamic size distribution of sulfate aerosols obtained by a low-pressure cascade impactor (LPI) (Bondietti and Papastefanou, 1993). In Figure 4.1a is the frequency distribution of sulfate aerosols, in which are indicated the Aitken nuclei mode (I), the accumulation mode (II) and the coarse mode (III), while in Figure 4.1b is the log-normal distribution of sulfate aerosols. This distribution was selected from 12 measurements made

Residence time of tropospheric aerosols in association with radioactive nuclides

81

(a)

(b) Fig. 4.1. Aerodynamic size distribution of sulfate aerosols obtained by a low-pressure cascade impactor (LPI) (12 measurements). Lower Dp limits are arbitrary. (a) Frequency distribution, I. Aitken nuclei mode, II. Accumulation mode, III. Coarse particles mode. (b) Log-normal distribution.

over an almost 6-month period (14 March–6 September 1985) at Oak Ridge, Tennessee, (35◦ 58 N, 84◦ 17 W). It shows a dominant submicron peak in the 0.23–0.52-µm region. On average, about 68.7% of the mass of sulfates was found to be associated with aerosols in the 0.08–1.4-µm size range of the low-pressure stages (accumulation mode). These particles originate and grow by a condensation and droplet-phase process (John et al., 1990) and dominate the surface area and usually the volume (mass) of the atmospheric aerosol. The mass of sulfates associated with aerosol particles smaller than 0.08-µm in diameter (Aitken nuclei mode) was 12.6%, while for aerosol particles larger than 1.4-µm in diameter (coarse mode) it was 18.7%. Evidence of both Aitken nuclei and the coarse mode aerosols is also shown in Figure 4.1a. The mass median aerodynamic diameter (MMAD) for the accumulation mode aerosols for the 12 measurements (Bondietti and Papastefanou, 1993) varied from 0.25 to 0.31 µm (average 0.28 µm). The mean value of the geometric standard deviation (σg ) was 2.06. The sulfate loadings varied from 3.2 to 22 µg m−3 (average 11.3 µg m−3 ). Atmospheric aerosol size distributions consist basically of three modes: nuclei mode (0.1µm diameter), accumulation mode (<2.0-µm diameter, but >0.1-µm diameter) and coarse particle mode (>2.0-µm diameter). Depending on their source there may be from one to three

82

C. Papastefanou

distinct maxima in the surface or volume (mass) distributions. Assuming that the fraction of the aerosol with particle diameter ranging from 0.08 µm up to 1.4 m, low-pressure stages (accumulation mode) is F1 and the fraction below 0.08 µm, back-up filter (Aitken nuclei mode) is F2 , then, according to the aerosol formation mechanism, the mean residence time τR of the aerosol particles is controlled by the ratio F1 /F2 , which was affected by homogeneous nucleation, source injection of very small particles, variabilities in the dominance of gas and liquid phase reactions, aerosol removal, etc., i.e., F1 /F2 = constant. τR

(4.11)

Therefore, if we consider the sulfate aerosols in terms of the naturally occurring radioactive aerosols (e.g., 210 Pb-aerosols), the following equation applies: F1 /F2 , SO2− 4 τR ,

SO2− 4

=

F1 /F2 , 210 Pb . τR , 210 Pb

(4.12)

Using the data on radon decay product aerosols, Papastefanou and Bondietti (1991) reported a mean residence time, τR , of 8 days for aerosols of 0.3-µm activity median aerodynamic diameter (AMAD) size as determined from 210 Bi/210 Pb activity ratios. From the decay scheme of 218 Po (T 1/2 = 3.05 min), because of the relatively short half-lives of the product radionuclides after two α-decays and two β-decays, 210 Pb-aerosols are produced. From 32 experiments for radon decay product aerosols (210 Pb-aerosols), Papastefanou and Bondietti (1991) calculated average values of fractions F1 and F2 of about 76.11 and 21.32, respectively. From 12 measurements of sulfate aerosols, Bondietti and Papastefanou (1993) calculated in the same manner average values of fractions F1 and F2 of about 68.67 and 12.63, respectively. According to Equation (4.12) and the above mentioned data, a mean residence time, τR , SO2− 4 , of about 12 days would apply to sulfate aerosols of 0.3-µm mass median aerodynamic diameter (MMAD) size. If it is assumed that the particle diameter growth rate during the spring and summer seasons for a continental area is about 0.001 µm h−1 , according to the theory of aerosol growth by condensation and coagulation (McMurry and Wilson, 1982, 1983), then the value of the mean residence time τR can be calculated by dividing the difference between the mean value of the mass median aerodynamic diameter (MMAD) of about 0.3 µm for sulfate aerosols as a result of the mean size distribution and the size of the Aitken nuclei peaking at 0.01–0.015-µm size range by the growth rate of 0.001 µm h−1 . From this calculation a mean residence time, τR , of about 11 days results for sulfate aerosols.

References Ahmed, A.A., Mohammed, A., Ali, A.A., El-Hussein, A., Barakat, A. (2000). A study on aerosol residence time in El-Minia, Egypt. J. Aerosol Sci. 31 (S1), 470–471. Balkanski, P.J., Jacob, D.J., Gardner, G.M., Graustein, W.C., Turekian, K.K. (1993). Transport and residence times of tropospheric aerosols inferred from a global three-dimensional simulation of Pb. J. Geophys. Res. 98 (011), 20573–20586. Baskaran, M., Shaw, G.E. (2001). Residence time of arctic haze aerosols using the concentrations and activity ratios of 210 Po, 210 Pb and 7 Be. J. Aerosol Sci. 32, 443–452.

Residence time of tropospheric aerosols in association with radioactive nuclides

83

Bondietti, E.A., Brantley, J.N. (1986). Characteristics of Chernobyl radioactivity in Tennessee. Nature 322, 313–314. Bondietti, E.A., Papastefanou, C. (1989). Large particle nitrate artifacts in the aerodynamic size distributions of ambient aerosols. J. Aerosol Sci. 20, 667–670. Bondietti, E.A., Papastefanou, C. (1993). Estimates of residence times of sulfate aerosols in ambient air. Sci. Total Environ. 136, 25–31. Bondietti, E.A., Papastefanou, C., Rangarajan, C. (1987). Aerodynamic size associations of natural radioactivity with ambient aerosols. In: Radon and Its Decay Products: Occurrence, Properties, and Health Effects. In: ACS Symposium Series, vol. 331. American Chemical Society, Washington, DC, pp. 377–397. Burton, W.M., Stewart, N.G. (1960). Use of long-lived natural radioactivity as an atmospheric tracer. Nature 186, 584–589. Dibb, J.E., Jaffrezo, J.-L. (1993). Beryllium-7 and lead-210 in aerosol and snow in the Dye 3 Gas, Aerosol and Snow Sampling Program. Atmos. Environ. 27A, 2751–2760. Dingle, A.N. (1965). Stratospheric trapping by intense convective storms: Implication for public health in the United States. Science 148, 227–229. Ehhalt, D.H. (1973). Turnover times of 137 Cs and HTO in the troposphere and removal rates of natural particles and water vapor. J. Geophys. Res. 78, 7076–7086. Francis, C.W., Chesters, G., Haskin, L.A. (1970). Determination of 210 Pb mean residence time in the atmosphere. Environ. Sci. Technol. 4 (7), 586–589. Fry, L.M., Menon, K.K. (1962). Determination of the tropospheric residence time of lead-210. Science 137, 994–995. Gäggeler, H.W., Jost, D.T., Baltensperger, U., Schwikowski, M. (1995). Radon and thoron decay product and 210 Pb measurements at Jungfraujoch, Switzerland. Atmos. Environ. 29 (5), 607–616. Gavini, K.B., Beck, J.N., Kuroda, P.K. (1974). Mean residence times of the long-lived radon daughters in the atmosphere. J. Geophys. Res. 79, 4447–4452. Jaworowski, Z. (1963). The determination of lead-210 and bismuth-210 in biological materials. Nukleonika 8 (5), 333–340. John, W., Wall, S.M., Ondo, J.L., Winklmayr, W. (1990). Modes in the size distribution of atmospheric inorganic aerosol. Atmos. Environ. 24A, 2349–2359. Junge, C.E. (1963). Air Chemistry and Radioactivity. Academic Press, New York. Koch, D.M., Jacob, D.J., Graustein, W.C. (1996). Vertical transport of tropospheric aerosols as indicated by 7 Be and 210 Pb in a chemical tracer model. J. Geophys. Res. 101 (013), 18651–18666. Kuroda, P.K. (1982). Use of radioactive isotopes as atmospheric tracers. In: Labrecht, R.M., Morcos, N. (Eds.), Applications of Nuclear and Radiochemistry. Pergamon Press, New York, pp. 457–481. Kuroda, P.K., Daniel, P.Y., Nevissi, A., Beck, J.N., Meason, J.L. (1978). Atmospheric residence times of strontium-90 and lead-210. J. Radioanal. Chem. 43, 443–450. Lambert, G., Buisson, A., Sanak, J., Ardouin, B. (1979). Modification of the atmospheric polonium-210 to lead-210 ratio by volcanic emissions. J. Geophys. Res. 84 (011), 6980–6986. Lambert, G., Polian, G., Sanak, J., Ardouin, B., Buisson, A., Jegou, A., Leroulley, J.C. (1982). Cycle du radon et de ses descendants: Application a l’etude des echanges troposphere–stratosphere. Ann. Geophys. 38, 497–531. Lambert, G., Sanak, J., Polian, G. (1983). Mean residence time of submicrometer aerosols in the global troposphere. In: Pruppacher, H.R., Semonin, R.G., Slinn, W.G.N. (Eds.), Precipitation Scavenging, Dry Deposition and Resuspension. Elsevier, New York, pp. 1353–1359. Lehmann, L., Sittkus, A. (1959). Bestimmung von Aerosolverweilzeiten aus den RaD und RaF—Gehalt der atmospharischen Luft und des Niederschlages. Naturwissenschaft 46, 9–10. Marley, N.A., Gaffney, O.S., Drayton, P.J., Cunningham, M.M., Orlandini, K.A., Paode, R. (2000). Measurement of 210 Pb, 210 Po and 210 Bi in size-fractionated atmospheric aerosols: An estimate of fine-aerosol residence times. Aerosol Sci. Technol. 32, 569–583. Martell, E.A. (1970). Transport patterns and residence times for atmospheric trace constituents vs altitudes. In: RadioNuclides in the Environment. In: ACS Symposium Series, vol. 93. American Chemical Society, Washington, DC, pp. 138–157. Martell, E.A., Moore, H.E. (1974). Tropospheric aerosol residence times: A critical review. J. Rech. Atmos. 8, 903– 910. McMurry, P.H., Wilson, J.C. (1982). Growth laws for the formation of secondary ambient aerosols: Implications for chemical conversion mechanisms. Atmos. Environ. 16, 121–134.

84

C. Papastefanou

McMurry, P.H., Wilson, J.C. (1983). Droplet phase (heterogeneous) and gas phase (homogeneous) contributions to secondary ambient aerosol formation as functions of relative humidity. J. Geophys. Res. 88 (C9), 5101–5108. Milford, J.B., Davidson, C.I. (1987). The sizes of particulate sulfate and nitrate in the atmosphere: A review. J. Air Pollut. Control Assoc. 37, 125–134. Moore, H.E., Poet, S.E., Martell, E.A. (1972). Tropospheric aerosol residence times indicated by radon and radondaughter concentrations. In: Adams, J.A.S., Lowder, W.M., Gessel, T.F. (Eds.), Natural Radiation Environment II. Technical Information Center/U.S. Department of Energy, Washington, DC, CONF-720805-P2, pp. 775–786. Moore, H.E., Poet, S.E., Martell, E.A. (1973). 222 Rn, 210 Pb, 210 Bi and 210 Po profiles and aerosol residence times versus altitudes. J. Geophys. Res. 78 (30), 7065–7075. Moore, H.E., Poet, S.E., Martell, E.A. (1980). Size distribution and origin of lead-210, bismuth-210 and polonium210 on airborne particles in the troposphere. In: Gesell, T.F., Lowder, W.M. (Eds.), Natural Radiation Environment III. Technical Information Center/U.S. Department of Energy, Washington, DC, CONF-780422, pp. 415– 429. National Research Council, NRC (1979). Airborne Particles. University Park Press, Baltimore, MD. Papastefanou, C. (2006). Residence time of tropospheric aerosols in association with radioactive nuclides. Appl. Radiat. Isotopes 64, 93–100. Papastefanou, C., Bondietti, E.A. (1987). Aerodynamic size associations of 212 Pb and 214 Pb in ambient aerosols. Health Phys. 53, 461–472. Papastefanou, C., Bondietti, E.A. (1991). Mean residence times of atmospheric aerosols in the boundary layer as determined from 210 Bi/210 Pb activity ratios. J. Aerosol Sci. 22, 927–931. Papastefanou, C., Ioannidou, A. (1995). Aerodynamic size association of 7 Be in ambient aerosols. J. Environ. Radioact. 26, 273–282. Peirson, D.H., Cambray, R.S., Spicer, G.S. (1966). Lead-210 and polonium-210 in the atmosphere. Tellus 18, 427– 433. Poet, S.E., Moore, H.E., Martell, E.A. (1972). Lead-210, bismuth-210 and polonium-210 in the atmosphere: Accurate ratio measurement and application to aerosol residence time determination. J. Geophys. Res. 77 (33), 6515–6527. Porstendörfer, J., Gründel, M. (2003). Comparison of the activity size distribution of the radionuclide aerosols in outdoor air. In: Book of Abstracts, Dresden Symposium on Radiation Protection, Dresden, Germany, March 3–7, 2003. Dresden University of Technology, Dresden, p. 24. Rangarajan, C. (1992). A study of the mean residence time of the natural radioactive aerosols in the planetary boundary layer. J. Environ. Radioact. 15, 193–206. Reineking, A., Porstendörfer, J. (1995). Time variations of size distribution of aerosol-attached activities of 212 Pb, 210 Pb and 7 Be in the outdoor atmosphere. In: Book of Abstracts, Natural Radiation Environment VI, Montreal, Canada, 5–9 June 1995. Clarkson University, Potsdam, NY, p. 199. Sanak, J., Gaudry, A., Lambert, G. (1981). Size distribution of 210 Pb aerosols over oceans. Geophys. Res. Lett. 8 (10), 1067–1069. Shapiro, M.H., Forbes-Resha, J.L. (1976). Mean residence time of 7 Be-bearing aerosols in the troposphere. J. Geophys. Res. 81 (15), 2647–2649. Warneck, P. (1988). Chemistry of the Natural Atmosphere. Academic Press, San Diego, CA, pp. 360–373. Winkler, R., Dietl, F., Franck, G., Tschiersch, J. (1998). Temporal variation of 7 Be and 210 Pb size distribution in ambient aerosol. Atmos. Environ. 32 (6), 983–991. Yu, K.N., Lee, L.Y.L. (2002). Measurement of atmospheric 7 Be properties using high efficiency gamma spectroscopy. Appl. Radiat. Isotopes 57, 941–946.

85

Chapter 5

Radioactive particles and human subjects

1. Introduction Human studies with radioactive aerosols are described here with particular reference to those relevant to exposure to atmospheric pollutants. The Task Group on Lung Dynamics (TASK, 1966) considered the respiratory tract as consisting of three compartments (Figure 5.1) (Chamberlain, 1991): 1. Naso-pharyngeal (NP) region, or upper respiratory tract. This comprises the nose, mouth, naso-pharynx and oro-pharynx. 2. Tracheo-bronchial (TB) region. This begins at the larynx and comprises the trachea, bronchi and bronchioles. The branching airways are smaller and more numerous at each division, ending with the terminal bronchioles which are about 0.5 mm in diameter. 3. Pulmonary (P) or alveolar region. The respiratory bronchioles and alveoli perform the function of the lung in exchanging oxygen and CO2 . There are about 5 × 108 alveoli, each on average about 150 µm in diameter. The walls of the naso-pharyngeal and tracheo-bronchial regions are covered with cilia, which propel the mucous layer out of the respiratory tract. Particles caught in the mucous are brought up and eventually swallowed. The alveoli are not ciliated, and particles deposited in the pulmonary region remain there until they dissolve in lung fluid, or are engulfed by wandering cells called macrophages, which transport them to the ciliated bronchioles or to the lymphatic system of the lung.

2. Radioactive dose from inhalation of radon decay product aerosols The largest proportion of the natural radiation dose to humans results from the inhalation of the short-lived radon decay products 218 Po, 214 Pb, 214 Bi and 214 Po. During respiration these radioactive decay products of radon are deposited within different regions of the lung and can lead to substantial local doses there. For this reason, the natural radiation dose receives strong consideration in radiation protection. Thus, the surveillance of radon exposure in homes and in workplaces has emerged as an important operation. In all dosimetric models, the dominant parameter related to dose is the activity size distribution of the radon decay products in air as the original deposition destination and amount of RADIOACTIVITY IN THE ENVIRONMENT VOLUME 12 ISSN 1569-4860/DOI: 10.1016/S1569-4860(07)12005-2

© 2008 Elsevier B.V. All rights reserved.

86

C. Papastefanou

Fig. 5.1. Diagram of the respiratory tract.

the inhaled activity deposited in the lung depend on the particle size. For the determination of radiation dose, the relationship between the potential alpha energy concentration, PAEC, of the radon decay product exposure, Cp t, and dose, the so-called dose concentration factor, DCF, is needed and is defined as DCF = Dose/Cp t,

(5.1)

where Cp is the potential alpha energy concentration (PAEC) of the short-lived radon decay products and t is the time of inhalation (Porstendörfer, 2001). In the estimation of the dose conversion factor, DCF, by dose model calculations, the activity size distribution in terms of potential alpha energy concentration (PAEC) is an important input parameter. For practical reasons related to measurements, the activity size distribution should be divided into three parts as follows: (i) the size distribution of the unattached radon decay products, (ii) the size distribution of the radon decay product aerosol, and (iii) the unattached fraction in terms of potential alpha energy concentration, fp . These quantities depend strongly on the different interaction processes of the decay products in air like attachment to aerosol particles, cluster formation, neutralisation of the cluster, and plateout on surfaces and on composition and behaviour of the atmospheric air (number concentration and size distribution of the aerosol, humidity, trace gases, turbulence).

Radioactive particles and human subjects

87

Fig. 5.2. Effective dose per unit exposure of the thoracic regions as a function of the particle diameter (wTL = 0.12, wR = 20). —, 218 Po, - - -, 214 Pb, · · ·, 214 Bi/214 Po. V = 0.75 m3 h−1 (nose breathing).

Fig. 5.3. Effective dose per unit exposure of the thoracic regions as a function of the particle diameter (wTL = 0.12, wR = 20). —, 212 Bi/212 Po, - - -, 212 Pb.

Figures 5.2 and 5.3 show the effective dose per unit exposure of the tracheo-bronchial (bronchial and bronchiolar) and pulmonary or alveolar region as a function of aerosol particle diameter calculated for the radon decay products 218 Po,214 Pb, 214 Bi and 214 Po and the thoron decay products 212 Pb, 212 Bi and 212 Po. A tissue weighting factor of the lung and the radiation weighting factor of 0.12 and 20, respectively, are taken into account. The effective dose from a radioactive mixture can be deduced by adding the effective doses of each decay product. The nuclei of the secretory cells and the basal cells are the critical target cells for the induction of lung cancer (Zock, 1996; Zock et al., 1996). In contrast to the ICRP 66 lung model (ICRP, 1994b), the depth-dose distribution is weighted with the relative number distribution of basal and secretory cells in the epithelium. Therefore, because of the greater number of basal cells in the deeper epithelium tissue, the dose values to the secretory + basal cells of the bronchial region are significantly lower than the dose values to the secretory cells alone (Zock, 1996). The dose contribution of each part of the lung depends strongly on the particle

88

C. Papastefanou

Fig. 5.4. Effective dose by the inhalation of radon decay products as a function of the particle diameter at different relative cancer sensitivity distributions between the bronchial, wBB , bronchiolar, wbb , and alveolar, wAI , region of the thoracic lung (wR = 20, wTL = 0.12). v = 0.75 m3 h−1 (basal cells + secretory cells).

size of the inhaled decay product. The differences between the doses of the short-lived radon decay products are small (Figure 5.2) in contrast to thoron decay products (Figure 5.3). The dose per exposure unit from 212 Bi/212 Po is a factor of up to five times higher than the dose contribution by 212 Pb. The bronchial region gives the dominating contribution to total lung dose in the size range of the cluster mode (smaller than 5 nm in diameter), while the bronchiolar dose is the highest in the size range of the aerosol-attached activity (larger than 10 nm in diameter). Even in the size range 10–100 nm with a higher deposition particle efficiency in the alveolar region the effective dose contribution of this region is more than five times lower than the effective dose of the bronchial and bronchiolar region. In Figure 5.4 the dose fractions of the lung regions are added for a total lung dose. In addition, the dose fractions are weighted with the apportionment factor which reflects the relative cancer sensitivity of the three regions of the lung (Porstendörfer and Reineking, 1999). Taking into account the data about cancer of the human respiratory tract, then the relative sensitivity between bronchial, WB , bronchiolar, Wb , and alveolar–interstitial, WA , lung region is WB : Wb : WA = 0.80 : 0.15 : 0.05.

(5.2)

This partitioning of detriment leads to higher effective doses by the decay product clusters (diameters <5 nm) and lower values by the aerosol fraction than the partitioning WB = Wb = WA = 0.33 (33%) recommended by ICRP 66 (ICRP, 1994b). In addition, Figure 5.4 shows the great differences between nose and mouth breathing for small particle diameters (<5 nm), the size range of the unattached decay products in air. The dose functions for the different organs by inhalation of the thoron decay products are illustrated in Figure 5.5. Also, in the case of these decay products, the dominant contribution to the effective dose is caused by the contribution of the lung. In air, the radon decay products exist in two forms: (i) as unattached fraction and/or (ii) attached to the surface of aerosol particles. After their formation, the unattached decay products are predominantly positively charged. They form charged and neutral clusters by reactions with water vapour and atmospheric trace gases in air. The dominant parameter which influences the fraction of the unattached radon decay products is the attachment

Radioactive particles and human subjects

89

Fig. 5.5. Effective dose contribution to the different organs by inhalation of 212 Pb as a function of the particle diameter (wR = 20, wT after ICRP 60).

rate which depends on the number concentration of the aerosol, Z (cm−3 ) (Porstendörfer and Reineking, 2000). The unattached fraction in terms of potential alpha energy concentration is estimated by the equation fp,Rn =

k 1 λ 1 + k2 λ 2 r 414 = , βZ Z

(5.3)

where k1 = 0.105, k2 = 0.516 and r = 0.8, the recoil factor, which is the fraction of desorption from the particle surface after the alpha decay of 218 Po, β = 1.4 × 10−6 cm3 s−1 being the average attachment coefficient for atmospheric aerosol particles (Porstendörfer and Mercer, 1978). Figure 5.6 shows the unattached fraction of radon (fp,Rn ) and thoron (fp,Tn ) decay products as a function of the particle concentration of atmospheric aerosols. The fp values as a function of the particle concentration, Z, are measured by means of a condensation nuclei counter (CNC). Many working places have aerosol sources due to human activities and combustion and technical processes with a high particle concentration, Z > 4 × 104 particles cm−3 , and therefore fp values below 0.01. The fp values are higher than 0.1 for places with particle concentrations <4 × 103 particles cm−3 . This is the case in poorly ventilated rooms (ventilation rate <0.5 h−1 ) without additional aerosol sources, rooms with an operating air cleaner and poorly ventilated underground caves. For the unattached thoron decay products in indoor air, the unattached fraction is estimated by the equation fp,Tn =

150 . Z

(5.4)

90

C. Papastefanou

Fig. 5.6. Unattached fraction of the radon, fp,Rn , and the thoron, fp,Tn , decay products as a function of the aerosol particle concentration.

For a typical indoor air quality with an aerosol particle concentration in the range of (1–5) × 10−3 particles cm−3 , the fp,Tn values range between 0.01 and 0.03. Similar values are measured in room air (Reineking et al., 1992). The activity size distribution of the unattached radon decay product clusters as a function of the cluster diameter, d, depends on the concentration of water vapour, trace gases and the electrical charge distribution of the radionuclides in air. The cluster size is determined by measurements of the diffusion coefficient of the decay product clusters and is therefore described by the diffusion equivalent diameter. The measurement of the activity size distribution of the unattached decay product clusters (d < 10 nm) under realistic environmental conditions is difficult. Some data from measurements in a few indoor locations with higher radon levels such as dwellings, water supply stations and therapy-mines are presented in Table 5.1. In most cases, the activity size distribution of the unattached decay product clusters in terms of potential alpha energy concentration obtained from measurements under normal conditions of humidity and radon concentration can be approximated by three log-normal distributions (Figure 5.7). Each mode is characterised by the activity median aerodynamic diameter (AMAD), the geometric standard deviation (σg ) and the fraction of the potential alpha energy concentration of the radon decay products, fp . The AMAD values are 0.60, 0.85 and 1.3 nm. In places with high radon concentration and/or high humidity, the fraction with the greatest AMAD value, 1.3 nm, was not registered (Figure 5.8). An explanation is that almost all of the clusters are neutral in air under such conditions, because it is known from chamber studies that the neutralisation rate increases with higher humidity and radon concentration in air (Porstendörfer et al., 1998).

Radioactive particles and human subjects

91

Table 5.1 Average values of the relative size distribution of the unattached radon decay products in terms of potential alpha energy concentration (PAEC). The measured values were approximated by a sum of i log-normal distributions, characterised by the activity median aerodynamic diameter, AMAD, noted as AMDui , geometric standard deviation, σg ui , and activity fraction, fp ui . n = number of measurements; Z = aerosol particle concentrations; C0 = radon concentration; RH = relative humidity Place Z (cm−3 )/RH (%)

n

C0 Mode 1 Mode 2 Mode 3 (Bq cm−3 ) AMDu1 σg u1 fp u1 AMDu2 σg u2 fp u2 AMDu3 σg u3 fp u3 (nm) (nm) (nm)

Dwellings 500–3000/60–70 Water supply station 11000/60 Water supply station 18000/100 Therapy-mine <200/∼98

26

800–1500

0.6

1.1

0.13

0.85

1.1

0.35

1.25

1.1

0.52

4

10000

0.60

1.2

0.38

0.80

1.1

0.45

1.3

1.1

0.17

4

490000

0.52

1.1

0.38

0.80

1.2

0.62 −

−

−

5

8000

0.57

1.2

0.39

0.9

1.1

0.61 −

−

−

Fig. 5.7. Relative size distribution in terms of potential alpha energy concentration, PAEC, of the unattached radon decay product clusters measured in indoor air.

Besides cluster formation, the radon decay products attach to the existing aerosol particles within 1–100 s, forming the radioactive aerosols of the radon decay products. Results of the activity size distribution measurements carried out at different places in outdoor air, dwellings and workplaces are presented in Table 5.2. In general, the activity size distribution of the radon

92

C. Papastefanou

Fig. 5.8. Relative activity size distribution in terms of potential alpha energy concentration, PAEC, of the unattached radon decay product clusters measured in a water supply station with high radon concentration and high humidity.

Table 5.2 Parameters of the activity size distribution of the aerosol-attached short-lived radon decay products in air at different locations. Activity median aerodynamic diameter, AMAD, noted as AMD; geometric standard deviation, σg ; fraction of the mode, fpi . The indices i = n, a and c represent the nucleation (Aitken nuclei), accumulation and coarse particles modes, Z = aerosol particle concentrations Place Z (103 cm−3 )

Outdoor air 10–70 Dwelling 2–500 Workplaces 10–500

Radon decay products Nucleation mode

Accumulation mode

AMDn (nm)

σgn

30–40

1.9–2.2 0.3 (0.2–0.4) 1.7–2.1 0.3 (0–0.4) 1.6–2.2 0.3 (0.2–0.5)

20–40 15–40

fpn

Coarse mode

AMDa (nm)

σga

fpa

AMDc (nm)

310 (250–450) 210 (120–350) 300 (150–450)

2.1 (1.8–3.0) 2.2 (1.6–3.0) 2.5 (1.8–4.0)

0.67 (0.4–0.8) 0.7 (0.6–1) 0.60 (0.3–0.8)

3000 1.7 0.03 (2000–6000) (1.6–2.0) (0–0.10)

σgc

fpc

5000 1.8 0.10 (3000–8000) (1.1–2.8) (0–0.3)

decay product aerosols can be described by three modes, as illustrated by an example for outdoor air in Figure 5.9. There is the nucleation or Aitken mode (AMADn : 30–40 nm), the accumulation mode (AMADa : 250–450 nm) and the coarse mode (AMADc : 2000–6000 nm). The greatest activity fraction of the outdoor air is in the accumulation mode with an average fpa = 0.63. There is a variation of the AMADi , σgi , and fpi values (Table 5.2) obtained

Radioactive particles and human subjects

93

Fig. 5.9. Activity size distribution of the radon decay products in outdoor air.

Fig. 5.10. Typical relative activity size distribution of the potential alpha energy concentration, PAEC, of the radon decay product aerosols in room air. Ventilation <0.5 h−1 , without aerosol sources. AMAD noted as AMDn = 40 nm, - - -, δgn = 1.7, fpn = 0.2; AMDa = 200 nm, · · ·, δga = 2.2, fpa = 0.8.

from continuous measurement in outdoor air over three weeks. However, the correlations with the weather parameters, aerosol concentrations, and differences between day and night time hours were not significant (Porstendörfer et al., 2000). Compared with outdoor air, the AMADa of the accumulation mode in poorly ventilated rooms (ventilation rate <0.5 h−1 ) and

94

C. Papastefanou

Fig. 5.11. Relative activity size distribution of the radon decay product aerosols in air containing a high particle concentration of combustion aerosol from diesel engines and cigarette smoke. —, mine air (working + diesel engine) AMDa = 201 nm. - - -, room air (+ cigarette smoke) AMDa = 270 nm.

without additional aerosol sources, like cigarette smoking or cooking, is shifted to smaller sizes (200 nm, Table 5.2). In addition, under these conditions, the coarse mode of the activity size distribution is not significant, which can be explained by the greater plateout rate of large aerosol particles on room surfaces (Figure 5.10). In most places with one dominating aerosol source, e.g., cigarette smoking or combustion aerosols from diesel engines, the measured activity size distributions can be approximately described by a single log-normal distribution, as is demonstrated by the examples of mine air and air with cigarette smoke in Figure 5.11. The contributions of the potential activity energy concentration in the size ranges of nucleation are generally small. The activity size distributions of the 212 Pb aerosols obtained from measurements in indoor and outdoor air are presented in Table 5.3 and illustrated in Figure 5.12. The lower activity fractions in the nucleation mode in comparison to the values of the radon decay products (Table 5.2) are remarkable. The dose conversion factor, DCF, in effective dose per exposure unit is calculated by taking into account the dose function of the particle diameter (Figure 5.4) and the radon decay product characteristics. The dose conversion factors for living and work places with typical activity size distributions, as a function of the unattached fraction, fp , are illustrated in Figures 5.13 and 5.14, respectively. The value of the dose conversion factor, DCFae , for fp = 0 represents the dose contribution of the radon product aerosols. The values of dose conversion factor in Figure 5.14 are based on aerosol conditions which are typical for many workplaces.

Radioactive particles and human subjects

95

Table 5.3 Parameters of the activity size distribution of the aerosol attached 212 Pb activity in indoor and outdoor air. Activity median aerodynamic diameter, AMAD, noted as AMD; geometric standard deviation, σg ; fraction of the mode, fpi . The indices i = n, a and c represent the nucleation (Aitken nuclei), accumulation and coarse particle modes Place/number of measurements

Thoron decay product 212 Pb Nucleation mode AMDn (nm)

Outdoor air/44 30–50 Indoor air/10

30–50

σgn

Accumulation mode fpn

1.9–2.2 0.11 (0–0.35) 1.9–2.1 0.14 (0.06–0.2)

Coarse mode

AMDa (nm)

σga

fpa

AMDc (nm)

330 (149–519) 217 (175–273)

2.1 (1.5–3.6) 1.8 (1.5–2.1)

0.87 4240 1.6 0.02 (0.65–1.0) (1580–6460) (1.1–2.5) (0–0.23) 0.86 (0.8–0.94)

σgc

fpc

Fig. 5.12. Relative activity size distribution of 212 Pb measured in indoor air (- - -) and outdoor air (—).

Most of the workplaces can be divided into three groups regarding particle number concentration and activity size distribution: (1) workplaces with coarse particles generated by human activities and dispersion processes, (2) workplaces in rooms without coarse particles, and (3) workplaces without coarse particles and nucleation mode generated by one dominant aerosol source (combustion and condensation aerosols).

96

C. Papastefanou

Fig. 5.13. Dose conversion factor, DCF, as a function of the unattached radon decay product clusters in indoor air (—) and outdoor air (- - -). The DCFae value is the dose fraction by the aerosol. Indoors, v = 0.75 m3 h−1 (no coarse particles mode) (30% nucleation or Aitken nuclei mode). Outdoors, v = 1.2 m3 h−1 (3% coarse particles mode) (30% nucleation or Aitken nuclei mode). Nose breathing, wBB : wbb : wAI = 0.8 : 0.15 : 0.05.

Figure 5.14 shows that not only the unattached fraction but also the different aerosol size distributions have a great influence on the values of the dose conversion factor. The contributions of the dose conversion factor of the radon decay products, DCFae , can vary between 5.4 and 10.6 mSv per WLM. The values of the dose conversion factors of the unattached fraction, DCFu , and of the aerosol-attached radon decay products, DCFae , and their sum, DCF (DCF = DCFu + DCFae ) for outdoor air, dwellings and workplaces are summarised in Table 5.4. Typical aerosol conditions such as aerosol concentration and activity size distribution for these places were assumed. Nose breathing and the relative cancer sensitivity distribution of the thoracic lung was taken to be WB : Wb : WA = 0.80 : 0.15 : 0.05 (Porstendörfer and Reineking, 1999). Caused by the different aerosol conditions of the places with regard to particle number concentrations and activity size distributions, the dose conversion factors of the radon decay products vary over the range 4.2–11.5 mSv WLM−1 . In homes and workplaces with high particle concentrations (>6 × 104 particles cm−3 ), the values of the dose conversion factor range between 4.2 and 7.1 mSv WLM−1 . In the case of normal aerosol conditions in homes, at workplaces and outdoors, the values of the dose conversion factor are significantly higher. The values of dose conversion factor vary between 8.0 and 11.5 mSv WLM−1 . The obtained values of the dose conversion factor for homes and workplaces with higher aerosol particle concentrations are comparable with DCF = 3.8 mSv WLM−1 (homes) and DCF = 5.0 mSv WLM−1 (workplaces) derived from epidemiological studies and recommended by ICRP 65 (ICRP, 1994a) but, for places with normal aerosol conditions, there still

Radioactive particles and human subjects

97

Fig. 5.14. Dose conversion factor, DCF, as a function of the unattached radon decay product clusters for working places with different aerosol conditions. The DCFae value is the dose fraction by the aerosol. (- - -) 10% coarse particles mode, (· · ·) no coarse particles mode, (—) no coarse particles mode + no nucleation or Aitken nuclei mode. Nose breathing: v = 1.2 m3 h−1 , wBB : wbb : wAI = 0.8 : 0.15 : 0.05.

exists a discrepancy between the results of the epidemiological and the dosimetric approaches. The calculations give about a factor of 2 higher values for the dose conversion factor. The dose conversion factor for the thoron decay products, obtained from calculation with the data for the dose function (Figure 5.5) and the radon decay product characteristics (Table 5.3) are summarised in Table 5.5. The values of the dose conversion factor vary between 2 and 3 mSv WLM−1 . In industrial exposure, continuously operated filter samplers can be used to measure the potential alpha energy concentration, PAEC, but in houses passive dosimeters are most commonly used and these measure concentrations of radon, but not radon decay products. In Figure 5.15, the bronchial dose is shown as a function of the number of nuclei. For a typical indoor concentration of nuclei of 1010 m−3 , averaged over day and night, the dose conversion factor is 10 nSv per Bq h m−3 . Possible variations in the number and size of nuclei are by a factor of 2. Chamberlain and Dyson (1956) only took into account the bronchial dose for the unattached fraction and estimated a dose conversion factor of 6.2 nSv per Bq h m−3 for a mouth-breathing subject. The radiation dose to the alveolar region depends only slightly on the unattached component. For particles with diameter of 100 nm, the dose conversion factor for the alveolar region was estimated to be 0.4 nSv per Bq h m−3 of 222 Rn. In mines, nuclei are numerous and fp is low and, for example, a median value of 0.01 can be derived from the measurements of George and Hinchliffe (1972). With fp = 0.01, the value of the dose conversion factor becomes 8.8 mSv per WLM. A number of other factors

98

C. Papastefanou

Table 5.4 Average dose conversion factor, DCF, for the inhalation of unattached, DCFu , and aerosol-attached, DCFae , radon decay products in air of different locations arranged according to aerosol conditions. WBB : Wbb : WAI = 0.80 : 0.15 : 0.05 is the relative cancer sensitivity distribution of the bronchial, wbb , and alveolar, wAI , regions of the thoracic lung, v = inhalation rate, Z = aerosol particle concentration. The indices i = n, a and c represent the nucleation (Aitken nuclei), accumulation and coarse particle modes Place

Aerosol/sources i-mode

Work places

DCF = DCFu + DCFae (mSv WLM−1 )

Nose breathing v (m3 h−1 )

DCFu (mSv WLM−1 )

20 (10–40)

1.2

1.5

9.1

10.6

No sourcesa 30% n-mode 70% a-mode + sources 100% a-mode

8 (3–20)

0.75

1.9

6.1

8.0

80 (20–500)

0.75

0.2

4.0

4.2

No sourcesa 30% n-mode 70% a-mode + sources 30% n-mode 60% a-mode 10% c-mode + sources 100% a-mode

8 (3–20)

1.2

3.1

8.4

11.5

40 (20–60)

1.2

0.6

10.6

11.2

80 (50–500)

1.2 1.7

0.3 0.6

5.4 6.5

5.7 7.1

Outdoors No sourcesa 30% n-mode 67% a-mode 3% c-mode Homes

DCFae (mSv WLM−1 )

Particle concentration Z (103 cm−3 )

a Aerosol particles are partly hygroscopic: size was doubled.

Table 5.5 Fraction of the dose conversion factor, DCFi for the inhalation of 212 Pb (mouth breathing, wBB : wbb : wAI = 0.33 : 0.33 : 0.33). Fraction of the mode = fpi . The indices i = u, n, a and c represent the unattached cluster, nucleation (Aitken nuclei), accumulation and coarse particle mode Place

fpTh,i (%)

Outdoor air u: 0.5 n: 11 a: 87 c: 2 Indoor air

u: 2 n: 14 a: 86

Inhalation rate, v (m3 h−1 )

212 Pb

Cluster DCFu (mSv WLM−1 )

DCFn

DCFa (mSv WLM−1 )

DCFc

0.75

0.15

0.40

1.27

0.20

2.0

1.2

0.25

0.65

2.03

0.33

3.3

0.75

0.61

0.52

1.22

Aerosol

DCF = DCFu + DCFn + DCFa + DCFc (mSv WLM−1 )

2.4

Radioactive particles and human subjects

99

Fig. 5.15. Dose to bronchial tissue from radon decay products in terms of radon concentration.

enter into the comparison of radiation dose. The average miner has a higher ventilation rate than a person at home. Against this, the bronchial dose depends on the particle size of the nuclei as shown in Figure 5.16 (James, 1987), and George et al. (1975) found this to average 0.17 µm in mines compared with 0.12 µm in dwellings. The US National Research Council Committee (NRC, 1988) have suggested a dose conversion factor of 8 mSv per WLM for miners, with considerable variation possible depending on ventilation rate and other factors. It is accepted that the bronchial epithelium is the critical tissue. An important parameter in evaluating indoor exposure is the mean life of 218 Po before attachment to nuclei as suggested by Chamberlain and Dyson (1956). Since the number of nuclei is usually less in dwellings than in uranium mines, the bronchial radiation dose, per Bq m−3 of 222 Rn, is greater in the domestic than in the industrial situation. The potential alpha energy of the decay products of thoron, 220 Rn, i.e. mainly 212 Pb, in indoor air is approximately half that of the decay products of radon, 222 Rn (Schery, 1985). Moreover, the radioactive half-life of 212 Pb is long enough for considerable absorption into the bloodstream and movement in mucous flow to occur before the alpha-emitting 212 Bi is formed, and the radiation dose to the bronchial epithelium is only about a third of that from 222 Rn decay products (James, 1987). Measurements of the radiation dose exposure to radon and its decay products is carried out by an electronic radon gas personal dosimeter. Such a dosimeter is that named DOSEman (Sarad, Dresden, Germany) (Grundel and Porstendörfer, 2003). Using the dosimeter, the radon concentration is measured in the environment, e.g. in dwellings, mines, caves, etc., and can be converted to a personal dose. The entire measuring system is accommodated in a handy housing, so that it can be carried comfortably on the body.

100

C. Papastefanou

Fig. 5.16. Dose rate to alveolar (A) and bronchial (B) regions of lungs by inhalation of radon decay products (James, 1987). Range of values for bronchial dose reflects uncertainties in breathing pattern, airway size and clearance.

Fig. 5.17. Functional diagram of the electronic radon gas personal dosimeter DOSEman. MCA: multi-channel analyser, EEPROM: electrically erasable programmable read only memory, IrDA: infrared interface.

The radon gas diffuses through a leather membrane into a cylindrical measurement chamber. In this measurement chamber the charged radon decay products are collected by an electrical field onto a semiconductor silicon-detector and the alpha (particle) radiation is measured spectrometrically. In addition, alpha decays of the radon gas are registered. The alpha decays

Radioactive particles and human subjects

101

Fig. 5.18. Radon spectrum of the electronic radon gas personal dosimeter DOSEman with the chosen region of interest, ROI, and the tailing function of the peaks (counts/channel, cts/chn).

are processed as electronic pulses via a multichannel analyser and the counts in five adjusted regions of interest (ROIs) are stored for selected measurement intervals. A functional diagram of the DOSEman device is shown in Figure 5.17. This dosimeter internally calculates the activity concentration CRn from the ROI counts and the effective dose from the totals of the individual measuring intervals, absorbed since the start of measurement. To calculate the dose values, the equilibrium factor, F , and the dose conversion factor, DCF, must be available and can be changed by the user to adapt to local conditions. The input of these values takes place as a product of the numbers, using PC software. The stored ROIs and the cumulative spectrum can be transferred by a standard infrared interface to a PC and can be processed with a commercial standard spreadsheet analysis program. In Figure 5.18, a spectrum for a radon gas source is shown. Fifty channels are available for the spectrum. The maximum of the 218 Po peak is situated in channel 24 and the maximum of the 214 Po peak is situated in channel 40. The energy difference between the peaks is 1.685 MeV, giving a resolution of 105 keV per channel. The energy cut-off at channel 0 of the spectrum amounts to 3.58 MeV. The spectrum is divided into five ROIs, which are stored and available for analysis. These five ROIs allow a separation of the radon decay products 218 Po, 214 Po and the thoron decay products. In Figure 5.18, an estimated peak tailing for the two radon decay product peaks 218 Po and 214 Po has been drawn. It is seen that the main peak intensity is situated well inside the chosen ROI. Two modes are available for the calculation of the activity concentration of the radon gas. The fast mode uses the peak areas of 222 Rn and 218 Po and the “slow mode” uses in addition the 214 Po peak.

102

C. Papastefanou

Fig. 5.19. Statistical analysis of data on deposition of inhaled aerosol particles in the head (Yu et al., 1981).

3. Deposition of radioactive aerosol particles in the lung Different mechanisms of deposition of particles in the lung are effective in the three regions of the respiratory tract. In the naso-pharyngeal and tracheo-bronchial regions, flow velocities are relatively high and transit times short, so impaction is of major importance, though diffusion and Brownian motion are also important for gases and very small aerosol particles. In the pulmonary region, velocities are very low, and the stay time of air in the lung during each breath allows sedimentation and Brownian movement to operate. Hounam et al. (1971) measured deposition of radioactive aerosol particles in the naso-pharyngeal region by drawing the aerosol in through the nose and out through the mouth of subjects who held their breath. Lippmann (1977), Stahlhofen et al. (1980) and Chan and Lippmann (1980) used collimated gamma scintillation counters to measure activity deposited in the heads of subjects inhaling via the nose. A correction was made for the activity removed by the mucociliary mechanism, swallowed and transferred to the stomach in the short interval between inhalation and measurement of the activity in the head. Yu et al. (1981) analysed these and other measurements of deposition in the naso-pharyngeal region in terms of the parameter ρp dp2 q, where ρp and dp are the density and diameter of the aerosol particles, respectively, and q is the flow rate of the inspired air. The parameter ρp dp2 is the square of the aerodynamic diameter of the aerosol particles, and, assuming that the linear velocity in the airways is proportional to the flow rate, q, it is a function of the stopping distance of the particles. Figure 5.19 shows the correlation of the fractional deposition in the head with ρp dp2 q. A typical value of flow rate q in normal inspiration is 500 ml s−1 , and for a 10-µm diameter unit density aerosol particle this makes ρp dp2 q equal to 5 × 104 g µm2 s−1 . Figure 5.19 shows 90% deposition of 10-µm aerosol particles in the naso-pharyngeal region with nose breathing, and 40% deposition with mouth breathing, at this value of ρp dp2 q. If dp is 1 µm, naso-pharyngeal region deposition is about 10% for nose breathing, and negligible for mouth breathing, at normal inspiratory flows. Aerosol particles in the micrometre size range can be deposited by impaction in the tracheobronchial region, particularly at the carina where the two main bronchi diverge. Using hollow casts of the trachea and main bronchi, Schlesinger et al. (1977) and Chan and Lippmann

Radioactive particles and human subjects

103

Table 5.6 Root mean square Brownian displacement and distance fallen in one second by unit density aerosol particles Diameter (µm)

Brownian displacement (µm)

Distance fallen (µm)

10 5 2 1 0.5 0.2 0.1

1.7 2.4 4.0 5.9 9.0 17 30

3000 730 130 35 9.4 2.2

(1980) found that the efficiency of deposition correlated with the impaction parameter ρp dp2 q. In the pulmonary region, air velocities are too low to impact particles small enough to reach that region, and the mechanisms of deposition are sedimentation and Brownian diffusion. The efficiency of both processes depends on the length of the respiratory cycle, which determines the stay time in the lung. If the cycle is 15 breaths min−1 , the stay time is of the order of a second. Table 5.6 shows the distance fallen in one second and the root mean square distance travelled by Brownian diffusion in one second by unit density particles (Fuchs, 1964). Sedimentation velocity is proportional to particle density, but Brownian motion is independent of density. Table 5.6 shows that sedimentation of unit density particles is more effective in causing deposition than Brownian diffusion when dp exceeds 1 µm, whereas the reverse is true if dp is less than 0.5 µm. For this reason, it is appropriate to use the aerodynamic diam1/2 eter Dp equal to ρp dp when this exceeds 1 µm, but the actual diameter for submicrometre aerosol particles. It is difficult to distinguish particles in the tracheo-bronchial and pulmonary regions by using collimated detectors, but deposition in the two regions can be separated by reference to the different rates of clearance from the two regions. Figure 5.20 shows typical lung clearance curves of the response of counters over the chests of subjects who have inhaled aerosol particles of 2- and 5-µm diameter. By curve stripping, the short and long components can be separated, though sometimes more than two components are found (Morrow and Yu, 1985). The intercepts of the dashed lines in Figure 5.20 give the deposition in the pulmonary region as a fraction of that in the lung, that is pulmonary/(tracheo-bronchial + pulmonary). The fraction deposited in the tracheo-bronchial and pulmonary regions varies between subjects, but in a single subject, pulmonary/(tracheo-bronchial + pulmonary) varies systematically with particle size (Lippmann et al., 1980; Stahlhofen et al., 1980). The first measurements of deposition of radioactive aerosols and lung clearance were those of Albert and Arnett (1955), who used heterogeneous particles of iron oxide labelled with 59 Fe (T 1/2 = 44.6 d). In later work, to obtain monodisperse particles, Albert et al. (1964, 1967) fed a dilute resuspension of iron oxide sol onto a spinning disk. In the outlet duct, the droplets were dried by infrared irradiation, leaving solid particles of size depending on the concentration of the original suspension. The Fe2 O3 particles were labelled with 51 Cr (T1/2 = 27.71 d) or 198 Au (T1/2 = 2.696 d), and no leaching of tracer was observed after storing the particles for several weeks in saline solution or in rat muscle. Because they were aggregates of primary sol particles with diameters of about 0.01 µm, the density of the

104

C. Papastefanou

Fig. 5.20. Typical lung clearance curves: (A) 2-µm aerosol particles, P /(T B + P ) = 0.8; (B) 5-µm aerosol particles, P /(T B + P ) = 0.5.

particles inhaled was found to be 2.5 g cm−3 , which is less than that of solid Fe2 O3 . Numerous measurements of lung clearance using iron oxide particles have been summarised by Lippmann et al. (1980). Booker et al. (1967) used a spinning disk to make monodisperse polystyrene particles. Polystyrene was dissolved in xylene, at a concentration of 0.2%, and chromium acetyl acetonate, labelled with 51 Cr, was added. The spinning disk was operated at 3 × 104 rpm to produce 40-µm droplets of xylene which evaporated to give 5-µm polystyrene spheres. Few et al. (1970) adapted this method to produce particles of polystyrene labelled with 99m Tc (T1/2 = 6.02 h). This radioisotope has only very slight beta emission, so the dose to the lung is low, and though the radioactive half-life is only 6 h, this is adequate for estimation of the ratio pulmonary/(tracheo-bronchial + pulmonary) and for analysis of the kinetics of the mucociliary clearance. To inhale the aerosols labelled with radioisotopes (radioactive aerosols), subjects breathe through a mouth piece. The volume of air in the mouth and tracheo-bronchial region is respiratory dead space, and the activity reaching the pulmonary region depends on the volume of air inhaled as well as the deposition in the tracheo-bronchial region. To obtain reproducible results, subjects take breaths of predetermined volume, and the breathing cycle is also standardised, usually at 14 or 15 breaths/min. Exhaled aerosol is collected on a filter, and the deposition in the mouth also estimated, so that the deposition in the lung (tracheo-bronchial

Radioactive particles and human subjects

105

Fig. 5.21. Experimental results of deposition of aerosol particles in alveolar region (open symbols) and whole respiratory tract (closed symbols) (14 or 15 breaths/min by mouth, tidal volume 1.0 to 1.5 l). Error bars are 1 S.E. Lines are theoretical calculations of Yu and Diu (1982).

+ pulmonary) is known as a fraction of the activity inhaled. The distribution between tracheobronchial and pulmonary regions is then determined by analysis of the clearance curves as in Figure 5.20. Figure 5.21 shows the fraction of the inhaled activity deposited in the lung and in the pulmonary region as a function of the particle size (aerodynamic diameter if greater than 0.5 µm, otherwise actual diameter). The distinction in clearance kinetics from the naso-pharyngeal and pulmonary regions, illustrated in Figure 5.20, has, however, been challenged by Stahlhofen et al. (1986). These authors arranged for subjects to inhale Fe2 O3 particles, labelled with 198 Au, in a bolus of air of defined volume. The bolus was introduced at various stages of the respiratory cycle in an attempt to label different parts of the tract with deposited particles. When a bolus of only 45 cm3 volume was inhaled near the end of an inspired breath, the particles were expected to deposit only in the naso-pharyngeal region. However, a long-term component in the clearance curve was found, implying either some penetration to the pulmonary region, or else slow clearance from some sites in the upper respiratory tract. The lines in Figure 5.21 are the results of theoretical calculations, using models of the respiratory tract (Yu and Diu, 1982). The points are measurements with radioactive aerosols. Numerous other determinations of fractional deposition in the whole tract have been made, using non-radioactive methods to count the number of particles in the inhaled and exhaled air (Heyder et al., 1986; Schiller et al., 1988). Fractional deposition is least for particles of about 0.2 to 0.5 µm diameter. Table 5.6 shows that the combined effect of sedimentation and Brownian motion is then at a minimum. Hygroscopic particles grow as water vapour is condensed on them as they pass down the respiratory tract. Tu and Knutson (1984) found 50% deposition of 0.35-µm NaCl particles inhaled orally. Minimum deposition of NaCl was found at a particle size of 0.075 µm. The results of Figure 5.21 refer to oral breathing, with a ventilation rate of about 1 m3 h−1 . If the ventilation rate is increased, by breathing more rapidly, more particles will be inhaled, but a

106

C. Papastefanou

Fig. 5.22. Respiratory tract clearance model used in calculating dose from inhaled radioactivity. Code: A series of computer programs to perform dosimetric calculations for the ICRP Committee. The values for the removal half-times, T a–j , and compartmental fractions, F a–j , are given in the tabular portion of the figure for each of the three classes of retained materials. The values given for DN–P , DT–B and DP (left column) are the regional depositions based on an aerosol with an AMAD of 1 µm. The schematic drawing identifies the various clearance pathways in the model, a–j, in relation to the depositions DN–P , DT–B , DP and the three respiratory tract regions, N–P, T–B and P. The entry n.a. indicates not applicable.

smaller proportion will reach the pulmonary region because more will be lost by impaction in the upper respiratory tract. Deposition of particles of less than about 3 µm diameter is mainly in the upper respiratory tract, and the proportion penetrating to the lung is reduced if the flow velocity is increased. Breathing through the nose instead of the mouth substantially reduces the fraction of particles in the micrometric range which penetrate to the lung (Tu and Knutson, 1984), as can be deduced from Figure 5.19.

4. Risk assessment due to inhalation of radon decay product aerosols The allowable limit on intake, ALI, for inhaled radioisotopes is based on a multi-compartment lung model. The model accounts for the fact that particulate deposition in the respiratory tract is governed by the activity size distribution of the inhaled aerosol, and that the clearance rate of deposited particulates is governed by the deposition site as well as by the chemical and physical properties of the particulates. Figure 5.22 is a graphical representation of the lung model and shows three regions where inhaled radioactive aerosol particles may be deposited, namely in the naso-pharyngeal region, NP, the tracheo-bronchial region, TB, and the pulmonary region, P, representing the deep respiratory tract where gas exchange occurs (Watson and Ford, 1980). The naso-pharyngeal region is divided into two compartments, (a) and (b). Compartment (a) represents that part of the aerosol particles deposited in the naso-pharyngeal region that dissolves and is absorbed directly into the blood, while compartment (b) represents the aerosol particles that are cleared from the naso-pharyngeal region into the gastrointestinal tract, G.I., by swallowing. The tracheo-bronchial region is also represented by two compartments, (c) and (d), from which deposited particulates are cleared by the same

Radioactive particles and human subjects

107

Fig. 5.23. ICRP dust deposition model. The radioactive or mass fraction of an aerosol which is deposited in the N–P, T–B and P regions is given in relation to the activity or mass median aerodynamic diameter (AMAD or MMAD) of the aerosol size distribution. The model is intended for use with aerosol size distributions having an AMAD or MMAD between 0.2 and 10 µm and whose geometric standard deviations are less than 4.5. Provisional deposition estimates further extending the size range are given by the dashed lines. For the unusual aerosol size distributions having an AMAD or MMAD greater than 20 µm, complete N–P deposition can be assumed. The model does not apply to aerosols with AMAD or MMAD below 0.1 µm.

two mechanisms, dissolution and absorption into the blood and mechanical transfer into the gastrointestinal tract, in this case by way of the ciliary escalator to the throat and into the gastrointestinal tract by swallowing. The pulmonary region is modelled by four compartments. One of these compartments, (e), represents dissolution and absorption into the blood. Compartments (f) and (g) represent transfer of undissolved particulates into the gastrointestinal tract via the upper respiratory tract, the tracheo-bronchial region. Compartment (f) is cleared by mechanical transport, presumably by unbalanced forces during respiratory excursions, and compartment (g) is cleared by alveolar macrophages that migrate into the tracheo-bronchial region. Compartment (h) empties into the pulmonary lymph nodes. The pulmonary lymph nodes are represented by two compartments, (i) and (j). Compartment (i) empties into the blood stream after the particulates have dissolved, while compartment (j) permanently retains some highly insoluble particulates. The exact fraction of the deposited aerosol particles that is cleared by each route and the respective clearance rates are governed by the chemical composition of the aerosol. However, since it is not practical to determine each of these parameters for every compound of every radioactive element, the different compounds of all the elements have been assigned, according to the criteria shown in Figure 5.22, into one of three classes, D, W and Y. Class D aerosols are rapidly cleared from the deep respiratory tract with a clearance half-time of the order of

108

C. Papastefanou

Fig. 5.24. Model usually used to describe the kinetics of radionuclides in the body (from ICRP Publication 30).

one day or a fraction of one day. Class W aerosols are cleared with a clearance half-time of the order of weeks, while class Y aerosols are retained in the lungs with a clearance half-time of the order of years. Of the different aerosol particles that may be transported to the lymph nodes, only class Y aerosols are permanently retained in the lymph nodes. For risk assessment purposes, the lung and the pulmonary lymph nodes are considered as a single organ. That is, the activity in the lung and the lymph nodes is added together, and the total weight of the lungs and pulmonary lymph nodes is used to estimate the dose from inhaled radioactive aerosol particles. ICRP recommendations for inhaled aerosols are based on inhalation and deposition of an aerosol whose activity median aerodynamic diameter, AMAD, is 1 µm and the standard deviation, σg , is 4. This assumed log-normal distribution leads to deposition of 30% of the inhaled aerosol particles in the naso-pharyngeal region, 8% in the tracheo-bronchial region and 25% in the pulmonary region. The balance, i.e. 37%, is exhaled. Deposition for other size distributions is shown in Figure 5.23 (Cember, 1987). Radioactive aerosol particles brought up from the lungs and swallowed enter into the gastrointestinal tract, from where it may subsequently be eliminated in the faeces, while irradiating the different parts of the gastrointestinal tract and other organs during its passage. It may also undergo dissolution in the gastrointestinal tract, and the dissolved portion may be absorbed into the body fluids and transfered to organs where it may be deposited (Figure 5.24). Thus, for example, inorganic mercury, Hg, is deposited mainly in the kidneys, iodine, I, in the thyroid, strontium, Sr, and radium, Ra, in the skeleton, and plutonium, Pu, in the liver and skeleton. The gastrointestinal tract is modelled by four distinct regions, each with its own kinetic parameters, so that organ doses during passage of radioactive aerosol particles can be estimated. Similarly, the organs where radioactive aerosol particles are deposited are also

Radioactive particles and human subjects

109

modelled by appropriate equations, usually one or more first-order linear differential equations, that allow the organ doses to be estimated. To illustrate the application of this lung model, let us estimate the allowable limit on intake, ALI, by inhalation, for 137 Cs, using particulates of 1-µm activity median aerodynamic diameter (AMAD) and a standard deviation (σg ) less than 4.5. From Figure 5.23, we find that 63% of the inhaled aerosol particles is deposited in the respiratory tract, distributed as follows: naso-pharyngeal region: 30%, tracheo-bronchial region: 9% and pulmonary region: 25%, the balance, 37% is exhaled. For risk assessment purposes represented by the allowable limit on intake, ALI, particulates of caesium compounds are assigned to clearance class D aerosols, since they have been found to be cleared rapidly from the lungs. The committed dose equivalent to a target from intake of 1 Bq of 137 Cs is estimated by determining how many becquerels, Bq, of 137 Cs may be inhaled before reaching the limiting committed dose equivalent. First, the dose to the lung from 137 Cs deposited in the lung is estimated, and then that from the 137 Cs distributed in the rest of the body and which also irradiates the lungs. Next, the effective whole body dose from the inhaled activity is estimated. In making these estimations, it should be noted that the radioactive aerosol particles deposited in the naso-pharyngeal region do not contribute directly to the intra-pulmonary dose. They contribute to the lung dose only by virtue of their presence in the rest of the body. The intrapulmonary dose from 1 Bq of 137 Cs intake is due only to the 0.08 Bq of 137 Cs deposited in the tracheo-bronchial region and the 0.25 Bq of 137 Cs deposited in the pulmonary region. The International Commission on Radiological Protection (ICRP, 1981) estimated the likelihood of incurring cancer from irradiation (excluding genetic effects) as 1.25 × 10−2 per Sv of effective dose. Subsequently, they considered the risk from lifetime indoors exposure to radon decay products as 2 × 10−8 per Bq h m−3 equilibrium equivalent concentration, EEC, in each year (ICRP, 1988). These two estimates are broadly compatible, as follows: The potential alpha energy concentration, PAEC, is 5.54 nJ m−3 per Bq m−3 of equilibrium equivalent concentration, EEC, and the 50-y effective dose to the bronchial tissue is 5.54 × 4.5 × 50 nSv = 1.25 µSv. Hence, the lifetime risk, per Bq h m−3 in each year is 1.25 × 10−6 × 1.25 × 10−2 = 1.6 × 10−8 . Jacobi and Paretzke (1985) estimated that uranium miners in Colorado accumulated an average radiation dose exposure of 820 WLM in the years 1950–1977. Using the factor 16 mSv per WLM, the average bronchial dose would have been 13 Sv, giving a chance of 20% of cancer on the basis of the ICRP (1981) estimate. The Biological Effects of Ionizing Radiation, BEIR IV, report of the US National Research Council (NRC, 1988) recorded 256 deaths from lung cancer among the Colorado miners. The total radiation dose exposure was 73600 person-years, and 58 deaths would have been expected if there were no carcinogenic effects. For exposure indoors of the general population, it is more useful to work in terms of the concentration of radon than the concentration of radon decay products. Brown et al. (1986) found the average concentration of 222 Rn in Cornish, UK houses to be 300 Bq m−3 , and Nero (1988) has estimated that 2% of homes in the USA have similar or higher radon concentrations. Assuming an annual effective dose of 80 µSv per Bq m−3 , the lifetime dose is 1.2 Sv. Thus, occupants of many homes receive doses which are a tenth of those received by the Colorado miners, and on the above analysis carry a 2% chance of lung cancer. Taking into account a higher cancer risk factor and also the synergistic effect of smoking, a 4.5% risk of lung cancer to UK residents in homes with 300 Bq m−3 was estimated (NRPB, 1990). As

110

C. Papastefanou

Nero (1988) has pointed out, such a high level of risk greatly exceeds those normally associated with environmental hazards. The radon concentration indoors in terms of radiation dose exposure is expressed in WL (Working Level, the radiation level of 100 pCi per litre or 3700 Bq per m3 of Rn in equilibrium with its decay products). Effects of radon are given in terms of WLM (Working Level Months), which is the exposure at 1 WL for one working month, or 170 h. Since there are 365 × 24 = 8760 h per year and 80% of them (7000 h) are spent indoors, the annual time of exposure is 7000/170 = 41 “working months”. A world-wide “representative value” adopted by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR, 1977) is that 1 pCi per litre or 37 Bq per m3 is the average 222 Rn concentration indoors, with an equilibrium factor (ratio of Rn decay product concentration to their concentration in radioactive decay equilibrium with Rn gas) of 0.5. This corresponds to an average concentration of 0.005 WL or 5 mWL. The annual indoor exposure is then 0.005 WL×41 WM = 0.205 WLM. References Albert, R.E., Arnett, L.C. (1955). Clearance of radioactive dust from the lung. Am. Med. Assoc. Arch. Ind. Health 12, 99–106. Albert, R.E., Petrow, H.G., Salam, A.S., Spiegelman, J.R. (1964). Fabrication of monodisperse lucite and iron oxide particles with a spinning disk generator. Health Phys. 10, 933. Albert, R.E., Lippmann, M., Spiegelman, J., Strehlow, C., Briscoe, W., Wolfson, P., Nelson, N. (1967). The clearance of radioactive particles from the human lung. In: Davis, C.N. (Ed.), Inhaled Particles and Vapours, II. Pergamon Press, Oxford, pp. 361–377. Booker, D.V., Chamberlain, A.C., Rundo, J., Muir, D.C.F., Thomson, M.L. (1967). Elimination of 5 µm particles from human lung. Nature 215, 30–33. Brown, L., Green, B.M.R., Miles, J.C.H., Wrixon, A.D. (1986). Radon exposure of the United Kingdom population. Environ. Int. 2, 45–48. Cember, H. (1987). Introduction to Health Physics. Pergamon Press, New York. Chamberlain, A.C. (1991). Radioactive Aerosols. Cambridge University Press, Cambridge. Chamberlain, A.C., Dyson, E.D. (1956). The dose to the trachea and bronchi from the decay products of radon and thoron. Br. J. Radiol. 29, 319–325. Chan, T.L., Lippmann, M. (1980). Experimental measurements and empirical modelling of the regional deposition of inhaled particles in humans. Am. Ind. Hyg. J. 41, 399–409. Few, J.D., Short, M.D., Thomson, M.L. (1970). Preparation of 99m Tc labelled particles for aerosol studies. Radiochem. Radioanal. Lett. 5, 275–277. Fuchs, N.A. (1964). The Mechanics of Aerosols. Pergamon Press, Oxford. George, A.C., Hinchliffe, L. (1972). Measurements of uncombined radon daughters in uranium mines. Health Phys. 23, 791–803. George, A.C., Hinchliffe, L., Sladowski, R. (1975). Size distribution of radon daughter products in uranium mine atmospheres. Am. Ind. Hyg. J. 34, 484–490. Grundel, M., Porstendörfer, J. (2003). Characteristics of an electronic radon gas personal dosimeter. Radiat. Prot. Dosim. 107, 287–292. Heyder, J., Gebhart, J., Rudolf, G., Schiller, C.F., Stahlhofen, W. (1986). Deposition of particles in the human respiratory tract in the size range 0.005 µm–15 µm. J. Aerosol Sci. 17, 811–825. Hounam, R.F., Black, A., Walsh, M. (1971). The deposition of aerosol particles in the nasopharyngeal region of the human respiratory tract. J. Aerosol Sci. 2, 47–61. International Commission on Radiological Protection, ICRP (1981). Recommendations of the International Commission on Radiological Protection from Its 1980 Meeting. ICRP Publications, vol. 26. Pergamon Press, Oxford. International Commission on Radiological Protection, ICRP (1988). Lung cancer risk from environmental exposure to radon daughters. Ann. ICRP 17 (1).

Radioactive particles and human subjects

111

International Commission on Radiological Protection, ICRP (1994a). Protection Against Radon-222 at Home and at Work. ICRP Publications, vol. 65. Pergamon Press, Oxford. International Commission on Radiological Protection, ICRP (1994b). Human Respiratory Tract Model for Radiological Protection. ICRP Publications, vol. 66. Pergamon Press, Oxford. Jacobi, W., Paretzke, H.G. (1985). Risk assessment for indoor exposure to radon daughters. Sci. Total Environ. 45, 551–562. James, A.C. (1987). Lung dosimetry for radon and thoron daughters. In: Radon and Its Progeny. In: Nazaroff, W.W., Nero, A.V. (Eds.), Indoor Air. Wiley, New York, pp. 259–309. Lippmann, M. (1977). Regional deposition of particles in the human respiratory tract. In: Lee, D.H.K. (Ed.), Handbook of Physiology—Reaction to Environmental Agents. The American Physiological Society, Bethesda, MD, pp. 213–232. Lippmann, M., Yeates, D.B., Albert, R.E. (1980). Deposition, retention and clearance of inhaled particles. Br. J. Ind. Med. 37, 337–362. Morrow, P.E., Yu, C.P. (1985). Models of aerosol behaviour in airways. In: Newhouse, M.T., Dolovich, M.B. (Eds.), Aerosols in Medicine Principles, Diagnosis and Therapy. Elsevier, Amsterdam. National Radiological Protection Board, NRPB (1990). Human exposure to radon in homes. Doc. NRPB 1, 17–32. National Research Council Committee on the Biological Effects of Ionizing Radiations, NRC (1988). Health Risks of Radon and Other Internally Deposited Alpha Emitters (BEIR IV). National Academy Press, Washington, DC. Nero, A.V. (1988). Estimated risk of lung cancer from exposure to radon decay products in U.S. homes: A brief review. Atmos. Environ. 22, 2205–2211. Porstendörfer, J. (2001). Physical parameters and dose factors of the radon and thoron decay products. Radiat. Prot. Dosim. 94 (4), 365–373. Porstendörfer, J., Mercer, T.T. (1978). Influence of nuclei concentration and humidity upon the attachment rate of atoms in the atmosphere. Atmos. Environ. 12, 2223–2228. Porstendörfer, J., Reineking, A. (1999). Radon: Characteristics in air and dose conversion factor. Health Phys. 76 (3), 300–305. Porstendörfer, J., Reineking, A. (2000). Radon characteristics related to dose for different places of the human. In: Proceedings of 10th IRPA Congress, Hiroshima, Japan, May 14–19, 2000. Porstendörfer, J., Dankelmann, V., Reineking, A. (1998). Neutralization of 218 Po-cluster in air. J. Aerosol Sci. 29 (Suppl. 1), S1017–S1018. Porstendörfer, J., Zock, C., Reineking, A. (2000). Aerosol size distribution of the radon progeny in outdoor air. J. Environ. Radioact. 51, 37–48. Reineking, A., Butterweck, G., Kesten, J., Porstendörfer, J. (1992). Thoron gas concentration and aerosol characteristics of thoron decay products. Radiat. Prot. Dosim. 45, 353–356. Schery, S.D. (1985). Measurements of airborne 212 Pb and 220 Rn at varied indoor locations within the United States. Health Phys. 49, 1061–1067. Schiller, C.F., Gebhart, J., Heyder, J., Rudolf, G., Stahlhofen, W. (1988). Deposition of monodisperse insoluble aerosol particles in the 0.005 to 0.2 µm size range within the human respiratory tract. Ann. Occupational Hyg. 32, 41–49. Schlesinger, R.B., Bohning, D.E., Chan, T.L., Lippmann, M. (1977). Particle deposition in a hollow cast of the human tracheobronchial tree. J. Aerosol Sci. 8, 429–441. Stahlhofen, W., Gebhart, J., Heyder, J. (1980). Experimental determination of regional deposition of aerosol particles in the human respiratory tract. Am. Ind. Hyg. J. 41, 385–398. Stahlhofen, W., Gebhart, J., Rudolf, G., Scheuck, G. (1986). Measurement of lung clearance with pulses of radioactively-labelled aerosols. J. Aerosol Sci. 17, 333–336. Task Group of Lung Dynamics, TASK (1966). Deposition and retention models for internal dosimetry of the human respiratory tract. Health Phys. 12, 173–207. Tu, K.W., Knutson, E.O. (1984). Total deposition of ultrafine hydrophobic and hygroscopic aerosols in the human respiratory tract. Aerosol Sci. Technol. 3, 453–465. United Nations Scientific Committee on the Effects of Atomic Radiation, UNSCEAR (1977). Sources and Effects of Ionizing Radiation. United Nations, New York. Watson, S.B., Ford, M.R. (1980). A User’s Manual to the ICRP Code: A Series of Computer Programs to Perform Dosimetric Calculations for the ICRP Committee 2. Report ORNL/TM-6980, February 1980.

112

C. Papastefanou

Yu, C.P., Diu, C.K. (1982). A comparative study of aerosol deposition in different lung models. Am. Ind. Hyg. J. 43, 54–65. Yu, C.P., Diu, C.K., Soong, T.T. (1981). Statistical analysis of aerosol deposition in the nose and mouth. Am. Ind. Hyg. Assoc. J. 42, 726–733. Zock, C. (1996). Die Messung der Aktivitatsgrossenverteilung des Radioactiven Aerosols der Radonzerfallsprodukte und deren Einflub auf die Strahlendosis beim Menschen. Dissertation, Georg-August-University, Gottingen, Germany (in German). Zock, C., Porstendörfer, J., Reineking, A. (1996). The influence of biological and aerosol parameters of inhaled short-lived radon decay products on human lung dose. Radiat. Prot. Dosim. 63, 197–206.

113

Chapter 6

Aerosol sampling and measurement techniques

1. Introduction The great diversity of application, the size range of atmospheric aerosol particles, the physical and chemical concentration variations, and the variety of measurement principles available imply many different combinations of application and measurement methods and procedures. Therefore, this chapter is focused on the most important methods in use. The methods applied for atmospheric aerosol sampling include filters and cascade impactors which collect the aerosol particles onto a surface. The collected sample must therefore be evaluated for size and composition. Because accumulation mode aerosols (fine aerosol particles) contain a substantial fraction of liquid material at normal temperatures and humidities, these fine aerosol particles must be sized in situ without precipitation. In some extreme cases, such as in Los Angeles smog, the liquid content may be as high as 75% or 80% of the total mass (Ho et al., 1974). Many aerosol-measuring techniques, such as the condensation nuclei counter, CNC, and filter collectors, integrate some response function of the instrument with the particle size distribution. In the particle-counting method, the response is essentially equal to the integral of the number distribution weighting. By weighing the filter, then the integrating measurement is the integral of the mass weighting of the distribution. Cascade impactors, such as Lundgren impactors, designed with inlet cutoffs well above those used for atmospheric aerosol sampling, showed that particles up to several hundred micrometres in diameter were present in the atmosphere under certain conditions (Lundgren, 1973). Most mass sampling methods, including the high-volume sampler (Stevens et al., 1976), truncate the distribution, thereby giving concentrations less than those actually existing. Due to their simplicity of construction and use and the relatively sharp cut-off characteristics, cascade impactors have been widely used for the size classification and size-classified chemical analysis of aerosols. Table 6.1 lists the most important integrating sampling methods with their main characteristics. Table 6.2 gives the most important differential, size-resolving methods used to sample and measure atmospheric aerosol particles. The section of the particle size distribution and the modes that dominate the sensitivity of the methods are indicated. The upper and lower size limits are nominal values for the most commonly used forms of the techniques. Cost, complexity of operational requirements, calibration problems, and the demands of the particular evaluation to be used also affect the choice of methods. For example, chemical analysis usually requires that a sample be collected, then taken to the evaluation device. RADIOACTIVITY IN THE ENVIRONMENT VOLUME 12 ISSN 1569-4860/DOI: 10.1016/S1569-4860(07)12006-4

© 2008 Elsevier B.V. All rights reserved.

114

Table 6.1 Major integral sampling and measurement methods used for measuring atmospheric aerosols Method or instrument

Effective size limit (µm) Upper

Lower concentration limit

Measurement principle

None

25

2 µg/m3

Filter-weighing

0.15

2

Light-scattering

0.0035

0.05–0.2

1% of air-scattering (about 1 µg/m3 ) 1 cm3

0.007–0.02

2–25

1 µg/m3

0.01

1

100 cm3

Charging and precipitation Condensation—low supersaturation

None

Unknown

10−5 /L

Ice nucleation

0.01 electrostatic; 0.3 impactor

10; 20

1 µg/m3

Frequency charge of crystal

None (on filter)

20

2–5 µg/m3

β attenuation

10

1000

100 µg/m3

5

None

None

Impact and charge measurement Settling and weighing

5

100

1/L

Condensation—high supersaturation

Acoustic pulse from particle transit

C. Papastefanou

Hi-vol filter and other filters—measures all of the nonvolatile mass in accumulation mode and most of that in the coarse particle mode depending on sampling efficiency Integrating nephelometer—essentially responsive to the in situ mass in the accumulation mode only Condensation or Aitken nuclei counter (ANC)—for combustion-dominated aerosols, measures the nuclei mode; for aged aerosols, measures number in accumulation mode Electrostatic charger—with diffusion charging upper limit is about 2 µm; senses both nuclei and accumulation modes Cloud condensation nuclei counter (CNC)—lower size limit depends on supersaturation and design; measures part of nuclei mode and most of accumulation mode Ice nuclei counter (INC)—measures only those few particles capable of causing supercooled water drops to freeze; responsive to part of the accumulation mode sizes and most of the coarse particle mode sizes Quartz crystal—electrostatic version measures nuclei and accumulation modes and part of coarse particle mode; impaction version measures two-thirds of accumulation mode and most of the coarse particle mode β attenuation—performance dependent on counting period, collection surface mass Contact electrification—useful only for high concentrations; sensitivity varies over 50/L range for different materials Dust fall—relation to concentration is very dependent on location, wind velocity, and collector design Acoustic counter—has been used primarily to count ice crystals as a sensor on ice nuclei counter

Lower

Table 6.2 Major size-resolving methods used for measuring atmospheric aerosols Method or instrument

Effective size limit (µm) Upper

Maximum number of size intervals practical

Lower concentration limit (µm)

Measurement principle

Cascade impactors—measures one-half to two-thirds of accumulation mode, all of coarse particle mode; particle bounce is a serious problem in some applications Dichotomous virtual impactors—one filter collects fine particles and one collects coarse particles; no bounce problem

0 (filter) 0.3 (last stage)

30

8

1 µg/m3 /stage

Inertial impaction

0 (filter) 2 (for cut on second filter)

30

2

1 µg/m3 /stage

Virtual impaction

Crystal microbalance impactor— measures one-half of nuclei mode, all of accumulation mode and coarse particle mode; subject to particle bounce on dry aerosols Spiral centrifuge—measures one-half of nuclei mode, all of accumulation mode, and part of coarse particle mode; sensitivity depends on method of evaluation; mass determination is difficult

0 (filter) 0.06 (last stage)

30

10

0.02 µg/stage 1 µg/m3 /stage

Impaction, quartz crystal sensing

0.06

10

20

10 µg/m3

Centrifuge onto filter

Optical particle counters—single counter useful for only I decade of size; requires calibration; sensitive to shape and refractive index

0.3 (commercial) 0.1 (research)

3–100

>200 possible 20 useful

1 particle/L

Light-scattering from single particles

Electric mobility—can measure nuclei and accumulation modes; can resolve σg of 1.4 without correction and 1.25σg with correction

0.006

1

15

1 µg/m3

Diffusion charging and electric mobility measurement

Aerosol sampling and measurement techniques

Lower

115

116

C. Papastefanou

This normally requires a collection-type method, such as the high-volume filter collectors. Simplicity during sampling may mandate the use of a particular collection method despite its other problems. The rapid development of modern electronics and inexpensive acquisition systems has led to in situ, automatic, and complex monitoring techniques. Collection methods are being replaced with both physical and chemical analysis in situ, automatic methods and digital electronic devices.

2. Cascade impactors Cascade impactors are instruments which have been extensively used for sampling and separating airborne particles to determine the aerodynamic size classification of aerosol particles. There are three kinds of cascade impactors: inertial impactors, virtual impactors and particle trap impactors. Inertial impactors have been used extensively to measure aerosol size distribution and to collect samples for further chemical analysis. They consist of an acceleration nozzle and a flat plate. In inertial impactors, particles larger than the cutoff diameter of the impactor will impact on the collection plate, while particles smaller than the cutoff diameter will follow the streamlines and not be collected on the plate. Many researchers have investigated inertial impactors to collect and separate aerosol particles and found inherent problems which were particle bounce and re-entrainment (Dzubay and Rickel, 1978; Markowski, 1984; Marple, 1970; Marple and Liu, 1974). Virtual impactors have been designed to eliminate problems of bounce and re-entrainment and to allow the collection of larger particles (Chen et al., 1986; Marple and Chein, 1980). In a virtual impactor, the intake air is typically divided into two air streams: the major and the minor flows. The major flow carries most of the fine particles smaller than the cutoff diameter and the minor flow carries all the coarse particles above the cutoff diameter with a small fraction of the fine particles. Various studies on virtual impactors have been carried out for particle collection and concentration since the late 1980s. Meyer and Lee (1994) used a virtual impactor as a means of concentrating metal aerosol suspended in the air to enhance the performance of an inductively coupled plasma analyser. Kim et al. (2000) designed and tested multi-nozzle virtual impactors for high-volume particle collection. Virtual impactors, however, have an inherent problem that a small fraction of the fine particles is collected with the large particles. A particle trap impactor (cup impactor) has been designed by Biswas and Flagan (1988), who intended to collect particles in the trap rather than to sample aerosol particles. Tsai and Cheng (1995) investigated the effect of solid particle loading on the collection efficiency of an impactor with different impaction surface designs. The particle trap impactor was applied to a sampling inlet with a cutoff diameter of 10 or 2.5 µm (Kim et al., 1998, 2003). Major advantages of the particle trap impactors are their freedom from particle bounce-off and the problem of overload. The designs of particle trap impactors are based on those of inertial impactors and virtual impactors. The particle trap impactors are similar to inertial impactors in that they use inertial force and do not have a minor flow rate. Also, the particle trap impactors are somewhat similar to virtual impactors in that they have virtual space to collect the particles.

Aerosol sampling and measurement techniques

117

Fig. 6.1. A schematic of an aerosol cascade impactor with trajectories shown for aerosol particles of three different sizes.

Even though inertial and virtual impactors have been widely used for aerosol collection and separation, particle trap impactors have not been used extensively. So far, studies on particle trap impactors were not carried out sufficiently. In addition, most particle trap impactors employ the design of inertial impactors for their design parameters, though actually the design of particle trap impactors is different from that of inertial impactors. Therefore, there still remains the necessity for study of the accurate design parameters for particle trap impactors. 3. Inertial impactors The basic configuration shown in Figure 6.1 (Mercer et al., 1970; Andersen, 1976; Marple and Willeke, 1979; Raabe, 1979) is simple. A jet of aerosol from a nozzle is directed at an

118

C. Papastefanou

Fig. 6.2. Efficiency curve for a typical impactor.

impaction surface: the high inertia, large particles cross the air streamlines and strike the impaction plate, while the smaller particles with less inertia are not collected but remain airborne and pass onto a second nozzle-impactor plate arrangement. The second nozzle is of a smaller diameter than the first nozzle, which increases the velocity and inertia of all the remaining aerosol particles. The larger of the previously uncollected particles thus have sufficient inertia to be collected by the second stage. This process continues as the aerosol stream enters successive stages of the cascade impactor. The final element of a cascade impactor is usually a filter which collects all particles too small to be collected by the last impaction stage. The nozzles of each stage can be of any shape, but are normally rectangular or round. The nozzles of each stage of the Andersen cascade impactors are round. Many cascade impactors have several nozzles of an identical size at a given stage with the aerosol jets from the nozzles impinging upon a common impaction surface. The most important characteristic of an impactor stage is the collection efficiency curve, which gives the fraction of particles of a given size collected from the incident aerosol stream as a function of particle size. Ideally, an impactor would collect all particles larger than a certain size and none of the smaller ones, which would correspond to the sharp ideal cutoff illustrated in Figure 6.2. The efficiency curve of a typical real impactor, also shown in Figure 6.2, spans a range of particle sizes but still has a good sharpness of cut. It has been shown (Marple and Liu, 1975) that ideal impactor behaviour can be obtained if two requirements are met: (1) in the region between the jet exit plane and the impaction plate, the Y -component of fluid velocity is a function of Y only (Y is parallel to the centre line of the nozzle, Figure 6.3), and (2) the Y -component of the velocity of the particles at the jet exit plane is uniform across the jet. These two requirements are reduced to only the first requirement, if the velocity of the particles is equal to the fluid velocity at the exit plane of the jet. Such requirements are nearly met in a real impactor, as can be seen by the velocity profiles in Figure 6.3 for a rectangular impactor. It can be seen that the velocity component perpendicular to the jet exit plane, Vy , is independent of X over a major portion of the jet, where X is the distance from the jet centre line. However, this criterion is not met in the boundary layer near the wall. Particles passing through this region are the cause of the non-ideal cutoff characteristics at the larger efficiencies of the real impactor performance curve shown in Figure 6.2. A schematic diagram of the Andersen cascade impactor stages is shown in Figure 6.4 (Marple and Willeke, 1979). This figure shows how impaction occurs at the orifice–collector interfaces. The use of an impactor with a single circular orifice for each jet provides a convenient method for determining the

Aerosol sampling and measurement techniques

Fig. 6.3. Flow fields in a rectangular impactor (S/W = 1, T /W = 1, Re = 3000).

Fig. 6.4. A diagram of an aerosol cascade impactor stage.

119

120

C. Papastefanou

aerodynamic size distribution of an easily assayed aerosol such as one of radioactive material (Raabe, 1979). Submicrometre aerosols, or condensation nuclei, are most readily measured for number concentration and size distribution by diffusion batteries and condensation nuclei counters, CNC (Sinclair et al., 1979). The impaction characteristics of the cascade impactors are governed by the fluid’s velocity flow field, which in turn is specified by the Navier–Stokes equations. A solution of the Navier– Stokes equations (Marple et al., 1974) reveals that the velocity flow field is a function of the physical configuration of the impactor and the Reynolds number, Re, of the flow passing through the nozzle. The Reynolds number, based on the hydraulic diameter of the nozzle, is defined as Re =

ρVo W μ

(6.1)

for round cascade impactors, and Re =

ρVo 2W μ

(6.2)

for rectangular cascade impactors, where ρ is the fluid density, Vo is the mean velocity at the nozzle throat, W is the nozzle diameter or width and μ is the fluid viscosity. Furthermore, the trajectory of the aerosol particles in the flow field, which determines whether or not the particles are collected, is found by solving the particles’ equation of motion (Marple and Liu, 1974), the solution of which is a function of the particles’ Stokes number and of the parameters which characterise the flow field. The Stokes number is defined as the ratio of the aerodynamic particle stopping distance to the radius, or half-width, of the impactor throat: Stk =

ρp CVo Dp2 /18μ W/2

,

(6.3)

where ρp is the particle density, C is the Cunningham slip correction factor and Dp is the particle diameter. The square root of the Stokes number, (Stk)1/2 , is a dimensionless particle size. The collection efficiency of a cascade impactor is, therefore, a function of the Stokes number, the Reynolds number, and the physical configuration of the impactor. It has been found that the physical configuration for either a round or rectangular nozzle can be sufficiently specified by the value of the nozzle’s diameter or width, W , the nozzle-to-plate distance, S, and the nozzle throat length, T (Marple, 1970). Since the Stokes number and the Reynolds number are dimensionless, it is convenient to express the physical dimensions in dimensionless form by expressing all dimensions in units of the jet width, such as S/W and T /W . The inertial cascade impactors are multi-stage cascade impactors. The Mercer-style cascade impactor is a seven-stage cascade impactor operating at low flow rate, e.g. 50– 150 cm3 min−1 (Mercer et al., 1970; Raabe, 1979). Other inertial impactors are the Andersen cascade impactors, such as the 1 ACFM eight-stage cascade impactor operating at a flow rate of 28 L min−1 which is described in the next section (Andersen, 1976), the Lundgren-type eight stage cascade impactor operating at a flow rate of 1 L min−1 , with the samples at stages collected on Teflon, and the final stage being quartz (Lundgren, 1967; Marley et al., 2000), and the Berner-type nine-stage low-pressure cascade impactor with a

Aerosol sampling and measurement techniques

121

nominal volumetric flow rate of 28.9 L min−1 (1 ACFM), with the samples at stages collected on aluminium foil (Berner and Lurzer, 1980; Winkler et al., 1998). In this category also is classified the dichotomous sampler which is used for air pollution studies and utilises a virtual impactor with two emerging aerosol flows to separate the coarse and fine fractions of atmospheric aerosol particles and operates at 1 m3 h−1 (16 L min−1 ) (Hinds, 1999). The aerosols are collected on impaction substrates stamped from Teflon, aluminium foils, polycarbonate membranes from all stages of the cascade impactors of any type and the smallest fraction on the backup glass fibre filter. After the end of collection, the collecting substrates are carefully removed from the impactor and are reweighed for evaluation of the size distribution, i.e. the mass or volume distribution of the particles. Then the substrates are folded and compressed to provide suitable geometry for measurement of the activities of the radionuclides associated with the aerosol particles using a high resolution and high efficiency low-background gamma-ray spectrometric system incorporating a high purity Ge detector for the determination of the activity size distribution of the radioactive aerosols.

4. Andersen cascade impactors The design concept of the Andersen cascade impactors (Andersen Samplers, Inc., 4214 Wendell Drive, Atlanta, GA 30336) evolved from the following facts: The human respiratory tract is an aerodynamic classifying system for airborne particles (Hamilton and Walton, 1952; Hatch, 1959). The sampling device is used as a substitute for the respiratory tract as a particle collector. As such, it should reproduce to a reasonable degree the particle-collecting characteristics of the human respiratory system (Hamilton and Walton, 1952; Hatch, 1955) so that lung penetration by airborne particles can be predicted from the sampling data. The sampling instrument should, therefore, classify the aerosol particles collected according to the aerodynamic dimension, which as Wells (1955) states, is the true measure of lung penetrability. The fraction of inhaled particles retained in the respiratory system and the site of deposition varies with size, shape and density and all the physical dimensions (Brown et al., 1950; Hamilton, 1952; Harber and Morton, 1953; Hatch, 1959) (Figure 6.5). Methods that employ light-scattering or filtration and microscopic sizing of aerosol particles do not reckon with density and some other properties that affect the movement of the particles in air, and therefore, do not give the desired information (Hamilton and Walton, 1952). Because the lung penetrability of unit density spherical particles is known (Brown et al., 1950; Morrow, 1964) and since the particle sizes that are collected on each stage of the sampler are used according to the standard operating procedure, the stage distribution of the collected material will indicate the extent to which the sample would have penetrated the respiratory system (Andersen, 1958). With this information and with knowledge of the chemical, biological, and/or radiological properties of the material collected, the exact nature and extent of the health hazard can be assessed (Andersen, 1966). The earliest and most fundamental work in inertial impaction theory was conducted in the early 1950’s (Ranz and Wong, 1952). In this work, Ranz and Wong (1952) showed that the collection of a particle by an obstacle is a function of what is defined as the inertial impaction

122

C. Papastefanou

Fig. 6.5. Andersen aerosol cascade impactor simulates human respiratory tract.

parameter: K=

CρU Dp2 18μDc

,

(6.4)

where U is the relative velocity, ρ is the particle density, Dp is the particle diameter, μ is the gas viscosity, Dc is the diameter of the round jet and C is the Cunningham slip correction factor. Data from inertial impactors are normally presented as 50% effective cutoff diameters. For the Andersen cascade impactors containing round jets and flat collection surfaces, the 50% effective cutoff diameter would yield a value of 0.14 for the inertial impaction parameter K. The Cunningham slip correction factor is equal to C = 1 + 0.16 × 10−4 /Dp

(6.5)

for normal temperatures and pressures. This factor corrects for the fact that, as particle diameters approach the mean free path length of the gas molecules, they tend to “slip” between gas molecules more easily and are therefore more easily able to cross the bulk flow stream lines. The collection efficiency is therefore slightly greater than would be predicted by inertial impaction theory for particle diameters of the order of 1 or 2 µm. The overlapping of particle size between stages, which is naturally inherent in all cascade impaction devices, is minimised in Andersen’s cascade impactors by design. Ranz and Wong (1952) stated that, as a particle passes through a jet, its nearness to the axis of the jet is one of the factors that determines whether or not the particle will reach the impaction surface. In contrast to competitive cascade impactors which have larger rectangular jets in each stage, the Andersen cascade impactors

Aerosol sampling and measurement techniques

123

have 400 small round jets. Travel of the particle is thus confined near the axis of the jets. The average distance of the particles from the axis of the jets is very much less than in other cascade impactors. Ranz and Wong (1952) also stated that round jets have sharper cutoffs than rectangular jets. The Andersen cascade impactors, therefore, on a theoretical basis, should have a sharper cutoff. Another inherent advantage of the Andersen cascade impactors over other cascade impactors is that single circular orifice and multiple rectangular orifice cascade impactors by design must operate with high orifice velocities. This results in more turbulent flow, greater re-entrainment, and a skewing of the size distribution towards the lower end (i.e. the indicated size distribution being smaller than it really is). The Andersen cascade impactors include: (1) the 1 ACFM ambient cascade impactor, (2) the low pressure cascade impactor (LPI), and (3) the high volume cascade impactor (HVI). 4.1. 1 ACFM ambient cascade impactor The 1 ACFM (CFM = cubic feet per minute) is operated at 28 L min−1 (1 ft3 min−1 ) constant flow rate ensured by a continuous duty vacuum pump and is comprised of eight aluminium stages (Figure 6.6). The stages have effective cutoff diameters (ECDs) of 0.4, 0.7, 1.1, 2.1, 3.3, 4.7, 7.0 and 11.0 µm, and are numbered 0, 1, 2, 3, 4, 5, 6, 7 and F. Stage 0 is an orifice stage only. Stage F contains the collection plate for stage 7 and the backup filter. Each impactor stage contains multiple precision-drilled orifices. When air is drawn through the impactor, multiple jets of air in each stage direct any airborne aerosol particles towards the surface of the collection plate for that stage. The size of the jets is constant for each stage, but is smaller in each succeeding stage. Whether an aerosol particle is impacted on any given stage depends on its aerodynamic dimension. The range of aerosol particle sizes collected on each stage depends on the jet velocity of the stage and cutoff of the previous stage. Any particle not collected on the first stage follows the air stream around the edge or passes on to the succeeding stage, and so on until the jet velocity is sufficient for impaction. Stages 0–6 have integral air inlet sections that contain 400 orifices. Stage 7 contains 201 orifices. The inlet sections are approximately 3.125 inches in diameter. The orifices are progressively smaller from top to bottom stages, ranging from 0.0625 inch diameter in stage 0 to 0.0100 inch diameter in stage 7. Each stage has a removable stainless steel or glass (3.25 inches diameter) collection plate. The exhaust section of each stage is approximately 0.75 inch larger in diameter than the collection plate, allowing unimpacted particles to go around the plate and into the next stage (Figure 6.7). The cascade impactor assembly begins by placing stage F on the base plate stage which is bolted to the base plate. A pre-weighed glass fibre backup filter is inserted into stage F and the large diameter O-ring placed into position around the periphery of the filter. Next, preweighed collection plate 7 is placed onto stage 7, so that the plate rests on the three raised, notched metal seats to prevent plate movement. The stainless steel plates are placed with the curved lip down so that a raised smooth surface is exposed for particle impingement. This is followed by stage 7, collection plate 6, stage 6, and so on until the inlet cone is positioned last. Special collection substrates other than stainless steel plates, e.g. glass plates, can be used

124

C. Papastefanou

Fig. 6.6. Andersen 1-ACFM aerosol cascade impactor.

because of their lighter tare weight and/or specific analytical requirements. These collection materials consist of glass fibre, cellulose, aluminium foil, vinyl metricel and other materials which must be used within the inverted stainless steel collection plate. The length of each collection period varies from 3 h for 214 Pb-aerosols, to 30–40 h for 212 Pb-aerosols and 1–14 d for 7 Be-, 210 Pb-, 35 S- and 32 P-aerosols. 4.2. Low-pressure cascade impactor The Andersen low-pressure cascade impactor (LPI) is an adaptor assembly which is used with stages 0 through 7 of an Andersen 1 ACFM ambient cascade impactor (Section 4.1) to extend the range of particle classification below 0.4 µm. Standard aerosol cascade impactors are nor-

Aerosol sampling and measurement techniques

125

Fig. 6.7. A schematic cross-section of the non-viable impaction stages. Progressively smaller orifices increase the orifice velocity in eight successive stages causing impaction of smaller aerosol particles onto the collection discs of each succeeding stage.

mally used to classify aerosol particles down to approximately 0.4 µm aerodynamic diameter. However, many aerosols such as cigarette smoke and diesel exhaust are predominantly sub 0.4 µm in diameter. The 50% effective cutpoint of a cascade impactor is given by 9μStk50 Dj 1/2 Dp,50 = (6.6) , ρp CVj where μ Stk50 Dj ρp C Vj

is the gas viscosity, in poise, is the Stokes number, dimensionless, Equation (6.3), is the jet diameter, in cm, is the particle density, in g cm−3 , is the Cunningham slip correction factor, Equation (6.5), is the jet velocity, in cm s−1 .

For any given geometry, the gas viscosity, Stokes number and particle density are fixed. Therefore, to get a smaller cutpoint the diameter of the jet must be reduced, the velocity through the jet increased or the Cunningham slip correction factor increased. Reducing the jet diameter to less than 0.0254 cm is difficult because the jet-to-collection surface spacing should be less than 10 jet diameters. Increasing the jet velocity causes unacceptable levels of particle bouncing. That leaves the Cunningham slip correction factor as the best variable. The Cunningham slip correction factor is a measure of how easily an aerosol particle can slip in between gas molecules when crossing the bulk flow stream to reach the impaction surface. Reducing the

126

C. Papastefanou

Fig. 6.8. A schematic of a low-pressure cascade impactor, LPI complete system.

gas pressure will increase the Cunningham slip correction factor and thereby allow collection of smaller particles by a given impactor geometry. The Andersen low-pressure cascade impactor is capable of pulling 3 L min−1 (regulated flow rate) at 27 inches Hg pressure drop and includes five low-pressure stages for the submicron region and eight atmospheric-pressure stages for separating aerosol particles above 1.4 µm (Figure 6.8). The complete assembly consists of a low-pressure filter stage, LF, the low-pressure cascade impactor stages L5, L4, L3, L2 and L1, a critical orifice stage with brass pressure tap, atmospheric pressure stages 0, 1, 2, 3, 4, 5, 6, 7, inlet cone, 7 collection plates, 7 collection plates with open centres, a vacuum pump and a pressure-measuring device which is an absolute pressure gauge or a 16 inches–0–16 inches U-tube mercury manometer in conjunction with a mercury barometer to determine the absolute pressure downstream of the critical orifice. A bleed valve is included and adjusted to obtain an absolute pressure at stage L1 of 114 mm Hg (0.15 atm or 4.49 inches Hg). At least 400 mm Hg (16 inches Hg) pressure drop across the critical orifice is needed in order to maintain the critical flow of 3 L min−1 . This means that critical flow must be maintained at elevations where the atmospheric pressure is greater than approximately 525 mm Hg. The effective cutoff diameters (ECDs) of the five low-pressure stages are 0.08, 0.11, 0.23, 0.52 and 0.90 µm, whereas for the eight upper stages, the ECDs are 1.4, 2.0, 3.3, 6.6, 10.5, 15.7, 21.7 and 35.0 µm. Either polycarbonate or glass-fibre backup filters are used to collect all aerosol particles below the 0.08-µm collection plate. The length of each collection period varies from 3 h for 214 Pb-aerosols to 30–40 h for 212 Pb-aerosols. The amount of air pulled with the low-pressure cascade impactor is usually less than 10 m3 .

Aerosol sampling and measurement techniques

127

Fig. 6.9. A block diagram of a four-stage low-pressure aerosol cascade impactor.

4.3. Four-stage low-pressure cascade impactor The four-stage low-pressure cascade impactor incorporates both the impactor and wire screen methods (Tokonami et al., 1997). This system can measure the activity size distribution of radon decay products in a low level environment within 90 min. Figure 6.9 shows a block diagram of the activity-weighted size distribution instrument. In the first air inlet, unattached radon decay products are collected on a metal wire screen (300 mesh; openings 118.2 cm−1 ; wire diameter 3.75 × 10−3 cm). A silicon semiconductor detector, SSD, is set opposite the metal wire screens where both collection and detection are concurrent. Output signals from the silicon semiconductor detector are sent through a preamplifier, PA, and the internal amplifier of a multichannel analyser, MCA, and then to the multichannel analyser. The detection efficiency for unattached fractions is obtained from the product of the geometric efficiency (11.6%), the collection efficiency on the screen (86.1%) with a face velocity 37.1 cm s−1 (Thomas and Hinchliffe, 1972; Cheng et al., 1980) and the counting efficiency for unattached fractions on the up-stream side of the screen (80.1%). The 50% cutoff diameter

128

C. Papastefanou

of the screen is 4.4 nm. As all unattached fractions are not collected on the screen, a second screen holder (effective diameter of the screen is 40 nm) is connected in series with the first holder to prevent the invasion of unattached fractions into the cascade impactor. As the cutoff properties may change, the flow rate must be set at 7 L min−1 through the nozzle of the impactor when this system is in operation. Attached fractions are introduced into the impactor after removing unattached fractions completely. A microanalysis particle sampler is used for the activity size distribution measurements. The cascade impactor consists of four stages. At a particle density of 1 g cm−3 , the particle size cut-points are >2100, 700, 210 and 70 nm, respectively. A silicon photodiode (sensitive area is 10 mm × 10 mm; bias voltage is 5 V) is installed in each stage. The ceramic window is removed from the detecting surface of the photodiode. As radon decay products are directly collected on the surface, alpha particles are effectively detected. Each detecting surface is uniformly coated with a thin layer of silicon grease to prevent particle bounce-off. For alpha particles the detection efficiencies for attached fractions are 45.3% and 46.0% for 218 Po and 214 Po, respectively, taking into consideration the thickness of the dead layer in the photodiode. As no alpha particles are detected in the first stage, the photodiode is detached from the first stage. An inline type filter holder with a silicon semiconductor detector is used for the collection and detection of alpha particles below 70 nm. A membrane filter is used as a collection filter. As shown in Figure 6.9, the output signals from each detector are amplified by a pre-amplifier, PA, a linear amplifier, LA, and then analysed by a multichannel analyser, MCA, through a mixer/router. In order to determine radon decay product concentrations, the build-up and decay methods are used (Cliff, 1978; Trembley et al., 1979). The following timetable is used: the collection period is 28 min, the first waiting period is 2 min and there are four time-measurement periods of 10 min with a 2-min interval for the readout between them. Using the counting data from the resulting measurements, radon decay product concentrations are estimated by the least square method. 4.4. High volume cascade impactor The high volume cascade impactor (HVI) is a multi-stage cascade impactor which attaches to any standard high volume air sampler and fractionates aerosol particles into as many as six size fractions, nominally: 7.2 µm and greater (stage 1), 3.0 µm (stage 2), 1.5 µm (stage 3), 0.95 µm (stage 4), 0.49 µm (stage 5) and 0 to 0.49 µm (backup filter) with an air flow rate of 40 CFM (1.13 m3 min−1 ), and as seven fractions, nominally: 10.2 µm and greater (stage 1), 4.2 µm (stage 2), 2.1 µm (stage 3), 1.4 µm (stage 4), 0.73 µm (stage 5), 0.41 µm (stage 6) and 0 to 0.41 µm (backup filter) with an air flow rate of 20 CFM (566.32 L min−1 ). A winddirectional cyclone separator fits over the impactor which efficiently collects large aerosol particles greater than 5.5 µm for the 40 CFM high volume cascade impactor and greater than 11 µm for the 20 CFM high volume cascade impactor before they enter the impactor, eliminating any potential particle bounce problems in the impactor. The high volume cascade impactors are slotted impactors and feature the ultimate in precision, predictability, and sharpness of cut-off, combined with compactness, ease-of-use, and the lowest internal losses. Aerosol particles enter the high volume cascade impactor through the parallel slots in the first impactor stage, and particles larger than the particle cut-off size of the first stage im-

Aerosol sampling and measurement techniques

129

pact on the slotted collection substrate. The air stream then passes through the slots in the collection substrate, accelerates through narrower slots in the second impactor stage, and the remaining particles greater than the particle cut-off size of the second stage impact on the second collection substrate, and so on. The width of the slots is constant for each impactor stage, but is smaller for each succeeding stage, and most of the smaller particles eventually acquire sufficient momentum to impact on one of the collection substrates. After the last impactor stage, the remaining fine particles are collected by the backup filter in the high volume sampler. The collection substrates and backup filter are weighed before and after aerosol sampling to determine the particle size distribution and the activity size distribution. Since all aerosol particles are collected, the impactor also yields total particle mass concentration. 4.5. Activity size distribution of radioactive aerosol particles There are two methods that can be used to make activity-weighted size distribution measurements: screen diffusion batteries (Hopke et al., 1992; Cheng et al., 1994) and all-type cascade impactors. Although screen diffusion batteries have been widely used, the activity size distribution cannot be easily obtained with them due to technical and calculation complexities. Even when conventional equipment and the impactor method are used it is difficult to obtain a result that is not affected by the decay of the radioisotopes. The stainless steel collecting plates of the Andersen cascade impactors are leached after the end of the aerosol sampling with a 2-ml solution of 1 M (molar) HNO3 in the case of 214 Pb and 212 Pb. The leachate is rapidly evaporated on 5.08-cm (2-inch) diameter stainless steel plates using a hot plate, and the activity of the radioisotopes is measured by α-, β- and γ -counting systems. Less than 30 to 40 min are needed to prepare the samples for counting after sampling is stopped. Seven ZnS(Ag) α-scintillation counters, each with an effective counting diameter of 5.16 cm (2 inches), are used to determine the α-activity of the samples. The time interval between measurements are 2 to 5 min over the first 2 h after the stop-sampling time and 15 to 30 min for the next 3 to 4 h in order to separate decay rates controlled by the half-lives of 214 Pb (26.8 min) and 212 Pb (10.6 h), respectively. In the decay scheme of 222 Rn and 220 Rn, the α-particles from 218 Po and 216 Po are not registered in the measurements because they are very short-lived radioisotopes. For 218 Po, the fraction of the total α-decays on a collection surface that is at radioactive equilibrium with ambient air is much less than 5% (Evans, 1969). Furthermore, a 15-min minimum delay before the measurements results in almost complete (97%) decay of 218 Po. Therefore, the measured α-decays are from 214 Po and a combination of 212 Bi + 212 Po. Alphas from 210 Po and other radionuclides are negligible as might be confirmed by repeated examination of the plates for the presence of long-lived (>1 d) α-emitters that would bias the 212 Pb measurements. The counting-rate data must be examined after background correction to determine the contribution of 214 Pb and 212 Pb at the stop-sampling time. The calculated activities of 214 Pb and 212 Pb are then used to determine the activity size distribution of each radionuclide. For 210 Pb and 210 Bi, the stainless steel plates are leached with 300 cc of 2 N HCl. The solution is then filtered. The high leaching efficiency of 2 N HCl, better than 99%, can be confirmed by counting 212 Pb and 212 Bi in: (a) the solution after leaching and (b) the leached filter residue, using Ge γ -detectors. The filtrate is then passed through Dowex-1 anion ex-

130

C. Papastefanou

change resin to retain the Bi and Pb. The Pb is next eluted using 300 cc of 0.05 N HCl and finally the Bi using 300 cc of 1.8 M H2 SO4 . The 0.05 N HCl is treated with six drops of Bi carrier in 8 M NaOH solution and is then filtered through Millipore paper filters. The precipitate is placed in standard sample holders, sealed using thin plastic adhesive tape, and counted in a low background phoswitch scintillation detector having a background of 1–2 cpm for counting of the β-radiation. The phoswitch detector consists of a thin CaF(Eu) primary crystal with a decay time of 0.9 µs and a NaI(Tl) guard crystal with a decay time of 0.23 µs. The 210 Pb is estimated by precipitating Bi, or by allowing 210 Bi to grow to an equilibrium value in about six half-lives (i.e. 30 days) and precipitating the 210 Bi. In the latter case, 210 Pb is precipitated and the equilibrium value compared with values obtained by direct 210 Bi precipitation. In many cases, such parallel analyses have been carried out and good agreement observed between the two values. The purity of precipitation can be tested by following the decay (or in-growth) of radioactivity with a half-life of 5.0 days. This is necessary in the samples stored for 210 Bi to attain equilibrium with 210 Pb, in case the Bi precipitate contained a long-lived component, possibly 210 Pb, which was not eluted by the 0.05 N HCl wash. In most cases the bismuth precipitate decays with a half-life of about 5 days. In counting the samples, sufficient time has to be allowed for complete decay of the 212 Bi (T1/2 = 60.6 min) and the 212 Pb (T1/2 = 10.64 h) as they are present in much higher concentrations. Generally, a time of 10 h for 212 Bi and 5 days for the 212 Pb sample is adequate for the complete decay of 212 Bi and 212 Pb activities. Lead-212 is useful for checking the possible presence of Pb and Bi (Papastefanou and Bondietti, 1991). For 7 Be measurements with high volume cascade impactors (HVI), even with 1 ACFM cascade impactors, the plates are leached with 0.1 M HCl and the activity of 7 Be is measured with coaxial Ge γ -detectors through the gamma photons of energy 477 keV. For 35 S measurements with cascade impactors of all types, the 0.1 M HCl leachate solution from the plates is purified using an alumina column (Veljkovic and Milenkovic, 1958). After eluting the sulfate from the alumina column with NH4 OH, it is converted to H2 SO4 using a small Dowex-50 (H+ ) column. The resulting solution of dilute sulfuric acid is concentrated and the β-activity (Eβ − , max = 167.5 keV) is determined using liquid scintillation techniques. The purity of the 35 S is checked by following the decay rate of the isolated β-activity. For 32 P measurements with cascade impactors of all types, the radionuclide is isolated from the 0.1 M HCl leachate solution from the plates by precipitation of zirconium phosphate (Mullins and Leddicotte, 1962) and the β-activity is determined using liquid scintillation techniques. The purity of the 32 P is checked by following the decay rate of the isolated β-activity.

5. Online α-impactor The online cascade impactor is a low-pressure cascade impactor (Section 4.2) with pi = 220 mbar in the lowest stage (Kesten et al., 1993). The impactor consists of nine stages on which the aerosol particles are deposited (Figures 6.10a and 6.10b; SARAD GmbH, Wiesbadener Str. 20, 01159 Dresden, Germany). Each stage consists of a jet plate, a rotatable foil holder with a thin foil (4 µm) serving as impaction plate, and a passivated-implanted planar silicon (PIPS) detector with an active area of 2000 mm2 (Figure 6.11). The central circular area (corresponding to the detector area) of each holder was eroded, the remaining material

Aerosol sampling and measurement techniques

131

(a) Fig. 6.10. (a) Gottingen SARAD online α-impactor. (b) A sketch of an online α-impactor with electronics.

forming a network which supports the foil. The distance S between the jet plate and the impaction plate of each stage was chosen as three times the diameter W of the jets, i.e. S/W = 3, and the relation between length T and diameter W of the jets, T /W , ranges from 2 to 3. To reduce energy loss of the α-particle penetrating the aerosol layer and the foils, the jets in each jet plate are arranged not on one distinct radius, but on different cycloids and the foil holders can be rotated by an electric motor at one revolution per 50 min to provide a homogeneous and thin aerosol layer on the foils. Furthermore, four backup filter holders, supplied

132

C. Papastefanou

(b) Fig. 6.10. (continued)

with cellulose nitrate membrane filters of 50-mm diameter and 8-µm pore size, are mounted in parallel below the last stage to minimise the pressure drop.The holders are supplied with PIPS detectors. With the detectors mounted to the stages and the filter holders, the α-decay of 218 Po (T1/2 = 3.05 m) and 214 Po (T1/2 = 164 µs) is measured online during air sampling. If after the end of air sampling the decay of 214 Po is measured, the 214 Pb (T1/2 = 26.8 m) activity distribution can also be evaluated. The signals of each detector are amplified by a preamplifier–amplifier combination and the spectra are collected with a 16k multichannel analyser, controlled by a PC, which starts and stops the α-spectroscopy and stores the spectra. The PC can also be used to start and stop the pump for air sampling automatically, thus allowing a continuous operation of 1–2 weeks depending on the aerosol concentration. For a longer operation time, the aerosol layer on the foils gets too thick and thus the tailing and, as a consequence, the uncertainty on the 218 Po counts becomes too large. As the main advantage of the online cascade impactor, the sampling time is not restricted by the decay of the radioactive isotope investigated. Therefore, the detection limit can be improved by increasing the sampling time. Furthermore, the online α-cascade impactor was constructed for a flow rate of 5.1 m3 h−1 , thus improving the detection limit again. The flow rate must be compared to the 1.8 m3 h−1 of the Berner 30 cascade impactor (Berner, 1978). Owing to the absorption of α-particles by the network of foil holders, the efficiencies of the detection of α-particles in the different stages range from only 13% to 16%. For a statistical error of 10% and a sampling time of 1 h, the detection limit is about 2 Bq m−3 for the shortlived radon decay products. By penetrating the aerosol layer, the foil and the air between foil and detector, the α-particles suffer energy losses, thus leading to a broadening of the peaks to lower energies, called tailing. The tailing is more pronounced for the upper stages, operating at normal air pressure (FWHM about 6%). In the lowest stage (p = 220 mbar), this value is

Aerosol sampling and measurement techniques

133

Fig. 6.11. A sketch of two single impactor stages of an online α-impactor.

reduced to 4%. Nevertheless, even for the upper stages, the energy resolution of the α-peaks is sufficient for the separation of 218 Po (Eα = 6.00 MeV) and 214 Po (Eα = 7.69 MeV) (Figure 6.12). As shown by Schumann et al. (1988), bounce-off effects do not play any role for particles smaller than 1 µm in diameter. For larger aerosol particles, the bounce-off effect can be minimised by coating the impaction plates or foils with oleic acid solutions. Additionally, the bounce-off problem is more severe for mass size distributions than for activity size distributions with median diameters between 200 and 600 µm. Nevertheless, the influence of blow-off effects on the activity size distribution for longer operation periods has still to be investigated. Aerosol distributions are usually described by log-normal distributions (Section 2 of Chapter 1). Therefore, the ratio of the 50% cutoff diameters of subsequent stages should be constant. Furthermore, the pressure below the last (lowest pressure) stage should be so low that this stage serves as a critical orifice. In this case, the pressure in the other stages, and as a consequence their 50% cutoff diameters, become independent of this pressure. With these conditions the number and diameter of the jets for each stage can be calculated iteratively.

134

C. Papastefanou

Fig. 6.12. The α-spectrum of 218 Po and 214 Po of an online α-impactor stage. The lined area represents the 214 Po counts below the 218 Po peak.

The internal impaction of aerosol particles with an aerodynamic diameter, dp , on a specific stage i can be described by the Stokes number St(dp ): St(dp ) =

C(dp )dp2 ρp vi 9μWi

,

(6.7)

where ρp is the particle density, vi is the velocity of air inside the jets, μ is the dynamic viscosity of air, Wi is the diameter of the jets, and C(dp ) is the Cunningham slip correction factor: 2λ(pi ) C(dp ) = 1 + (6.8) 1.257 + 0.4e−0.55dp /λ(pi ) , dp where λ(pi ) is the mean free path length of air at the pressure pi in the jets of stage i. At a certain value St0 (dp50 ) of the Stokes number, 50% of particles are impacted on the plate below the jets, thus defining the 50% cutoff diameter Dp50 : 1/2 9St0 μWi . Dp50i = (6.9) C(dp50i )ρp vi As an example, the activity size distributions of the aerosol-attached fraction of all shortlived radon decay products are shown in Figure 6.13. The errors (dashed lines) included the statistical errors of the counting rates and the uncertainties in the efficiencies of the detector of the on-line α-cascade impactor in the several stages. The fitted distribution and its parameters are also shown in Figure 6.13. The AMAD of 218 Po is about 10% smaller and σg is about

Aerosol sampling and measurement techniques

135

Fig. 6.13. Measured and fitted activity size distributions of the short-lived radon decay products obtained by an online α-impactor.

10% larger than the values for 214 Pb and 214 Po. As can be seen, a relative fraction of the activity (about 30%), deposited on the back-up filter, is not represented by the log-normal distribution. This activity is attached to aerosol particles smaller than 56 nm in diameter and larger than 1 nm in diameter. Reineking et al. (1988) has shown that particles smaller than 1 nm in diameter were deposited by diffusion processes in the impactor entrance and the walls of the first two stages after the entrance. To improve the size resolution of this fraction, diffusion screens below the last stage were used. To obtain a more precise lower cutoff for the smallest particles, a diffusion battery with 50% cutoff diameter of 4 nm was used in front of the entrance stage of the impactor.

6. A micro-orifice uniform deposit cascade impactor (MOUDI impactor) The MOUDI impactor has several features normally not found in other cascade impactors (Marple et al., 1991). These include collection of particles as small as 0.056 µm in aero-

136

C. Papastefanou

dynamic diameter with a moderate pressure drop, uniform particle collection on the impaction plates, interchangeable impaction plates to allow for the plates to be easily and quickly changed in the field, and covers for these substrates to provide a means of safe storage and/or transport of the collected particles to the laboratory for analysis. The primary parameter which governs the collection of particles in an inertial cascade impactor is the Stokes number, St, defined as St =

ρp CVo Dp2 9μW

,

(6.10)

where ρp is the particle density, C is the Cunningham slip correction factor, Vo is the average air velocity at the nozzle exit, and Vo = q/π(W/2)2 , Dp is the particle diameter, μ is the air viscosity, W is the nozzle diameter and q is the volumetric rate through the nozzle. The Stokes number is a dimensionless parameter that can be used to predict whether a particle will impact on an impaction plate of a stage or will follow the air steamlines out of the impaction region and remain airborne. It is common practice to use (St)1/2 as a dimensionless particle size and to define (St50 )1/2 corresponding to Dp50 , which is the value of the particle diameter, Dp , collected with 50% efficiency. Thus 9μW 1/2 Dp50 = (6.11) (St50 )1/2 . ρp CVo Inspection of the terms in Equation (6.11) reveals the operational or design alternatives that can be employed to obtain small values of Dp50 . For example, (St50 )1/2 , the particle density, ρp , and the air viscosity, μ, are constants and cannot be varied to decrease Dp50 . The average air velocity at the nozzle exit, Vo , cannot be increased beyond sonic speed in a converting nozzle, and to avoid particle bounce problems the jet should not even approach sonic conditions. The only remaining parameters in Equation (6.11) are the Cunningham slip correction factor, C, and the nozzle diameter, W , both of which can be adjusted to collect small particles. Small particles can be collected if the particle slip correction is increased by operating the impactor at low pressures as reviewed by Hering and Marple (1986). A pressure as low as 3 kPa (0.03 atm) was reported for the collection of 0.05-µm diameter particles. The other method for collecting small particles, and the method used in the MOUDI cascade impactor, is to employ very small nozzles as described by Kuhlmey et al. (1981). The small dimensions of the nozzle allow for small particles to be collected at relatively low jet velocities, and consequently low pressure drops. By the use of multiple micro-orifice nozzles in the MOUDI cascade impactor (2000 nozzles of 52 µm in diameter in the final stage), the cut-size of the final stage can be as low as 0.056 µm with a total flow rate through the impactor of 30 L min−1 . The uniform deposit feature of the MOUDI cascade impactor was designed to facilitate the elemental composition determination of the deposits by X-ray fluorescence analysis (Dzubay and Rickel, 1978). The uniform deposit is achieved by using multiple nozzles at each stage and rotating the impaction plate beneath the nozzles (Marple et al., 1981). By placing the nozzles at specific distances from the centre of rotation, a uniform deposit is obtained upon the rotating impaction plate. A schematic diagram of a stage of the MOUDI cascade impactor is shown in Figure 6.14. Each stage contains the impaction plate for the nozzles above and the nozzles for the im-

Aerosol sampling and measurement techniques

137

Fig. 6.14. A diagram of a stage of a micro-orifice uniform deposit impactor (MOUDI) showing its relation to stage above and stage below.

paction plate below. By rotating this stage relative to the stage above and the stage below, the impaction plate is rotated relative to the upper nozzle plate and the nozzle plate is rotated relative to the lower impaction plate. Rotation of alternative stages of the impactor, while the other stages are held stationary, results in every nozzle plate having rotation relative to its corresponding impaction plate. The usage of the word “stage” here is somewhat different than in the general description of the cascade impactor, where stage referred to a nozzle and its respective impaction plate. The MOUDI cascade impactor consists of two basic assemblies. One is the cascade impactor and the other is the rotator. Figure 6.15 shows the cascade impactor in the rotator (MSP Corporation, 5910 Rice Creek Parkway, Shoreview, MN 55126). Gears on alternative stages of the cascade impactor mesh with gears on a rotating shaft of the rotator. The hooks on the remaining stages mesh with the bearings on the shaft, restraining these stages from rotating. Although the primary purpose of the rotator is to provide rotation of the alternate stages of the impactor, the rotator also contains a control valve for controlling the flow through the MOUDI cascade impactor and two pressure gauges. The upper pressure gauge monitors the pressure drop across the first five stages to indicate the airflow through the impactor. The lower pressure gauge monitors the pressure drop across the lower stages. Because the lower stages use small nozzles, particle deposition on the nozzle walls can reduce the nozzle diameter, causing the pressure drop to increase. This indicates that the final stages of the impactor should be cleaned before further tests are conducted. The MOUDI cascade impactor consists of seven stages and the inlet which are set in the background, while the eighth stage and after-filter (in the base) are set in the foreground. Next to the eighth stage are the components of a replaceable impaction plate assembly. This consists of an impaction plate base, substrate clamping ring and substrate. The impaction plate

138

C. Papastefanou

Fig. 6.15. Micro-orifice uniform deposit impactor (MOUDI).

assembly is held on the body by the use of two four-pole magnets: one attached to the body and the other to the back of the impaction plate. Next to the two-piece base is a replaceable after-filter holder, consisting of a filter support and a filter clamp ring. Although the cascade impactor is considered an eight-stage cascade impactor, there can also be an impaction plate placed on top of the first stage. This impaction plate along with the inlet tube constitutes a ninth stage. By removing particles at the inlet, an upper size limit for particles collected on the first stage is obtained. Table 6.3 shows the critical design and operating parameters for the inlet and ten available stages of the impactor. Any eight of these ten stages can be used in an eight-stage MOUDI cascade impactor. The number of nozzles ranges from 1 for the inlet to 2000 for the lower stages with the nozzle diameters ranging from 1.71 cm to 52 µm. Nozzles whose diameters are 90 µm and larger are drilled while the two smallest sizes are chemically etched. In the design of MOUDI cascade impactors, it is desirable to have S/W values of at least 1/2 and Reynolds numbers of airflow in the nozzles from 500 to 3000, where S is the jet-to-plate distance and W is the nozzle diameter. The values of S/W and Reynolds numbers of the nozzles in the MOUDI cascade impactor are close to these values. The large S/W values at the lower stages are a consequence of the small nozzle diameters and the practical requirement of all pressures at the exit of each stage relative to the inlet pressure are also tabulated in Table 6.3. These values range from essentially 1.00 for the inlet and the first four stages to 0.53 at the exit of

Aerosol sampling and measurement techniques

139

Table 6.3 Design and operation parameters of a micro-orifice uniform deposit impactor (MOUDI) Stage

Nominal cut-pointa (µm)

As calibrated cut-pointa (µm)

Nozzle diameter (cm)

Number of nozzles

S/W b

P /P0 c

Nozzle Reynolds number

Inlet 1 2 3 4 5 6 7 8 9 10

18 10.0 5.6 3.2 1.8 1.00 0.56 0.32 0.18 0.100 0.056

18 9.9 6.2 3.1 1.8 1.00 0.56 0.32 0.18 0.097 0.057

1.71 0.889 0.380 0.247 0.137 0.072 0.040 0.0140 0.0090 0.0055 0.0052

1 3 10 10 20 40 80 900 900 2000 2000

0.75 0.5 1.0 1.0 1.0 1.0 1.5 4.1 6.4 10.6 11.1

1.00 1.00 1.00 1.00 1.00 0.99 0.97 0.95 0.89 0.76 0.53

2420 1560 1090 1680 1510 1440 1340 350 580 500 750

a Based on flow rate of 30 L/min at standard atmospheric temperature and pressure. b S = jet-to-plate distance; W = nozzle diameter. c P = absolute pressure at stage exit with all upstream stages present; P = pressure at MOUDI inlet. 0

stage 10. A typical mass size distribution of aerosols as measured by the MOUDI cascade impactor in a coal mine using diesel-powered equipment where the concentration was much higher and the run times shorter (90 min) is shown in Figure 6.16. These data indicate a bimodal distribution with the lower and the upper modes consisting of diesel exhaust particles and coal and rock dust, respectively.

7. Particle trap impactor A particle trap impactor with three different sets of geometrical parameters is presented in a schematic diagram in Figure 6.17. The design of the particle trap impactor is based on the particle cup (trap) impactor configuration of Kim et al. (2002). The sampling flow rates of Systems I and II are set at 5 L min−1 , and the acceleration nozzle diameter is calculated from the Stokes number, Stk. The Stokes number of a particle having a 50% probability of separation, Stk50 , is defined as follows: Stk50 =

2 U ρp Cc d50 , 9μD1

(6.12)

where U is the jet velocity, d50 is the diameter of the particle having a 50% probability of collection, ρp is the particle density, μ is the air viscosity, D1 is the nozzle diameter and Cc is the Cunningham slip correction factor. So far, the value of 0.24 is used for the Stk50 of particle trap impactors and cup impactors (Kim et al., 1998, 2002). The acceleration nozzle diameter, D1 of System I, 2.6 mm, is determined from the Stk50 , and that of System II, 2.2 mm, is decided to increase collection

140

C. Papastefanou

Fig. 6.16. Mass size distribution as measured by a MOUDI impactor in a coal mine with diesel-powered equipment.

Fig. 6.17. A schematic of a particle trap impactor.

Aerosol sampling and measurement techniques

141

Table 6.4 Design parameters and test conditions of a particle trap impactor Parameter

System I

System II

System III

D1 (mm) S (mm) S/D1 D0 (mm) Q (L/min)

2.6 3.1 1.2 7 5

4.5 5.4 1.2 8.6 25

Reynolds No.

2660

2.2 2.6 1.2 7 5 (3, 5 and 8) 3140 (1890, 3140 and 5030)

7690

efficiency and to obtain the low cutoff diameter (Kim et al., 2002). The nozzle-to-trap distance, S, of System I can easily be varied and its particle trap diameter, D0 , is 7 mm which is large enough to collect particles. It is found that the collection efficiency of particle trap impactors decreases when particle trap diameter decreases (Tsai and Cheng, 1995). The low collection efficiency for a small diameter of particle trap is probably caused by quiescent air in the cavity that stops the impacting particles effectively. The intake angle of an acceleration nozzle, θ , is very important to effect particle focusing and reduce coarse particle loss. Intake angle of all the acceleration nozzles is 45◦ . The design of System III is the same as that of System I except for the acceleration nozzle diameter and the particle trap diameter. The sampling flow rate of System III is set at 25 L min−1 and its nozzle diameter 4.5 mm is determined from the Stk50 like System I. The S/D1 value of all the systems is set at 1.2. In general, the smaller the S values, the smaller the cutoff diameter (Kim et al., 2003). The design parameters and Reynolds numbers of the particle trap impactors are listed in Table 6.4. The Reynolds number of each system is less than 7700. McFarland et al. (1978) reported that a Reynolds number from 2000 to 9000 has little effect on the shape of the collection efficiency curves. The particle trap impactors, therefore, do not have any problem as PM separators. An experimental set-up used for characterising a particle trap impactor is shown in Figure 6.18. Figure 6.19 shows the penetration efficiency curves for the different nozzle diameters. Systems I and II have the acceleration nozzle diameter of 2.6 and 2.2 mm, respectively, while the sampling flow rates and S/D1 values of both systems are the same: 5 L min−1 and 1.3, respectively. Penetration curves are expressed as a function of the aerodynamic particle diameter (a) and the square root of the Stokes number (b), respectively.

8. The diffusion battery method for aerosol particle size determination It is known that health hazards are caused by inhalation of sub-micron aerosol particles. Drinker and Hatch (1936) stated that vast numbers of particles below 0.25-µm radius are generated by industrial processes and they considered this sub-micron aerosol to be quite dangerous. Particles below 0.1-µm radius are too small for direct microscopic observation. Likewise, the standard light scattering techniques cannot be used. Consequently, the methods in use depend on observation of the particles with the ultramicroscope or electron microscope.

142

C. Papastefanou

Fig. 6.18. Experimental apparatus for characterising the particle trap impactor.

(a)

(b)

Fig. 6.19. Penetration efficiency curves for the different nozzle diameters. Systems I and II have acceleration nozzle diameters of 2.6 and 2.2 mm, respectively. Penetration curves are expressed as functions of aerodynamic particle diameter (a), and the square root of the Stokes number (b), respectively.

These methods are tedious and are not suited for use in filter paper efficiency investigations. LaMer et al. (1950) have developed a unique method in which the particles are grown by use of an appropriate solvent, the grown size measured by standard methods and the original size then calculated from the known growth factor. Submicrometre aerosols or condensation nuclei in the size range 0.002 to 0.2 µm diameter are most readily measured for number concentration and size distribution by diffusion bat-

Aerosol sampling and measurement techniques

143

Fig. 6.20. A sketch of a typical diffusion battery.

teries and condensation nucleus counters (Sinclair et al., 1979). The diffusion battery method has several outstanding advantages: (1) the aerosol particle is measured without collection or alteration of the individual particles, (2) the method can be made rapid, provided only that the aerosol mass concentration may be measured rapidly, and (3) there is no lower limit to the size which is measurable. However, the diffusion battery method does not seem to have ever been given a direct experimental test against the light scattering method. A sketch of a typical diffusion battery is shown in Figure 6.20 (Thomas, 1953). For a diffusion battery having these dimensions, the effect of gravity is negligible for particles under 0.3-µm radius. The operation of a diffusion battery is based on the Brownian motion of the aerosol particles. As the aerosol particles move in streamline flow through the long narrow channels, the random Brownian movement of the particles causes them to displace from their original position in the air flow streamline. The most probable displacement from a streamline is zero, but the average displacement is proportional to the square root of the travel time. Consequently, some of the particles are displaced sufficiently to reach the walls of the diffusion battery. It is assumed that once the particles reach the diffusion battery walls, they will stick to the walls. The efficiency of filter papers increases with time when using solid test aerosol. If particles were dislodged from the surface of the filter, the efficiency would tend to decrease with increasing testing time. In this manner, some of the influent aerosol particles are removed by the diffusion battery as they travel along the channels, and a fraction F of the influent particles appears in the

144

C. Papastefanou

diffusion battery effluent. It is easy to visualise in a qualitative way that, since the magnitude of the Brownian movement increases with decreasing particle size, the smaller the particles the more sorption by the walls and the lower the value of F . This dependence of F on the particle size of the aerosols is the basis of the diffusion battery method. The usual procedure has been to use equations relating F and the diffusion constant of the aerosol particles and then to calculate the particle size from the diffusion constant. This is done by use of the Einstein relation with modifications for the slip of particles between air molecules. In the diffusion battery, the gaseous ions are sorbed by the wall of the diffusion tube while en route through it, analogous to the removal of aerosol particles by the battery wall. The fraction penetrating the battery, F , is given by the equation (Townsend, 1900) F = 0.78e−α + 0.097e−6.1α ,

(6.13)

where α=

3.66πDZ , Q

(6.14)

which is dimensionless, Q is the volumetric flow rate, in cm3 s−1 , Z is the battery length, in cm, and D is the diffusion coefficient, in cm2 s−1 . The diameter of the tube does not enter into the above equation. This does not seem reasonable. Tripling the diameter of a tube of fixed length, Z, at a given volumetric rate, Q, would make it necessary, on average, for each particle to diffuse three times as far before contacting the wall, but this would also reduce the linear air velocity, in cm s−1 , by a factor of 9, giving the particle nine times as much time for diffusion. Since the diffusing distance is proportional to the square root of the time, the net result is the same mass transfer to the wall, leaving the fraction penetrating the battery, F , unaffected. The diffusion battery method is good for particle sizes under 0.5-µm radius. With a sensitive method for the determination of mass concentration, it should be possible to determine approximate particle sizes from 0.5 µm down to gas molecule size, 0.0003-µm radius. Measurement of the number concentration can be made with a condensation nucleus counter, CNC, which requires constant flow rate. The condensation nucleus counter operates either by intermediate sampling or varying flow rate and adiabatic expansion of a humid atmosphere to cause cooling and condensation of water droplets on the nuclei. A continuous flow condensation nucleus counter which utilises thermo-electric elements to obtain continuous cooling and ethyl alcohol as the condensate is shown in Figure 6.21 (Sinclair and Hoopes, 1975). Aerosol is drawn through the alcohol tube and the cold tube in succession at 4 L min−1 . The decrease in light transmission through the cold tube with increase in aerosol particle concentration may be measured with a microammeter or a millivolt recorder connected to the photocell. The cold tube is usually run at −20 ◦ C. Measurement of the Aitken nuclei or Aitken particles can be made by a pneumatic expansion type particle counter. Aitken particles is the name given to those atmospheric particles whose size lies between a few angstroms (1 angstrom is 10−8 cm or 10−4 µm) and 0.1 µm. They constitute the greatest contribution to the spectrum of sizes in the atmospheric aerosol. The measurement of the concentration of these particles is based on their ability to act as condensation nuclei for water vapour, when this is supersaturated.

Aerosol sampling and measurement techniques

145

Fig. 6.21. Continuous flow condensation nuclei counter, side view. N = nuclei inlet; A = alcohol tube; AP = alcohol pool; O = orifices; C = cold tube; LB = light bulb; L = lens; PC = photocell; E = exhaust; F = flow meter; P = pump; I = thermal insulation; D = drain.

A schematic drawing of an Aitken particle counter is shown in Figure 6.22 (Ruiperez et al., 1984) which consists of the following elements: (i) an aluminium chamber (2), 30 cm long, which has a transverse section limited by a circumference in its 3/4 superior part and by three small consecutive target circumferences in its inferior part; this configuration is designed to store water in the aluminium chamber to ensure vapour saturation of the air, (ii) two electrovalves (1) and (6), two logicvalves (3) and (5) and a membrane pump, which delivers energy to the system, (iii) two electronic programmers P1 and P2 whose purpose consists in controlling the consecutive stages of the operational cycle, (iv) a security valve (4) whose purpose is to attain the pressure necessary to activate the tandem cylinder (8) and to safeguard the pump membrane, and (v) a drier, not represented in Figure 6.22, behind the pump and whose purpose is to protect its membrane. The air intake is through valve (1) crossing chamber (2) and the valve (6). The air expelled by the pump acts on the logicvalves (3) and (5) and on the security valve (4). This opens when the pressure reached is >2 atmospheres. This sweeping operation is maintained long enough to permit an equal air volume to circulate through chamber (2) at least eight times its volume. At the end of the sweeping stage, the programmer P2 gives the command to open logicvalve (5) which allows the double piston of the tandem cylinder (8) to displace from the extreme right position to the extreme left. Subsequently, the programmer P2 sends to electrovalve (6) an order to change its position and to electrovalve (1) to close, blocking the air entrance to the aluminium chamber (2). A new order from the programmer P2 closes the logicvalve (5) and opens the logicvalve (3), causing a piston movement, which in turn causes the air volume in the aluminium chamber (2) to compress isothermically. Following that, the electrovalves (1) and (6) open, which cause an adiabatic expansion in the aluminium chamber (2), that

146

C. Papastefanou

Fig. 6.22. A schematic of an Aitken (nuclei) particle counter.

originates cloud droplets, one for each particle. The counting is done by the detecting system and the cycle is completed. The detection limit is 300 Aitken particles cm−3 . 9. Condensation particle counter battery An important phenomenon controlling the number concentration of aerosol particles in the atmosphere is new particle formation from gaseous precursors, which involves the production of nanometre-size aerosol particles by nucleation (Aitken nuclei particles) and their subsequent growth to detectable sizes. The formation and growth of fresh atmospheric aerosol particles might be investigated using a condensation particle counter battery, CPCB, which is a matrix of four separate condensation particle counters, CPCs, which differ in the combination of both cut-off size and working liquid (water; n-butanol). The counting efficiencies and cut-off sizes of the condensation particle counters are characterised under different conditions for different condensing vapours, temperature differences between condenser and saturator, and aerosol types. The condensation particle counter battery represents a novel application to infer information on the chemical composition of aerosol particles between 2 and 20 nm, and based on the principle of particle detection by vapour activation in a condensation particle counter. This principle consists of three processes: (1) creation of a supersaturated vapour (working fluid); (2) growth of aerosol particles by condensation of the supersaturated vapours; and (3) optical detection of the aerosol particles after their growth.

Aerosol sampling and measurement techniques

147

Fig. 6.23. A block diagram of the principle of the condensation particle counter battery.

The condensation particle counter battery, CPCB, is composed of four individual condensation particle counters, CPCs, which are operated in parallel. Figure 6.23 shows the arrangement of the four condensation particle counters, which represent a 2 × 2 matrix of different cut-offs (3 and 11 nm), as well as different working liquids (n-butanol and water) (Kulmala et al., 2007). The pair of condensation particle counters, CPCs, as well as the pair of ultrafine condensation particle counters, UCPCs, were designed so that the same count rates are expected for particle materials that show affinity neither to water nor to n-butanol. Owing to the increased activation probability in water vapour, hygroscopic particles will be detected down to lower aerosol particle sizes in condensation particle counters operating with water. Therefore, an increased count rate will be measured in comparison to the condensation particle counters operating with n-butanol. The differential signal between the water and n-butanol condensation particle counters of each pair is then expected to be related to the presence of hygroscopic particles in the size range of the corresponding cut-off regions. The cut-off sizes of the two n-butanol condensation particle counters were determined to be 3 and 11 nm, respectively. The cut-off sizes of the two water condensation particle counters were adjusted so that they matched very closely the cut-off of the corresponding n-butanol condensation particle counters. The particular temperature differences are 70 K for the water ultrafine condensation particle counter (UCPC 3786, Figure 6.23) and 25 K for the water condensation particle counter (CPC 3785, Figure 6.23) and are used as operational values during the measurements. Besides detecting aerosol particles, the condensation particle counters can also be used to detect small molecular clusters of aerosol particles (Kulmala et al., 2005). When the saturation ratio is increased above a certain limit by increasing the temperature variation in a condensation particle counter, ion clusters will be activated first. Further increase of the saturation ratio will then also activate neutrally charged clusters and finally new aerosol particles are formed via homogeneous nucleation. By varying the temperature variation and the type of

148

C. Papastefanou

condensation vapour, information on the size and chemical composition of an aerosol particle population can be obtained. The condensation particle counter battery measurements are complemented by aerosol particle size distribution measurements using a dual differential mobility particle sizer system covering a size range of 3–900 nm, and an aerodynamic particle sizer covering aerosol particle sizes between 0.7 and 20 µm. In addition, air ions are detected using a balanced scanning mobility analyser and an air ion spectrometer. During the period of measurements, several new particle formation (nucleation) events occur in tropospheric air.

10. The Mercer-style impactor The use of an impactor with a single circular orifice for each jet provides a convenient method for determining the aerodynamic size distribution of an easily assayed aerosol, such as of a radioactive aerosol. This type of impactor is preferred when small sample rates and short sampling periods are desired since the stage characteristics can be accurately prescribed and particle bounce or re-entrainment can be minimised with the use of an appropriate liquid film on the collectors. For this purpose, Mercer et al. (1962) designed a miniature five-stage cascade impactor using four individual circular jets and a filter stage (back up filter) for collecting small aerosol particles that negotiate the impaction stages. Each successive jet hole was smaller than the previous jet and the ratio of collector separation to hole size was maintained similar for each jet. This impactor was specifically designed to sample radioactive aerosols in an inhalation research laboratory operated by Lovelace Foundation, Albuquerque, New Mexico. Continuing in this line, Mercer et al. (1970) described the improved version having seven impaction stages. This design and its various modifications are commonly called the Mercer-style impactor. The collection efficiency of each stage of any cascade impactor for various particle sizes depends on a number of geometric and operational factors including (1) the linear velocities of the air streams, Uo , which is equal to the volumetric flow rate divided by the area of the opening, (2) the jet opening shape and size, W , which is the jet width for slits or the jet diameter for holes, (3) the aerosol particle shape, (4) the particle diameter, Dp , which is the geometric diameter, if the particle is spherical, (5) the particle density, ρ, (6) the Cunningham slip correction factor, C, which is important for smaller particles, (7) the air viscosity, η, (8) the jet-to-collector separation distance, S, (9) the jet throat length, T , and (10) the character of the jet airflow as can be conveniently described by the Reynolds’ number, NRe . Aerodynamic separation for a given particle based upon inertial momentum in an airstream which changes direction can be characterised by the particle stopping distance. This is defined as the distance travelled in the forward direction before coming to rest with respect to the surrounding air for a particle travelling initially at a speed Uo , in the forward direction, with respect to the air. The stopping distance for a spherical particle of geometric diameter, Dp , can be expressed as L=

Uo ρCDp2 18η

=

ρ ∗ Uo Dae , 18η

(6.15)

Aerosol sampling and measurement techniques

149

Fig. 6.24. A schematic of the cross-section of the Mercer-style cascade impactor.

where L is the stopping distance, ρ is the physical density of the aerosol particle, assuming the density of air is essentially nil, ρ ∗ is the standardising density which is equal exactly to 1 g cm−3 , C is the Cunningham slip correction factor for correcting Stokes’ law of viscous resistance for smaller particles, η is the viscosity of air and Dae is the aerodynamic (resistance) diameter which is equal to Dp [ρC/ρ ∗ ]1/2 . The flow Reynolds’ number, NRe , for a circular jet is given by NRe =

ρair Uo W , H

(6.16)

where ρair is the physical density of air (ρair = 0.001293 g cm−3 ) and W is the jet diameter. The design of the Mercer-style impactor is illustrated in Figure 6.24 (In-Tox Products, L.L.C., 101 Wilderness Ct. S., Moriarty, NM 87035). Each stage of the impactor has a collector cover-slip holder which can accommodate a standard 22 mm diameter glass cover slip

150

C. Papastefanou

as is commonly used in light microscopy. The cover slip on each stage must be coated with a suitable liquid film to prevent particle bounce during impaction. Anti-foam spray has been found to be a satisfactory coating for impactor stages. The cover slip holders are spring-fitted with wire legs so that they are automatically positioned against the metal pins which act as jet-collector spacers. Each stage is O-ring-fitted into a metallic cylindrical body, so that interstage leaks are prevented. The nose inlet cap of the device contains the first impactor jet orifice. The nose inlet is chamfered and O-ringed to fit a one-half inch diameter opening from which samples can be taken. The back end-cap contains a built-in filter-holder to accommodate a 25 mm diameter filter which serves as the eighth collection stage. The back end-cap also has a pipe thread outlet to be connected to the flow control and vacuum system used to draw samples through the impactor. Nominal pressure drop through the impactor varies up to about 1.5 psi depending upon the chosen flow rate. The so-called 100 cm3 min−1 impactor is normally operated at flow rates from 50 to 150 cm3 min−1 . The effective aerodynamic cutoff diameters are 0.23, 0.50, 0.75, 1.12, 1.53, 2.13 and 2.90 µm. The 1 L min−1 unit is normally operated from 300 to 1000 cm3 min−1 . The effective aerodynamic cutoff diameters are 0.37, 0.81, 1.18, 1.78, 2.38, 3.34 and 5.10 µm. Operation of the impactor depends critically on maintenance of a constant sampling rate. Normally, the impactor should be operated for no more than 10 min to avoid alteration of the effectiveness of the liquid film on the stage collectors in preventing bounce and to avoid re-entrainment caused by sample build-up depending on the aerosol concentration. The samples are analysed by determining the relative mass of material collected on each stage and the back-up filter. For radioactive aerosols, this is conveniently accomplished by counting the emissions. Since the dead-space in the impactor is about 15 cm3 from the inlet to the filter, a dead-space correction should be made for each sample on each stage if the dead space volume is a significant fraction of the total sample volume. The data consist of the mass (or corrected mass) and the radioactivity of material on each of eight stages including the filter. Using a method of weighed least-squares, a log-normal function describing the aerosol mass and activity size distribution with respect to aerodynamic (resistance) diameter has been routinely fitted to the data. It is also possible to approximate a log-normal function for the mass and activity distribution with the use of log-probability paper. The distribution is then describable by a mass or activity median aerodynamic diameter, MMAD or AMAD, and associated geometric standard deviation, σg .

11. Berner-type impactor The Berner-type impactor is a nine-stage low-pressure cascade impactor with a nominal volumetric flow rate of 28.9 L min−1 (1 ACFM) (LPI 30/0.06/2, Hauke KG, A-4810 Gmunden, Austria; Berner et al., 1979; Berner and Lurzer, 1980). The effective cutoff diameters are 0.06, 0.12, 0.25, 0.5, 1.0, 2.0, 4.0, 8.0 and 16 µm. Besides the entrance stage, which has a single orifice in order to precipitate particles larger than 16 µm, each stage has several equal size orifices lined up equidistantly on a circle centred on the axis of the stage (Figure 6.25). This arrangement warrants symmetrical flow patterns and symmetrical particle deposition, and therefore

Aerosol sampling and measurement techniques

151

Fig. 6.25. Berner-type impactor stage (partly cut away in order to show the arrangement of the stage elements) with the orifice plate, OR. The collection aluminium foil (not shown) covers the stagnation plate, SP, completely and is held in place by the spacer, S.

the deposition spots beneath the orifices contain equal amounts of mass, within the limits of a few percent of the average spot mass. The aerosol jet coming from the orifices is deflected by the stagnation plate, which has a large centre hole for the flow to pass to the subsequent stage. This design is advantageous in so far as particles which are carried away from their location of deposition are not transferred into the deposits of the subsequent stages. The aerosol particles are collected on impaction substrates stamped from Apiezon-greased aluminium foils in order to prevent the particles from bouncing or dry aluminium foils of about 10 µm thickness. In order to avoid overloading, less than about 8 mg of particulate material is collected at any stage within the sampling period. Evidently, a certain amount of coarse particles is missing when dry foils are used, but they do not occur in the finer particle fractions as the unnormalised mass size distributions are identical in the size range of 0.70 µm aerodynamic diameter. The distance between the orifice and stagnation plates is made by a removable circular spacer. A critical orifice, which is placed behind the final impactor stage, controls the total volume flow rate and keeps the impactor flow rate from fluctuations.

12. Dichotomous sampler The dichotomous sampler is a virtual impactor with two emerging aerosol flows. The air beyond the virtual interface is withdrawn at a slow rate, thus rendering this impactor a twoflow or dichotomous impactor. In both flows aerosol particles can be collected on a filter or analysed in a real-time aerosol instrument. A continued suspension of the size-fractionated particles in the effluent air flows is the principal advantage of these virtual or dichotomous impactors (Chen et al., 1985). The Andersen PM 10/2.5 dichotomous sampler accurately monitors ambient air for coarse and fine particulates (Figure 6.26; Andersen Samplers Inc., 4215 Wendell Drive, Atlanta, GA 30336). The dichotomous (dichot) sampler separates aerosol particles into two distinct size fractions, 2.5 and 10 µm (“coarse” particles) and less than 2.5 µm (“fine” particles). The virtual

152

C. Papastefanou

Fig. 6.26. An Andersen dichotomous sampler and a diagram of the inertial collection of aerosol particles.

Aerosol sampling and measurement techniques

153

impactor ensures that the particulate deposit is uniformly collected on two Teflon membrane filters which eliminate artifact formation and possible problems of particle bounce and reentrainment often experienced in other particle size impactor methods. The dichotomous sampler operates at a flow rate of 1 cm3 h−1 (16.7 L min−1 ). The coarse (PM 10 to 2.5) particle receiver tube has a flow rate of 0.1 cm3 h−1 (1.67 L min−1 ) for collection of the coarse aerosol particles, while the fine particles, less than 2.5 µm, follow a flow rate of 0.9 cm3 h−1 (15 L min−1 ) to the fine particle filter.

13. Compact multistage cascade impactor A fully assembled compact multistage cascade impactor, CCI, and its internal parts are shown in Figure 6.27 (Demokritou et al., 2004). The sampling device consists of eight impactor stages and an after-filter. The 50% cutoff points of the eight stages are 9.9, 5.3, 3.3, 2.5, 1.7, 1.0, 0.47 and 0.16 µm (aerodynamic diameter), with pressure drops of 0.02, 0.02, 0.04, 0.06, 0.08, 0.27, 1.57 and 5.73 kPa. These pressure drops are considerably lower than those obtained using flat rigid impaction substrates with comparable cutoff points. Each stage consists of a slit-shaped acceleration nozzle and a rectangular polyurethane foam, PUF, impaction substrate (density = 20 kg m−3 , Merryweather Foam, OH). The PUF substrates are easily inserted and removed from the substrate base (Figure 6.27), and securely transported for gravimetric or chemical analysis. The acceleration nozzle width and length dimensions are shown in Table 6.5. The optimal thickness of the polyurethane foam for all stages is 0.64 cm, which ensures that aerosol particles do not penetrate through the PUF to the substrate holder. The major feature of this aerosol sampler is its ability to both fractionate by size and collect relatively large amounts of aerosol particles (mg quantities) onto the inert polyurethane foam impaction substrates. Even though the impaction substrates are not coated with adhesives, such as grease or mineral oil, particle bounce and re-entrainment losses were found to be insignificant. Ultrafine particles, less than 0.16 µm, are collected on a 47 mm diameter 2 µm pore Teflon membrane filter. The filter material has both a high collection efficiency and a relatively low pressure drop. Other filter materials, such as polypropylene fibre, can also be used to collect the ultrafine aerosol particles. Particle losses for each stage are less than 10% for aerosol particles smaller than 7 µm and less than 20% for particles larger than 7 µm. Comparison of the compact multistage cascade impactor with the collocated micro-orifice cascade impactor showed that the mass concentrations measured by the latter are considerably lower than those measured by the former and the average ratio of total mass concentration by micro-orifice impactor to compact multistage impactor is 0.86, with the size distribution measured by the compact multistage impactor being closer to that measured using the real-time particle-sizing instruments. According to the impaction theory, the dimensionless Stokes number, Stk, is the governing parameter for impaction and is defined as follows: Stk = ρp dp2 U Cc /9μW,

(6.17)

where μ is the dynamic viscosity of air, dp is the diameter of an aerosol particle, ρp is the particle density, W is the nozzle diameter, U is the jet velocity and Cc is the Cunningham slip

154

C. Papastefanou

Fig. 6.27. Compact multistage cascade impactor.

correction factor. The slip correction factor is given by the equation (Hinds, 1999) Cc = 1 + Kn(2.34 + 1.05e−0.195Kn )/2,

(6.18)

where Kn is the dimensionless Knudsen number, 2λ/dp , defined by the ratio of two times the mean air molecule free path, λ, to the particle diameter, dp .

14. K-JIST 5-stage cascade impactor A cross-sectional view of the K-JIST 5-stage cascade impactor is shown in Figure 6.28 (Kwon et al., 2003). Each impactor stage consists of the identical stage wall and impaction plate but

Aerosol sampling and measurement techniques

155

Table 6.5 Physical characteristics, flow parameters and experimental results of a compact multistage cascade impactor Stage No.

1

2

3

4

5

6

7

8

0.058 5.00 1.0

0.058 1.78 1.0

0.028 1.91 2.0

Physical characteristics Acceleration nozzle W (cm) L (cm) S/W Substrate Width × length (cm) Material

0.432 7.19 1.0

0.236 7.01 1.0

0.160 6.32 1.0

0.137 5.92 1.0

0.094 5.59 1.0

1.8 × 8.1 1.8 × 8.1 1.0 × 6.4 1.0 × 6.4 0.64 × 6.4 0.64 × 6.4 0.64 × 6.4 0.64 × 6.4 PUF

PUF

PUF

PUF

PUF

PUF

PUF

PUF

Re U (m/s)

926 1.6

950 3.0

Flow parameters 1053 1125 1191 4.9 6.2 9.5

1330 17.1

3744 48.1

3495 93.9

d50 (nm) √ Stk σ P (kPa)

9.9 0.47 1.38 0.02

5.3 0.47 1.33 0.02

Experimental results 3.3 2.5 1.7 0.46 0.43 0.44 1.32 1.38 1.33 0.04 0.06 0.08

1.0 0.47 1.43 0.27

0.47 0.39 1.81 1.57

0.16 0.32 1.74 5.73

Note. W , nozzle width; L, nozzle length; S, substrate-to-nozzle distance; Re, Reynolds number; U , nozzle air veloc√ ity; d50 , 50% cut point; Stk, square root of Stokes number: σ , collection efficiency curve sharpness; P , pressure drop.

differs in terms of nozzle diameter, the number of nozzles and the thickness of the nozzle throat, T , which was adjusted by modifying the thickness of the nozzle plate. The cutoff point diameters of the stages are 10, 5, 2.5, 1.0 and 0.7 µm, with a nominal flow rate of 30 L min−1 at atmospheric pressure and an air temperature of 20 ◦ C. It is known that aerosol particles tend to focus towards the centre in a straight orifice which may result in their enrichment near the axis of the nozzle. In the absence of a gradual entrance region, strong aerodynamic focusing occurs which is responsible for the aerosol particle concentration enrichment that occurs in the axial region of the nozzle. This enrichment in turn enhances the collection of small aerosol particles and results in a tail at the low efficiency end. Therefore, a gradual inlet is introduced to minimise aerodynamic focusing effects when manufacturing the impactor nozzles. O-rings are used to prevent an air leak from the interstage. Fully assembled impactors are tightened and sealed in the impactor housing. In this design, aerosol particles are drawn through a series of progressively narrowing nozzles, each of which is followed by an impaction plate. After traversing the last stage, stage 5, the stream is drawn through an after-filter as it exits the device. The mesh screen is installed on the outlet to support the after-filter. When an impactor is calibrated using the counting method, the converging outlet is used to reduce particle loss during transportation to the particle counter. Each impactor stage, an inlet and an after-filter stage are tightened to prevent air leakage from the impactor. A stage of an impactor is composed of a nozzle plate, an impactor plate and a stage wall. The nozzle plate plays the most important role amongst the impactor components, in terms

156

C. Papastefanou

Fig. 6.28. Cross-sectional view of the K-JIST 5-stage cascade impactor.

Table 6.6 Design and operation parameters of the K-JIST 5-stage cascade impactor Stage No.

Number of nozzles

W (mm)

V (m/s)

Re

S/W

T /W

Cut-point diameter (µm)

1 2 3 4 5

3 12 12 20 40

9.04 3.24 2.04 0.93 0.58

2.60 5.06 12.76 36.52 46.63

1563 1091 1732 2270 1814

0.5 1.0 2.1 5.4 8.2

0.4 1.5 2.9 5.5 8.8

10 5.0 2.5 1.0 0.7

St50 = 0.49 except for stage 1 which was 0.42, T = 293 K, flow rate = 30 L/min.

Aerosol sampling and measurement techniques

157

of determining the particle collection efficiency. The number of nozzles, N , and the nozzle diameter, W , are major design parameters in a multi-nozzle impactor. Marple and Willeke (1976) investigated the effects of the Reynolds number on the particle collection efficiency by controlling the number of nozzles used per stage in an impactor. However, the results were used as guidelines when analysing or designing impactors with Reynolds numbers of 500, 3000 and 10,000. Marple (1970) and Marple and Liu (1974) recommended a Reynolds number from 500 to 3000. Using the square root of Stokes number at 50% collection efficiency, 1/2 St50 , the number of nozzles at each stage can be determined based on Reynolds number criteria with a nominal flow rate of 30 L min−1 . After the number of nozzles has been determined, the nozzle diameter is calculated. Table 6.6 summarises the design and operation parameters of the K-JIST 5-stage cascade impactor.

References Andersen, A.A. (1958). New sampler for the collection, sizing and enumeration of viable airborne particle. J. Bacteriol. 76, 471–484. Andersen, A.A. (1966). Sampler for respiratory health hazard assessment. AIHA J. 27 (April–May). Andersen, A.A. (1976). 1-ACFM Ambient Particles Sizing Samplers. Andersen Inc., Atlanta, GA, 21 pp. Berner, A. (1978). Chem. Ing. Tech. 50, 399. Berner, A., Lurzer, C. (1980). Mass distributions of traffic aerosols at Vienna. J. Phys. Chem. 84, 2079–2083. Berner, A., Lurzer, C., Pohl, F., Preining, O., Wagner, P. (1979). The size distribution of the urban aerosols in Vienna. Sci. Total Environ. 13, 245–261. Biswas, P., Flagan, R.C. (1988). The particle trap impactor. J. Aerosol Sci. 19, 113–121. Brown, J.H., Cooke, K.M., Ney, F.G., Hatch, T.F. (1950). Influence of particle size upon the retention of particulate matter in the human lung. J. Am. Public Health Assoc. 40, 450–458. Chen, B.T., Yeh, H.C., Cheng, Y.S. (1985). A novel virtual impactor: Calibration and use. J. Aerosol Sci. 16 (4), 343–354. Chen, B.T., Yeh, H.C., Cheng, Y.S. (1986). Performance of a modified virtual impactor. Aerosol Sci. Technol. 5, 369–376. Cheng, Y.S., Keating, J.A., Kanapilly, G.M. (1980). Theory and calibration of a screen-type diffusion battery. J. Aerosol Sci. 11, 549–556. Cheng, Y.S., Yu, C.C., Tu, K.W. (1994). Intercomparison of activity size distributions of thoron progeny by alphaand gamma-counting methods. Health Phys. 66, 72–79. Cliff, K.D. (1978). The measurement of low concentrations of radon-222 daughters in air, with emphasis on RaA assessment. Phys. Med. Biol. 23, 55–65. Demokritou, P., Lee, S.J., Ferguson, S.T., Koutrakis, P. (2004). A compact multistage (cascade) impactor characterization of atmospheric aerosols. J. Aerosol Sci. 35, 281–299. Drinker, S.B., Hatch, T. (1936). Industrial Dust. McGraw–Hill, New York, pp. 55–59. Dzubay, T.G., Rickel, D.G. (1978). In: Russel, P.A., Hutchings, A.E. (Eds.), Electron Microscopy and X-Ray Application. Ann Arbor Science, Ann Arbor, MI, pp. 3–209. Evans, R.D. (1969). Engineer’s guide to the elementary behavior of radon daughters. Health Phys. 17, 229–255. Hamilton, R.J. (1952). The Falling Speed and Particle Size of Airborne Dusts in Coal Mines. National Coal Board, Scientific Research Establishment, Rep. No. 137. Hamilton, R.J., Walton, R.J. (1952). The Selective Sampling of Airborne Dust. National Coal Board, Scientific Department, Central Research Establishment, Rep. No. 139. Harber, G.J., Morton, J.D. (1953). The respiratory retention of bacterial aerosols; Experiment with radioactive spores. J. Hyg. 51, 372–385. Hatch, T.F. (1955). Developments in the sampling of airborne dust. Arch. Ind. Hyg. Occup. Med. 11, 212–217. Hatch, T.F. (1959). Respiratory dust retention and elimination. In: Processing of the Pneumonoconiosis Conference, Johannesburg. J. & A. Churchill, London, pp. 113–132.

158

C. Papastefanou

Hering, S.V., Marple, V.A. (1986). In: Lodge, J.P., Chen, T.L. (Eds.), Cascadec Impactor: Sampling and Data Analysis. American Industrial Hygiene Association, Akron, OH, pp. 103–127. Hinds, W.C. (1999). Aerosol Technology. Properties, Behavior, and Measurement of Airborne Particles. John Wiley & Sons, New York. Ho, W., Hidy, G.M., Govan, R.M. (1974). Water content of atmospheric aerosols. J. Appl. Meteorol. 9, 828–829. Hopke, P.K., Strydom, R., Ramamurthi, M., Knutson, E.O., Tu, K.W., Scofield, P., Holub, R.F., Chung, Y.S., Su, Y.S., Winklmayr, W., Strong, J.C., Solomon, S., Reineking, A. (1992). The measurement of activity-weighed size distributions of radon progeny: methods and laboratory intercomparison studies. Health Phys. 63, 560–570. Kesten, J., Butterweck, G., Porstendörfer, J., Reineking, A., Heymel, H.-J. (1993). An online α-impactor for shortlived radon daughter. Aerosol Sci. Technol. 18, 156–164. Kim, D.S., Kim, M.C., Lee, K.W. (2000). Design and performance evaluation of multi-nozzle virtual impactors for concentrating particles. Part. Part. Syst. Char. 17, 244–250. Kim, D.S., Hong, S.B., Kim, H.T., Lee, J.H., Lee, K.W. (2003). Design and performance evaluation of low-volume PM2.5 inlet for analyzing chemical composition. Part. Part. Syst. Char. 20, 130–134. Kim, H.T., Kim, Y.J., Lee, K.W. (1998). New PM10 inlet design and evaluation. Aerosol Sci. Technol. 29, 350–354. Kim, H.T., Han, Y.T., Kim, Y.J., Lee, K.W., Chun, K.J. (2002). Design and tests of 2.5 µm cutoff size inlet on a particle cup impactor configuration. Aerosol Sci. Technol. 36, 136–144. Kuhlmey, G.A., Liu, B.Y.H., Marple, V.A. (1981). Am. Ind. Hyg. Assoc. J. 42, 790–795. Kulmala, M., Lehtinen, K.E.J., Laakso, L., Mordas, G., Hammeri, K. (2005). On the existence of neutral atmospheric clusters. Boreal Environ. Res. 10, 79–87. Kulmala, M., Mordas, G., Petaja, T., Gronholm, T., Aalto, P.P., Vehkamaki, H., Hienola, A.I., Herrmann, E., Sipila, M., Riipinen, I., Manninen, H.E., Hameri, K., Stratmann, F., Bilde, M., Winkler, P.M., Birmili, W., Wagner, P.E. (2007). The condensation particle counter battery (CPCB): A new tool to investigate the activation properties of nanoparticles. J. Aerosol Sci. 38, 289–304. Kwon, S.B., Lim, K.S., Jung, J.S., Bae, G.N., Lee, K.W. (2003). Design and calibration of a 5-stage cascade impactor (K-JIST cascade impactor). J. Aerosol Sci. 34, 289–300. LaMer, V.K., Inn, E.C.Y., Wilson, I.B. (1950). The methods of forming, detecting and measuring the size and concentration of liquid aerosols in the size range of 0.01 to 0.25 microns diameter. J. Colloid Sci. 5, 471. Lundgren, D.A. (1967). An aerosol sampler for determination of particle concentration of size and time. J. Air Pollut. Control Assoc. 17, 225. Lundgren, D.A. (1973). Mass distribution of large atmospheric particles. Ph.D. thesis, University of Minnesota, Minneapolis, MN, 161 pp. Markowski, G.R. (1984). Reducing blow off in cascade impactor measurement. Aerosol Sci. Technol. 3, 431–439. Marley, N.A., Gaffney, J.S., Drayton, P.J. (2000). Measurement of 210 Pb, 210 Po, and 210 Bi in size-fractionated atmospheric aerosols: An estimate of fine-aerosol residence times. Aerosol Sci. Technol. 35, 569–583. Marple, V.A. (1970). A fundamental study of inertial impactors. Ph.D. dissertation, University of Minnesota, Particle Technology Laboratory, Minneapolis, MN, Publication No. 44. Marple, V.A., Chein, C.M. (1980). Virtual impactors: A theoretical study. Environ. Sci. Technol. 14, 976–984. Marple, V.A., Liu, B.Y. (1974). Characteristics of laminar jet impactors. Environ. Sci. Technol. 8, 648–654. Marple, V.A., Liu, B.Y.H. (1975). On fluid flow and aerosol impaction in inertial impactors. J. Colloid Interface Sci. 53, 31–34. Marple, V.A., Willeke, K. (1976). Impactor design. Atmos. Environ. 10, 891–896. Marple, V.A., Willeke, K. (1979). Aerosol impactors. In: Aerosol Measurement. University Presses of Florida, Gainesville, FL, pp. 90–107. Marple, V.A., Liu, B.Y.H., Whitby, K.T. (1974). On the flow fields of inertial impactors. ASME J. Fluid Eng. 96, 394–403. Marple, V.A., Liu, B.Y.H., Kuhlmey, G.A. (1981). J. Aerosol Sci. 11, 333–337. Marple, V.A., Rubow, K.L., Behm, S.M. (1991). A microorifice uniform deposit impactor (MOUDI): Description, calibration, and use. Aerosol Sci. Technol. 14, 434–446. McFarland, A.R., Ortiz, C.A., Bertsch, R.W.J. (1978). Particle collection characteristics of a single-stage dichotomous sampler. Environ. Sci. Technol. 12, 679–682. Mercer, T.T., Tillery, M.I., Ballew, C.W. (1962). A cascade impactor operating at low volumetric flow rates. Rep. LF-5, Lovelace Foundation, Albuquerque, NM, pp. 1–19.

Aerosol sampling and measurement techniques

159

Mercer, T.T., Tillery, M.I., Newton, G.J. (1970). A multi-stage, low flow rate cascade impactor. J. Aerosol Sci. 1, 9–15. Meyer, G.A., Lee, K.W. (1994). Real time determination of metal hazardous air pollutants in the gas emissions. Process Control Quality 6, 187–192. Morrow, P.E. (1964). Evaluation of inhalation based upon the respirable dust concept and philosophy and application of selective sampling. Am. Ind. Hyg. Assoc. J. 25, 213. Mullins, W.T., Leddicotte, G.W. (1962). The Radiochemistry of Phosphorus. Rep. NAS-NS 3056, National Technical Information Service, Springfield, VA. Papastefanou, C., Bondietti, E.A. (1991). Mean residence times of atmospheric aerosols in the boundary layer as determined from 210 Bi/210 Pb activity ratios. J. Aerosol Sci. 22, 927–931. Raabe, O.G. (1979). Design and use of the Mercer-style impactor for characterization of aerosol aerodynamic size distributions. In: Aerosol Measurement. University Presses of Florida, Gainesville, FL, pp. 135–140. Ranz, W.E., Wong, J.B. (1952). Jet impactors for determining the particle size distribution of aerosols. Arch. Ind. Hyg. Occup. Med. 5, 464–477. Reineking, A., Becker, K.H., Porstedörfer, J. (1988). Measurements of the activity size distributions of the short-lived radon daughters in the indoor and outdoor environment. Radiat. Prot. Dosim. 24, 245–250. Ruiperez, L.G., Garcia, B.A., Uria, J.M.J., Iglesias, J.M.P. (1984). A new counter of Aitken particles. Atmos. Environ. 18 (8), 1711–1714. Schumann, T., Gysi, H., Kaelin, S. (1988). J. Aerosol Sci. 19, 993. Sinclair, D., Hoopes, G.S. (1975). A continuous flow condensation nucleus counter. J. Aerosol Sci. 6, 1. Sinclair, D., Countess, R.J., Liu, B.Y.H., Pui, D.Y.H. (1979). Automatic analysis of submicron aerosols. In: Aerosol Measurement. University Presses of Florida, Gainesville, FL, pp. 544–563. Stevens, R.K., McFarland, A.R., Wedding, J.J. (1976). Comparison of the virtual dichotomous sampler with the high-volume sampler. Proc. Am. Chem. Soc. 16 (1), 84–85. Thomas, J.W. (1953). The diffusion battery method for aerosol particle size determination. ORNL Report 1648, Oak Ridge National Laboratory, Oak Ridge, TN, 68 pp. Thomas, J.W., Hinchliffe, L.E. (1972). Filtration of 0.001 µm particles by wire screens. J. Aerosol Sci. 3, 387–393. Tokonami, S., Takahashi, F., Iimoto, T., Kurosawa, R. (1997). A new device to measure the activity size distribution of radon progeny in a low level environment. Health Phys. 73, 494–497. Townsend, J.S. (1900). The diffusion of ions into gases. Trans. Roy. Soc. 193A, 129–158. Trembley, R.J., Leclerc, A., Mathieu, C., Pepin, R., Townsend, M.G. (1979). Measurement of radon progeny concentrations in air by α-particle spectrometric counting during and after air sampling. Health Phys. 36, 401–411. Tsai, C.J., Cheng, Y.H. (1995). Solid particle collection characteristics on impaction surfaces of different designs. Aerosol Sci. Technol. 23, 96–106. Veljkovic, S.R., Milenkovic, S.M. (1958). Concentration of carrier-free radioisotopes by sorption on alumina. In: Proceedings of the 2nd UN Conference on Peaceful Uses of Atomic Energy, vol. 20. United Nations, New York, pp. 45–49. Wells, W.F. (1955). Airborne Contagion and Air Hygiene. Harvard University Press, Cambridge, MA, p. 105. Winkler, R., Dietl, F., Frank, G., Tschiersch, J. (1998). Temporal variation of 7 Be and 210 Pb size distributions in ambient aerosol. Atmos. Environ. 32 (6), 983–991.

This page intentionally left blank

Appendices

162

C. Papastefanou

(a) Appendix 1. Particle size range of aerosol particle properties and measurement instruments: (a) Application range for aerosol particle size measuring instruments, and (b) size range of aerosol particle properties.

Appendices

(b) Appendix 1. (continued)

163

164

C. Papastefanou

Appendix 2 Frequently used aerosol particle properties at 20 ◦ C and 1 atm Particle diametera (µm)

Slip correction factor

Settling velocity (cm/s)

Diffusion coefficient (cm2 /s)

Mobility [cm/(s dyn)]

rms Brownian displacement (cm)b

0.001 0.002 0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10 20 50 100

226.20 113.33 45.605 23.039 11.770 5.0506 2.8658 1.8405 1.3067 1.1516 1.0758 1.0303 1.0152 1.0076 1.0030 1.0015

6.82E−07 1.37E−06 3.44E−06 6.95E−06 1.42E−05 3.81E−05 8.65E−05 2.22E−04 9.86E−04 3.48E−03 0.0130 0.0777 0.306 1.21 7.57 30.2

5.37E−02 1.35E−02 2.17E−03 5.47E−03 1.40E−04 2.40E−05 6.81E−06 2.19E−06 6.21E−07 2.74E−07 1.28E−07 4.89E−08 2.41E−08 1.20E−08 4.76E−09 2.38E−09

1.33E+12 3.33E+11 5.36E+10 1.35E+10 3.46E+09 5.93E+08 1.68E+08 5.41E+07 1.54E+07 6.76E+06 3.16E+06 1.21E+06 5.96E+05 2.96E+05 1.18E+05 5.88E+04

1.036571 0.518800 0.208147 0.104612 0.052873 0.021905 0.011667 0.006612 0.003523 0.002339 0.001598 0.000989 0.000694 0.000489 0.000309 0.000218

a Displacement in 10 s. b Effective diameter of air molecule.

Appendices

165

(a)

(b) Appendix 3. Slip correction factor for standard and non-standard conditions: (a) Slip correction factor minus one versus aerosol particle diameter at standard conditions (curve A for aerosol particles of 0.1 µm diameter, curve B for aerosol particles of less than 0.1 µm diameter), and (b) slip correction factor versus aerosol particle diameter times pressure for temperatures from 233 to 893 K (−40 to 600 ◦ C).

166

C. Papastefanou

Appendix 4 Reference values for atmospheric properties at sea level and 293.15 K (20 ◦ C)a Property

Symbol

Value (SI units)

Value (cgs units)

Absolute temperatureb Acceleration of gravityc Atmospheric pressurec,d Avogadro constantc Boltzmann constantc Densitye Diffusion coefficientf Mean free pathg Molar volumee Molecular concentratione Molecular weightc Molecular collision diameterh Molecular velocity (mean)e Ratio of specific heats (cp /cv )i Speed of sounde Universal gas constantc Viscosity, dynamicj

T g p Na k ρg D λ vm n M dm c¯ κ Vs R η

293.15 K 9.8066 m/s2 1.0132 × 105 Pa 6.0222 × 1023 mol−1 1.3806 × 10−23 J/K 1.2041 kg/m3 1.99 × 10−5 m2 /s 0.066 µm 0.024053 m3 /mol 2.5036 × 1025 m−3 0.028964 kg/mol 3.7 × 10−10 m 462.90 m/s 1.400 343.23 m/s 8.3143 J/mol K 1.8134 × 10−5 Pa s

293.15 K (20 ◦ C) 980.66 cm/s2 1.0132 × 106 dyn/cm2 6.0222 × 1023 mol−1 1.3806 × 10−16 dyn cm/K 1.2041 × 10−3 g/cm3 0.199 cm2 /s 0.066 µm 24.053 L/mol 2.5036 × 1019 cm−3 28.964 g/mol 3.7 × 10−8 cm 46290 cm/s 1.400 34323 cm/s 8.3143 × 107 dyn cm/K mol 1.8134 × 10−4 Poise

a Values rounded off to five significant figures where available. b Value given is the standard temperature used in this book. It differs from the adopted sea level temperature of 288.15 K (15 ◦ C) used in USSA (1976). c USSA (1976) value. d Also 760 mm Hg exactly (USSA, 1976). e USSA (1976) value adjusted to the standard temperature of 293.15 (20 ◦ C). f Bolz and Tuve (1973) value. g Calculated by Equation (2.25) using die adopted value of 0.37 ran for the molecular collision diameter. (Note, this value of λ differs slightly from the USSA (1976) value corrected to 293.15 (20 ◦ C) because of a different adopted

value (0.365 nm) for molecular collision diameter.) h Adopted value. This value differs slightly from the adopted value used in USSA (1976) of 0.365 nm. (Note, this

affects the mean free path as calculated by Equation (2.25).) i USSA (1976) adopted value. j USSA (1976) value adjusted to the standard temperature of 293.15 (20 ◦ C) by the Southerland equation (see the

example problem in Section 4 of Chapter 2).

References USSA (1976). U.S. Standard Atmosphere, 1976. National Oceanic and Atmospheric Administration (NOAA), National Aeronautics and Space Administration (NASA), and United States Air Force (USAF), Washington, DC. Bolz, R.E., Tuve, G.L. (1973). CRC Handbook of Tables for Applied Engineering Sciences, second ed. CRC Press, Boca Raton, FL.

Appendices

167

Appendix 5 Common property values of air and water Air at 20 ◦ C and 1 atm (NTP) Density (ρg ) 1.205 × 10−3 g/cm3 = 1.205 g/l = 0.075 lb/ft3 1.832 × 10−4 P = 1.832 × 10−5 Pa s 0.0665 µm

Velocity (ν) Mean free path (λ)

Average molecular ¯ weight (M)

28.96 g/mol

Specific heat ratio (γ ) Diffusion coefficient (D)

1.40 0.19 cm2 /s

Water at 20 ◦ C

Composition of dry air by volume Gas

Content (vol%)

Molecular weight (g/mol)

N2 O2 A CO2 Other

78.08 20.95 0.934 0.033a <0.003

28.01 32.00 39.95 44.01

Viscosity Surface tension Vapour pressure

0.01002 dyn s/cm2 72.75 dyn/cm 17.54 mm Hg = 2.338 kPa

Water vapour at 20 ◦ C Diffusion coefficient Density

0.75 × 10−3 g/cm3

a CO concentration may vary from place to place. 2

Appendix 6 Dimensionless numbers dp U ν

Knudsen (Kn) = dλ

Mean free path of molecules/particle diameter

Reynolds (Re) =

Mach (Ma) = U U sonic

Flow velocity/sonic velocity

ν Schmidt (Sc) = D

Momentum diffusivity/mass diffusivity

Peclet (Pe) = LU D

Bulk mass transfer/diffusive mass transfer

Stokes (Stk) = τLU

Stopping distance/ characteristic flow dimension

Prandtl (Pr) = αν

Momentum diffusivity/ thermal diffusivity

p

Inertia force/viscous force

This page intentionally left blank

169

Subject Index

accelerators, high energy 11, 59, 51 accumulation mode 5–8, 13, 17, 23, 25, 27, 39, 46, 48, 50, 59, 73, 80–82, 92, 93, 95, 113, 115 ACFM ambient cascade impactors 123, 124 activity median aerodynamic diameter, AMAD 25, 28, 44, 53, 54, 73, 91, 92, 95, 108 activity size distribution 5, 13–15, 25, 28, 32, 38, 41, 42, 44, 45, 47–49, 52, 66, 72, 85, 86, 90– 96, 106, 121, 127–129, 133, 150 aerodynamic diameter 13, 14, 22, 25, 26, 28, 32, 38, 43–47, 53, 54, 72, 73, 90–92, 95, 102, 103, 105, 107–109, 125, 134, 150, 151, 153 aerosol sampling 13, 28, 41, 50, 113, 129 air sampling 60 air-to-vegetation transfer 59, 67, 68 Aitken nuclei 3, 5–8, 13, 16, 23, 27, 38, 39, 59, 73, 80–82, 96–98, 144, 146 Aitken nuclei mode 5–7, 13, 59, 80–82, 96, 97 alpha-recoil 32, 35 aluminium-26 11, 12 Andersen cascade impactors 46, 54, 118, 120–123, 129 argon-39 11, 12 ash 3, 61 attachment coefficient 20, 89 attachment rate 57, 111 barium-140 11, 44, 46, 50, 71, 75, 79 Berner-type impactor 150, 151 beryllium-7 11–15, 27, 28, 33, 43, 44, 50, 51, 59, 60, 63, 65–68, 71–74, 124, 130 beryllium-10 11, 12 bimodal distribution 5, 139 biomass density 68 biomass-normalised deposition velocity 68 bismuth-210 11, 33, 71, 72, 75, 76, 78, 79, 129, 130 Brownian diffusion 103

caesium-137 11, 40–48, 50, 60, 63, 65–68, 109 carbon-14 11, 12 cascade impactors 4, 14, 21–28, 31, 33, 35, 36, 38, 42, 43, 46, 52, 54, 72, 73, 80, 81, 113, 116– 130, 132, 135–139, 148–150, 153–157 Chernobyl accident 42, 45 chlorine-36 11, 12 chlorine-38 11 clusters 3, 6, 8, 9, 17, 18, 36, 86, 88, 90–92, 96–98, 147 coagulation 3, 6–8, 13, 27, 32, 33, 38, 39, 41, 42, 48, 59, 80, 82 coarse particle mode 6–8, 39, 81, 98 coarse particles 3–5, 8, 81, 95–97, 116, 151 cobalt-60 46, 47, 50, 60, 72 compact multistage cascade impactor 153–155 condensation 3, 6–9, 13, 15–17, 19, 23, 27, 30, 32, 33, 38, 39, 41, 44, 46, 51, 82, 89, 95, 113, 114, 120, 142–148 condensation, heterogeneous 8, 9 condensation nuclei 3, 6, 16, 17, 19, 51, 89, 113, 120, 142, 144, 145 condensation nuclei counter, CNC 51, 89, 113, 114, 120, 144, 145 condensation particle counter battery, CPCB 146– 148 cosmic-ray particles 11, 59 cosmogenic radionuclides 11, 12, 14, 27, 60 Cunningham slip correction factor, C 120, 122, 125, 126, 134, 136, 139, 148, 149, 153 deposition dry 12, 61, 67, 71, 151 in rain 16, 79 of particles 15, 59, 61, 102, 106, 137, 150 to foliage crops 68 to soil, ground 41, 61, 66 to surfaces 18, 66, 67

170

Subject Index

velocity 61–63, 68, 102, 103, 106 wet 12, 61, 63, 64, 67, 71 dichotomous sampler 121, 151–153 diffusion battery 37, 51, 120, 135, 141, 143, 144 diffusion coefficient 6, 7, 18–20, 36–38, 90, 144, 164, 167 diffusion, of particles 20, 37 diffusivity 18, 167 distribution Junge 4, 5, 23 log-normal 25, 45, 47, 48, 51–53, 80, 81, 90, 91, 94, 133, 135, 150 mass 4, 5, 7, 8, 23, 25, 27, 28, 32, 38–40, 45, 59, 80, 107, 113, 121, 129, 133, 139, 140, 143, 150, 153 number 4, 5, 86, 87, 95–97, 113, 120, 142, 153 surface 5, 7, 15, 23, 25, 27, 42, 72, 128 volume 5, 7, 8, 14, 23, 27, 28, 33, 38, 73, 105, 113, 121, 123, 129, 150 dose conversion factor 86, 94, 96–99, 101 dose effective 87–89, 94, 101, 109 radiation 33, 85–87, 97, 99, 109, 110 to respiratory tract 88 dust 3, 4, 39, 51, 52, 61, 76, 107, 114, 139 effective dose 87–89, 94, 101, 109 equilibrium equivalent concentration, EEC 109 filters, air sampling 60 fine particles 3, 5, 8, 116, 129, 153 fission product aerosols 41 fission product radionuclides 11, 40, 42–44, 46, 48, 60, 71, 75, 79, 80 fog 3, 7 four-stage low-pressure cascade impactor 127 fume 54 gas-to-particle conversion

8, 46, 59

haze 78 heterogeneous condensation 8, 9 high-volume cascade impactors 26, 28 homogeneous nucleation 8, 147 human respiratory tract 88, 121, 122 impaction, of particles 7, 140 impactors 1 ACFM 23, 24, 35, 36, 38, 46, 72, 73, 120, 121, 123, 124, 130, 150 Andersen type 46, 54, 118, 120–123, 129 Berner-type 120, 150, 151

cascade 4, 14, 21–28, 31, 33, 35, 36, 38, 42, 43, 46, 52, 54, 72, 73, 80, 81, 113, 115–130, 132, 135–139, 148–150, 153–157 compact multistage 153–155 dichotomous 115, 121, 151–153 four stage 128, 138 high volume 14, 23, 27, 33, 72, 123, 128–130 inertial 116, 117, 120–122, 136, 152 K-JIST 5-stage 154, 156, 157 low-pressure 21, 22, 27, 31, 38, 46, 72, 80, 81, 120, 124, 126, 127, 130, 150 Lundgren-type 120 Mercer-style 148, 149 MOUDI 135–140 on-line alpha-cascade 25, 134 particle trap 116, 117, 139–142 virtual 115–117, 121, 151 inertial impactors 116, 117, 120, 122 inhalation, of particles 141 iodine-131 11, 40–46, 50 ions 3, 6, 7, 16, 17, 19–21, 36, 41, 42, 144, 148 large 3, 6, 17 small 3, 6, 7, 16, 17, 19, 21, 36, 148 Junge distribution

4

K-JIST 5-stage cascade impactor krypton-81 11, 12

154, 156, 157

lead-210 11, 13, 14, 16, 25–28, 32, 33, 39, 44, 60, 63, 65, 66, 71–76, 78–80, 82, 124, 129, 130 lead-212 22, 26, 27, 32, 130 lead-214 11, 16, 17, 21–27, 29–33, 35–39, 60, 78, 85, 87, 124, 126, 129, 135 log-normal distribution 23, 47, 51–53, 80, 81, 90, 91, 94, 133, 135 low-pressure cascade impactors 21, 22, 27, 31, 80, 81, 120, 124, 126, 127, 130, 150 Lundgren-type impactor 120 Mercer style impactor 148, 149 micro-orifice cascade impactor 153 mine aerosols 11, 38, 39 mist 3 MOUDI impactor 135, 140 nuclear weapons testing 11, 52, 71, 79 nuclei, Aitken 3, 5–8, 13, 16, 23, 27, 38, 39, 59, 73, 80–82, 96–98, 144, 146 nuclei, condensation 3, 6, 16, 17, 19, 51, 89, 113, 120, 142, 144, 145 nuclei mode 5–7, 13, 17, 39, 59, 80–82, 96, 97, 114 on-line cascade impactors

25, 134

Subject Index particle deposition 59, 61, 137, 150 particle trap impactor 116, 117, 139–142 phosphorus-32 11, 12, 27, 28, 59, 130 phosphorus-33 11, 12, 59 plutonium aerosols 11, 52, 54 plutonium-239 52 polonium-210 11, 16, 22, 25, 26, 39, 71, 72, 75, 76, 78–80 polonium-214 25, 32, 33, 85, 87, 101, 128, 135 polonium-218 7, 11, 16, 17, 29–33, 36, 37, 39, 60, 82, 85, 87, 89, 99, 101, 128 potential alpha energy concentration, PAEC 86, 89–93, 97, 109 radiation dose 85, 86, 97, 99, 109, 110 radiolysis 6, 17 radon-220 1, 16, 24, 25, 36, 60, 99, 129 radon-222 1, 14, 16, 17, 24, 25, 33, 34, 36, 60, 71, 76–79, 97, 99, 101, 109, 110, 129 radon decay products 16, 18–20, 26, 29, 33, 34, 38, 60, 71, 79, 85–88, 91–94, 96–101, 109, 127, 128, 134, 135 recoil-alpha 32, 35 residence time 33, 36, 65, 71, 73–76, 78–80, 82 respiratory tract 85, 86, 88, 102, 105–107, 121, 122 resuspension 33, 52, 59, 60, 66, 67, 103 factor 66, 67 rate 60, 66, 67 Reynolds number 120, 139, 155, 157 risk assessment 106, 108 ruthenium-103 11, 40–46 scavenging coefficient 64 scavenging ratio 64–66 silicon-32 11, 12 size distribution 4–8, 13–15, 23, 25, 27–33, 38–42, 44, 45, 47–49, 51–54, 59, 66, 72, 73, 80, 81, 85, 86, 90–96, 106, 107, 113, 116, 120, 121,

171

123, 127–129, 133, 139, 140, 142, 148, 150, 153 smog 4, 5, 113 smoke 94, 125 sodium-22 11–13, 50, 59, 60 sodium-24 11, 50 sticking probability 8 Stokes number 120, 125, 134, 136, 139, 141, 142, 153, 157 strontium-89 11, 71, 75, 79, 80 strontium-90 11, 71, 75, 79, 80 sulfate aerosols 15, 32, 80, 81 sulfates 4, 6, 17, 26, 27, 42, 59, 80, 130 sulphur-35 11–14, 27, 28, 59, 124, 130 surface area 5–7, 17, 23, 27, 46 TASK group 85 tellurium-132 11, 40, 41, 43, 45, 46 thoron 11, 16, 25, 26, 71, 87–90, 95, 97, 99, 101 transfer, air-to-vegetation 59, 67, 68 trimodal distribution 13 tritium 11, 12 tropospheric aerosols 15, 71, 73–76, 79 unattached fraction 33–38, 86, 88–90, 94, 96, 97, 127, 128 uranium-238 39, 52 uranium mines 35, 99 velocity, deposition 61–63, 68 virtual impactors 116, 117, 121, 151 washout 6, 59, 63–66 ratio 64–66 wire screens 35, 37, 127 working level, WL 110 working level month, WLM 96–99, 109, 110

This page intentionally left blank