Non-Crystalline Chalcogenides (Solid-State Science and Technology Library, Volume 8) (Solid-State Science and Technology Library)

NON-CRYSTALLINE CHALCOGENIDES

SOLID-STATE SCIENCE AND TECHNOLOGY LIBRARY VOLUME 8 Editorial Advisory Board L. R. Carley, Carnegie Mellon University, Pittsburgh, USA G. Declerck, IMEC, Leuven, Belgium F. M. Klaassen, University of Technology, Eindhoven, The Netherlands

Aims and Scope of the Series The aim of this series is to present monographs on semiconductor materials processing and device technology, discussing theory formation and experimental characterization of solid-state devices in relation to their application in electronic systems, their manufacturing, their reliability, and their limitations (fundamental or technology dependent). This area is highly interdisciplinary and embraces the cross-section of physics of condensed matter, materials science and electrical engineering. Undisputedly during the second half of this century world society is rapidly changing owing to the revolutionary impact of new solid-state based concepts. Underlying this spectacular product development is a steady progress in solid-state electronics, an area of applied physics exploiting basic physical concepts established during the first half of this century. Since their invention, transistors of various types and their corresponding integrated circuits (ICs) have been widely exploited covering progress in such areas as microminiaturization, megabit complexity, gigabit speed, accurate data conversion and/or high power applications. In addition, a growing number of devices are being developed exploiting the interaction between electrons and radiation, heat, pressure, etc., preferably by merging with ICs. Possible themes are (sub)micron structures and nanostructures (applying thin layers, multi-layers and multi-dimensional configurations); micro-optic and micro-(electro)mechanical devices; hightemperature superconducting devices; high-speed and high-frequency electronic devices; sensors and actuators; and integrated opto-electronic devices (glass-fibre communications, optical recording and storage, flat-panel displays). The texts will be of a level suitable for graduate students, researchers in the above fields, practitioners, engineers, consultants, etc., with an emphasis on readability, clarity, relevance and applicability.

The titles published in this series are listed at the end of this volume.

Non-Crystalline Chalcogenides by

Mihai A. Popescu National Institute of Materials Physics, Bucharest – Magurele, Romania

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

eBook ISBN: Print ISBN:

0-306-47129-9 0-792-36648-4

©2002 Kluwer Academic Publishers New York, Boston, Dordrecht, London, Moscow Print ©2000 Kluwer Academic Publishers All rights reserved No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher Created in the United States of America Visit Kluwer Online at: and Kluwer's eBookstore at:

http://kluweronline.com http://ebooks.kluweronline.com

V

CONTENTS INTRODUCTION..................................................................................................................1

Chapter 1 THE CHALCOGENS AND THEIR COMBINATIONS 1.1 The Chalcogens................................................................................................................. 4 1.1.1 Sulphur (S)......................................................................................................... 4 1.1.2 Selenium (Se).................................................................................................... 9 1.1.3 Tellurium (Te).................................................................................................. 13 1.2 The General Crystallo-chemical Properties of the Chalcogens.................................... 15 1.3 The Chalcogen Alloys and Compounds....................................................................... 17 1.4 The Binary Compounds and the Alloys of Chalcogens with Pnictide Elements.......... 21 1.4.1 The phosphorus chalcogenides....................................................................... 21 1.4.2 The arsenic chalcogenides................................................................................ 24 1.4.3 The antimony chalcogenides............................................................................. 29 1.5 The Binary Compounds and Alloys of Chalcogens with Tetrahedral Elements ......... 31 1.5.1 The silicon chalcogenides................................................................................. 31 1.5.2 The germanium chalcogenides.......................................................................... 34 1.5.3 The tin chalcogenides....................................................................................... 39 1.6 Other Binary Chalcogenides........................................................................................ 41 1.6.1 The chalcogenides based on heavy elements.................................................... 4 1 1.6.2 The chalcogenides based on light elements ..................................................... 42 1.6.3 Alkali-chalcogenide systems............................................................................. 43 1.6.4 Halo-chalcogenide systems............................................................................... 44 1.7 The Ternary Chalcogenides......................................................................................... 44 1.7.1 The chalcogenide systems based on antimony, germanium and tin ................. 44 1.7.2 The phospho-chalcogenide systems ................................................................. 53 1.7.3 The chalcogenide systems with silicon............................................................. 56 1.7.4 The oxy-chalcogenide glasses........................................................................... 60 1.7.5 The halo-chalcogenide systems........................................................................ 61 1.7.6 The chalcogenide systems with metals............................................................. 65 1.7.7 Other ternary chalcogenide systems.................................................................. 86 1.8 Quaternary Chalcogenide Glasses and more Complex Glasses ................................... 89

Chapter 2 PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENS AND BINARY CHALCOGENIDE GLASSES 2.1 General properties...................................................................................................... 103 2.1.1 Metastability of glass ..................................................................................... 103 2.1.2 Preparation techniques ................................................................................... 108 2.1.3 Impurities and purification.............................................................................. 110 2.1.4 Mechanical properties.................................................................................... 111

vi 2.1.5 Thermal and thermo-electrical properties....................................................... 114 2.1.6 Electrical conduction...................................................................................... 115 2.1.7 Optical absorption ......................................................................................... 117 2.1.8 Dielectric properties ...................................................................................... 119 2.1.9 Magnetic susceptibility .................................................................................. 120 2.1.10 The influence of water adsorption .................................................................. 122 2.1.11 The additivity method for the determination of the properties of complex glasses............................................................................................................. 123 2.1.12 Survey of the general properties of the non-crystalline chalcogenides........... 126 2.2 Properties of Non-crystalline Chalcogens................................................................. 127 2.2.1 Sulphur........................................................................................................... 127 2.2.2 Selenium......................................................................................................... 128 2.2.3 Tellurium........................................................................................................ 134 2.2.4 The chalcogen-chalcogen systems ................................................................. 136 2.3 Properties of Non-crystalline Binary Chalcogenides................................................ 146 2.3.1 Phosphorus-chalcogen.................................................................................... 146 2.3.2 Arsenic-chalcogen .......................................................................................... 152 2.3.3 Antimony-chalcogen (Sb-Ch)......................................................................... 171 2.3.4 Silicon-chalcogen (Si-Ch)............................................................................... 176 2.3.5 Germanium-Chalcogen (Ge-Ch)..................................................................... 181 2.3.6 Tin-Chalcogen (Sn-Ch).................................................................................. 193 2.3.7 Thallium-Chalcogen....................................................................................... 193 2.3.8 Alkali-Chalcogenides...................................................................................... 195 2.3.9 Bismuth-Chalcogen (Bi-Ch)......................................................................... 195 2.3.10 Halogen-Chalcogen........................................................................................ 196 2.3.11 Heavy metal-Chalcogen ................................................................................ 197 2.3.12 Silver-Chalcogen (Ag-Ch)............................................................................. 199 2.3.13 Silver-selenium (Ag-Se)................................................................................. 199 2.3.14 Silver-(S,Se)................................................................................................... 199 2.3.15 Silver-(S,Te)................................................................................................... 199 2.3.16 Oxygen-Chalcogen (O-Ch)............................................................................. 199 2.3.17 Other binary chalcogenides ............................................................................ 201 Chapter 3 MODIFICATIONS INDUCED IN NON-CRYSTALLINE CHALCOGENIDES 3.1 Modifications Induced by Light................................................................................. 209 3.1.1 Irreversible modifications............................................................................... 211 3.1.2 Reversible modifications................................................................................ 224 3.1.3 Models for the photo-inducedprocesses........................................................ 242 3.1.4 Models for the photo-induced anisotropy....................................................... 255 3.1.5 Exotic photo-induced modifications............................................................... 264 3.2 Modifications Induced by other Electromagnetic Radiations in

Chalcogenide Glasses.................................................................................... 3.2.1 Modifications induced by ultraviolet radiations ............................................ 3.2.2 Modifications induced by X-rays................................................................... 3.2.3 Modifications induced by gamma rays..........................................................

266 267 269 270

vii

3.3 Modifications Induced by Irradiation with Particle Beams........................................ 272 3.3.1 Modifications induced by electron beams....................................................... 272 3.3.2 Modifications induced by ion beams.............................................................. 277 3.4 Modifications Induced by Electric and Magnetic Fields............................................ 277 3.5 Mechanically Induced Modifications......................................................................... 280 3.5.1 Modifications induced by pressure ................................................................ 280 3.5.2 Anisotropy induced by mechanical deformations .......................................... 282 3.6 Effects Induced by Ultrasounds................................................................................ 282 3.7 Thermally Induced Modifications.............................................................................. 283

Chapter 4 APPLICATIONS 4.1 Ovonic Devices.......................................................................................................... 293 4.1.1 Phase-change switches.................................................................................... 293 4.1.2 Hybrid amorphous/crystalline transistors....................................................... 299 4.1.3 Thermo-switches ............................................................................................ 301 4.2 Xerography and X-Ray Radiography...................................................................... 302 4.3 Holography................................................................................................................ 308 4.4 Photo-recording......................................................................................................... 315 4.4.1 Photo-thermal recording................................................................................ 315 4.4.2 Materials for recording with bubbles............................................................. 320 4.4.3 Thermo-plastic recording............................................................................... 321 4.4.4 Materials for photo- and electron-beam recording........................................ 323 4.5 Photolithography........................................................................................................ 328 4.6 TV Pick-up Tubes (Vidicon)..................................................................................... 333 4.7 Radiation Sensors and Measuring Devices................................................................ 338 4.7.1 Detectors of radiation and radiation amplifiers.............................................. 338 4.7.2 Radiometric Devices....................................................................................... 341 4.7.3 Solid state integrators .................................................................................... 345 4.8 Acousto-optical Devices............................................................................................ 347 4.9 Chemical Sensors and other Devices......................................................................... 348 4.9.1 Chemical remote sensors................................................................................ 348 4.9.2 Ion-sensitive electrodes................................................................................. 348 4.9.3 Electrical resistance elements......................................................................... 349 4.9.4 Optical fibre sensors....................................................................................... 349 4.10 Solid State Batteries................................................................................................... 350 4.11 Infrared Optical Media............................................................................................... 350 4.12 Components for Integrated Optics............................................................................. 352 4.12.1 Optical fibres.................................................................................................. 352 4.12.2 Planar waveguides ......................................................................................... 354 4.12.3 Kinoform optical elements............................................................................. 356 4.12.4 Chalcogenide materials in optoelectronics........................................................... 358 4.13 Applications of Heavy metal – Chalcogen Compounds and Alloys................................. 360

Subject Index..................................................................................................................... 368

This page intentionally left blank

1

INTRODUCTION The earliest experimental data on an oxygen-free glass have been published by Schulz-Sellack in 1870 [1]. Later on, in 1902, Wood [2], as well as Meier in 1910 [3], carried out the first researches on the optical properties of vitreous selenium. The interest in the glasses that exhibit transparency in the infrared region of the optical spectrum rose at the beginning of the twentieth century. Firstly were investigated the heavy metal oxides and the transparency limit was extended from (the case of the classical oxide glasses) up to wavelength. In order to extend this limit above the scientists tried the chemical compositions based on the elements of the sixth group of the Periodic Table, the chalcogens: sulphur, selenium and tellurium. The systematic research in the field of glasses based on chalcogens, called chalcogenide glasses, started at the middle of our century. In 1950 Frerichs [4] investigated the glass and published the paper: “New optical glasses transparent in infrared up to 12 . Several years later he started the study of the selenium glass and prepared several binary glasses with sulphur [5]. Glaze and co-workers [6] developed in 1957 the first method for the preparation of the glass at the industrial scale, while Winter-Klein [7] published reports on numerous chalcogenides prepared in the vitreous state. As compared to the oxide glasses, the mechanical strength and the thermal stability of the chalcogenide glasses are significantly lower, while the thermal expansion coefficient and the temperature coefficient of the refractive index are higher. The range of infrared transparency, controlled by higher atom masses and lower bonding force constants in the chalcogenide combinations is essentially broadened towards higher wavelengths (lower wave numbers). The decrease of the average bonding energy in solids, which is related to the increase of the mass of the atoms in the homologous rows of the Periodic Table has as a consequence the narrowing of the hole band in their electronic band structure. By thermal activation of the charge carriers a not too high electrical conductivity is obtained. Many glasses are known, which exhibit an electrical conductivity of at 298 K. The chalcogenide glasses are situated close to the oxide glasses with electronic conduction and form a new group of glasses with semiconducting properties. The first research program in the field of chalcogenide glasses was initiated in 1955 by two competing groups of physicists and chemists from Sankt-Petersburg: the group from “A.F. Joffe” Physico-Technical Institute, led by B.T. Kolomiets and N.A. Goryunova, who discovered the first semiconducting glass, [8], in the system and the group from the University of Sankt-Petersburg (former Leningrad University) led Prof. R.L. Myuller [9]. Later, in 1968, S.R. Ovshinsky and his co-workers [10] from Energy Conversion Devices (Troy, Michigan, USA),

2

discovered the memory and switching effects exhibited by some chalcogenide glasses. The unusual electrical effects in chalcogenides stimulated the development of the theory and applications of chalcogenide materials. From that time, new phenomena, properties and effects have been discovered in the non-crystalline chalcogenides. In parallel, the chalcogenides have been applied in xerography, holography, computer memories, radiometry and optical transmission of the information. The recent discovery of induced and of stable anisotropy, as well as the discovery of the optical non-linear properties of several chalcogenide materials, opened new horizons for optoelectronic applications. Although the work in the field of non-crystalline chalcogenides, along more than forty years, resulted in a huge amount of theoretical and experimental results, there are few reviews and books aiming to cover both the scientific and technological aspects of these advanced materials. A comprehensive work on the theory and properties of the non-crystalline materials including the chalcogenide ones has been published in 1971 by Sir N.F. Mott (Nobel Prize Winner for Physics, 1977) and E.A. Davis (“Electronic Processes in the NonCrystalline Materials”, Oxford University Press, Second Edition, 1979). A book which includes a chapter related to the chalcogenides was published in 1976 by K..K. Schwarts and co-workers in Soviet Union (“Optical Recording Media”). In the next years, there were published several books that treat the amorphous solids, in general, and the chalcogenide solids in particular: “The Chemistry of Glasses” (1982, A. Paul), “The Physics of Amorphous Solids” (1983, R. Zallen), “Physics of Amorphous Materials” (1983, S.R. Elliott), “Amorphous and Glassy Inorganic Solids” (1983, A. Feltz). The first book entirely dedicated to chalcogenide glasses, entitled “Chalcogenide Semiconducting Glasses”, was published in 1983 by Z.U. Borisova, the former co-worker of Prof. R.L. Myuller. One year later, G.Z. Vinogradova, from the Institute for General and Inorganic Chemistry in Moscow published the book: “Glass Formation and Phase Equilibrium in Chalcogenide Systems”. In Moldova, Academician A.M. Andriesh and co-workers published in 1988 a book on chalcogenides with special emphasis to applications: “Glassy Semiconductors in Photoelectric Systems for Optical Recording of Information”. The book entitled: “The Physics and Applications of Amorphous Semiconductors” published in 1988 by A. Madan and M.P. Shaw in USA is only partly dedicated to chalcogenide glasses. However, a concise and organised textbook on non-crystalline materials and applications is still desired as a valuable resource for graduate students and workers in the field. The present book aims to give a large, detailed account on the physical and technological aspects of chalcogenide systems including the applications. The basic information on materials, with special emphasis on preparation, the extent of the glassy domain and the crystal-chemistry of chalcogenides, is covered in chapter 1. The general physico-chemical properties of the chalcogenide glasses with a special look on the chalcogens and binary glasses are discussed in chapter 2. The most important properties of the chalcogenide glasses are those related to the effects induced by various external factors, in particular light and other radiations. These effects are of great interest in applications and are treated in detail, both from experimental and theoretical point of view, in chapter 3. Finally, chapter 4 is dedicated to applications, a continuously expanding field.

3

The topics include the author’s original works as well as his recent results, some of them not yet published, on the structure and properties of non-crystalline chalcogenide materials. The book is intended for use as a reference and research book for graduate students, scientists and engineers working in the field of materials science and device physics.

REFERENCES [1]. C.Shultz-Sellack, Ann. Phys., 139, 182 (1870). [2]. R.W. Wood, Phil. Mag., 3, 607 (1902). [3]. W. Meier, Ann. Phys., 31, 1017(1910). [4]. R. Frerichs, Phys. Rev., 78, 643 (1950). [5]. R. Frerichs, J. Opt. Soc. Amer., 43, 1153 (1953). [6]. F.W. Glaze, D.H. Blackburn, J.S. Osmalov, D. Hubbard, M.H. Black, J. Res. Nat. Bur. Standards, 59, 83 (1957). [7] A. Winter-Klein, Verres et Refractaires, 9, 147 (1955). [8]. N.A. Goriunova, B.T. Kolomiets, Zhurnal Tekhnicheskoi Fiziki (russ.), 25, 2069 (1955). [9]. R.L. Myuller, Zhurnal Prikhladnoi Himii (russ.), 28, 363, and 1077 (1955). [10]. S.R. Ovshinsky, Phys. Rev. Lett., 21, 1450 (1968); J. Non-Cryst. Solids, 2, 99 (1970).

4

CHAPTER 1

THE CHALCOGENS AND THEIR COMBINATIONS

1.1. The Chalcogens The chalcogen elements belong to the VI-A subgroup of the Periodic Table. These elements are: sulphur, selenium and tellurium. The VI-A subgroup contains also the oxygen and polonium. The chalcogens are the basic elements of the chalcogenides compounds. The chalcogenides are compounds of sulphur, selenium or tellurium with electropositive elements or with organic radicals. The name chalcogenide originates from Greek: and being given initially to the chalcogenide minerals that contain copper in combination with sulphur, selenium and tellurium. 1.1.1. SULPHUR (S) Sulphur was known from the old times of the humanity. Four thousand years ago Homer mentioned this element and later the Bible described "its power". For many centuries, sulphur was used for paintings, for whitening the linen, in cosmetics and medicine (about 1600 year b.C.). Berzelius discovered sulphur as chemical element in 1817. In 1822, Mitcherlich discovered the allotropy of sulphur. Sulphur has the order number Z=16 and its atomic mass is 32.064. On the scale of natural abundance, sulphur is situated at the position 15. In the Earth core the sulphur concentration is at. % and in the planetary ocean is at. %. The sulphur appears both as native element and in the compounds as e.g.: sulphides (Sb2S3 antimonite, - argentite, FeAsS - arsenopyrite, PbS - galena, CoAsS - cobaltine, HgS - cinnabar, - pyrite, - chalcopyrite, ZnS - würtzite or zinc blende, etc.). Much sulphur can be found in carbon, bituminous schist, oil, thermal springs and volcanic gases. The natural sulphur contains four stable isotopes: 32S(95.02%),33S(0.75%), 34 S(4.21%) and 36S(0.02%). The valence electronic shell of sulphur has 6 electrons with the disposal

The sulphur, can be found in the following oxidation states: -2, 0, +2, +3, +4, +5, +6. It is a typical non-metal. The maximum co-ordination number of sulphur is 6 and the most stable state is the hybrid state

THE CHALCOGENS AND THEIR COMBINATIONS

5

The sulphur exhibits several crystalline and non-crystalline forms. In order to explain the atomic scale structure in the solid and liquid phase we must to take into account some peculiarities of the chemical bonding. Sulphur forms di-covalent bonds. It has two unpaired p electrons and can form -type bonds. The p orbitals are oriented reciprocally at 90° angles. The angle between sulphur bonds is 105°. This value is very near to the characteristic angle for the hybridisation. Starting from these bonds it is possible to define two distinct positions in the series of four bonded atoms: the cis or eclipsed position and the trans or staggered position (figure 1.1)

Figure 1.1. The eclipsed (cis) and staggered (trans) configuration of the chalcogen bonds.

The bonding in the configuration cis leads to the formation of ring molecules

and the bonding in the configuration trans leads to the formation of chain-like molecules. The angles between the planes defined by the atoms of a given configuration are called dihedral angles. These angles are situated in the interval 90-100°. If only -bonds should exist then a random rotation angle around a common bond should be expected. The special situation of two types of configurations appears due to the contribution of the -bonds between the p-electron pairs on neighbouring atoms. The ring (crown) molecules give the most stable structural configuration in the solid state. Other molecules as e.g. and also long chains of atoms can be packed in the solid state of sulphur. The Crystalline Forms of Sulphur

A. Orthorhombic S The orthorhombic structure of sulphur is stable up to the temperature of 95.6 °C. The space group is and the unit cell contains 128 atoms configured in 8-fold rings (Figure 1.2). In the rings the atoms exhibit the trans configuration with the dihedral angles of 102°. The bond angle is 105°. The lattice parameters were determined with highest accuracy by Cooper et al. [1] at 24.8 °C: a = (10.4646 0.0001) Å, b = (12.8660 0.0001) Å; c = (24.4860 0.0003) Å; The average distance between two bonded sulphur atoms is (2.048 0.002) Å.

6

CHAPTER 1

Figure 1.2. The unit cell of the orthorhombic sulphur a the ring packing b, c. front view and side view of the

ring.

B. Monoclinic Sulphur The orthorhombic sulphur transforms enantiomorphically and irreversibly in monoclinic sulphur at 95.6 °C and is stable up to the melting point (119.25 °C). The transformation takes place only under slow heating regime. The space group of the monoclinic sulphur is and the unit cell contains six 8-fold rings. The lattice parameters have been determined by Burwell [2] at the temperature of 103 °C: a = 10.90 Å; b = 10.96 Å; c = 11.02 Å; C. Rhombohedral Sulphur

The rhombohedral modification of sulphur has been

reported as early as 1891 by Engel [3]. The rhombohedral sulphur is unstable at room

temperature and transforms into plastic sulphur, which gradually changes into the orthorhombic sulphur. The rhombohedral sulphur can be obtained by crystallisation from the saturated solution of sulphur in toluene. The probable space group is and the unit cell contains 36 sulphur atoms grouped in six-fold rings. The lattice parameters are: a = 6.45 Å; or a = 10.818 Å and c = 4.280 Å in the hexagonal reference frame [5]. The length of the sulphur-sulphur bond is 2.0 Å and S-S-S bond angles are around 100°. In the six-fold sulphur rings the bond angles are 104° and the dihedral angles are 71° (Fig. 1.3).

THE CHALCOGENS AND THEIR COMBINATIONS

7

Figure 1.3. The packing of the sulphur atoms in the unit cell of the rhombohedral structure. (The picture is valid, also, for rhombohedral selenium). a. View of the ring packing along the [001] axis. b. The six-fold ring (the distances are for selenium).

D. This is a special monoclinic structure obtained by crystallisation of sulphur dissolved in some solvents (e.g. in alcohols). The crystallites are needle-like. is metastable and slowly transforms into orthorhombic sulphur. The space group of is The unit cell contains 32 atoms grouped in four 8-fold rings (Fig. 1.4). The lattice parameters are [6]: a = 8.57 Å; b = 13.05 A: c = 8.23 Å;

Figure 1. 4. The unit cell of the monoclinic The packing of the rings viewed: a. along the axis c b. along the axis b.

E. Orthorhombic Sulphur This form of sulphur has been discovered in 1966 [7] in the powder obtained during the reaction of and (x+y=12) and up to day no complete structural studies have been conducted. There were reported orthorhombic crystals with a = 4.730 Å, b = 9.104 Å, c = 14.574 Å. It seems that this unstable form contains two 12-fold sulphur rings in the unit cell.

8

CHAPTER 1

F. Fibrous Sulphur The fibrous sulphur is obtained by the application of a stretching force to the plastic, non-crystalline sulphur. The space group seems to be The monoclinic cell contains 112 atoms and the lattice parameters are: a=26.4 Å; b=9.26 Å; c=12.32 Å; The density is 2.01 g/cm3. It is supposed that the fibrous sulphur is in fact a superposition of two forms of sulphur: one soluble form would be and the other one would be the so-called sulphur which consists from helical micelles with the periodicity of 10 sulphur atoms for 3 complete turns [9]. Sulphur is characterised by a dihedral angle of 85° and a bond angle of 106° [10]. The fibrous sulphur transforms slowly into the orthorhombic sulphur. The total number of allotropes of sulphur reported in literature is close to 50 but very few of them have been satisfactorily characterised. An excellent review of these allotropes is given in [11,12].

The Amorphous States of Sulphur During quenching the orthorhombic sulphur does not succeed to transform into monoclinic sulphur and melts at 112.8 °C as a viscous yellow liquid called sulphur or cycle-octasulphur. The liquid sulphur exhibits an anomalous dependency on temperature of the viscosity. By heating the liquid, its viscosity firstly decreases (down to poise*s at 155 °C), further increases and then the liquid colour becomes darkbrown. At about 187 °C the sulphur viscosity reaches 93.3 poise’s and, thereafter, gradually decreases so that at 300 °C the sulphur becomes again a soft, fluid mass. The liquid sulphur boils at 444.60 °C and its viscosity at the boiling temperature reaches poise*s. These phenomena can be explained by the dissociation of the sulphur molecules in the melt and the competitive recombination of the atoms. This lead to an equilibrium for the concentration of various kinds of rings. The fraction of rings is 90% mol., as deduced from the decrease of the melting point of the monoclinic sulphur by slow heating. The rest of the melt consists in 6-fold rings and other bigger cyclic molecules up to The macromolecule rings do not appear at low temperatures due to sulphur tendency to be bonded in cis (eclipsed) configuration. Nevertheless, when the temperature is raised, the probability for the bonding in trans configuration increases and at the critical temperature of 160°C this configuration becomes dominant. As a consequence, the viscosity of the melt strongly increases and this leads to the parallel orientation of the molecules. This orientation favours the formation of long, closed chains in equilibrium with the small rings. At high temperatures the sulphur melt contains only about 40% small molecular rings. The remnant part consists mainly in very large macromolecule rings (up to 5000 atoms). It was definitely established that liquid sulphur contains three main components: -sulphur, -sulphur and -sulphur. When the liquid sulphur is further heated the content increases. The viscosity is higher than that of because by heating the liquid becomes viscous. The component is at variance with the component as concerning the rate of dissolution in transforms spontaneously in The following equilibrium equation can be written:

THE CHALCOGENS AND THEIR COMBINATIONS

9

The researches have shown that both and are based mainly on ring configurations while contains long chains. would contain, additionally, a few sixfold rings, which explain the difference in solubility between and . The rings open at ~ 160 C. The analysis of the radial distribution curves indicated that in liquid and amorphous sulphur the molecules exhibit a tendency towards ordering based on lattice fragments with orthorhombic structure. The molecules interact by weak chemical bonds [13]. It is quite remarkable that both sulphur rings and chains play an important role for the amorphous state. The rings act as plastifiers in the process of stretching the amorphous plastic sulphur. Moreover, they prevent the closer approaching of the chains one to another. On the other hand the chains prevent the approaching and the ordering of the sulphur rings. Thus the amorphous, disordered state is stabilised. 1.1.2. SELENIUM (Se)

Selenium is the element with Z=34 in the Periodic Table, and its atomic mass is 78.96. In the Earth shell its abundance is at.%. The native selenium is very rare. Usually, selenium is found in sulphide minerals (more than 40 minerals), for example: CuSe, ZnSe, HgSe, PbSe, CdSe, etc.. The natural selenium consists of 6 stable isotopes: Se (0.98 %); Se(9.02 %); Se (7.58%); Se (23.52%); Se (49.82 %); Se(9.19%). The configuration of the valence electrons of selenium is The oxidation states of selenium are -2; 0; +2; +4; +6. The hybridisation is less stable than in sulphur. The Crystalline States of Selenium

The particular features of the crystalline states of selenium are based on the tendency of the selenium atom to have the trans configuration more expressed than sulphur atom. All the crystalline forms of selenium show configurations with two first order neighbours situated at distances between 2.32 ,and 2.40 Å. A. Hexagonal (grey or metallic) Selenium This is the most stable form of selenium. The lattice is made from parallel helical chains (Fig. 1.5). Every atom has two neighbours in its own chain situated at Å and four neighbours situated in neighbouring chains at The radius of the helix is 0.984 Å. Within the selenium chains the atoms are bonded by covalent bonds and between chains act molecular forces of the Van der Waals type with a weak metallic component. The hexagonal unit cell contains three atoms. The lattice parameters measured at the temperature of 18 °C are [14,15]: a = 4.3544 Å, c = 4.9496 Å, c/a = 1.1367. The bond angle is 103.1 0.2 ° and the torsion (dihedral) angle is 100.7 0.1°.

10

CHAPTER 1

Figure 1.5. The hexagonal selenium a. chain configuration in the unit cell, b. the atom chain (view along the c-axis),

B. Monoclinic Selenium This structure is obtained by slow evaporation of a saturated solution of selenium in carbon disulphide. Usually the monoclinic is accompanied by monoclinic The space group of is: and the lattice parameters are: a = 9.054 Å; b = 9.083 Å, c = 11.601 Å and consists of 8-atom rings (Figure 1.6a). The selenium atoms are situated in the corners of two superposed squares, rotated by 45° and shifted one to another so that the Se ring is no planar. In the ring every atom has 4 neighbours situated at a mean distance of 3.7 Å. differs from by the type of ring packing. The Se bond angle is 105.7 1.6 ° and the average torsion angle is 101.3 3.2°. In this type of Se the bond distance is significantly shorter than in metallic selenium.

C. Monoclinic Selenium This form exhibits prismatic, transparent, dark-red crystals. The unit cell (Figure 1.6 b) has the parameters: a = 12.58 Å, b = 8.07 Å, c = 9.31 Å, (space group ). Both monoclinic forms are obtained by heating of the amorphous selenium. By slow heating of the amorphous Selenium films with the thickness 600-800 Å, at 35-40 °C does appear the crystalline form and around 65 °C does appear [17,18]. D. Rhombohedral Selenium. This is a relatively recently discovered form of selenium [19]. The selenium crystals have been obtained from carbon sulphide solution saturated in selenium. The space group of this selenium is The unit cell contains six-atom rings (see Figure 1.3b). The lattice parameters are: a = 11.362 0.001 Å, c = 4.429 0.008 Å . The rhombohedral Se transforms irreversibly into at temperatures above 105 °C

E. Cubic Selenium. Andrievski et al. [17] have shown that by heating the amorphous layers at 150-160 °C a gradual crystallization occurs and after maintaining at 160 °C a new form does appear: the face centered cubic form with the lattice parameter:

THE CHALCOGENS AND THEIR COMBINATIONS

11

a = 5.755 Å. This form has been called -cubic selenium. The same authors have shown that by local heating of the films one gets an other cubic phase characterized by a = 2.790 Å, called -cubic selenium. All the crystalline phases of selenium transform into hexagonal selenium by heating at 180-220 °C.

Figure 1.6 a. Monoclinic selenium: Iring II - ring packing in the unit cell III – structure projected along the b axis IV - structure projected along the c axis.

The Amorphous Phases of Selenium

The non-crystalline (amorphous and vitreous) selenium is a dark-grey solid. There was suggested that both amorphous and vitreous selenium would be built from disordered chains and rings of di-covalent selenium atoms. The covalent distance, the valence angle, the dihedral angle and the second order distance in non-crystalline selenium seem to be

12

CHAPTER 1

Figure 1.6 b Monoclinic selenium: Iring II - ring packing in the unit cell III - structure projected along the b axis IV - structure projected along the c axis.

very similar to those from hexagonal selenium crystal. The same Van der Waals forces act between the chains. The density deficit of ~ 10% in the amorphous phases suggests that the packing of the structural units be far from close packing. Nabitovitch (see [20]) has studied the amorphous selenium in detail. He concluded that selenium exhibits two amorphous phases: the first one appears during deposition at temperatures around 20 °C and the other one appears at deposition temperatures of

THE CHALCOGENS AND THEIR COMBINATIONS

13

From the structural point of view the transition from the first to the second phase takes place by gradual deformation of the structural elements (8-atom rings) with the simultaneous formation of the chain molecules. Kaplow and Averbach [21] have studied the vitreous selenium and concluded that its structure consists of 95% 8-atoms rings and 5% chain-like, strongly deformed trigonal units. Richter [22] pointed out three non-crystalline forms of selenium he called: Se I, Se II and Se III. Se I is obtained by evaporation and is stable at liquid air temperature. By heating up to room temperature it transforms in Se II. Se III is the structural form obtained by evaporation under large current pulses with a method developed by Weber [23], followed by condensation on substrates. In order to establish firmly the constitution of amorphous selenium (is it built only from chains or consists in a mixture of chains and rings?) the diffraction methods are not enough. Chemical methods for analysis are needed. Brieglieb [24] dissolved the vitreous selenium in and demonstrated on the basis of selective dissolution that this type of selenium is made of a mixture of chains and rings. The proportion between chains and rings are closely dependent on the preparation conditions of the amorphous selenium. The selenium vapours, at 900 °C, consist of molecules with atoms. Between 900 °C and 1000 °C the main part is formed by molecules. Above 1500 °C selenium is totally dissociated in atoms. The association state of selenium atoms in the vapour phase is essential for the quality of the non-crystalline phase deposited on solid substrates. 1.1.3.

TELLURIUM (Te)

Tellurium has the atomic number Z=52 in the Periodic Table, the atomic mass 127.60 and the configuration of the valence shell: It is a hard solid with metallic aspect. The oxidation states in compounds are +2, +4 and +6. The natural tellurium consists in 8 isotopes of mass 120, 122, 126, 128 and 130. The most stables are the isotopes 128 and 130. Tellurium belongs to rare element category and in the earth crust is contained in the proportion of %. Tellurium is fixed, especially, in minerals with Au, Ag, Pt, Cu, Fe, Pb, Bi and in sulphide minerals: (calaverite), AgTe (empresite), (mancheite), (kostovite), (frohbergite), (bilibinskite), (silvanite), (ricardite), PdTe (katulskite), NiSeTe (kitkaite), (ingreite), (tellurite), (joseite), etc... The Crystalline Phases of Tellurium Because the cis configuration is not favoured in tellurium, it exists only one crystalline state of tellurium at normal pressure. This form is called -tellurium, exhibits hexagonal symmetry and is analogous to -(grey) selenium. The crystal structure of tellurium consists in long spiral atom chains parallel to the crystallographic axis c and closely packed. The hexagonal cell has the parameters: a = 4.4570 Å; c = 5.9290 Å. when one passes from -Se structure to -Te structure the

14

CHAPTER 1

atom packing changes a little bit by favouring a Na-Cl type configuration with 6 first order neighbours. The unit cell contains three atoms disposed in a characteristic chain. For the dextrogire crystals the space group is ). Figure 1.7 shows the structure of -Te.

Figure 1.7. a. The structure of -tellurium projected down the c-axis. The bonding distances (solid lines) are 2.834 Å and the interchain distances (dashed lines) are 3.494 Å. b. The same structure shown as a distorted simple cubic structure (after [11]).

In an ideal molecular crystal with chain structure the bonds within the chains are purely covalent while those acting between chains are Van der Waals bonds. Based on geometrical considerations Lucovsky [25] discussed the possibility to describe the chemical bonding in the hexagonal crystals of Se and Te in the frame of an idealised model. In the hexagonal crystal every atom has two first order neighbours situated at the bonding distance and four second order neighbours in the neighbouring chains at the distance In the same chain two neighbours of third order are at the distance The numerical values of these distances and their ratios for Se and Te are given in Table 1.1.

THE CHALCOGENS AND THEIR COMBINATIONS

15

The equal angles between the bonds in Se and Te lead to the same ratio The ratio in Se is higher than in Te and this fact proves weaker bonding between chains in selenium. The closeness to the ideal molecular crystal can be estimated by comparing the ratio to the ratio between the Van der Waals radius Rv and the covalent radius For a typical molecular crystal (e.g. orthorhombic sulphur) the following relation is valid: while for Se and Te this ratio is 0.89 and 0.76, respectively. Moreover, for selenium and while for tellurium and Therefore, the chemical bond in selenium is nearer to a bond characteristic of a molecular ideal crystal than in tellurium. In the same time the bonds within the chains of tellurium are weaker than the pure covalent bond, while in selenium the bonds within the chains are nearer to a pure covalent bond. Two high-pressure forms of tellurium have been observed: -Te, stable between 40 and 70 kbar at room temperature of unknown structure and -Te (rhombohedral) structure stable over ~70 kbar at room temperature. -tellurium is isostructural with -polonium, exhibit the space group and one atom per unit cell at (000) [26]. The lattice constants are a = 3.002 0.015 Å, Each tellurium atom is thus equally distanced to another at 3.002 Å. The structure may be thought as a modification of the normal -tellurium structure to one of higher symmetry. The Non-crystalline State of Tellurium

Tellurium cannot be obtained in glassy state by melt quenching. The amorphous state is obtained by evaporation and deposition on solid substrates maintained at very low temperatures. In the amorphous state the interatomic bond distance (2.80 Å) and the co-ordination number are lower than in the crystal. Stuke [27] suggested that amorphous tellurium should have a distorted chain structure where the interchain bonding is weaker than in the hexagonal tellurium but the bonds within the chains are longer and nearer to the covalent bond.

1.2. The General Crystallo-chemical Properties of the Chalcogens

In the following paragraphs the main properties of the chalcogens are presented according to the data from literature. Table 1.2 shows the crystallo-chemical data on chalcogens both in the crystalline and non-crystalline (amorphous and liquid) states. Table 1.3 presents mainly the chemical, mechanical and physical properties of these materials. Figure 1.8 is a schematically drawing of the transformations occurring in chalcogens. Due to small fluctuations of the parameters which characterise the structural phases, and, also, due to the large scattering of the data in the literature we present in the table 1.2 mainly the averaged values of the structural parameters. For a correct judgement the readers are counselled to consult the original papers indicated in the references.

16

CHAPTER 1

THE CHALCOGENS AND THEIR COMBINATIONS

17

Figure 1.8. The crystallo-chemical transformations of sulphur, selenium and tellurium.

1.3. The Chalcogen Alloys and Compounds The System Se-S

Several compositions in the system Se-S have been prepared and investigated. Calvo et al. [31] have described the compound and also the compositions and crystallise as sulphur isomorphs wherein the sulphur positions are occupied at random by Se and S. The crystalline exhibits the symmetry of -Se. Mixed crystals can exist in a large interval of compositions. Such crystals ( - yellows crystals and - orange crystals) have been prepared and studied by Boudreau and Haedler [32]. The space group of these crystals is and the lattice constants are: a = 8.34 Å, b=13.11 Å, c = 9.30 Å, and a = 8.40 Å, b = 13.26 Å, c = 9.37 Å, respectively. They are isostructural with and their unit cell contains 8-atom rings. These structures can be found up to the equi-atomic composition SeS.

Equi-atomic compositions Se+S have been evaporated at 200 °C in vacuum ( Torr) on substrates heated at 55 °C. There were obtained red colour amorphous films [33] with the composition enriched in selenium Se-S (36at.%S) and Se-S (30 at.% S). The films with the thickness of exhibit softening points between 76 °C and 100° C. At these points a partial melting and crystallisation occurs. The complete melting occurred only at 140 °C. There were obtained also amorphous films with lower sulphur concentration (23 at.% S and 14.5 at.% S) by evaporation from two different sources followed by deposition on solid substrate. Shilo [34] has shown that glasses can be obtained in all the concentration range if the melted mixture Se+S is heated at ~ and, thereafter, quenched in cold water

18

CHAPTER 1

Jecu et al. [28] have prepared several compositions of Se-S as non-crystalline thick films (0.1 mm thickness) by rapid cooling of the melt on a metallic block and between two glass plates: and the eutectic composition The X-ray diffraction measurements and the extended X-ray absorption edge measurements (EXAFS) have shown that the structure of the Se-S alloys can be described as the result of breaking of Se chains by sulphur atoms and the subsequent formation of rings with alternated Se and S [35]. At high sulphur concentrations a partial separation of sulphur takes place and rings of sulphur are produced. For very high sulphur concentrations the compositions seem to exhibit structural configurations with ionic bonds. The System Se-Te Selenium and tellurium form a continuous solid solution consisting in mixed chains of atoms [36]. The analysis of the Mössbauer data suggests that the covalent bond Se-Te is stronger than the mean value of the Se-Se and Te-Te bonds [37]. Therefore the alloys tend to maximise the number of Se-Te bonds. As a consequence, the weak bonds between chains are attenuated. After Grison [36] the lattice parameters c and a show only a small deviation from the Végard law.

The glasses can be obtained by synthesis in evacuated ampoules (at ~500 °C) and quenching in water in the composition range at.% [34]. By spraying the melt in cold water Sarrach et al. [38] succeeded to get glasses up to 50.at % Te.

In the amorphous and liquid states of the system Se-Te the spiral chains dominate up to the concentration of ~ 50 at.% Te. The proportion of rings decreases rapidly with the increase of the tellurium content and with the temperature [39]. Above 50 at.% Te more and more atomic configurations exhibiting the co-ordination number three are formed. In the chains the Se-Te atom pairs are preferred. The System S-Te While the system Se-Te does not exhibit special properties due to the resemblance of the atoms (similar size and properties) the S+Te alloys are very interesting because, on one hand, the tendency to cis or trans bonds are totally different in sulphur and in tellurium and, on the other hand, the size of the sulphur atom is considerably smaller than that of the tellurium atom. Few papers exist on the S-Te system. Surprisingly, at the end of the XIX-th century the chemists were interested in this system. Three types of compounds were described: TeS, and and there were reported that they are stable below -20 °C. The less stable might be Berzelius [40] has shown that Te and S can form homogeneous mixture in all the concentration range. For high sulphur content the alloys are yellowreddish. The rise of tellurium content makes the mixture firstly red and transparent and then, when more Te is added, the alloy becomes dark-red and opaque. It was shown that in Te-S system does not exist a definite compound and usually mixed crystals are obtained. Some of these are isomorphous with the orthorhombic sulphur. The crystals with 0.557 at % Te exhibit the axial ratios: a:b:c=0.8108:l:1.9005. Hawes [41] tried to demonstrate that in the series the only homogeneous phase is Nevertheless, this phase seems to be too unstable to be isolated by re-crystallisation from the melted

THE CHALCOGENS AND THEIR COMBINATIONS

products and decomposes in

19

and Te. In this unstable phase the tellurium atoms enter

into the sulphur atom rings (Te+7S). Sulphur crystals with about 2 wt.% Te have been found in some volcano regions in Japan and, therefore, the supposed combination seems

to be stable. An amorphous solid solution Te-S, not soluble in has been obtained. The melted dissolves around 10 at. % Te and dissolves ~20 at.% Te [42]: In the system Te-S exists an eutectic at the composition [43]. Geller [44] has shown that the melting of the mixture Te+S, without a preliminary thermal annealing, leads to the separation of the elements. Therefore, the preparation procedure must take into account the solid state reaction and this requires the maintenance of the mixture for a long time at high pressure and at temperatures below the melting points of the two elements. Thus, it was possible to prove [44] the formation of the crystalline phase: Geller has shown firstly that the crystal phase cannot be or He brought the samples with these compositions at normal pressure and observed the appearance of a new phase and of an exceeding amount of sulphur. If this phase would contain mixed molecular species, then a large range of solid solutions would be formed. In the TeS sample has been observed the appearance of excess tellurium and, therefore, the crystalline phase was not TeS. Other sample, of composition contained still a small amount of free tellurium. Due to the fact that the Te-S phase induced by pressure is not a valence compound and does not have a solid solution range, it is possible to conclude that no bonds between tellurium and sulphur atoms exist. It results that only the crystallisation of the tellurium chains with the sulphur chains in the ratio 1:1 is possible. If one compares the tellurium chains from the hexagonal phase of Te with the sulphur chains from the fibrous sulphur one observes that a segment with seven Te atoms is equal in length with one segment with 10 sulphur atoms and therefore the lattice constant of the mixed phase is: Å or Å, where , Å is the periodicity of the helical chain of sulphur built along the pseudo-orthorhombic axis (see Figure 1.9a). As a consequence, the chemical formula of the homogeneous crystalline phase must be The modelling of this structure has indicated that the most probable arrangement of the helices corresponds to the structure of fibrous sulphur i.e. left hand

sulphur helix and right hand tellurium helix, both alternating along the orthorhombic baxis. It was shown that the phase has orthorhombic symmetry (space group ) and its lattice parameters are: a = 41.49 Å, b = 31.64 Å, c = 9.25 Å. While the existence of the crystalline compounds in the Te-S system is still subject of controversial discussions, the possibility to get non-crystalline, homogeneous phases is without any doubt. Non-crystalline compositions TeS, and were prepared by quenching the melt between two thick metal plates (splat-cooling technique) [27,45]. The softening temperature is +5 °C for and ~ 10 °C for The glassy samples are not stable: the crystallisation occurs at room temperature, in several days after preparation. Considering the values of the first co-ordination sphere, as resulted from the X-ray diffraction measurements, the authors from [42] advanced the hypothesis that in the amorphous solid are formed mixed chain configurations built by Te or S atoms with statistical alternation.

20

CHAPTER 1

Figure 1.9. The dimensional relation between the sulphur and the tellurium chains. a. chain compared with chain b. ideal packing of the Te and S helices. The circles represent the projection of the Van der Waals cylinders where the helices are inscribed.

The effect of tellurium consists in the breaking of the 8-atom sulphur rings and the formation of Te-S chains whose length is dependent on the tellurium concentration in

material. In sulphur rich compositions the ring configuration seems to predominate. The tellurium participates in the structure with very short chains. For compositions with minor sulphur concentration chain configurations with mixed Te+S composition are formed. Zegbe et al. [46] have performed X-ray, Mössbauer and optical studies in the system. The results provide evidence for the formation of rings for and the formation of copolymer chains up to the limit composition x=0.5 In the tellurium solid solution range (0.85<x<1) sulphur atoms are included in chains (limit composition ). The System S-Se-Te The glassy alloys of the three chalcogens have been obtained by synthesis from elements at 500° C and quenching in cool water [37]. The boundaries of the vitrification domain depend on the cooling regime [50,51]. No ternary, nor eutectic compounds have been evidenced. For the composition takes place the stratification in the melt. Figure 1.10 shows the vitrification diagram for the system.

THE CHALCOGENS AND THEIR COMBINATIONS

21

Figure 1.10. The vitrification domain in the system S-Se-Te. -------- slow cooling [50] -------- quenching

1.4. The Binary Compounds and the Alloys of Chalcogens with Pnictide Elements

In order to facilitate the general discussion of the chalcogenide systems we shall use in this book the notation “Ch” for a non-specified chalcogen.

1.4.1. THE PHOSPHORUS CHALCOGENIDES The System P-S The composition domain for the glass formation in the system P-S is not well established. Specker [49] reported a vitrification domain up to 40 at.% P. Heyder and Linke [50]

obtained glasses with at.% P. The maximum temperature for synthesis was and for these glasses raises with the phosphorus content from 25 °C to 110 °C. Krebs [51] has shown that some compositions, for example cannot be obtained in the vitreous state and explains this fact by the easiness of formation of the molecules and the corresponding crystalline compounds. can be obtained in two polymorphous modifications one of which, is stable up to 40 °C and exhibits rhombic symmetry with a = 9.660 Å, b = 10.597 Å, c = 13.671 Å, and Z = 8 [52]. is a monoclinic compound (space group ) (as well as ) and has the parameters: a = 8.87 Å; b = 17.35 Å; c = 6.85 Å; The compound has the triclinic symmetry with: a = 9.07 Å; b = 9.18 Å; c = 9.19 Å; and Z = 2.

22

CHAPTER 1

exhibits cubic symmetry [53]. Crystals and glasses in the system P-S hydrolyse very easily in humid air. Pearson [54] has indicated the domain of compositions able to form glasses in the range 70 - 100 at % S and Gutenev [55] has prepared glasses of composition:

The System P-Se The formation of the glasses in the system P-Se has been studied by Robinson and Scott [56]. For the synthesis the melt must be heated at 700° C and the glasses are obtained by quenching the melt in air. The compositions with easily form glasses. Above x = 0.52 the glass formation is apparently inhibited by the growth of the molecular crystals whose basic structural unit is shown in figure 1.11a. Nevertheless, glasses are formed also in the region x>0.63 [57]. Zigheli and Orlova [58] have prepared glassy compositions with the phosphorus content 10-40 at.% The glasses with up

Figure 1.11. The system P-Se. a. The structure of the molecule

b. Phase diagram of the system.

THE CHALCOGENS AND THEIR COMBINATIONS

23

to 70 at % phosphorus (i.e. up to the composition ) have been investigated [59] by nuclear magnetic resonance. The phase diagram of the system is given in Figure 1.11b. The compound melts incongruently and, therefore, in the glass network exists several types of structural units with P-Se and P-P bonds characterised by small bond energy difference. The participation of the electrons d in the formation of the P-Se bond proves that this bond is a -bond. The most advanced character is realised in and this feature allows supposing a tetrahedral co-ordination of selenium around phosphorus, as opposite to the trigonal co-ordination in In do appear P-P bonds and their number rises when the composition is shifted towards and then to The molecules and are formed without difficulty in these alloys. In the system P-Se have been obtained amorphous films with the thickness of in a large compositional range, by deposition from vapour phase on solid substrate [60]. The crystalline state is well represented in the system P-Se: (has rhombic symmetry, and is cubical, and and also show rhombic symmetry [61]. In the P-Ch systems do exist chalcogen atoms with double bond and this infers the existence of a new structural entity: with tetrahedral geometry besides the well-known entity. If selenium is added to phosphorus, then in the first stage P-Se bonds appear, the selenium chains become branched and increases. If more and more P is added then the double P=Se bonds will be gradually destroyed. A second polymer built from pyramidal units is now formed and continues to increase. At ~ 40 at % P the glass will consists of polymeric sheets For higher phosphorus concentration, do appear P-P bridges, which link the sheets in a three-dimensional network. The process will be finished at 50 at % P when the glass can be considered as a network of cages which, in average are linked one to another by two P-Se-P bridges. Again the equilibrium is established between the isolated molecules and the polymerised ones and this equilibrium can be shifted towards the molecular composition by thermal treatment or by adding phosphorus. In both cases the glass easily recrystallises. The System P-Te

Although several authors [54] claim that phosphorus reacts which tellurium and forms the phase there was not possible to state clearly the existence of this compound. There was suggested the formation of an unstable compound with the chemical formula PTe. In the phase diagram there was observed a large stratification domain and a monotectic transform at ~500 °C. No data are known on the glass formation domain. In the system PS-Se have been not observed ternary compounds. The vitrification domain is very broad (see figure 1.12) [62,63]. In the system P-Se-Te the glasses with a phosphorus content up to 40 at.% are stable in air but, above this concentration become hygroscopic. The softening temperature, raises with the tellurium concentration up to a maximum value situated around 100° C. During crystallisation, the solid solution Se-Te is separated. The vitrification domain is shown in figure 1.12b [61].

24

CHAPTER 1

Figure 1.12. The glass formation domains in the chalcogenide system with phosphorus. a. The system P-S-Se b. The system P-Se-Te

1.4.2. THE ARSENIC CHALCOGENIDES

The Systems As-S and As-Se The best studied are the compounds In the crystalline state and are isostructural with monoclinic lattice (space group ) (Figure 1.13.). The unit cell has 4 molecules. The structure consists of extended layers of interconnected 12 atom rings. This configuration was proved to be the densest packing possible for the chalcogens atoms linked with arsenic. Every arsenic atom has five valence electrons. Three electrons are used for valence bonds with three neighbouring chalcogens and the other two electrons form non-bonding orbitals. The chalcogen has 6 valence electrons: two are used for bonding with arsenic and the other four electrons form two non-bonding orbitals [64]. As a result, the arsenic atoms show strong covalent bonds with three chalcogens and the chalcogen with three arsenic atoms. The valence state of arsenic is a hybrid state between s + 3p and The electronegativity difference between arsenic and chalcogens corresponds to a maximum value of the bond ionicity of ~ 6%. We are dealing, therefore, with predominantly covalent As-Ch bonds. Due to the difference in the hybridisation of the arsenic electrons, the bonds with chalcogens are not equivalents from the point of view of strength. Therefore, after the Gordy rule [65], the distances between the arsenic atom and the bonded neighbours will be also not equivalents. Between layers act Van der Waals forces with a minor covalent component. The minimum interlayer distance (4.785 Å) exceeds considerably the distance between the first neighbours within the layer [66]. The arsenic chalcogenides could be considered as molecular crystals where the molecules are extended to infinite in two spatial directions. The interaction forces between layers are hundred times weaker than the binding forces within the layers.

THE CHALCOGENS AND THEIR COMBINATIONS

25

Figure 1.13. The structure of orpiment viewed in two projections. (the figures indicate the position of the atoms above the plane of the paper)

Currently, the

structures are described in terms of pyramidal configurations

Thus, in six pyramids are inter-linked by the intermediary of sulphur and form a ring. The arsenic atom is situated in the top of the pyramid while sulphur atoms, which link pairs of neighbouring arsenic, form its base. The bond angles in the pyramid are 99°.

26

CHAPTER 1

As opposite to in exists two kinds of pyramids with different distances to the first order neighbours and different valence angles on arsenic.

The average value of the force constant for the interaction between atoms is in around 25 % lower than in The crystal is less covalent than and this fact leads to the formation of more shifted and more densely packed layers than in The lattice constants of the crystal are: a = 12.053 Å, b = 9.890 Å, c = 4.277 Å and [67]. The arsenic trisulphide belongs to the group of compounds that are hardly obtained in the crystalline state. By cooling the melt one gets a glassy material. Firstly are formed stable clusters with non-crystalline structure. By diminishing the mobility of the atoms and clusters in liquid the long-range order is prevented. The arsenic tri-chalcogenides can be obtained in amorphous state by deposition on solid substrates starting from the evaporation or sputtering of bulk targets. The tendency toward the glassy state is more evidenced in than in In the systems As-S and As-Se several compounds in the crystalline state are known besides and The compounds and were described but up today no reliable structural data are available. is known as a mineral under the name of realgar. It crystallises in the monoclinic system (space group ) with the lattice parameters a = 9.325 Å, b = 13.671 Å, c = 6.587 Å, and Z = 4 (Fig. 1.14).

Figure 1.14. The structure of

(realgar).

In the realgar structure the molecules are individualised and their conformation (cradle - like) is very near to that of the molecules in gas phase. The arsenic atoms occupy the corners of a tetrahedron and the sulphur atoms are disposed in a square, which cuts the middle of the tetrahedron. The As-As distances are 2.57 Å and the distances As-S are situated between 2.23 and 2.26 Å. The S-As-S angles are 95° while the As-As-S and As-S-As angles are ~ 100°. The Van der Waals intermolecular forces are acting for distances As-S not exceeding 3.85 Å.

THE CHALCOGENS AND THEIR COMBINATIONS

The compound

27

is known as mineral under the name of duranusite. It exhibits a

layered crystalline structure related to that of arsenic. Its elementary cell is orthorhombic

with the lattice parameters: a=3.576 Å, b=6.759 Å, c=10.074 Å [68]. Although and are isostructural, it is remarkable that is very stable in the glassy state while cannot be obtained in the glassy state even by very rapid quenching; only rapid quenching accompanied by high pressure can determine the formation of glass. The domains of glass formation for As-S and As-Se are shown in figure 1.15 and will be discussed in the section dedicated to systems:

Figure 1.15. The vitrification domains in the system

28

CHAPTER 1

The System As-Te In this system the crystalline compound, a, corresponds to the stoichiometric composition The crystals are monoclinic (space group ) with the lattice parameters: a = 14.339 Å, b = 9.873 Å, c = 4.006 Å and At high pressures and high temperatures is formed a different structure, described by Jakusev and Kirkinskii [69]. From the structural point of view is a bridge between the orpiment structure with cations having trigonal-pyramidal co-ordination and the tetradymite structure with octahedrally co-ordinated cations. The co-ordination polyhedra form double chains "coated" by the remaining arsenic atoms (Figure 1.16).

Figure 1.16. The structure of

(viewed along the double chain of octahedra).

The arrangement of the atoms in the

lattice is, grosso modo, the same to

those observed in structures with dense packing of spheres. The neighbourhood of the tellurium atoms is a distorted cubical arrangement of atoms. The compound has metallic-type bonds as proved by the average bond distance As-Te that is nearer to the sum of the ionic radii than of the covalent radii proper to the other compounds of arsenic [70].

X-ray diffraction investigations of the As-Te binary system carried out by several authors (see [71 ]) determined X-ray reflection sets, which did not belong to As, Te, or monoclinic To explain these reflections some metastable compounds were proposed: AsTe [72], [73] and the metastable modification, of [74]. An important remark is that all these metastable compounds form from liquid phase under

THE CHALCOGENS AND THEIR COMBINATIONS

29

quenching only, and then transformsby annealing into stable For example, the modification appears when the liquid is quenched with a cooling rate up to a critical cooling rate of the melt, which prevents the crystallisation [75]. The metastable

compounds have a more symmetrical structure compared to centred cubic [72], while is rhombohedral 75].

AsTe is face-

During the cooling of the melt, the compound manifests a weak tendency to form glass but, by deposition from the vapour phase, thin amorphous films are easily obtained. The Mixed Systems

The mixed crystalline compounds and are isomorphous with and are obtained relatively easily in all the range of concentration. The mixed system As-S-Se is characterised by an important domain of vitrification, which is delimited by the chemical compounds: and (Figure 1.15) [76]. In the region situated near the section S-Se exists a glass formation domain where the glasses cannot be crystallised. For the other areas, during crystallisation is separated the arsenic, solid solutions AsS-AsSe, then and solid solutions based on different forms of sulphur and selenium. In the system As-S-Se the vitrification domain is very large. This domain can be further enlarged if, before cooling, the samples are subjected to high pressure (30-90 kbar), then heated to temperatures above 200 °C

and, finally, quenched. Thus, in the binary system the vitreous range can be extended up to 70 at. % As. In the system As-S-Te the glasses are easily obtained by air cooling the melts (simply taking the samples out of the oven at room temperature). The samples with S/As = 6.85 3.50 form glasses up to a concentration of 40 at.% Te. For the ratio S/As ~ 2.53 are obtained glasses up to 2 at.% Te. For S/As>>1.5 the tellurium forms firstly Te-S bonds and then Te-Te bonds, while the S-S bonds still remain. When the ratio S/As 1.5 tellurium prefers the direct Te-Te bonds and for S/As < 1.5 tellurium breaks the As-As bonds and forms As-Te bonds. In the system As-Se-Te the most stable glasses are and During crystallisation are separated Te, and solid solutions In other parts of the vitrification domain are separated the phases: As, and solid solutions based on and In the pseudo-binary system have been obtained glasses with the composition extending up to by quenching (with the rate of 200 °/s) (Figure 1.15) [77]. 1.4.3. THE ANTIMONY CHALCOGENIDES

The Systems Sb-S and Sb-Se The stoichiometric compounds orthorhombic symmetry (space group

and

are isostructural [78] and exhibit ) and Z = 4. The unit cell parameters are:

30

CHAPTER 1

for a= 11.20 Å, b = 11.28 Å, c = 3.83 Å and for c = 3.962 A [79].

a = 11.62 Å, b = 11.77 Å,

The structural configuration consists in double bands extended in the directions of the crystallographic axes a and c (Figure 1.17) along [001 ] direction [80]. Strong covalent forces are acting only within the chains. The chains are held together by weak forces [81]. The chalcogen atoms show the following valence states: the side atoms of the chains: the atoms in the main chains: the atoms in the branched sites: Therefore, the formula of the compound could be written:

Figure 1.17. The structure of stibnite

The

bands are extended along the axis b.

Therefore, one can affirm that the antimony chalcogenides are characterised by polymer bonds and their correct chemical formula is: Repeating equivalent

positions build the crystal: two for the antimony atoms and three for the chalcogen atoms. The two-chalcogen positions are essentially different from the two crystallographic sites of arsenic in An antimony atom is surrounded by chalcogens situated

in the corners of a strongly deformed octahedron. The other atom five chalcogens in a deformed square pyramid.

is surrounded by

The electron saturation of the valence shells is possible only if one electron passes from the three-valence sulphur atom to the single-valence sulphur atom. The increase of

THE CHALCOGENS AND THEIR COMBINATIONS

31

the co-ordination number as compared to the arsenic chalcogenides occurs as a consequence of the increase of p-orbital contribution to the chemical bonds. The chalcogen atom can have up to six neighbours of different type on the account of the nonlocalised p bonds. The different strength of the chemical bonds between Sb and Ch atoms is a major feature of the structure [82]. Only the glass with the composition was prepared in the system Sb-S. Meleh et al. [83] have reported the glassy state of minute samples (1.5-2 g) by cooling the melt from 599.5 °C (melting point) down to 0 °C in water + ice mixture. The material partially crystallises during storing for 8 h at 375 °C. In the system Sb-Se have not been obtained bulk glasses. The non-crystalline states of some compositions have been, nevertheless, easily obtained as thin solid films by deposition from vapours in vacuum on substrates held at room temperature. The system Sb-Te The stoichiometric compound crystallises in the tridymite structure. This is a rhombic lattice, where nine tellurium layers are packed and 2/3 of the octahedral sites are occupied by antimony atoms. The layers can be regarded as packed in groups wherein covalent and covalent-ionic bonds are acting. The bonds between groups are of Van der Waals type. From NMR measurements there was stated that the valence state of antimony is In the vicinity of this ion a high density of electrons exists and, as a consequence, occurs the superposition of the electron orbitals of with those of the chalcogen, which are strongly polarised. cannot be obtained in the glassy state but forms easily amorphous films by vacuum deposition on cold substrates. The thermally treated samples transform into tridymite structure. As to the non-annealed films, one believes that they contain the nonequilibrium phase where the Sb and Te atoms are placed without ordering, and exhibit a high number of bonds between atoms of the same type [84]. Several non-stoichiometric amorphous compositions have been prepared and investigated. The antimony chalcogenides belong to that category of compounds, which don’t form or form with difficulty glasses by solidification from melt. This feature is related to the fact that in its compounds with chalcogens the antimony shows bonds with more metallic character than the arsenic compounds do with chalcogens.

1.5. The Binary Compounds and Alloys of Chalcogens with Tetrahedral Elements 1.5.1. THE SILICON CHALCOGENIDES The Systems Si-S and Si-Se In these systems, two crystalline phases are known: SiCh and The silicon monochalcogenide SiS (SiSe) is usually called silicon poly-sulphur (poly-selenium). The crystals are fibrous [85,86]. In the polymeric structural configuration, the poly-ion -S-Sbinds two silicon atoms from the same layer (Figure 1.18).

32

CHAPTER 1

Figure 1.18. The structure of the silicon monochalcogenides. a, b, c - the three possible configurations in the layers.

From the geometrical point of view it is possible to substitute sulphur by Se or by Te. By substitution the angle Ch-Ch-Si decreases to 109° for S and to 103° for Te while

for the ideal angles on silicon, the Ch-Ch-Si angle must be 90°. The other crystalline phase is The crystals show rhombic symmetry (space group ) Z = 4 and the unit cell parameters are: for [87]:a = 5.60 0.01 Å; b = 5.53 0.01 Å, c = 9.55 0.01 Å for [88]: a = 6.03 0.01 Å; b = 5.76 0.01 Å; c = 9.76 0.01 Å The crystalline configuration is shown in figure 1.19.

Figure 1.19. The basic chain in the

structure.

THE CHALCOGENS AND THEIR COMBINATIONS

33

The figure illustrates the difference between chalcogenides and oxides, and this fact can be explained by a lower ionic component of Si-Ch bond as compared to the Si-O bond. In the the tetrahedra are bonded by edges and form long chains. The Si-S bond has the length of 2.16 Å, a value very near to that for a single bond (2.17 Å). The electron charge on silicon, calculated by the difference of electronegativity is +0.44. The repulsion of the charges leads to a deformation of the tetrahedra so that the bonding edges are shorter (3.32 Å) than the non-bonding edges (3.56 and 3.70 Å) [89]. In the compound [54] the Si-Si distance in neighbouring tetrahedra is 2.926 Å [90]. All Si-Ch compounds hydrolyse in humid atmosphere and form and Ch. In the system Si-S the glasses are prepared with difficulty. The glass is difficult to get because the temperature range for the existence of the melt is very narrow (1090-1130 °C). Glasses have been obtained in the composition interval when the samples were melted under pressure (25-80 kbar) and rapidly cooled [91]. The silicon chalcogenide glasses are much more stable in humid environment than the crystalline compounds. In the system Si-Se have been prepared glasses in the concentration domain 60-100 at.% Se. The structure of these glasses has been explained by Johnson et al. [92] by a model with clusters of composition in the selenium matrix. The nature of the structure changes firstly at the composition and then at the composition The Si-Te system In this system has been described the compound There was shown that it is possible to get compounds of the type and SiTe [54]. crystallises in the tetragonal system (space group Z = 4, a = 7.430 Å; c = 13.482 Å [93]. The details of the structure are not well known. Two structural models have been advanced. In both models, cluster of four silicon atoms form a trigonal pyramid with the Si-Si distance: 2.5-2.6 Å which is a function of the flattening of the pyramid. After Dittmar [94] the structure is a deformed structure of the type with the Si-Si pairs situated in the octahedral voids oriented along the three crystallographic directions (Figure 1.20). The symmetry would be triclinic (space group ), Z = 4 and a = b = 7.428 Å, c= 13.488 Å, After Grigoriadis et al. [95] the structure of is of trigonal symmetry with the space group The units are located at the centres of slightly distorted octahedra. Each tellurium atom is bonded to only two silicon atoms, the bond angles are around 93° and the Si-Te average distance is around 2.55 Å. Thus, an electron lone pair is formed on the Te atom. The system Si-Te forms an eutectic at 85 at.% Te. This eutectic exists also for the system Ge-Te and for the systems Al-Te, Au-Te, Hg-Te. In the system Si-Te the glass domain corresponds to 15-25 at.% Si and 85-75 at.% Te and within these limits are distinguished two composition domains, which have different electrical properties [96]. During glass crystallisation one forms the eutectic and crystals. Bartsch and Just [97] have prepared non-crystalline

34

CHAPTER 1

Figure 1.20. Structural models for The section (101) and the projection along the axis [001] for: a. model with 1/4 of the silicon atoms in the octahedral voids, b. model with all silicon atoms situated in the tetrahedral voids.

compositions and have shown that their atomic scale structure is characterised by silicon neighbouring with Te and While in tellurium the atoms are disposed in unique helices, in amorphous Si-Te occurs an inter-digitation of right and left-hand oriented helices, which are built from silicon and tellurium atoms [98]. 1.5.2. THE GERMANIUM CHALCOGENIDES The Systems Ge-S and Ge-Se The differential analysis and the radio-crystallographic analysis have shown that in both Ge-S and Ge-Se systems exist only two compounds with related structures: GeS (GeSe) and

THE CHALCOGENS AND THEIR COMBINATIONS

35

There was shown that the mono-chalcogenides of silicon are isostructural, with rhombic symmetry (space group ) (Figure 1.21) and Z=4: for GeS: a = 4.301 Å, b = 3.649 Å, c = 10.45 Å [99] for GeSe: a = 4.403 Å, b = 3.852 Å, c = 10.82 Å [100].

Figure 1.21. The structure GeS (GeSe) viewed in projection on the (a, c) plane.

The germanium and the chalcogen atoms form double layers oriented perpendicular to the c-axis of the crystal. The germanium atoms have three nearest neighbours in their own double layer and three more distanced from which two belong to the own double layer and one belongs to the neighbouring layer. The interatomic distances from the co-ordination octahedron are situated in the range 2.47 - 3.00 Å for GeS and 2.56 - 3.37 Å for GeSe. The valence angles in these compounds are respectively 91 - 96° and 96 - 103°. Between the double layers act weak forces, as evidenced by the cleavage surfaces (001). The crystallochemistry of the germanium mono-chalcogenides was explained in the model proposed by Krebs [101]. In this model it is considered the resonance of the trigonal pyramidal bonds between two groups of three atoms belonging to neighbouring double layers. Therefore, it seems to be more correct to imagine these compounds as atom systems with metallic-covalent bonds stabilised by resonance.

36

CHAPTER 1

The germanium dichalcogenides

and

are isostructural and crystallise

in the rhombic system (space group ) with Z=4 and with the lattice parameters: for a = 6.87 Å; b = 11.67 Å; c = 22.38 Å [61] for a = 6.953 Å; b = 12.220 Å; c = 23.04 Å [61]. In the germanium dichalcogenides every germanium atom is surrounded by four chalcogens and every chalcogen by two germanium atoms (Figure 1.22):

Figure 1.22. The structure of viewed along the normal to the atom layers, a. the tetrahedron configuration above and below the layer plane, b. the bonding configuration between the atoms of the upper layer.

The angles between the valence bonds into the slightly deformed tetrahedron are 98°, 104° and 107° (the average value is 103°) and the bonding distances are situated between 2.07 Å and 2.26 Å, much lower than those of the nearest neighbours in GeS. Moreover, the mean inter-atomic distance in the tetrahedron (2.19 Å) is smaller than the sum of the covalent radii (Ge+S) which is 2.25 Å. A similar situation is in GeSe. The glasses in the system Ge-S were obtained in a large range of compositions, but the limits were not clearly established. Hilton et al. [102] have shown that glasses can be obtained for the concentrations of at.% Ge. Glass formation domains in the range 29-38 at.% Ge [103] and 15-30 at.% Ge [104] were also reported. Kawamato and Tsuchihashi [105] have established two glass formation domains: in the range at.% Ge and at.% Ge. and Hrubý [106] reported three glass domains in the range Of course, no exact boundaries can

THE CHALCOGENS AND THEIR COMBINATIONS

37

be obtained. In the intermediary regions, during quenching is obtained a solid, which consists of a glassy matrix where microcrystals are dispersed [107]. In the thin amorphous films deposited from vapours the amorphization domain is not limited. In the system Ge-Se the researches have established the existence of an enough large glass formation domain: at % Ge [108]. It was shown that the application of a rapid quenching regime (quenching in water + ice mixture), for small amounts of melted alloys, has allowed to get glasses up to the composition [109]. Nevertheless, it was not possible to prepare glassy stoichiometric alloys Later it was shown that in the

system Ge-Se the glass formation domain is situated between and [110]. It is possible that in the domain microcrystallite are included in the glassy matrix. Recently, Senapati and Varshneya [111] published a ternary phase diagram for Ge-Sb-Se, which shows that the glass-formation region in the Ge-Se system is interrupted in the interval In the Ge-Ch systems the glassy compositions are much more extended when thin films are prepared. There were prepared and investigated Ge-Se films out of the glass formation domain [112].

As regarding the mixed chalcogenides, the chemists tried to prepare glasses in the pseudo-binary system [113]. They remarked a deep crystallisation tendency in the vicinity of the composition By analysing the structural features, they

concluded that the compositions from this system are nearer to than to This is similar to the case of the glasses in the system During crystallisation of the glasses in the system a significant decrease of the electrical resistivity was produced [114]. The solid solutions in these systems are supposed to consist of ordered clusters with average size of several nanometers, having a layer structure [115]. The system Ge-Te In the system Ge-Te has been obtained only one crystalline compound: GeTe. There was established the existence of two crystalline forms: the high temperature form (above 400 °C) and the low temperature form (down to room temperature). The low temperature form, -GeTe, exhibits rhombohedral symmetry, corresponds to a distorted NaCl-type lattice (Figure 1.23) The high temperature form ( -GeTe) has the NaCl structure. The -GeTe structure is built by double layers where the upper half is occupied by Ge atoms and the bottom half is occupied by Te atoms [116]. By heating the -GeTe crystals the rhombohedral angle gradually increases up to 90° at around 400 °C, where the stabilisation of the NaCl type modification takes place. The homogeneity domain of the system Ge-Te is situated between 50.2 and 50.9 at.% Te. In the interval at.% Te a new crystalline low tempe-rature form does appear and this form was called -GeTe. The unit cell of -GeTe and that of the other Ge chalcogenides exhibits the same symmetry [117]. As concerning the glassy state, the researches have shown that the system Ge-Te exhibits one main composition domain where glasses can be easily prepared [118]. This

38

CHAPTER 1

domain is situated around the eutectic composition ranges the glass formation is much hindered.

In other concentration

Figure 1.23. Chalcogenide layers derived from the NaCl structure,

a. the structure of GeTe and As b. the structure of SnS and black P.

In the glasses of the system Ge-Te the covalent bonds predominate. As opposite to crystalline GeTe, where ionic-covalent bonds are formed, the glassy composition GeTe is characterised by covalent bonds. A similar situation has been encountered in GeS (GeSe).

Figure I 24. The glass-formation domains in the system Ge-Te-Ch.

THE CHALCOGENS AND THEIR COMBINATIONS

39

In the mixed chalcogen systems, many preparations of glasses were carried out. In the Ge-S-Te system it was found a large vitrification region. No definite ternary compound has been observed. The glass formation domain is shown in Figure 1.24. For Ge-Te it extends from to [119]. In the system Ge-Se-Te the vitrification domain is limited to the composition with more than 5 at.% Se and an additional region is extended from up to (Figure 1.24) [120]. 1.5.3. THE TIN CHALCOGENIDES The Systems Sn-S and Sn-Se In these systems two crystalline compounds are known: SnCh and The structure SnCh is isomorphous with the structure GeCh and can be looked as a NaCl-type deformed structure. Nevertheless, as a consequence of the differences between the ionic radii of Ge and Sn and also due to the more expressed p-character of their chemical bonds, the level of deformation and the type of the deformation are somewhat different. The effect of the hybridisation of the s and p electrons in the tin mono-chalchogenides is small as proved by the bond angles near to 90° and also by the theoretical calculations of the valence configuration of tin in its divalent state. SnS (SnSe) crystallises in the orthorhombic system (space group [121] (Figure 1.25a) with the lattice parameters: SnS: a = 4.349 Å, b = 3.988 Å, c = 11.202 Å [122] SnSe: a = 4.46 Å, b = 4.19 Å, c = 11.57 Å [110] The ratios of the lattice parameters (a/c; b/c; c/c) illustrate the modification of the chemical bond, the diminishing of the s-p hybridisation and thus the approaching to the NaCl lattice: NaCl: 0.353; 0.333; 1 GeS: 0.412; 0.349; 1 SnS: 0.388; 0.356; 1 The tin dichalcogenides hexagonal system (space group

GeSe: 0.408; 0.355; 1 SnSe: 0.386; 0.353; 1 and

are isostructurals and crystallise in the (Figure 1.25b) with the parameters:

SnS2: a = 3.648 Å; b = 5.899 Å; c/a = 1.617 [123] SnSe2: a = 3.799 Å; b = 6.131 Å; c/a = 1.614 [124] Similar to the case of the germanium dichalchogenides we can imagine the structure of SnCh2 as consisting from structural units The ionisation energy of the Sn-S and Sn-Se bonds of the tetrahedra is somewhat lower than the ionisation energy of the As-S (As-Se) bonds in the pyramids. One can, therefore, conclude that the larger contribution of the p-orbital to the bond formation between tin and chalcogen

40

CHAPTER 1

determines the decrease of the covalent component of the metal chalcogen bond within tetrahedra. As a confirmation, the inter-atomic distances in the tin tetrahedra

Figure 1.25. The structure of the tin chalcogenides. a. the structure of SnS viewed in projection parallel to the plane of the atom layers b. the (110) section of the structure.

are closer to the sum of the ionic radii (Sn+Ch) than the sum of the covalent radii, as opposite to the case of the germanium tetrahedra. The tin dichalcogenide structure is close to the structure while the germanium dichalcogenides crystallize in the strongly deformed structure. It is expected that many physical properties of the tin dichalcogenides be different from those of the germanium dichalcogenides. The existence of the compound formed by peritectic reaction at [125] was proved by using DTA measurements. Nevertheless, it was shown by gamma nuclear resonance that the Mössbauer spectrum of is a superposition of two spectra: that of SnSe and that of On the basis of the experimental data it was concluded that compound exists only as a metastable structure in thin solid films [127]. Much work has been carried out in order to get glasses in the systems Sn-S and Sn-Se. It is well known that the glass formation tendency is low in materials with low per cent of covalent bond. A confirmation of this fact is given by the germanium monochalcogenides where the glass formation is still controversial. As opposite to the bulk glasses, the non-crystalline Sn-Ch alloys are easily obtained as thin solid films by deposition from vapour phase on various substrates [127].

THE CHALCOGENS AND THEIR COMBINATIONS

41

The Sn-Te System In this system has been obtained only one compound in the crystalline state: SnTe. SnTe crystallises in the NaCl-type lattice (space group At 5 K the structure becomes orthorhombic with the parameters a = 6.325 Å, and at 16 K has been reported an orthorhombic structure with a = 6.274 Å, b = 6.288 Å, c = 6.309 Å [128]. The interatomic distances are 3.15 Å while the sum of the covalent radii of the two types of atoms is 2.77 Å and the sum of the ionic radii is 3.33 Å. At 77 K takes place the transition towards the rhombohedral, low temperature form, while for GeTe a similar transition is observed at 673 K [129]. The investigation of the structural transformation at high pressures has shown that the cubic structure of SnTe (and also of the SnTe-GeTe alloys) transforms at 18 kbar in the orthorhombic GeS-type structure [130]. Based on the above considerations we can conclude the tin and germanium monotellurides must have very similar physical properties. Finally, it is important to remark that, due to the metallic-like properties of SnTe the absence of glasses in the system Sn-Te receives a simple explanation. 1.6. Other Binary Chalcogenides

1.6.1. THE CHALCOGENIDES BASED ON HEAVY ELEMENTS

The Systems Tl-Ch In the system thallium-sulphur (Tl-S) the phase diagram suggests the existence of the compounds TlS and crystallises in the hexagonal system, exhibits a monoclinic structure and is characterised by chains of tetrahedra linked by ions. TlS and have tetragonal symmetry. In the TlS structure no direct Tl-Tl bonds exist. The Tl atoms occupy two different sites with the co-ordination 4 and 6. Tetrahedral chains linked through edges are extended along the axis c. In the system Tl-S the glass formation domain is situated between TlS and (Figure 1.26) [131].

Figure 1.26. The glass formation regions in the systems Tl-S and Tl-Se.

42

CHAPTER 1

In the system Tl-Se two compounds have been described: and TlSe, which are and T1S. The glass formation domain is situated in the range TlSeIf one compares the glass formation domains in the systems Te-S and Tl-Se on the basis of the structural data, one remarks a common feature: the development of tetrahedral chains for the compositions rich in chalcogens. In Tl-Se, as opposite to Tl-S, it exists a larger region of glass formation due to the existence and stability both of chains and rings of selenium. In Tl-Se system occurs an interaction between the polymeric selenium chains and the tetrahedra while in the system Tl-S there are sulphur rings which, due to their low stability in the non-crystalline configuration, lead, in the presence of the thallium atoms, to liquid-like structures around the composition (by liquid-like structures we understand those structures without short covalent Tl-S bonds). In the system Tl-Te have been found the compounds TlTe and Te. Glassy compositions of [132] have been obtained by vacuum synthesis from elements (amount: 4g) at 900 °C, cooling in air down to the state of viscous flow and final quenching in water. Such glasses crystallise rapidly at 20 °C. isostructural with

The Systems Ag-Ch In the system Ag-S only the compound is known. No glasses have been obtained. In the system Ag-Se has been obtained the compound which crystallises in two stable forms: the rhombic structure (low temperature form) and the cubic structure (high temperature form). Glasses have been obtained by synthesis at 1000 °C (2h), cooling at 500 °C and then quenching in a mixture water + ice + NaCl. Thus, it was possible to introduce in glass up to 10 at.% Ag. At the maximum amount of silver, the softening temperature is only 105 °C [133]. No glasses in the system Ag-Te have been obtained. The Systems Re-Ch The rhenium chalcogenides and are isostructural. They crystallise in the triclinic system. The crystallisation takes place in the deformed -type structure [134]. Traore and Brenet [135] has shown that the compound Re2S7 crystallises into a structure characterized by Z=5. In the non-crystalline state have been obtained the compositions and by chemical way, as powder [136]. There was shown that in the non-crystalline state, too, the alloys with sulphur and selenium have similar structures.

1.6.2. THE CHALCOGENIDES BASED ON LIGHT ELEMENTS The Systems B-Ch

In the system B-S are known the compounds [137] and [138]. The preparation of or it is a very difficult task because the crystalline boron exhibits a very poor reactivity with sulphur (much better results are obtained with amorphous boron) [139], melts at 567 °C. This compound has been obtained in the glassy state by cooling the melt from 700-800 °C by simply switching off the furnace power [140]. The glasses are

THE CHALCOGENS AND THEIR COMBINATIONS

43

hygroscopic. No studies on the extension of the glass formation region have been carried out up to day. In the system B-Se has been obtained and studied the compound The glass formation domain extends from Se to (for higher boron concentration no investigations were performed) [141]. The glasses have been obtained by slowly cooling the melts: they are, nevertheless, unstable and easily hydrolyse in air. In the B-Ch systems B-Te glasses were no reported. The system Te-O Glasses of stoichiometric composition were prepared. Telluria, consists of network in which Te is fourfold co-ordinated and oxygen is twofold co-ordinated. Trigonal bypiramid units with a lone pair in the equatorial plane are the building blocks of respective crystalline and glassy networks [142,143]. 1.6.3. ALKALI-CHALCOGENIDE SYSTEMS

The first data on the glass formation in the alkali-chalcogenide systems were published in 1982 for the case of Cs-Se binary system [144]. Minaev [146] has calculated the ability to form glasses in the alkali metal-chalcogen systems on the basis of an original theory of glass function: the so-called "energostructural theory". The theory comprises a structural model of the glass and allows for drawing conclusions on the relation between various chemical-bonding forces, which act, into the network. For calculations have been used the phase diagrams which furnish information on the energy of the system at the transition liquid-solid. It is considered that the relation between the chemical bond energy of the atoms and the energy related to the melting at the liquidus temperature is a convenient parameter for the characterisation of the glass formation ability in the system metal-chalcogen.

44

CHAPTER 1

The vitrification domain in the system containing sulphur increases with the atomic number of the alkali metal (Table 1.4). In the system alkali metal - selenium the glass formation domains are shifted to higher chalcogen concentration. In the systems with tellurium and for the cooling rates exceeding 180 K/s, glasses are obtained by alloying with all the alkali metals. In Cs-Te system the glass is obtained with difficulty. 1.6.4. HALO-CHALCOGENIDE SYSTEMS.

The system Br-Se Glasses are formed for at.% Se. The transition temperature is situated between 70 and 90°C [145]. The system Br-Te Glasses are formed in the range 58-68 at.% Te [145].

1.7. The Ternary Chalcogenides 1.7.1. THE CHALCOGENIDE SYSTEMS BASED ON ANTIMONY, GERMANIUM AND TIN. The Systems As-Sb-Ch

The well-known crystalline compound in this system is AsSbS3. It exhibits monoclinic symmetry (space group with Z = 8 and the lattice parameters of the natural crystall are: a = 11.85 Å, b = 8.99 Å, c = 10.16 Å, = 116.45° [128]. For the crystals grown in laboratory the following parameters have been determined: a = 11.875 Å, b = 9.015 Å, c = 10.194 Å, = 116.37° [147]. The structure consists of deformed rings linked by sulphur bridges, thus forming sheets parallel to the axis [001] (Figure 1.27). The co-ordinations of As, Sb and S correspond to those proper for and bonds with small s contribution. The metallic atoms are three fold co-ordinated by sulphur and the units form trigonal pyramids with the bond angles situated in the range 84.1°-102.8°. The pseudo binary system has been prepared and studied in glassy state. The system exhibits the tendency to separate the components in liquid state. By melting the mixture of the two components at 900 °C and then quenching the melt in cold water, were obtained glasses in the composition range 0.15 1.0 [148]. The whole domain of glass formation in the system As-Sb-S has been established by Kawamoto and Tsuchihashi [149] (Figure 1.28). In the pseudo-binary system the glassy phase was obtained by melting the binary components at K for hours followed by air quenching. The glassy compositions thus obtained are situated in the range 0.6 1.0 [150].

THE CHALCOGENS AND THEIR COMBINATIONS

45

The system As-Sb-Se has been also studied by Sawan et al. [151]. When antimony is added to the basic structure is not drastically changed because Sb is an element analogous to As. Antimony raises the metallic character of the chemical bond in the system. The glass formation domain (figure 1.28) shows that Sb enters into glassy up to 19 at.%. The largest domain appears in alloys with the As/Se ratio:

Figure 1.27. The structure of

Figure 1.28. The glass formation domain in the system As-Sb-Ch.

46

CHAPTER 1

During crystallisation of the compositions there was observed the appearance of the compound. As regarding the As-Sb-Te domain, after our knowledge, no studies on the glass phase formation were reported. The Systems Ge-As-Ch In the systems Ge-As-Ch is known the crystalline compound GeAsSe [128]. The crystal symmetry is orthorhombic. The solid belongs to the class of Mooser-Pearson phases and

exhibits layered structure [152]. For the saturation of the bonds in this structure direct As-As bonding occurs. The phase diagram analysis in the systems with S and Se for the glassy domains has proved the formation of the ternary compounds GeAsS and GeAsSe [153]. There was established that, in these compounds, the As atoms are surrounded by Ge atoms and the Ge atoms are linked only to chalcogens. No direct bonds As-Ch exist. The glass formation region for the system Ge-As-S has been reported by Myuller et al. [154] (Figure 1.29).

Figure 1.29. The glass formation domain in the system Ge-As-Ch.

For low sulphur concentration the alloys exhibit a strong tendency toward crystallisation. For high As content, it is possible to appear As crystallites. Very easily one gets glasses with low sulphur and arsenic content, practically in any synthesis regime. These glasses are transparent in the visible region of the spectrum and show a red-orange colour. In the sulphur glasses the difference between the values of the bonding energies for Ge-S and As-S is 7 kcal/mol, higher than that for Ge-Se and As-Se in the selenium glasses (~4kcal/mol) and, therefore, for low chalcogen content (especially for alloys with sulphur), it is expected to be formed firstly structural units (or and (or and only thereafter, will appear structural units with weaker bonds: (or and (or Thus, for an enough large content of sulphur, takes place the interaction of sulphur with Ge and As and the formation of structural units

THE CHALCOGENS AND THEIR COMBINATIONS

47

(or and (or with an energy gain of 100 kcal/mol and 66 kcal/mol, respectively. As a consequence, glasses will be easily formed for an excess of 60-90 mol. % of one of the reaction products or For a content of 40-50 mol.% As and therefore for the case of the equilibrium of these structural units, a homogeneous glass will be obtained with difficulty. The tetrahedral bonds of germanium stabilise the vitreous state even for a significant sulphur excess due to the structural units The alloys with the As content exceeding 45 at.% cannot be obtained in the glassy state. No glasses can be obtained for the alloys with low content of structural units (below 50 %) and this is a proof for the glass formation property of these units [155]. Andreichin et al. [156] have studied in detail the glassy system -Ge and have shown that the structural units postulated on thermodynamic basis by Myuller et al. [157] and which act as building blocks of the glass network are, in general, confirmed. The same authors [156] improve essentially the compositional formulas of these glasses and point out that between the measured physical parameters and the calculated ones by the molecular additivity rule the agreement is nearly perfect. The sulphur glasses are different from the selenium glasses from the crystallochemical point of view. The sulphur glasses are characterised by a larger tendency towards crystallisation for low sulphur content and by the tendency towards stratification in melt for high sulphur content. The sulphur glasses are more difficult to prepare than the selenium glasses. The vitrification domain in the system Ge-As-Se differs from that of Ge-As-S. It is generally admitted that at the basis of the vitreous Ge-As-Se alloys lay the structural units and linked in a double chain for the case of more Se than Ge atoms but no more than the double of the number of Ge atoms. Thus the direct Ge-Se bonds are statistically less probable. It is also possible to suppose the formation of the structural units where the corresponding Ge atoms would have a pair of non-bonding electrons and, therefore, will exhibit the co-ordination 2 but the vitreous compositions do not exist for less than 25 at.% Ge so that such units would not play an important role. Suchet [158] supposed that the insertion of the third element, arsenic, is realised by the simultaneous substitution of two neighbouring selenium atoms with the formation of a bridge between two double chains:

Then, when the composition corresponds to the association with (i.e. for an average of Se atoms between two successive Ge atoms) the As-As bridges

become less probable within a double chain than between two different double chains, because in that positive they correspond better to a higher mobility in the liquid phase. The glass formation domain of the system is better developed also in the region of the binary system As-Se where do exist structural units and linked in double chains of the type:

48

CHAPTER 1

or

whose formation requires an equal or higher number of Se atoms than of As in order that the frequency of appearance of As-As bonds along a chain be statistically less probable. Suchet supposes that the insertion of the third element, Ge, is made here by Se substitution too, every atom preserving the non-bonding electron pair. There would be really impossible for the germanium to substitute arsenic in the same network of bonds without introducing an unsaturated valence whose presence has never been proved. The system Ge-As-Te is significantly different from the other systems [159]. The glasses in this system belong to the group of memory glasses, used in electronics (the socalled Ovshinsky's glasses). In the system does not appear the ternary compound of the type discussed above, but a large range of solid solutions based on GeTe are formed. The homogeneity domain extends towards We have shown above that the large domain of glass formation [160] is conditioned in the case of the compositions with S and Se by the formation of ternary compounds. In the Ge-As-Te there are two narrow glass formation domains (Figure 1.29), a feature probably related to the existence of two ternary eutectics that drastically decreases the crystallisation rate of the alloys. The alloy with small amount of tellurium substituted by arsenic is the easiest to vitrify. There was shown that this composition corresponds to an eutectic. Such alloys crystallise easily by heating or by intense illumination and two phases are separated: tellurium and GeTe. It is supposed that in these alloys there exist short chains of trigonal tellurium inter-linked by Ge atoms situated at the ends of these chains [161]. Minaev et al. [162] have shown that the highest tendency to form glasses is characteristic to the compositions situated on the crystallisation curves of the binary and ternary eutectic alloys. The glass formation domains in the neighbourhood of these curves must intersect the weakening lines of the binary eutectics. In the system exist two quasi-binary sections: the section GeTewith a large homogeneity domain of the solid solution based on GeTe and the section GeTe - GeAs [163]. The systems Ge-Sb-Ch In the system Ge-Sb-Ch few compounds are known. No Ge-Sb-S ternary compound has been identified. It was shown that during the crystallisation of the glasses of composition Ge-Sb-Se, one gets the compound by peritectic reaction at 463 °C [164]. This

THE CHALCOGENS AND THEIR COMBINATIONS

49

is a compound with f.c.c. cubic structure and Å. In the ternary system with tellurium three compounds have been identified: and The first one has trigonal symmetry (space group z = 1, the structure is layered and the unit cell parameters are: Å, Å or Å, (figure 1.30). The crystals and have the same symmetry (space group z = 1) and the lattice constants are a = 4.201 Å, c = 16.96 Å [165] and a = 4.21 Å, c = 23.65 Å [166], respectively. In the system Ge-Sb-S the glass formation region is not very large. For higher cooling rates (water quenching) the vitrification domain is significantly enlarged than for slow cooling regime (switching-off the furnace power) (figure 1.31a) [167].

Figure 1.30. The projection in the plane (110) of the

structure.

The glass of composition situated on the boundary of the glass formation domain for low quenching rates, has been used in the experiment “Crystal” placed on the spatial complex “Soyuz-Salyut 6”. There was demonstrated that the glass prepared in zero gravitation is more resistant to crystallisation than the glass prepared in normal gravitation conditions and, in the first case, a considerable enlargement of the vitrification region it is expected [168]. In the system Ge-Sb-Se the glass formation region is situated in the neighbourhood of the selenium rich compositions. The glass domain is smaller than that of the compositions with As lacking Sb. The glass domain in Ge-Sb-Se system is larger than the domain for the system with sulphur (Ge-Sb-S). Borisova and Pazin [169] have studied

50

CHAPTER 1

how the vitrification domain changes as a function of the cooling rate from the molten phase (figure 1.29b) [170].

Figure 1.31. The glass formation region in the system Ge-Sb-Ch. a. The subsystem Ge-Sb-S (---- rapid cooling — slow cooling) b. The subsystem Ge-Sb-Se. (1 - slow cooling - air cooling [170]; 4 - water quenching[170]) c. The general diagram Ge-Sb-Ch (1 ---- Ge-Sb-S ; 2 — Ge-Sb-Se ; 3 Ge-Sb-Te).

The glass-forming region in the Ge-Sb-Se system has been also studied by Afifi et al. [171]. The well-expressed glass-forming tendency in this system has been recently discussed topologically in terms of the chemical bands expected to be present in these materials by Fouad [172]. Using simple considerations based on co-ordination numbers (m) and bond energies, the average number of near neighbours of each type

expected to surround an atom has been estimated. The average number of bonds has been used to estimate the cohesive energies of these glasses assuming simple additive rule of

bond energies.

THE CHALCOGENS AND THEIR COMBINATIONS

51

According to Thorpe [173], in the range of glass formation composition, the system should contain rigid and floppy regions. The composition, corresponds to an average co-ordination m = 2.4. This represents a percolation threshold at which a transition from floppy polymeric glass to a rigid amorphous solid takes place. This means that the chemical stability of the system is optimum in the vicinity of the mechanical percolation threshold: m = 2.4. In the system Ge-Sb-Te the glass formation domain is very narrow [174]. The glass has the glass transition temperature and the crystallisation temperature 229 °C [175]. The comparison between the vitreous regions for the three chalcogen-alloying element is shown in figure 1.31c. It is considered that the Ge-Sb-Ch glasses are built by polymeric chains based on chalcogen atoms interconnected with Ge and/or Sb. In the systems Ge-Sb-Ch is confirmed the general feature of the glass formation in the alloys with the chalcogens: when the atomic number of the chalcogen increases, the ability to form glasses diminishes. The systems Sn-As-Ch The formation of the glasses in the system Sn-As-S has been studied by Goryunova et al. [176,177]. The glasses have been obtained by vacuum synthesis. The maximum temperature of the mixture of chemical element was 900 °C and the melts were quenched in water with the cooling rate of

By this procedure it was possible to introduce in the glassy state of this system up to at.% Sn (figure 1.32). In the system Sn-As-Se, the tin has been introduced in glassy alloys up to at.% after [176,177]. Shkolnikov [178] succeeded to introduce at.% Sn, and Borisova [179] has shown that the maximum ability to form glasses is exhibited by the alloy, where it is possible to introduce up to 11 at.% Sn (see figure 1.32). It was established that in the system takes place a micro-stratification for the tin content exceeding 3-5 at.%. Shutov [180] have introduced 0.1 - 3.5 at.% tin in and 1-10 at.% tin in AsSe. The synthesis was conducted in evacuated quartz ampoules at the temperature of 1100 °C followed by slow cooling with the furnace. Seifert and Frischat [181] have shown that the glasses Sn-As-Se are inhomogeneous at the microscopic scale even if they are formed at high cooling rates (above 100 K/s). It is supposed that in glass takes place the segregation of a super-saturated component with high crystallisation tendency. It is admitted that the glasses in this system do not form a continuous non-crystalline network but they are built by different structural groupings as there were established for other chalcogenide glasses [156], In the system are formed the configurations and The Se-Se bond has a strong “s” component and this feature explains the existence of the hybridisation. Thus, one can understand why in these glasses the structural modification does not exist, and why does not precipitate during crystallisation. In the system Sn-As-Te few papers were published on the glass-formation ability. The glass-formation region is very narrow and corresponds to a small amount of tin in the alloy [182] (figure 1.32).

52

CHAPTER 1

Figure 1.32 The glass-formation region in the system Sn-As-Ch.

The systems Sn-Sb-Ch Few data were reported on the crystalline state of these systems. In the system Sn-As-Te three compounds exist, which are isomorphous with the compounds where germanium substitutes tin: and The lattice constants of the first two compounds are a = 4.37 Å, c = 23.79 Å and a = 4.37 Å and c = 17.35 Å, respectively [183]. For the last compound have been reported the lattice parameters: a = 4.294 Å, c = 41.548 Å (space group [128,184]. After our knowledge no glasses have been prepared in these systems. The systems Ge-Sn-Ch Taking into account the observation that the binary chalcogenides with Sn and with Ge show remarkable common features it is interesting to see if the GeS component can be substituted by SnS component in the composition and if GeSe can be substituted by SnSe in the composition To this purpose, several investigations have been carried out. Feltz [185] has shown that in the system SnS-GeSthe substitution series is situated out of the glass formation domain. In the system Ge-Sn-S it is known only one crystalline compound, it is a monoclinic crystal (space group z = 4 and its lattice parameters are: a = 7.269 Å, b = 10.220 Å, c = 6.873 Å, [186]. In the system Ge-Sn-Te, it is known only one crystalline compound, of chemical formula [128]. In the system Ge-Sn-S Rowland et al. [187] have prepared and studied glassy compositions with Sn (up to 14-17 at.% Sn) and sulphur concentration situated near the eutectic composition (58 at.% S) (figure 1.33).

THE CHALCOGENS AND THEIR COMBINATIONS

53

Figure 1.33. The glass formation domains in the systems Ge-Sn-Ch.

In the glass it was possible to introduce up to 20 at.% Sn. During crystallisation are formed and GeS crystals. In the case of glasses situated in the zone rich in during crystallisation is formed the high temperature modification of In the system Ge-Sn-Se the glass domain is situated at low tin concentrations [188] (Figure 1.33). By quenching the melts from 950 °C in air it is possible to get glasses with a maximum content of 13 at.% Sn. The most stable glasses are those with selenium concentration of 70-80 at.%. In the system Ge-Sn-Te Minaev [189] has shown that the glass-formation region is situated in the zone: Sn (0 3 at.%), Ge (13 20 at.%), Te (80 85 at.%) (Figure 1.33). The evolution of the properties in the system Ge-Sn-Se-Te as a function of concentration of various chalcogenide components suggests that SnSe and SnTe behave as network modifiers in the glass based on the interconnection of the units. The glass-formation domain in the complex system Ge-Sn-Se-Te has been studied by Feltz [190]. Starting from the glasses GeSethe glassy domain extends in the direction of the eutectic GeTe-Te (15 at.% Ge). The maximum Sn concentration in these glasses is ~ 12at.%. 1.7.2. THE PHOSPHO-CHALCOGENIDE SYSTEMS

The Systems P-As-Ch In the system P-As-S it is known only one crystalline compound synthesised as early as 1898 [191]. The formula is and the symmetry is hexagonal: a = 12.85 Å; c = 7.25 Å. The crystal is built from quasi-molecular symmetric units [192] which consist of two subunits, interconnected by a P-S-P bridge and of two subunits, linked to two subunits by As-S-P bridges (Figure 1.34).

54

CHAPTER 1

Tverianovich et al. [193] have studied the system section and have shown that two crystalline phases do appear and The crystallisation of was not observed. The compositions which contain up to 10 at.% are crystalline.

Figure 1.34. Structural configurations in the system P-As-Ch

Blachnik and Hoppe [194] have determined the vitrification domain when the melts are cooled in air. The domain corresponds to a sulphur content of 0 - 36 at.% for the ternary composition - S and the maximum reaches 191 °C. The glasses in the system P-As-S hydrolyse easily in humid atmosphere. For the system P-As-Se the glass-formation region has been determined by Goryunova et al. [176] (Figure 1.35). The glassy compositions are extended in the domain 0-57 at.% P for the section - Se where reaches 184 °C. In the section glasses are formed in all the compositional range, even for slow cooling rates of the melts. In the system P-As-Te, after our knowledge, no glassy compositions have been prepared.

Figure 1.35. The glass formation region in the system P-As-Se [176].

THE CHALCOGENS AND THEIR COMBINATIONS

55

The Systems P-Sb-Ch The glass-formation domain for the subsystem P-Sb-Se has been studied by Linke et al. [195] (Figure 1.36). The material synthesis was performed at 700-800 °C for 6-10 h.

Figure 1.36. The glass formation domain in the P-Sb-Se subsystem.

and the melt quenching was carried out in air or in water (for the glasses situated at the boundaries of the vitrification domain). The glasses with up to 10 at.% P are stable in air. Above 25 at.% P strong hydrolysis takes place due to the humidity and to the elimination of a very poisonous gas. By crystallising the glasses from this subsystem there were identified Se, and two compounds not yet confirmed: and The Systems P-Ge-Ch The formation of the glasses in these systems has been studied by Hilton et al. [102] and by Vinogradova et al. [197] (Figure 1.37). The vitrification domains are given for the air cooling of the melt starting from 900 0C and for different material mass (10g [91] ; 3-5g [182]). High differences between the results obtained by different authors are probably caused by the fact that some authors (e.g. Hilton et al. [102]) have not considered the composition change of the alloy as a consequence of the outgoing of the white phosphorus from the ampoule when the ampoule was open. This is a reason why Hilton et al. reported in the system P-Ge-Te a larger amount of phosphorus (up to 30 at.%) in the glassy composition, while Vinogrodova et al. [198] have reported that phosphorus enters into the glassy phase up to 1-2 at.%. The last authors carried out more precise estimation, by analysing the gas in the ampoule during melt expulsion. They found that the gas from ampoule was pure phosphorus gas and its amount has been rigorously determined [199]. In the compositions P-Ge-Ch the largest glass domain was found in the subsystem P-Ge-S and the narrowest one in the P-Ge-Te subsystem. The regions of the glass for-

CHAPTER 1

56

Figure 1.37. The glass formation domain in the systems P - (Ag, Ge) – Ch

mation are narrower in the system with phosphorus than in the systems with arsenic or antimony. This is due to the fact that germanium (and, also, silicon) forms compounds with As but not with P nor with Sb. Ge and Si have preference for the chalcogens for chemical bonding. The P-Ge-Se glasses are stable against acid or alkaline water solutions[196]. The Systems P-Sn-Ch

The formation of the glasses in the system P-Sn-S has been demonstrated by Serioghin et al. [185]. They succeeded to get in the glassy state the alloys and In the subsystem P-Sn-Se the same authors [200] have obtained only one vitreous alloy, of composition 1.7.3. THE CHALCOGENIDE SYSTEMS WITH SILICON In general, the tendency to form glasses in a given system decreases when the atomic mass of the elements increases. Therefore, the glasses containing silicon would exhibit higher ability to produce glasses than those containing germanium would. Usually, the glasses with silicon are obtained by cooling the melts in normal atmosphere.

THE CHALCOGENS AND THEIR COMBINATIONS

57

The Systems Si-P-Ch Hilton et al. [102,201] has studied the glass formation domains in these systems. The glasses have been obtained by vacuum synthesis (mass: 10 g) at 1000 °C for h, followed by air quenching. Unfortunately, the phosphorous analysis has been not carried

out and, due to its well known variability, the glass diagram is of low confidence. The approximated diagram can be seen in figure 1.38.

Figure 1.38. The glass formation region in the subsystem Si-P-Te (after [102]).

The Systems Si-As-Ch

Shkolnikov [202] has reported that in the glassy composition can be introduced up to 20 at.% silicon. The material was synthesised at 1000 °C for 100 hours. The cooling of the melt was carried out in slow cooling regime down to 700 °C and then the material was quenched in air. The glass hygroscopicity increases with the silicon content. The glasses with above 20 at.% silicon easily hydrolyse in air and give rise to The glass formation in the subsystem Si-As-Te is illustrated in figure 1.39 [203]. In the diagram is evidenced a large area of glass formation which can be related to the appearance of some compounds in the three binary fundamental sections [204]. The transition temperature of the glasses decreases when the tellurium content increases. in the subsystems Si-As-Te and Ge-As-Te have identical values for compositions with identical atom ratios. The highest are situated around the composition on the line SiTe-As. De Neufville [205] has proposed for these glasses a structural model where every atom grouping has an average of 5 electrons (as in the case of arsenic) and direct Si-As bonds are formed.

58

CHAPTER 1

Figure 1.39. The glass-formation region in the system Si-As-Te.

The Systems Si-Sb-Ch In these systems the glass formation area is narrower than in the systems with arsenic and diminishes significantly in the series sulphur selenium tellurium (Figure 1.40).

Figure 1.40. The glass formation domains in the systems Si-Sb-Ch.

The glasses Si-Sb-S have been reported by Hilton et al. [102,201] and by Pearson [54] by synthesis at 1000 °C (for and for a mass of 10 g) and by cooling in air. The glasses are unstable in normal atmosphere and hydrolyse easily in humid air with the elimination of The maximum is 280 °C. The same authors have prepared glasses in the system Si-Sb-Se in identical conditions. The sections and limit the glass domain. These glasses, too, are unstable in air and easily hydrolyse. In the subsystem Si-Sb-Te the glassy domain is situated within the limits: Si: at.%; Sb: at.%; Te: at.% [182] (Figure 1.40).

THE CHALCOGENS AND THEIR COMBINATIONS

59

The Systems Si-Ge-Ch The preparation of glasses in the system Ge-Si-S is very difficult. Stepanek and Hrubý [206,207] have determined the vitrification domains of these glasses. In the region I cannot be obtained glasses. In the region II (up to 8 at.% Si) the glasses can be obtained only for high cooling rates in small diameter ampoules. These glasses are characterised by very high internal stresses. In the region III (up to 4 at.% Si) the glasses are easily formed. By comparing the systems Si-Ge-S and Ge-S one concludes that silicon does not facilitate the glass formation but, on the contrary, reduces the mechanical and chemical stability of the glass. No glasses in the System Si-Ge-Se are known. In the subsystem Si-Ge-Te have been found two vitrification regions [189] (Figure 1. 41): I - Si(7 25 at.%), Te(75 89 at.%), Ge(0 9 at.%) and II- Si(0 9 at.%), Ge(11 21 at.%), Te(75 85 at.%).

Figure 1. 41. Glass formation domain in the system Si-Ge-Ch, Si-Sn-Te and for Si-Ge-Te: 1 - Si (7-25 at.%), Te (75-89 at.%), Se (0-9 at. %); 2 - Ge (0-9 at.%), Ge (11-21 at.%), Te (75-85 at.%). I, II, III Ge-Si-S (see text).

The Systems Si - Sn - Ch In the system Si-Sn-Te exists only one vitrification domain extended within the limits: Si (12 15 at.%), Sn (0 9 at.%), Te (75 88 at.%), (Figure 1. 41). No glasses are known in the systems Si-Sn-S and Si-Sn-Te.

60

CHAPTER 1

1.7.4. THE OXY-CHALCOGENIDE GLASSES The first chalcogenide glasses containing oxygen have been prepared by King and Kelly [208] with the composition The interest in these glasses is due to their transparency in the infrared spectrum. The oxy-chalcogenide glasses with tellurium have been prepared by Ulrich [209] in

the system Alkali oxide alloying in telluria increases the glass forming tendency when x increases (e.g. in [143]. In the last years there was studied the system This system shows a broad vitrification region. Andreichin et al. [210] have obtained vitreous compositions with up to 50 mol.% The synthesis was performed in quartz ampoules at 650 °C (20 h). The cooling of the melts was carried out by simply turning off the furnace powder. The following compositions have been synthesised: 1, 5, 10, 20, 33, 40 and 50 mol.% At higher concentrations a stratification effect does appear due to the low sublimation temperature of The vitreous samples are opaque, red or red-orange colour. The opacity is, probably due to non-homogeneous distribution of in The oxy-chalcogenide glasses with As shows a good stability and strong chemical bonds. They do not crystallise when are annealed. This fact can be explained by the stable bonds formed between the groupings, which ensure the stability of the vitreous state.

With the increase of the oxygen content the chemical stability of the glasses against HC1 and attack diminishes. The large compositional domain of vitrification may be due to the fact that arsenic oxide can appear in two crystalline modifications: the claudetite structure (monoclinic) with a network similar to (orpiment) and arsenolite (cubical) which consists of molecules and exhibits the adamantine structure In the systems it has been studied the influence of the PbO [211] and HgO and CuO [212] additives. Other chalcogenide systems with oxides have been investigated [212]:

The electrical properties of the glasses in the following systems were studied [213]:

The glass forming region of the system is situated in the range 5 at.%-33 at.% [214]. is situated between 560 and 500 K. In the system glasses were obtained by casting the melt in split-mould made of mild steel at 573 K for 1 h. Glasses are formed for mol.% [215].

THE CHALCOGENS AND THEIR COMBINATIONS

61

1.7.5. THE HALO-CHALCOGENIDE SYSTEMS

The investigations have shown that the doping by halogen improves the electrophotographic properties of the amorphous selenium. The studies have been carried out for doping with 0.1 at.% doping element. At higher concentrations of doping element, the glassy alloys are difficult to prepare, are unstable and the rigorous control of the halogen content is not easy to make. The bromide introduced in selenium contributes with only a covalent bond because it has 7 electrons into the valence shell. The bromide can remain unbounded in the structure, can be charged by a negative (-1) charge or can be situated, covalently bonded, to an end of a selenium chain. The doping of the selenium photoconducting films with ppm Br leads to high dark currents, and to the decrease of the potential created by the surface positive charge, and this decrease is much more rapid than for other photoconductive films based on selenium [216]. It is possible to have a glass formation in the systems Se-Br due to high supercooling tendency for the compositions with at.% Se concentration [217]. In the system exist two compounds: SeBr4 which melts at 123 °C and which melts at 0 °C. The eutectic composition corresponds to the concentration of 46 at.% Se and exhibits a melting temperature of -5 °C. In the system Se-I exists two compounds and [218]. Jecu et al. [219] have prepared the compositions The iodine is not completely bonded in the samples. The concentration of non-bonded iodine, determined by iodometric methods is 15 at.% in and 9 at.% in Melting again the alloy leads to the doubling of the concentration of the non-bonded iodine. Ignatiuk et al. [220] have studied the formation of the glassy alloys in the system Te-I. By synthesis at 850 °C and melt quenching in water + ice there was determined the glass formation region, which extends from 40 to 55 at.% I. In the system there exists two stable compounds: and the and the unstable phases and -Tel which during heating transform in stable phases, crystallises in the triclinic system (space group: P1): a = 9.957 Å, b = 7.991 Å, c = 8.214 = 102.9°, Z = 9 [221]. crystallises in the rhombic system, space-group a = 13.54 Å, b = 16.73 Å, c = 14.48 Å, Z = 16 and also in the tetragonal system (space group: [222] with a = 16.12 Å, c = 11.20 Å, Z = 16. The halo-chalcogenide glasses are interesting because they represent mixtures of compositions with both covalent and ionic contributions to bonding in the network. The pseudo-binary systems were studied only recently. The glasses with Br and I (Br-Se-Te, I-Se-Te) have been prepared and investigated by Ma Hong Li et al. [223]. The vitreous region is shown in figure 1.42. In the regime that implies the cooling in water + ice, Rudnev and Ignatiuk [224] have prepared the fiollowing compositions: and

62

CHAPTER 1

Figure 1. 42. The glass formation domains in the systems (I, Br) - Sb .

Figure 1. 43. The glass formation domain in the system (Cl,Br, I) - S – Te [145].

Figure 1. 44. The glass formation domain in the system I - P - Se.

THE CHALCOGENS AND THEIR COMBINATIONS

63

Figure 1. 45. The glass formation domain in the systems Br-As-Ch and I-As-Ch.

The glasses from the system Hal - S - Te have been prepared by Zhang et al. [225] for Hal = Cl, Br, I. The vitreous domains are shown in the figure 1.43. The synthesis was performed at 300 °C and the cooling was made in air. In the system Hal-P-Ch have been prepared glasses of composition P-Se-I [226] (figure 1.44). In the systems As-Ch-Hal have been prepared glasses with Hal = Cl, Br, I and

Ch = S, Se, Te. The vitreous domains for some systems are shown in figure 1.45. They illustrate the tendency of glass formation to increase with the halogen radius. Thus,

chlorine forms stable glasses only in the presence of sulphur, bromine forms stable glasses also in the presence of selenium, and iodine forms stable glasses with all chalcogens. A large vitreous region has been identified in As-S-Br [227]. The composition has the softening point = -60 °C. The vitrification would be due to mixing of small size molecules of composition not completely identified [51]. When the halogen content decreases while sulphur concentration increases, larger molecules are formed and, therefore, the will increase. For the composition the softening temperature is 90 °C. The compositions As-Ch-I were the first halo-chalcogenide glasses to be synthesised [227] (in 1960). Glasses have been prepared with up to 33 mol.% iodine (figure 1. 45b). The presence of large area glass formation domains is conditioned by the existence of network former ternary compounds of the type AsChI with a chain-like structure. It is supposed that in As-S-I glasses there are polymeric chains and the iodine plays the role of chain terminator. The structure can be regarded as a network of pyramids interconnected by simple or double sulphur bridges or by -S-S- units. The network would contain molecular species dissolved: etc ... Therefore we are dealing in fact with a phase separation at the molecular scale. Nevertheless, there is controversial regarding the existence of discrete molecules or intermediary compounds of the type

64

CHAPTER 1

In As-Se-I iodine enters up to 40 mol.%. The vitreous area is narrower than for the system As-S-I. In As-Te-I do appear structural units based on all three components, for iodine concentrations of 5-8 mol.%. The substitution of iodine by bromine does not modify significantly the vitrification domain (Fig. 1. 45b). In the systems with antimony the best studied is Sb-S-Br where the largest vitrification region was observed (Fig . 1.42). In the systems Si-Ch-Hal, Dembovsky and Popova [228] have prepared glasses of composition Si-Se-I. They proposed for these glasses a structural model made from chains of tetrahedra interconnected by selenium atoms situated in the comers. In glassess the iodine plays the role of network terminator, as in the example proposed for SiSeI2:

In the systems Ge-Ch-Hal, the presence of large vitrification domains is accompanied by the appearance of network formers compounds, of the type which have a chain-like structure [228], Ternary compounds (network formers) of the

type

which exhibit layer or band structure are also formed. In the systems Ge-S-I and Ge-Se-I (Fig. 1.46) the glass formation domain is larger than in the system with silicon substituted for germanium. Germanium is terra-coordinate. The sulphur atoms form bridges between tetrahedral germanium atoms and iodine atom acts as network terminator. The added iodine substitutes the bridge sulphur.

Figure 1. 46. The glass formation domains in the systems Ge-Ch-I.

More complex halo-chalcogenide glasses have been prepared. When Ge is introduced in As-Ch-Hal glasses, glasses are obtained up to 30-40 at.% Ge. Sb enters into glasses up to 10 at.%. In the halo-chalcogenide systems the glass

formation ability decreases [229] in the series

THE CHALCOGENS AND THEIR COMBINATIONS

65

There were prepared iodine glasses in the systems Cu (Ag, Si, Ge)-As-Te-I where Te and As are major components: [230] and the same compositions with Cu substituted by Ag, Si, Ge. A recent review paper on the researches related to the preparation of the halochalcogenide glasses has been published by Sanghero, Heo and Mackenzie [231]. The chalconitride systems Nitrogen was introduced in chalcogenide glasses, by using nitrides as Si3N4 and BN. The chalconitride glasses based on (0.32 wt.%) in and based on BN (0.5 wt.%) in the same glass were prepared by mixing the components h at and cooling in air [232]. Higher and microhardness than in the corresponding glass matrix were observed. 1.7.6. THE CHALCOGENIDE SYSTEMS WITH METALS

In the non-crystalline chalcogenide systems can be introduced metals. Metal microdoping, i.e. doping at concentrations under 0.1 at.% is easy to carry out. These impurities do not influence significantly the properties of the chalcogenide glasses. This surprising fact is in the same time disappointing from the point of view of the possible applications in electronics of the glasses. This feature was explained by the complete saturation of the valence bonds of the doping elements and, as a consequence, the doping elements cannot act nor as donors nor as acceptors of electrons. The addition of large amounts of metals at.%) confers specific properties to glasses. Several metals can be introduced in high concentration in the composition of the chalcogenide glasses at.%) and strongly modify their properties. At such concentrations we are dealing, in fact, with metallic components of the glasses. For every metallic element there exists a limiting concentration above which is triggered the formation of chemical compounds of the metal with the components of the chalcogenide glass. The metals, which interact with both components of an usual chalcogenide binary glass and forms thus ternary compounds, can be introduced in high amount in the glassy matrix. The main role in the glass formation is played by the unstable structural forms appeared during melt quenching, which are destroyed when the composition is brought to the equilibrium state [233, 234]. The problem of the modification of the electrical properties of the chalcogenide glasses by doping with various elements has been reopened, as a consequence of observation that, in very pure glasses, carefully prepared, the effect of the doping atoms can be evidenced [235, 236]. The Alkali-chalcogenide Glasses

The possibility to get glasses in the systems alkali-chalcogen has been analysed by Minaev [146]. He demonstrated the increase of the glass formation ability with the increase of the atomic number of the alkali metal. In 1982, Chuntonov [144] confirmed the formation of glasses in the system Cs-Te.

66

CHAPTER 1

Glassy compositions with Sb, with the formula

(A = Na, K, Rb), have been

prepared by Bazacutza [237]. In the section [220] the glass formation domain is situated in the interval mol.% In the system exists the compounds The compound shows a dark orange colour, is stable in air and melts congruently at 510 °C. In the system [238] has been

remarked a vitreous domain in the concentration interval 60-80 mol.% In the system it is possible to appear the compounds and In the pseudo-binary system have been observed two glass formation domains in the interval mol.% and mol.% There was evidenced only one compound, which melts congruently at 400 °C. Borisova [239] has shown that several metals from the group I of the Periodic Table (among these is potassium) interact during synthesis with the two components of the arsenic chalcogenide glass and form complex structural units of formula These ternary units, with high covalent component, do participate to the structural composition of the chalcogenide glass and modify essentially its properties.

In the glass can be introduced up to 10 at.% Na, 20-30 at.% K, 28 at.% Rb and 17 at.% Cs. The glasses with alkali metals are synthesised at Inherently, during preparation takes place the contamination with oxides and other elements from the ampoule walls due to the high reactivity of alkali metals. The glass formation domain in the system K-As-Ch is shown in figure 1. 47 [240]. Golovei et al. [243] have synthesised

the glassy phases: 210 °C.

with

78 °C and, respectively,

Figure 1. 47 The vitrification domain in the K-As-Se system.

The ternary alkali-chalcogenide systems with Si or Ge are interesting due to the possibility to use them in electrical batteries with solid electrolyte. As opposite to alkalioxide glasses, in the alkali-chalcogenide glasses, the chalcogen atoms exhibit a higher tendency than oxygen to form covalent bonds with Ge (Si). This leads to weaker electrostatic interactions with the alkali cations.

THE CHALCOGENS AND THEIR COMBINATIONS

67

In the system Li2S-GeS2 there was observed a large concentration interval for the glassy compositions [241]. The synthesis was performed at and the cooling was performed in air or in water + ice. The glass boundary is situated at 50 mol.% In the system is known the compound In the system [241] the glass region extends from up to 60 mol.% Here are formed the compounds and and The vitreous thio-germanates from the system have been prepared by rapid cooling of the melts made of sulphide mixtures from up to room temperature [242]. The formation domain in the system corresponds to the molar ratio In comparison with its oxide analogs

[243] the glass formation region extends well in the regions rich in metal. The crystalline phases in the system are and In the system M-Ge-Se-As have been prepared glasses with sodium (M = Na) by air cooling of a mass m = 3 g of melt [244]. The glassy compositions situated in the vicinity of the vitreous domain strongly hydrolyse in humid atmosphere. The most stable glasses are those from the section By increasing the Ge concentration in glasses the the molar volume, the activation energy and the electrical conduction increases (Fig. 1.48).

Figure 1. 48. The glass formation domain in the chalcogenide system with Ge and Na.

The chalcogenide systems with metals of the subgroup I B (Cu, Ag, Au) In the systems with Cu there were studied mostly the compositions Cu-As-Ch. Cu-As-S. By synthesis at 900 °C and air quenching there were prepared glasses of composition with up to 20 mol.% [245]. In the glasses of composition

we have succeeded to introduce the maximum amount of copper of x = 0.04. At higher concentrations we have observed crystalline inclusions of luzonite Among the crystalline compounds of the system exists a high temperature crystalline form of the rhombic luzonite, and the enargite (tetragonal structure) with the transition temperature of 275 °C. In the system appears, also, the ternary phase (tenantite).

68

CHAPTER 1

Cu-As-Se. The glass formation region in this system has been studied by Zhenhua [246] (figure 1.49). The maximum amount of copper in these glasses is 35 at.%. By crystallisation of the glasses with up to 3.8 at.% Cu one separates For higher concentrations, but under 16 at.% Cu, the cubic phase is separated [247]. The density of is 5.512(3) [248]. Another compound is It exhibits, also, cubic symmetry and the measured density is 5.203(3) [248].

Figure 1.49. The glass formation region in the system Cu-As-Ch.

Hunter et al. [249] have prepared vitreous samples of composition with x = 0.05 and x = 0.25. Cu-As-Te. The glasses in this system have been prepared and studied by Borisova et al. [250]. The glass formation domain for the synthesis temperature of 900 °C and quenching in water + ice is shown in figure 1.49. The copper enters in glass with the maximum concentration of 25 at.%. The softening temperature increases from 90 ° to 130°C with the increase of the copper concentration. In the maximum is situated at 160 °C. The glasses have a great tendency towards crystallisation. During crystallisation one separates copper and arsenic tellurides and also the elements As and Te. No stable compounds are known in the system. In the glassy chalcogenides with arsenic the valence state of copper is The studies carried out on glasses in the system [251] have indicated that the orbitals 4s, 4p and 4d of the copper ion form hybrid bonds with the electrons from the lone pair of the bridge atom Se. The tetrahedron is defined by the 4 hybrid bonds. Because the and tetrahedra are bonded by edges, the formation of CuSe4 increases somewhat the length of the AsSe bond and the angles on As. It exists a relation between the lone pair electrons and the glass formation ability. Above 35 at.% Cu, the number of lone pair electrons per atom is 1.0 and the composition cannot be obtained in the vitreous state. The chalcogenide glasses with tetrahedral elements (Ge, Si) and with Cu were obtained only for tellurium.

THE CHALCOGENS AND THEIR COMBINATIONS

69

Cu-Si-Te. In this system copper can be introduced up to 18 at.%. The vitrification domain is shown in Fig. 1.50 [252]. In this system is known the compound which exists in two crystalline modifications: cubical structure with a = 5.93 Å and monoclinic structure with a = 12.86 Å; b 6.07 Å, c = 8.61

Figure 1.50. The vitrification domain in the system Cu-(Ge-Si)-Te.

Cu-Ge-Te. In this system are formed glasses with maximum 9 at.% Cu, with at.% Ge and at.% Te [198] (fig. 1.50). Only one crystalline compound is known, in two structural varieties: cubic (a = 5.94 Å) and tetragonal (a = 5.956 Å; c = 5.926 Å). The last one is situated within the vitreous domain. Cu-P-Se. In this subsystem have been prepared and studied the glasses by Radautsan et al. [253]. The glasses are stable. In detail has been studied the compound [254] which melts at 640 °C and crystallises in the hexagonal lattice: a 3.853 Å, c = 6.392 Å. In the section the maximum concentration of copper in the glassy phase corresponds to the composition In the systems with silver, there were studied mainly the formation and the properties of glasses based on arsenic chalcogenides. Ag-As-S. The glass formation domain is controversial. In a very careful study performed by Kawamoto et al. [255] have been evidenced two domains of glass forming compositions during cooling of the melt in air or water starting from °C (Fig. 1.51). A somewhat different diagram has been published by Kazenkova [256]. During crystallisation does appear (smithit), (proustite), (xanthoconite), (billingsleite) and a not identified phase. From these compounds

only

is stable in the glassy state and, therefore, can be considered as a network

former. In the chalcogen glasses the silver can be introduced up to 30-35 at.%, by very rapid cooling. The structure of the compositions [258] with up to at.% Ag in i.e. havebeen investigated.

70

CHAPTER 1

Figure 1.51. The vitrification domains in the system As-Ch-Ag.

A strong doping with Ag of the glass was obtained by photo-reactive diffusion of Ag. The saturation was obtained at at.% Ag but the X-ray diffraction measurements have shown the presence of the crystalline phases Ag, and [259]. The local order in glasses consists probably from molecular fragments Accordingto the microcrystalline models for the chalcogenide glasses [260] and to the results of Fusukawa and White [261], the phases which crystallise the first in the glasses are structurally the most approached to the glass configuration from which they derive. Tanaka [262] has studied the composition dependence of the photo-doping rate of Ag in Ag-As-S glasses. He has shown that with the increase of the Ag content, y (at.%), the photo-doping shows a maximum at y 2 and then dramatically decreases for y 25. In samples with y 30 the Ag films form clusters upon illumination. In samples with y 35, photo-induced effects are hardly observed.

Ag-As-Se. In the system As-Se with Ag four ternary compounds can be formed. One of them is AgAsSe, the structural analogous of laudetite. The other is The structure of the crystal is isomorphous with the mineral proustite and pyrargyrite In this structure, the AsSe3 groups form trigonal pyramids, which are bridged by Ag atoms. Silver and selenium atoms form -Se-Ag-Se-Ag- infinite spiral chains parallel to the c-axis. It exists also and These compounds cannot be obtained in the glassy state. Nevertheless, the interaction of the components with the formation in melt of a great number of structural ternary units, which are

THE CHALCOGENS AND THEIR COMBINATIONS

71

different from the point of view of spatial configuration, make difficult the crystallisation of the reported phases and thus stimulates the glass formation. The vitreous region is large (Fig. 1.51) and in the section it is possible to introduce up to 20 mol.% Ag-As-Te. In the ternary glass with tellurium can be introduced up to 28 at.% Ag. The glass domain corresponds to at.% As and at.% Te. Figure 1.51 shows the domain after the data published by Borisova [263]. The stable and homogenous vitreous samples in the systems Ag-Se-S and Ag-Se-Te have been prepared with 5 at.% S (Te) and at.% Ag [264].

Ag-Sb-S. In this subsystem there was evidenced a narrow vitreous domain (Fig. 1.52a). The glass region intersects the section [257]. The compound has been obtained in glassy state. Ag-P-S. Glasses have been obtained in a very narrow compositional range [265] by synthesis of a mass of 4.5 g material at 600 °C and water quenching of the melt. The composition AgPS has been obtained in glassy state. The content in the glasses is situated between 45 and 55 mol.%. raises from 162 °C (for 45 mol.% to 182 °C (for 55 mol.% (Fig. 1.52 b). Two compounds exist in the system: and and they are situated in the vitreous region. Ag-P-Se. In this system have been obtained unstable glasses [253] by adding 5 at.% Ag to These glasses decompose in air. There were obtained, also, glasses of composition AgPSe [266]. Ag-P-Te. After our knowledge no glasses have been prepared in this system.

Ag-Ge-S. The glass formation domain for this ternary system is shown in figure 1.52c. Ag-Ge-Se. In this system two glass forming regions have been found by Mitkova et al. [345]. They are shown in figure 1.52d. The domains are situated in a region of the diagram characterized by low silver content and by the ratio germanium/sulphur not far from unity. Ag-Ge-Te. After our knowledge no glasses have been prepared in this system. Au-As-Ch. As regarded the glasses with gold, only the systems Au-As-S [267] and Au-As-Se [241] have been investigated. The synthesis was carried out at 900°C and was followed by air quenching. In glasses enter up to 8 mol.% in the section and up to 10 mol.% in the section

72

CHAPTER 1

Figure 1.52. The vitreous domain in the system Ag-Sb(P)-S, Ag-Ge-S and Ag-Ge-Se.

The Chalcogenide Glasses with Transition Metals (3d and 4d) In fact we cannot have vitreous-chalcogenide systems with the transition metals because the metals with 3d or 4d electrons exhibit low ability to interact with the network former elements. Thus, in the composition of the As-Se chalcogenide glasses the elements Fe, Co, Ni practically do not enter into the homogeneous glassy matrix. In the glasses of more complex composition (Ge-As-Se, Sb-Ge-Se) the iron can be introduced in an amount

THE CHALCOGENS AND THEIR COMBINATIONS

73

situated somewhat above 2 at.%. A systematic study of the introduction of the Fe with the chalcogenide glasses of various compositions has been carried out by Andreev et al. [268]. There exists a large category of glasses where iron introduced in a concentration not above 0.2 at.% is already in the polycrystalline state based on the compounds: FeS, This is in case of the glasses glasses which are situated in the vicinity of the eutectic composition where their ability to form a nano-structural state is higher. There exists also an other category of glasses situated near the eutectic: In these glasses, iron was introduced in the amounts of 0.2-2.2 at.%. These glasses, which receive more iron in network than the previous ones are remarkable by the high germanium content, the chalcogen being no more than 60 at.%. In these glasses the local neighbouring of the iron is analogous to that from Therefore, it is possible to appear this compound for which the noncrystalline state is not specific. The units are distributed statistically in the glass. When their number increases they associate and form collective phases. As a consequence, Fe does not enter a too large amount in glasses. It is well known from the literature that is a compound, which melts at 1045 °C and dissociates by melting [269]. In order to increase the Fe content in the glass on the account of the structural units characteristic to this compound, one half of antimony has been substituted by Ge [270]. Thus it was possible to introduce in glass 3 at.% Fe. From the Mössbauer analysis it was revealed that in glasses appear the structural units and For 4 at.% Fe do appear the additional phase [270]. The behaviour of Ni, Co and Mn in the chalcogenide glasses has been studied on two compositions in the system Ge-Sb-Se: (eutectic) and where (in the last one) has been introduced the maximum amount of iron: 2.2 at.%. The synthesis has been performed by melting the corresponding glasses with the metal in evacuated quartz ampoules for 5 h at 900÷950 °C, and the melt quenching was performed in air. Nickel has been introduced in glasses up to 0.4 and 0.3 at.%, respectively, without observing crystalline inclusions. For higher concentrations, selenium crystallites and binary Ni-Ge crystallites do appear. There was revealed also the presence of NiGeSe3. In the table 1.5 there are shown the maximum concentration of metals introduced in glasses that does not modify the homogeneous vitreous state of the matrix [239].

In the glasses, Fe is mainly in the valence state and nickel is in the There were identified also small amounts of and The concentration of the trivalent nickel increases with the total nickel concentration in samples. By deposition of

74

CHAPTER 1

films with the composition together with nickel (by rf. sputtering) at ~80 °C, it was possible to get amorphous, homogeneous films with up to 11 at.% Ni [273]. The crystalline chalcogenides of the transition metals 4d, from the groups IVB-VIB (Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W) form layered structures for the composition and chain structures for the composition Amorphous solids of several transition metal di-chalcogenides have been obtained by precipitation at low temperatures from non-aqueous solutions [128]. The amorphous chalcogenide with transition metals have been also prepared by thermal and chemical decompositions of the sulphur and selenium salts [272]. Liang et al. [271] have prepared and studied the compositions and In the glasses the co-ordination of the transition elements is complicated due to the presence of the d-electrons. In di- and trichalcogenides of the transition metals the co-ordination of the transition element by chalcogen is made by six atoms, which can be, disposed both octahedral-prismatically and prismatic-trigonally. Nevertheless, in many ionic and molecular chalcogenide clusters the transition elements can take various neighbouring. The electrical charge of molybdenum and of tungsten in the chalcogen clusters can vary between and having the Me-Ch length situated between 2.0 and 2.7Å. Liang et al. [271] have proposed for the amorphous state a model with clusters similar to those units characteristic to crystalline The Glassy Chalcogenides with Metals from the III A group (Al, Ga, In, Tl) Chalcogenide Glasses with Aluminium.

As-Se-Al. The ability to form glasses is very low in this system. In the it is possible to introduce up to 3 at.% aluminium.

glass

As-Te-Al. There were prepared glasses with aluminium by synthesis at 1050 °C in graphitized quartz ampoules in order to prevent the interaction of Al with quartz. It was possible to introduce in the chalcogenide composition up to 30 at.% Al. The increases from 78 °C for up to 175 °C for The glasses are formed in the region rich in tellurium (Fig. 1.53 a) [274]. Ge-Te-Al. In this system the glassy state has been obtained by melt quenching in values for the glasses with

liquid nitrogen for a large composition interval [275]. The mol.% Al are constants (~ 170 °C) (Fig. 1.53 b).

THE CHALCOGENS AND THEIR COMBINATIONS

75

Figure 1.53. The glass formation domain in the systems As-Te-Al and Ge-Te-Al.

Chalcogenide Glasses with Gallium. As-Ch-Ga. The introduction of gallium in the vitreous system can determine the formation of the coordinative valence bond of the type - Se - Se -, with transition from the trigonal structure to the tetrahedral one The multiplication of the tetrahedral units with the introduction of the gallium in glass leads to the density increase, hardness increase, mechanical strength increase, etc. Gallium interacts chemically both with arsenic and with selenium. With arsenic it forms GaAs, a semiconductor with zinc blende structure (ZnS). With selenium it forms semiconducting crystals of having also the zinc blende structure and low mobility of electrical charge carriers. is obtained in vitreous state only in the concentration range There were obtained glasses in slow cooling regime by turning off the furnace power (the cooling is from 900 °C to 700 °C during followed by air cooling from 700 °C in 10 min. down to room temperature. The vitreous domain is shown in Fig. 1.54. The glasses with Ga content up to 3.5 at.% have been obtained in the system As-S-Ga by water quenching of the melt starting from 900 °C. In the system As-Te-Ga the glasses have been obtained by quenching the melt in air down to the viscous state (400 °C) followed by throwing in cool water flow. In this cooling regime the maximum content of gallium in glasses has been raised up to 20 at.%. These glasses are stable. The most stable glass is The introduction of gallium in glasses diminishes the crystallization ability [276]. The glass formation domain can be seen in Fig. 1.54.

76

CHAPTER 1

Figure 1.54. The vitrification domain in the system Ga-As-(Se, Te).

Ge-Ch-Ga. The glasses from the system Ge-S-Ga show a narrow composition domain [277] (Fig. 1.55). The glass transition temperature varies from 150 °C for [278] to 360 °C for The glasses for this system crystallize easily.

The synthesis is carried out by mixing the corresponding sulphides, followed by firing at 1100 °C and finally quenching in water + ice mixture. The glasses from the system Ge-Se-Ga are chemically stable and easily crystallize. The softening temperature decreases when the Ga and/or Se content increases. For and for the glass

Figure 1.55. The glass formation domain in the system Ga-Ge-Ch.

In the system Ge-Te-Ga were revealed the compounds and with hexagonal symmetry (space group and having the lattice parameters: a = 4.02 Å, c = 16.82 Å and a = 4.05 Å, c = 16.87 Å, respectively.

THE CHALCOGENS AND THEIR COMBINATIONS

77

The Si-Te-Ga glasses have a narrow composition domain similar to that of the Ge-Te-Ga system (Fig. 1.56).

Figure 1.56. The glass formation domain in the system Si-Te-Ga.

Chalcogenide Glasses with Indium. As-Ch-In. In the glasses As-S-In obtained by synthesis at 900 °C and water quenching can be introduced up to 3 at.% In [245]. In the system As-Se-In have been obtained glasses in the same preparation regime with the same maximum indium content [245]. In the section Borisova and Babrov [279] have succeeded to introduce only 1.2 at.% In. During crystallization is formed firstly In the system As-Te-In the maximum In content in the glassy phase was 15 at.% accepted under water quenching of the material synthesized at 1000 °C. The increases with the indium concentration from 84 °C for up to 155 °C for [280]. The most stable against crystallization are the glasses and The glass formation region is shown in figure 1.57.

Figure 1.57. The glass formation domain in the system As-Te-In.

78

CHAPTER 1

Ge-Ch-In. In the system Ge-S-In, there was possible to introduce in the glassy network up to ~ 15 at.% In by quenching the melts with the rate of 10 K/s [281]. The vitrification region is shown in Fig. 1.58. In the system Ge-Se-In, in similar quenching conditions, a similar vitrification domain is obtained.

Figure 1.58. The glass formation domain in the systems Ge-Ch-In.

The glasses from this system have enough large

(from 200 °C for

up to 320 °C for Ge31.7Se63.3In5).Matsushita et al. [282] have shown that the glass formation domain is narrow. Its delimitation has been made on the basis of samples prepared as films by flash method. The deposition was carried out on glass substrates at 300 °C. Boncheva-Mladenova and Mitkova [283] have prepared and investigated several amorphous compositions in this system:

In the system Ge-Te-In has been introduced in the glassy phase up to 13 at.% In after [284] and up to 22 at.% In after [285]. Chalcogenide Glasses with Thallium Thallium can be introduced in the chalcogenide glasses up to 30 at.%. The binary compositions increase their ability to form glasses when thallium is added. Glasses have been prepared in the systems As-Ch-Tl, Ge-Ch-Tl, P-Ch-Tl. The presence of the III A group elements in the chalcogenide compositions, i.e. the presence of elements with one electron in the normal state situated on the last occupied orbital p, leads to the breaking of As-S and As-Se chains. The unique electron of the element III A is weakly bonded (mainly to thallium) and can break a Se chain thus building the configuration or, more probably, As-Ch-Tl. The vitreous domain in the system As-S-Tl is similar to that of the As-S binary alloy where the structural units are and The upper limit for the

THE CHALCOGENS AND THEIR COMBINATIONS

As concentration in the vitreous domain corresponds to 50% 27%. The scheme for the association of the units and the link of more than two sulphur atoms in a sequence, is:

79

and the lower limit is which does not allow

The sulphur tendency to form rings is the cause of the crystallisation of the sulphur rich compositions. The breaking of the S-S bond determines the rearrangement of the atoms and these bonds exist only in a narrow composition domain at.% As) and the compositions will crystallise easily. The glass formation region is shown in Fig. 1.59. The theory of the glass structure has been developed on the basis of the analogy with the crystalline structures. Thus, the lorandite, exhibits a monoclinic structure [286]. The pyramids AsS3 are bonded in spiral chains weakly bonded through (?) the ions. Every thallium atom is more closely attached to its chain than to neighbouring chains and this fact explains the lamella habitus and the excellent cleavage properties along the plane (100). In the vitreous compositions of the system As-Se-Tl exists Se structural units selenium arsenide units and also polar structural units of the type The ionic component of the chemical bond increases in the series The arsenic selenides with an intermediary ionic component are able also to react with the Tl structural units and Se structural units. As a consequence, in the thallium glasses appear two glassy phases: one phase rich in chalcogen and other phase rich in structural units based on thallium [287].

Figure 1.59. The glass formation domain in the system As-Ch-TI.

80

CHAPTER 1

In the glasses containing Se and also in the glasses with sulphur, the presence of large glass formation domains (Fig. 1.59) is conditioned by the formation of the ternary network former compounds with more than a chain structure and this fact puts a limits to the domain of stable glasses. In the system As-Te-Tl does not appear any ternary network former compound in the usual cooling conditions and, consequently, in the system does not exist a domain for stable glasses [288].

Ge-Ch-Tl. The vitreous domain for the system Ge-S-Tl has been studied by Linke [289]. In this system are formed the compounds: (monoclinic symmetry, space group (space group and (monoclinic system, space group The vitreous domain is shown in Fig. 1.60. In the system Ge-Se-Tl the glasses are formed in a larger domain (Fig. 1.60). The softening temperature, increases from 40 °C to 150 °C with the increase of the germanium content [290].

The Chalcogenide Glasses with Rare Earth Elements The most Ln-Ch systems are characterised by the formation of a high number of phases in the domain 50-75 at.%Ch and the absence of chemical compounds for high concentration of rare earths [291]. The compounds and belong to the category of the compounds of variable composition. As an example, a very large homogeneity domain has been found for the lutetium sulphide 43-54 at.% S). The homogeneity domain increases in the series La-Ch Gd-Ch. LaS 0.9-1.0; CeS0.88-1.0; NdS0.78-1.0.

Figure 1.60. The glass formation domain in the system Ge-Ch-Tl.

THE CHALCOGENS AND THEIR COMBINATIONS

81

The chalcogenides of divalent rare elements show similar effects. The ytterbium sulphide crystalline structure exists in the range YbS1.11-YbS1.14. When the sulphur content is increased, some defects appear in the metallic sublattice and a decrease of the crystalline lattice parameters occur. This effect is related to the transfer of a part of the ions in the valence state ions of smaller size. The Yb telluride exists in a slightly different domain: 50-52 at.% Te. Large homogeneity domains are encountered also to the selenide and telluride of europium. For Sm-Ch the homogeneity region corresponds to at.% S and at.% Te, respectively. Zachariasen [292] has established the boundaries of the homogeneity domain of the cerium sulphide: All the mono-chalcogenides of the rare earth elements crystallise in the NaCl structure.

The rare earths behave similarly to the transition elements in the glassy chalcogenide matrices. Thus, europium enters in small amounts in the sulphide glasses, e.g. in the gallium sulphide glass, where it forms structural units of the type which are

characteristic for the crystalline compound of the same formula which cannot be obtained in the glassy state [293]. Raspopova et al. [294] have shown that in the glassy chalcogenides with arsenic the amount of rare earth which enters in the composition is low. For and glass with Ce, Sm, Eu and Gd the concentration of the rare earth in glass can reach 2 at.%, but a very good homogeneity is obtained for 0.4 at.% cerium (perfect homogeneity is reached for 0.1 at.% Ce), 0.1 at.% for Sm and Eu and 0.07 at.% for Gd. The glass formation domain for the glasses with europium is shown in figure 1.61 [295]. The glassy chalcogenides with rare earts are obtained by quenching the melts in water (especially for germanium chalcogenides) and in air (the case of glasses based on the As chalcogenides). Because the formation of the glass is more efficient if the glass composition is more complex, there were used multic-component systems as e.g. In this system the vitrification region has been analysed in detail and is shown in Fig. 1.62.

Figure 1.61. The glass formation domain in the system As-S-Eu.

82

CHAPTER 1

Figure 1.62. The glass formation region in the system Ge-Ga-Se-Ce.

Very well studied is the system Yb-As-Se [296]. The glass formation region is given in Fig. 1.63.

In the section have been obtained glasses with a maximum content of Yb of 3.2 at.%. In the system is possible to get the compounds the last three being characterised by congruent melting temperatures of 725 °C, 880 °C and 750 °C, respectively. The solid solutions based on arsenic selenides and the eutectic between the network former compounds and the network modifier compounds are responsible for the glass formation. Transparent and homogeneous glasses were not obtained in the systems when and As, whereas the system was vitrificable in the mol.% composition range. Figure 1.64 shows the glass formation region in the ternary system in which about 35 mol.% of GeS2 can be substituted for [297].

Figure 1.63. The glass formation domain in the system As-Se-Yb.

THE CHALCOGENS AND THEIR COMBINATIONS

Figure 1.64. Glass formation of the ternary

vitrified by quenching in air vitrified by quenching in ice

83

system;

partially vitrified; water crystallised.

Chalcogenide Glasses with Heavy Metals Chalcogenide Glasses with Mercury. There were prepared and studied glasses in the systems As-Ch-Hg and Ge-Ch-Hg. As-Ch-Hg .In the system As-S-Hg have been obtained glasses with a maximum Hg amount of 5 at.% in the region of the binary compound AsS. In the system Ag-Se-Hg the glass formation region is larger due to the fact that Se and AsSe are network formers as opposite to S and [298]. The As-Se-Hg glasses are obtained by quenching the melts from 450 °C to 100 °C with the rate of 100 K/min. In the system has been found the stable f.c.c. crystalline compound HgSe with a By introducing mercury in arsenic glass takes place some distortion of the glass network due to micro-inhomogeneities in the distribution of the mercury selenide groupings. For more than 4 at.% Hg, the HgSe phase separation is produced. The metal-chalcogen interaction is very effective: even for low Se concentration appear Mg-Se bonds. Below 1 at.% Hg takes place a linking of the chain elements in a spatial network, and this fact leads to a certain network ordering. In the domain 1-4 at.% Hg occurs the intensification of the separation of the structural units Hg-Se which remain still strongly dispersed and this fact leads to the distortion of the network of the arsenic glass and to the formation of high amount of defects [299]. Ge-Ch-Hg. In the system Ge-Se-Hg the vitrification domain has been studied by Feltz et al. [300]. Two glassy domains have been observed. In the selenium rich domain, the main region is occupied by the glasses formed up to the compositions with 15 at.% Hg. The second domain comprises glasses with up to 25 at.% Hg. In the system are known the compounds and still two one-dimensional compounds [301]. crystallises in a structure derived from ZnS, where the

84

CHAPTER 1

cationic sublattice is only partially occupied. The mercury and germanium occupy the tetrahedral unoccupied sites following an ordered sequence. The co-ordination of selenium is 3. The vitrification domain is shown in figure 1.65 a.

Figure 1.65. The vitrification domains in the systems Hg-Ge-Se (a) and Hg-Ge-Te (b).

In the system Ge-Te-Hg the vitrification domain is smaller and is situated in the neighbourhood of the eutectic GeTe-Te (figure 1.65 b). Chalcogenide Glasses with Lead. As-Ch-Pb. These systems behave similarly to those with Hg [288]. The glasses As-S-Pb have been prepared by synthesis in vacuum from elements at 900 °C and quenching (quenching rate: 200 °/s) [267]. Thus, one can introduce in glass up to 3 at.% Pb. One knows a natural glass of composition which cannot be crystallised even for long thermal annealing [302]. The glasses As-Se-Pb have been described in the papers [245, 267]. They can take up to 8 at.% Pb. The vitreous domain is shown in Fig. 1.66. By slow cooling one can introduce in glasses up to 2 at.% Pb and in quenching regime 5 at.% Pb [303]. The glass formation region for the system As-Te-Pb has been found by Minaev et al. [182] (Fig. 1.66).

Figure 1.66. The glass formation domain in the system As-Se-Pb.

THE CHALCOGENS AND THEIR COMBINATIONS

85

Figure 1.67. The glass formation regions in the systems. a) Ge-Ch-Pb b) (Si, Ge)-Te-Pb.

Ge-Ch-Pb. The glasses with Ge and Pb are better studied. In the system Ge-S-Pb the glasses were obtained by Feltz et al. [304]. The vitreous domain is given in Fig. 1.67 a. In the glass can be introduced up to 57 mol.% PbS. In the system two compounds exist:

and

which cannot be obtained in the glassy state. In the system

Ge-Se-Pb [305] the boundary of the glass formation domain corresponds to the water quenching with the rate of 8 K/s through the softening interval. depends on the composition and is situated in the interval 236-244 °C. Only one compound has been found: The compositions with tellurium exhibit a somewhat narrower vitrification domain [207] (figure 1.67 b). The substitution of Ge by Si makes this domain narrower. Chalcogenide Glasses with Bismuth. Glasses in the system Ge-Ch-Bi have been prepared and investigated [305-308]. Tychý et al. [309] have published a detailed study on the vitrification region in the system Ge-S-Bi. The vitrification domain is situated around the axis up to

25 at.% Bi. The composition with the maximum Bi content is and exhibits a maximum in the system: 363 °C. In the paper [310] it is claimed that in the system the glasses are formed up to The alloy is of eutectic type. The eutectics appear when the clustering (i.e. the association of similar atoms or molecules) is preferred to the association of molecules (atoms) of different kind. Of course, the degree of association of the atoms or molecules (A-A or B-B) is limited, and both species A and B interact through the bonds between the atoms situated on the surface of the clusters in contact, because the total

association of the atoms would lead to the immiscibility (if the Van der Waals bonds between clusters would be negligible). There was supposed that in these glasses exists an association of tetrahedra and pyramids under the form of submicropscopic

86

CHAPTER 1

Figure 1.68. The glass formation domain in the system Ge-Ch-Bi.

the extension of the glass domain by increasing the cooling rate of the melts. air cooling; water ice cooling. air cooling.

agglomerations of

and

Tichý et al. [309] suggest that the interatomic bonds of

type Ge-S-Bi inhibit the connection of the clusters and thus preventing the formation of aggregates with enough large size to become nucleation sites. The maximum ability to form glasses, i.e. the centre of the domain, must appear to the composition with

the maximum probability of formation of the Ge-S-Bi bonds. For the system Ge-Se-Bi the vitrification domain is shown in figure 1.68. In the glassy state it is possible to have up to 14 at.% Bi [307].

A system containing bismuth, not very well studied, is the system I-Se-Bi [311] with great perspectives for highly sensitive xerography. 1.7.7. OTHER TERNARY CHALCOGENIDE SYSTEMS

The glassy state has been researched in a great variety of ternary systems. In this paragraph we shall discuss several ternary systems which are poorly studied. Only scarce and not confirmed results were obtained. Nevertheless, these systems can represent a departure point for new investigations and ideas. The Chalcogenide Glasses with Boron In the binary systems B-S and B-Se the vitrification domain has been not established.

There was possible to get glassy by synthesis of the elements at 590 °C and cooling with the furnace [312]. The samples are higroscopic. The glassy domain of the system B-Se extends from Se to For higher boron concentrations no studies have been carried out in B-Ch systems, after our knowledge [313]. The very well known crystalline compounds

and

are monoclinic.

THE CHALCOGENS AND THEIR COMBINATIONS

87

As-Ch-B. In the system As-S-B the vitrification domain is situated in the corner rich in As and S of the composition triangle (Figure 1.69 a) [314].The glasses have been synthesised at 800 °C and the melt has been cooled slowly (0.5 °/min).

Up to 10 at.% B the glasses are stable in air. For higher concentration they become unstable and easily hydrolyse in wet atmosphere. In the system As-Se-B the glassy domain is situated in the region rich in As and Se (Fig. 1.69a) [315]. The synthesis was performed by heating the mixture of components at 950 °C hours) and cooling with the furnace. Up to a content of 9 at.% B the glasses are stable in air. Above this concentration they decompose rapidly.

Figure 1.69. The glass-forming region in the systems. a) As-Ch-B, b) Ge-Ch-B, c) Tl-Se-B.

(Ge,Si)-Ch-B. In the system Ge-S-B it appears a large vitreous domain (figure 1.69 b) [317]. The synthesis was carried out at 1100 °C and was followed by quenching in water ice. The properties of these glasses have been not studied because they easily hydrolyse in air. The system Ge-Se-B behaves similarly (figure 1.69 b) [316].

88

CHAPTER 1

There was studied the system Si-Se-B [316] where the boundaries of the vitreous domain have been not established due to the presence of the effect of micro-stratification in melt. In the system Sn-Se-B [316] we are dealing with similar behaviour as in Si-Se-B: the glasses hydrolyse and the melt is stratified. Tl-Ch-B . In the system Tl-Se-B the glass forming domain is large [317]. The synthesis was carried out at 800 °C. As a function of the cooling regime the vitreous domain varies (Fig. 1.69 c). In the pseudo-binary system the compounds exhibit large deviations from stoichiometry. In the system are formed the compounds and The presence of large vitrification domains in these systems is conditioned by the formation of ternary (network former) compounds of type and [288].

Chalcogenide Glasses with Zn and Cd As-Ch-Zn. In these systems the maximum concentration of zinc is 3 at.% for As-S-Zn [177] and 10 at.% Zn for As-Se-Zn (synthesis at 900 °C and cooling in water + ice mixture). By cooling in air from 900 °C Doinikov and Borisova [318] succeeded to introduce in the selenium glasses up to 1.9 at.% Zn.

As-Ch-Cd. In this system the glasses have been obtained by Kolomiets et al. [177]. The maximum Cd content introduced by synthesis at 900 °C and melt cooling with the

rate 200 °/s is 40 at.%. In the system As-Se-Cd [246] the maximum Cd amount in glass was 5 at.% (Fig. 1.70). By exceeding this value one separates CdSe. The valence state of Cd is The electrostatic field of the reduces the flexibility of the lone pair electrons of bridging Se atoms and, moreover, destroys the As-Se and forms Cd-Se bonds.

Figure 1.70. Glass forming region of Cd-As-Se system. glass glass-crystal crystal.

•

Pb-Se-Cd. In this system it was introduced up to 5 at.% Cd in the composition [253]. The glasses are unstable. They decompose in the usual atmosphere, at room temperature.

THE CHALCOGENS AND THEIR COMBINATIONS

89

The Chalcogenide Glasses with other Elements Tl-Ch-Zr. Chernov et al. [319] have shown that by quenching can be obtained glasses of composition and when during preparation are used small masses of material.

Tl-Ch-V. In the system Tl-S-V the glassy domain is not too large [320]. In glasses can be introduced up to 5 at% V. The softening temperature increases from 6 °C for T1S2 at 40 °C for the composition with 15 mol% in the section The crystalline compound crystallises in the f.c.c. lattice [321]. In the glasses Tl-Se-V obtained by synthesis at 550-600 °C and quenching in cool water flow there was possible to introduce up to 5 at.% V [320]. Tl-Ch-Ti. In the system Tl-S-Ti titanium enters in glasses as Chernov et al. [322] have studied the glass formation in the section The ability to form glass is rapidly modified with the change in composition. The alloy 20% is obtained in the vitreous state by air cooling and the alloy of composition 60% is obtained in the vitreous state only by water + ice quenching of a film of material of thickness 1-2 mm.

The glasses with Be, Mg and Cd in arsenic triselenide have been and metals, at the maximum temperature 800 °C and cooling with the furnace. The maximum metal content reached in the glassy state is 3.8 at.% Be, 1.9 at.% Mg and 1.9 at.% Ca. The glasses with manganese and rare earth elements were prepared by melting pure elements at two temperature steps : 870-920 K and 1070-1120 K. The metal concentration introduced in was 0.1-0.5 at.% [323]. obtained by synthesis starting from

1.8. Quaternary Chalcogenide Glasses and more Complex Glasses

Many Chalcogenide glasses with four or more elements have been prepared. When the number of elements increases, the glasses are obtained more and more easily. The complex Chalcogenide glasses entered into the attention of the researchers due to the effect of electrical switching discovered by Ovshinsky [324-326]. The switching glasses or Ovshinsky (ovonic) glasses have a minimum one chalcogen element (usually Se or Te) combined with one or more elements as e.g. Ge, Si, I, As, Sb, Ga, In, Tl and P. Besides the binary systems Ge-Te, similar properties were evidenced in the ternary systems As-Te-I, Ge-As-Te and in several quaternary or more complex systems. There was shown that the ovonic glasses with memory, which contain the iodine as main element show, usually, a crystalline secondary phase in the glassy matrix. Other glasses as e.g. the glasses with arsenic, germanium, tellurium and gallium, show a separation of the

non-crystalline phases. The most important quaternary system for the threshold (nonstable) switching properties is As-Te-Si-Ge. The typical glasses in this system are [325,326] and [327]. Bunton [328] have prepared glasses in the same

90

CHAPTER 1

system and has drawn the most complete phase diagram. The glasses are formed in a region where the tellurium content is higher than 70 at.% but lower than 87.3 at.%. All the glasses exhibit phase separation. The second phase can be non-crystalline in the samples with low Te content but it is, very probable, a crystalline phase in the samples with higher Te content and there are indications that it exists even crystallised tellurium. A complex material, the most studied switching material, is [329]. A very good composition for the multi-component glasses is the

This material shows a minimum defect concentration under the form of domain with partial crystallisation or inclusions with chain structure [330]. The memory glass has been studied by Moss and de Neufville [331] in the same time with the binary memory glass from which it derives. They have shown that the mechanism of recovery of the initial state of the glass, after switching, is related to the inhibiting role of antimony and sulphur as to the formation of GeTe by re-

dissolution of the dendrite network of conducting tellurium that form a percolation network during the crystallisation, when the conducting state (“on” state) is thermally induced.

Ovshinsky [332] studied a large class of glassy compositions with switching effects comprising also quaternary compositions:

In comparison with the glasses based on Ge-Te-As-Si the switching glasses of the type As-Te-I show a rapid ageing due to their high crystallisation ability. The best characteristics from the point of view of the switching threshold are exhibited by the glassy compositions from the complex system: (Cu,Ag,Si,Ge)As-Te-I. The glass

formation domain in this system has been studied by Gurin et al. [333] (Fig. 1.71). The domain comprises a large volume on the metal part, and extends on the side faces As-Te-I and Me-As-Te of the tetrahedron M-As-Te-I. The example from the figure

1.71 is given for the section of the diagram been obtained by quenching the corresponding melts 200-400 °C/h.

The glasses have with the quenching rate of

The effect of the Te and iodine addition in the Ge-As-Se system glasses has been

systematically studied [334]. Addition of iodine improves the glass-forming ability and the addition of Te shifts the IR cut-off edge to longer wavelengths. The results have suggested that glasses from this system are promising for IR optical fibres materials for 8-12 radiation wavelength, with good stability against crystallisation and high

THE CHALCOGENS AND THEIR COMBINATIONS

91

Figure 1.71. The glass forming region in the complex system (Cu Ag,Si,Ge)-As-Te-I.

New glasses in the system As-Ge-Ag-Se-Te-I were studied [335]. The glass forming region of is shown in Fig. 1.72. The physical properties of typical glasses warrant the large-scale applications in the mid-infrared optics.

Figure 1.72. Glass-forming region of the - transparent, glassy; - crystalline.

•

system:

New glasses in the system Te-As-Ge-Si have been prepared recently [336]. The samples were obtained by alloying reagent grade arsenic, tellurium, germanium and silicon. Two quaternary compositions have been prepared: and Thje glassy state has been obtained by the melt-quenching technique. Appropriate amounts of 99.999 % pure constituent elements were put in an evacuated

silicon ampoule. The sample temperature was raised gradually in order to reduce the vapour pressure of the components. At first, it was raised from room temperature to

the melting point of each element and kept at each temperature for 2 hours.

92

CHAPTER 1

Figure 1.73. Glass forming domain of the system GeS2-Ga2S3-CdS.

Figure 1.74. Glass forming domain of the

Then the temperature was raised up to 1000 °C for a long time The long period of synthesis and the furnace periodic rocking ensures homogeneity of the material composition. The molten compositions were finally quenched in ice + water mixture in

THE CHALCOGENS AND THEIR COMBINATIONS

93

order to get the glassy phase sample. The mean rate of cooling in this case was about 300 °C/min. No phase diagram of these glasses has been obtained up to day. Exotic glasses were prepared in the system, and Multi-component glasses with and A1N are useful for sintering additives for the production of AlN [337]. Quaternary glasses of composition were prepared due to the interest in photoconductive and non-linear optical materials. The vitrification domain in this system is shown in fig. 1.73 [338]. Complex glasses with rare earths were prepared in the system These glasses are interesting because of their ionic properties. The glass forming domain is shown in figure 1.74 [339]. Recently, a new class of sulphide glasses based upon system has been investigated [340]. Glassy compositions were prepared in graphite or vitreous carbon crucibles, encapsulated in silica ampoules, and cooled from a melt temperature of 1000 °C. These glasses have good glass stability, softening temperatures between 400 and 500 °C, and infrared transmission toward the wavelengths of about These properties make these glasses good candidate materials for rare earth doped lasers and fibre amplifiers, infrared fibre sensors and optical components for infrared systems. The glass-forming domain is shown in fig. 1.75.

Figure 1.75. Glass forming region in the pseudo-ternary system

[340].

Polynary glasses based on Ge, S and Ga with Ag, Br, Cl, I, have been prepared: [341]. Vitreous samples were obtained by melting the mixture of components in the temperature range of 875-1250 K for 8 h. Melts were then quenched in air (cooling rate K/min) and with furnace (cooling rate The vitrification domains are shown in fig. 1.76.

94

CHAPTER 1

Figure 1.76. Glass forming regions in

Glasses in the system with were prepared and investigated [342]. The maximum temperature and the synthesis time were 960 °C and 20 hours, respectively. The glass forming domains in the quasi-ternary systems

and

were found by Olekseyuk et al. [343]. The

heating of the initial mixture of elements was carried out at 1270 K for 6-10 h and the

melt was cooled in 25 % solution of NaCl in water. Glasses were obtained in a narrow concentration interval around the binary eutectic.

Bulk glasses from the system Cu-As-Se-I were prepared and investigated due to possible applications in optical recording [344]. The following glasses were investigated: with

REFERENCES [1]. A. S. Cooper, W.L. Bond. S.C., Abrahams, Acta Cryst., 14, 1008 (1961). [2]. J.T., Burwell II, Z. Krist,. 97, 123 (1937). [3]. R.C. Engel, Compt. Rend., 112, 866 (1891). [4]. J.D.H. Donnay, Acta Cryst., 8, 215 (1955). [5]. J. Donohue, A. Caron, E. Goldish, J Amer. Chem. Soc., 831, 3748, (1961). [6]. Y.M. de Haan, Physica, 24, 855 (1958).

[7]. A. Caron, J. Donohue, Acta Cryst., 18, 562 (1965). [8]. K.H. Meier, Y. Go, Helv. Chim. Acta, 17, 1081 (1934). [9]. J.A. Prins, J. Schenk, L.H.J. Wachters, Physica, 23, 747 (1957). [10]. L. Pauling, The Nature of the Chemical Bond, Cornell University Press, Ithaca, (1960), p. 136. [11]. J. Donohue, The Structure of Elements, John Wiley & Sons, New York, (1974). [12]. M. Popescu, A. Andriesh, V. Chumash, M. Iovu, S. Shutov, D. Tsiulyanu, Fizica Sticlelor Calcogenice (roum), Ed. Stiintifica, Bucuresti – Ed. Stiinta, Chishinau, 1997.

[13]. C.W. Tompson, N.G. Gingrich, J. Chem Phys., 31, 1598 (1959). [14]. G.B. Abdulaev, D.S. Abdinov, Fizika Selena (russ.), ELM, Baku, (1975), p. 17 [15]. V Zacek, Acta Univ. Carol. Geol., 1988, 315 (1988). [16]. P. Cherin, P. Unger, Acta Cryst., B28, 313 (1972).

THE CHALCOGENS AND THEIR COMBINATIONS

95

[17]. A.T. Andrievski, I.D. Nabitovici, P.I. Kriniakevici, Dokl. Akad. Nauk. SSSR, 2, 124 (1964). [18] V. Detiareva, V. Sikorov, Fiz. Tverd. Tela (russ.), 19, 289 (1977). [19]. Y. Miyamoto, Jap. J. Appl. Phys., 16, 2257 (1977); 19, 1813 (1980) [20]. L.I. Tatarinova, Elektronografia amorfnîh veshcestv (russ.), Nauka, Moskva, 1972, p. 58.

[21]. R. Kaplow, T.A. Rowe, B.L. Averbach, Phys. Rev., 168, 1068 (1968). [22]. H. Richter, J. Non-Cryst. Solids, 8-10, 388 (1972).

[23]. J. Weber, Z. Naturforsch., 8a, 564 (1953). [24]. S. Brieglib, Z. Phys. Chem., A144, 321 (1929). [25]. G. Lucovsky, Phys. stat. sol (b), 49, 633 (1972). [26]. J.C. Jamieson, D.B. McWhan, J. Chem. Phys., 43, 1149 (1965). [27]. J. Stuke, J. Non-Cryst. Solids, 4, 1 (1970).

[28]. D. Jecu, Ph. D. Thesis, Bucharest, 1986. [29]. Entsiklopedia Neorganiceskih Materialov(russ.), Ed. Entsiklopednaia, Ukraine, Kiev,vol. 2, 1977. [30]. N. Ahmetov, Inorganic Chemistry, MIR Publ., Moskva, 1973. [31]. C. Calvo, R.J. Gillespie, J.E. Wekris, H.N. NG, Acta Cryst., B34, 911 (1978). [32]. R.A. Boudreau, H.M. Haendler, J. Solid Slate Chem., 36, 289 (1981). [33]. W.D. Gill, G.B. Street, J. Non-Cryst. Solids, 13, 120 (1973/74). [34]. V.P.Shilo, Ph. D. Thesis, AN SSSR, I.O.N.H., Moskva 1967. [35]. M.Popescu, E. Indrea, N. Aldea, Proc Intern. Conf. Amorph. Semicond. ‘82, Bucharest 1982, CIP Press, p. 135. [36]. E. Grison, J. Chem. Phys., 19, 1109 (1951). [37]. P. Boolchand, P. Suranyi, Phys. Rev., B7, 57 (1973). [38]. J. Sarrach, J.P. de Neufville, W.L. Haworth, J. Non-Cryst. Solids, 22(2), 245 (1976). [39]. W. Hoyer, B. Kunsch, M. Suda, E. Wieser, Z. Naturforsch., 36a, 880 (1981). [40]. J . J . Berzelius, Schwiegger’s Journal, 6, 311 (1812). [41]. L.L. Hawes, Nature, 198, 1267 (1963). [42]. J.W. Mellor, Inorganic Chemistry, V, XI, 111 (1959). [43]. C.J. Smithells, Metals Reference Book, Butterworths, 1978, p.773. [44]. S. Geller, Science, 161, 290 (1968). [45]. D. Jecu, J. Jaklowszky, A. Trutia, I. Apostol, M. Dinescu, I. N. Mihailescu, G. Aldica, M. Popescu,

N. Vlahovici, S. Zamfira, E. Indrea, J. Non-Cryst. Solids, 90, 319 (1987). [46]. G. Zegbe, J. Olivier-Fourcade, J.C. Jumas, I. Deszi, G. Langouche, Europ. J Solid St. Inorg. Chem.,

30(1-2), 165 (1993). [47]. J. Jerger, Patent (USA) 2883295, from 21.04.1959. [48]. L.N. Suvorova, Z.U. Borisova, G.M. Orlova, Izv. Akad. Nauk SSSR, Neorg. Mat., 3(10), 441 (1974). [49]. H. Specker, Angew. Chem., 65(11), 299 (1953). [50]. F. Heyder, D.Linke, Ztschr. Chem., 13(12), 480 (1973). [51]. H. Krebs, Fundamentals of Inorganic Crystal Chemistry, McGraw-Hill, 1968, p. 351. [52]. Y.C. Leung, J. Waser, S. van Houten et al., Acta Cryst., 10(9), 574 (1957). [53]. A. Vos, A. Wiebanga, Acta Cryst., 8(4), 217 (1955). [54]. A.D. Pearson, Lecture at NATO Summer Course on Amorph. Semic., University of Gent, Belgium,

1969. [55]. M.S. Gutenev, Proc. Intern. Conf. “Amorph. Semic., ‘89”, Uzhgorod, URSS, 1989, vol. 1, p. 181. [56]. P.L. Robinson, W.E. Scott, Ztschr. anorg. allg. chem., 210, 57 (1933). [57]. D.L. Price, M. Misawa, S. Susman, T. I. Morrison, G.K. Shenoy, M. Grimsditch, J. Non-Cryst. Solids, 66, 443 (1984). [58]. V.V. Zigheli, G.M. Orlova, Proc. Intern. Conf. “Amorph. Semic ‘74”, 1974, Reinhardsbrunn, RDG, vol. 2, p.272. [59]. L.A. Bajdakov, C.K. Novosilov, Proc. Intern. Conf. “Amorph. Semic. ‘74”, Reinhardsbrunn, RDG, 1974, vol. 3, p. 156. [60]. B.T. Kolomiets, S.S. Lantratova, V.M. Lyubin, V.P. Shilo, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1978, vol. 2, p. 669. [61]. G.Z. Vinogradova, Stekloobrazovanie i Fazovîe Ravnovesia v Halkoghenidnîh Sistemah (russ.), Nauka, Moskva 1984, p.32. [62]. R. Blachnik, A. Hoppe, J. Non-Cryst. Solids, 66, 443 (1984).

96

CHAPTER 1

[63]. R. Blachnik, A. Hoppe, J. Non-Cryst. Solids, 34, 191 (1979). [64]. R. Loudon, J. Phys. (Paris), 26, 677 (1965). [65]. W. Gordy, J. Chem. Phys., 14, 305 (1946). [66]. A.A. Vaipolin, E.A. Porai-Koshits, Fiz. Tverd.Teta (russ.), 5, 683 (1963). [67]. D.W. Henderson, D.O. Ast, J. Non-Cryst. Solids, 64, 43 (1984). [68]. Z. Zohan, C. Laforet, P. Picot, J. Féraut, Bull. Soc. Franc. Mineral. Crist., 96, 131 (1973). [69]. V.T. Iakusev, V.A. Kirkinskii, Dokl. Akad. Nauk SSSR., 186(4), 882 (1969). [70]. K.K. Shvarts, Fizika Opticeskoi Zapisi v Dielektr. i Poluprovod. (russ.), Riga, Zinatne, 1986. [71]. A. Tverjanovich, M.Yagodkina, V. Strykanov, J. Non-Cryst. Solids, 223, 86 (1998). [72]. R.K. Quinn., Mater. Res. Bull., 9(6), 803 (1974). [73]. V.A. Kirkinskiy., V.G. Yakushev, Neorg. Mater., 10(8), 1431 (1974). [74]. S.Toscani, J. Dugue, R. Ollitrault, R. Ceolin, Thermochim. Acta, 186(2), 247 (1991). [75]. H.V. Shu, S. Jaulmes, J. Flahaut, Mat. Res. Bull., 21(12), 1600 (1996). [76]. A.S. Hvorostenko, Halkogenidnîe Mîshiaka (russ.), Obzor, Moskva 1972, State Committee for

Standards, Moskva, 1972. [77]. T. Minami, M. Hattori, F. Nakamachi, M. Tanaka, J. Non-Cryst. Solids, 3, 327 (1970). [78]. G. Lucovsky, Bull. Amer. Phys. Soc. 17, 113 (1972). [79]. I.P. Grigas, A.S. Karpus, Fiz. Tverd. Tela (russ.), 9, 2882 (1967). [80]. N.W. Tideswell, F.H. Kruse, G.D. McCullough, Acta Cryst., 10(2), 99 (1957). [81]. E. Mooser, W.A. Pearson, J. Phys. Chem. Solids, 7, 65 (1958). [82]. V.P. Zaharov, V.S. Gherasimenko, Strukturn. Osobenost. Amorfnîh Poluprovod. (russ.), Ed. Naukova Dumka, Kiev 1976. [83]. B.T. Meleh, Z.V. Maslova, M.S. Ablova et. al., Fiz. Him. Stekla, 2(2), 189 (1976). [84]. S.A. Semiletov, “Kristalografia”, 1, 403 (1956). [85]. E. Henggl, G. Olbrich, Z. anorg. allg. Chem., 365, 321 (1969). [86]. H. Emons, L. Theisen, Z. anorg. allg. Chem., 361, 321 (1968). [87]. A.S. Berejnoi, Kremnii i evo Binarnîe Sistemî (russ.), Kiev, Ed. Akad. UKR SSSR, 1958. [88]. A. Weiss, Z. Naturf. 7(8), 483, 1952. [89]. L. Pauling, The nature of the Chemical Bond, Cornell University Press, Ithaca, 1960, p.444. [90]. J. Peters, B. Krebs, Ada Cryst., B38, 1270 (1982). [91]. M.S. Silverman, J.R. Soulen, Inorg. Chem., 4(1), 125, (1965). [92]. R.W. Johnson, S. Susman, D.L. Price, J. Non-Cryst. Solids, 75, 57 (1985). [93]. K. Ploog, W. Stetter, A. Nowitzki, E. Schönherr, Mat. Res. Bull., 11, 1147, (1976). [94]. G. Dittmar, Ph. D. Thesis, Techn. Hochsch., Darmstadt, 1976. [95]. P. Grigoriadis, G. L. Bleris, J. Stoemenos, Acta Cryst., B39, 421 (1983). [96]. S.A. Altunian, V.S. Minaev, M.S. Minajdinov, B.K. Scacicov, Fiz. Tehn. Poluprovod. (russ.), 4(2), 2214 (1970). [97]. G.E. Bartsch, T. Just, Z. Metallkunde, 63, 360 (1972). [98]. F.R.L. Schoening, J. Mat. Sci., 14, 2397 (1979). [99]. A.V. Nonoselova, M.K. Todria, I.N. Odin, B.A. Popovkin, Neorg. Mat., 7, 1125 (1971). [100]. S.N. Dutta, G.A. Jeffrey, Inorg. Chem., 4, 1363 (1965). [101]. H. Krebs, Acta Cryst., 9, 95 (1956). [102]. A.R. Hilton, C.E. Jones, M. Brau, Infrared Phys., 4, 213 (1964). [103]. M. Imaoka, Asahi Garasu Koguo Gijutsu Shoreikai, Kenkou Hokoku, 13, 421 (1967). [104]. J.A. Savage, S. Nielson, Proc. 7-th Intern. Congress on Glass, Brussel, Belgium, 1965, p. 105. [105]. Y. Kawamoto, S. Tsuchihashi, J. Amer. Ceram. Soc., 52, 626 (1969). [106]. L. Cervinka, A. Hrubý, 5-th Intern Conf. Amorf. Liq. Semic., Garmisch-Partenkirchen, Germany, 1973, Abstr., B22. [107]. Y. Kawamoto, S. Tsuchihashi, J. Amer. Ceram. Soc., 52, 626 (1969). [108]. A. Gmelin-Kraut, Germanium, 8-th edition, 1948, p.52. [109]. Z.U. Borisova, Himia Stekloobraznâh Poluprovodnikov (russ.), Leningrad Univ. Press, SSSR, 1972. [110]. D. Henderson, J. B. Ortenburger, J. Phys.C-Solid State Phys., 6, 631 (1973). [111]. U. Senapati, A.K. Varshneya, J. Non-Cryst. Solids, 185, 289 (1995). [112]. R.W. Fawcett, C.N.J. Wagner, G.S. Cargill, J. Non-Cryst. Solids, 8-10, 64 (1972). [113]. M.Robbins in 4-th Intern Conf. Amorf. Liq. Semic., Ann Arbor, USA, 1971.

THE CHALCOGENS AND THEIR COMBINATIONS

97

[114]. G.H. Ivanov, B.T. Kolomiets, V.T. Lyubin, V.P. Shilo, Proc. Intern. Conf. “Amorph. Semicond. ‘72”, Sofia, Bulgaria, p.I, 88 (1972). [115]. D.I. Bletskan, V.S. Gherasimenko, D.V. Chepur, Proc. Intern. Conf. “Amorph. Semicond. ‘82”,

Bucharest, Romania, 2, 68 (1982). [116]. K. Schuberik, H. Fricke, Z. Metallkunde, 44, 457 (1953). [117]. S.G. Karbanov, V.P. Zlomanov, A.V. Novoselova, Izv. Akad. Nauk SSSR, Ser. Neorg. Mat., 5, 1171 (1969).

[118]. J.A. Savage, J. Non-Cryst. Solids, 11, 121 (1972). [119]. S. Manéglier-Lacordaire, P. Besançon, J. Rivet, J. Flahaut, J. Non-Cryst. Solids, 18, 439 (1975). [120]. D.J. Sarach, J.P. de Neufville, W.L. Haworth, J. Non-Cryst. Solids, 22, 245 (1976). [121]. A. Okasaki, J. Ueda, J. Phys. Soc. Japan, 11, 470 (1956). [122]. M.I. Karahanova, A.S. Pasinkin, A.V. Novoselova, Izv. Akad. Nauk SSSR, Ser. Neorg. Mat., 2, 991 (1966). [123]. R.S. Mitchell, Y. Fujiki, Y. Ishizawa, Nature, 247, 537 (1974). [124]. H.P. B. Rimmington, A.A. Balchin, phys. stat. sol., 6a, K47 (1971). [125]. W. Blitz, W. Macklenburg, Z. anorg. Chem., 64, 226 (1909). [126]. B.I. Boltaks et al., Izv. Akad. Nauk SSR, Ser. Neorg. Mat., 6, 818 (1970). [127]. M.S. Palatnik, V.V. Levitin, Dokl. Akad. Nauk SSSR, 96, 975 (1954). [128]. F. Hulliger, Structural Chemistry of Layer-Type Phases, Ed. F. Levy, D. Reidel, Publ. Co., vol. 5, 1976. [129]. S.I. Novikova, L.I. Selimova, Fiz. Tverd. Tela, 9, 1336 (1967). [130]. S.S. Kabalkina, I.R. Serebrianaia, L.F. Veresciaghin, Fiz. Tverd. Tela (russ.), 10, 733 (1968). [131]. L. Cervinka, A. Hrubý, J. Non-Cryst. Solids, 34, 275 (1979). [132]. Z.U. Borisova, V.R. Panus, N.N. Apihtin, T.S. Fokina, Fiz. Him. Stekla (russ.), 5(3), 308 (1979). [133]. K. Taneva, Z. Boncheva-Mladenova, Montagsh. Chem., 109(4), 911 (1978). [134]. L.H. Slaugh, Inorg. Chem., 3, 920 (1964). [135]. K. Traore, J. Brenet, C.R. Acad. Sci. (Paris), 249, 280 (1959). [136]. E. Diemann, Z. anorg. Allg. Chemie, 431, 273 (1977). [137]. H. Moissan, C.R. Acad. Sci. (Paris), 115, 271 (1892). [138]. J. Economy, V.J. Matkovich, R.F. Gieser, Z. Krist., 122, 248 (1965). [139]. S.W. Martin, D.R. Bayer, J. Amer. Ceram. Soc., 73, 3481 (1990). [140]. Z.G. Jukov, Ia.H. Grinberg, Izv. Akad. Nauk SSR., Ser. Neorg. Mat., 5(9), 1646 (1969). [141]. V.A. Boriakova, Ia. H. Grinberg, Z.G. Jugov, Izv. Akad. Nauk SSR, Ser. Neorg. Mat., 5(3), 477 (1969). [142]. O. Lindquist, Acta Chem-Scand., 22, 977 (1968). [143]. M. Zhang, P. Boolchand, Science, 266, 135 (1994). [144]. K.A. Chontonov, A.N. Kuznetsov, V.M. Fedorov, S.P. Iatsenko, Izv. Akad. Nauk SSR, Ser. Neorg. Mat., 18(7), 1108 (1982).

[145]. W. J. Bresser, J. Wells, M. Zhang, P. Boolchand, Z. Naturforsch., 51a, 373 (1996). [146]. V.S. Minaev, Electron. Tehn. -Ser. Mat., 9(208), 44 (1985). [147]. T.R. Guillermo, B.J. Wünsch, Acta Cryst., B29, 2536 (1973). [148]. N.A. Goriunova, B.T. Kolomiets, V.P. Shilo, Jurn. Eksp. Teoret. Fiz. (russ.), 28, 981 (1958). [149]. Y. Kawamoto, S. Tsuchihashi, Yogyo Jyokai Shi, 77(890), 328 (1969). [150]. G.C. Das, N.S. Platakis, M.B. Bever, J. Non-Cryst. Solids, 15, 30 (1974). [151]. Y. Sawan, F.G.Wakim, M. EI-Gabaly, M.K. El-Rayess, J. Non-Cryst. Solids, 41, 319 (1980). [152]. K.E. Pachali, J. Ruska, H. Thurn, Inorg. Chem., 15, 881 (1976). [153]. G.Z. Vinogradova, S.A. Dembovski, J. Neorg. Him., 16, 2036 (1971). [154]. R.L. Myuller, G.M. Orlova, V.N. Timoteeva, G.I. Ternova, Solid State Chemistry, Ed. Z.U Borisova,

Consultants Bureau, New York 1966, p. 232. [155]. R.L. Myuller, V.N. Timofeeva, Z.U. Borisova, Him. Tverd. Tela (russ.), Ed. Leningrad Univ., 1965, p. 75. [156]. R. Andreichin, M. Nikiforova, E. Skordeva, L. Yurukova, R. Grigorovici, R. Manaila, M. Popescu, A. Vancu, J. Non-Cryst. Solids, 20, 101 (1976). [157]. R.L. Myuller, Solid State Chemistry, ed. Z.U. Borisova, Consultants Bureau, New York 1966, p. 1. [158]. J.P. Suchet, Mat. Res. Bull., 6, 491 (1971). [159]. G.Z. Vinogradova, S.A. Dembovski, A N. Kopeikina, N.A. Lujnaia, J. Neorg. Him., 20, 1367 (1975).

98

CHAPTER 1

[160]. S.A. Dembovski, N.P. Popova, J. Neorg. Mat., 6, 138 (1970). [161]. R. Grigorovici, Amorphous Semiconductors, Ed. J. Tauc, Plenum Press, 1974. [162]. V. S. Minaev, Iu. N. Kuznetov, V.Z. Petrova, A.N. Boguslavski, Proc. Intern. Conf. “Amorph. Semicond. ‘78”, Pardubice, Czechoslovakia, 1, 79 (1978). [163]. G.Z. Vinogradova, S.A. Dembovski, A.N. Kopeikina, N.P. Lujnaia, Proc. Intern. Conf. “Amorph. Semic ‘74”, Reinhardsbrunn, Germany, 2, 268 (1974). [164]. G.M. Orlova, I.I. Kojina, V.G. Korolenko, Fiz i Him.(russ.), 4(1), 90 (1973). [165]. B.K. Vainstein, R.M. Imamov, A.G. Talibov, Kristallografia (russ.), 14, 597 (1970). [166]. I.I. Petrov, R.M. Imamov, Z.G. Pinsker, Kristallografia (russ.), 13, 339 (1968). [167]. M. Frumar, H. Tycha, M. Bures, L. Koudelka, Z. Chem., 15, 199 (1975). [168]. C. Barta, L. Štourac, A. Triska, J. Kocka, M. Závetová, J. Non-Cryst. Solids, 35&36, 1239 (1980). [169]. Z.U. Borisova, A.V. Pazin, Solid State Chemistry, Ed. Z.U. Borisova, Consultants Bureau, New York 1966, p.63.

[170]. R.W. Maistry, H. Krebs, Angew. Chem., 80, 999 (1968). [171]. M.A. Afifi, H.H. Labib, M.H. El-Fazary, M. Fadel, Appl. Phys., A55, 167 (1992). [172]. S.S. Fouad, Physica B, 215, 213 (1995). [173]. M.F. Thorpe, J. Non-Cryst Solids, 57, 350 (1983). [174]. V.S. Minaev, Proc. Intern. Conf. “Amorph. Semicond. ‘78”, Pardubice, Czechoslovakia, 1, 71 (1978). [175]. T. Katsuyama, H. Matsumura, J. Non-Cryst. Solids, 139, 177 (1992).

[176]. N.A. Goriunova, B.T. Kolomiets, V.P. Shilo, J. Tehn. Fiz (russ.), 28(5), 981 (1958). [177]. N.A. Goriunova, B.A. Kolomiets, V.P. Shilo, Fiz. Tverd. Tela (russ.), 2(2), 280 (1960). [178]. E.V. Shkolnikov, Him. Tverd. Tela (russ.), Ed. LGU, Leningrad 1965, p. 199. [179]. Z.U. Borisova, Halkogenidnîe Poluprovodnikovîe Stekla (russ.), Leningrad Univ. Press, SSSR, 1983, p.214. [180]. Iovu M, Shutov, S., J. Optoel. Adv. Mat., 1 ( 1 ) , 27, 1999. [181]. A. Seifert, G.H. Frischat, J. Non-Cryst Solids, 49, 173 (1982). [182]. V.S. Minaev, V.A. Feodorov, Proc. Intern. Conf. “Amorph. Semicond, ‘80”, Khishinau, Moldova, vol. I, p. 106. (Structure of Amorphous Semiconductors.). [183]. R.M. Imamov, S.A. Semiletov, Z.G. Pinsker, Kristalografia, 15, 239 (1970). [184]. T.B. Jukova, A.I. Zaslavskii, Kristallografia, 16, 796 (1972). [185]. A. Feltz, E. Schlenzig, D. Arnold, Z. anorg. allg. Chem., 403, 243 (1974). [186]. J. Fenner, D. Mootz, Z. anorg. allg. Chem., 427(2), 123 (1976). [187]. S.C. Rowland, F. Rittland, D. Haferburns, A. Bienenstock, preprint USA, 1976. [188]. V.S. Minaev, Iu. N. Kuznetov, V.A. Fedorov, Proc. Intern. Conf. “Amorph. Semic. ‘82”, Bucharest,

Romania, 2, 100 (1982). [189]. V.S. Minaev, V.M. Glazov, I.G. Aliev, V.A. Fedorov, Proc. Intern. Conf. “Amorph. Semic. ‘84”,

Gabrovo, Bulgaria, 1, 204 (1984). [190]. A. Feltz, B. Künzel, I. Linke, Z. anorg. allg. chem., 437, 237 (1977). [191]. E. Glatzel, Z. anorg. Chem., 4, 186 (1983). [192]. C. Wibbelman, W. Brockner, Z. Naturf., A36, 836 (1981). [193]. A. Tverianovici, M. Krüger, M. Soltvisch, D. Quitmann, J. Non-Cryst. Solids, 130, 236 (1991). [194]. R. Blachnik, A. Hoppe, J.Non-Cryst. Solids, 34, 191 (1979) [195]. D. Linke, H. Hey, Zeitschr. Chem., 16(10), 412 (1976). [196]. V.V. Bakulina, Z. U. Borisova, B.E. Kasatkin, Vestnik Leningradsk Univ., 16, 101 (1974). [197]. G.Z. Vinogradova, N.G. Maisashvili, I.Z. Babievskaia, O.I. Djaparidze, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 83 (1978). [198]. G.Z. Vinogradova, Stekloobrazovanie i Fazovîe Ravnovesia v Halkoghenidnîh

Sistemah, Ed.Nauka, Moscova, 1984, p. 104 [199]. G.Z. Vinogradova, N.G. Maisashvili, J. Neorg. Him., 24(4), 1 1 1 6 (1979). [200]. P.P. Serioghin, L.N. Vasiliev, A.A. Pronkin, Izv Akad. Nauk SSSR-Ser. Neorg. Mat., 8(2), 376 (1972) [201]. A.R. Hilton, C.E. Jones, M. Brau, Phys. Chem. Glasses, 7, 105 (1966). [202]. E.V. Shkolnikov, in Himia Tverdovo Tela (russ,), Ed. LGU. Leningrad, SSSR, 1965, p. 115.

[203]. D. Linke, Proc. Intern. Conf. “Amorph. Semic. ‘72”, Sofia. Bulgaria, 1, 19 (1972). [204]. H.E. Anthonis, N.J. Kreidl, W.H. Ratzenboeck, J. Non-Cryst. Solids, 13, 13 (1973/1974).

THE CHALCOGENS AND THEIR COMBINATIONS

99

[205]. J.P. de Neufville, J. Non-Cryst.Solids, 8-10, 85 (1972). [206]. B. Stepánek, A. Hrubý, Proc. Intern “Conf. Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 88 (1978).

[207]. B. Stepánek, A. Hrubý, J. Non-Cryst.Solids, 37, 343 (1970). [208]. B.W. King, G.D. Kelly, J. Amer. Ceram. Soc., 41, 367 (1958). [209]. D.R. Ulrich, J. Amer. Ceram. Soc. 47, 595 (1964). [210]. R. Andreichin, M. Nikiforova, E. Skordeva, Proc. Intern. Conf. “Amorph. Semic. ‘74”, Reinhardsbrunn, Germany, 1, 224 (1974). [211]. V.P. Shilo, B.T. Kolomiets, Izv. Akad. Nauk SSSR-Ser.Him., 28, 1285 (1964). [212]. B.T. Kolomiets, Stekloobraznoe Sostoianie (russ.), Ed. Nauka, Moskva, 1971, p. 82 [213]. T. Minami, M. Hihino, M. Tanaka, J. Non-Cryst.Solids, 15, 141 (1974) [214] M. Zhang, P. Boolchand, Science, 266, 1355 (1994). [215] M. A. Sidkey, R. El. Mallawany, R. I. Nakhla, A. Abd El-Moneim, J. Non-Cryst. Solids, 215, 75 (1997).

[216]. V. Bogus, Ph. D. Thesis, Bucharest University, 1986. [217]. G.V. Golubkova, E.S. Petrov, A.N. Kanev, Izv. Akad. Nauk SSSR-Ser.Him. Nauk (russ.), 5, 39 (1975). [218]. D Negoiu, Tratat de Chimie Anorganica (roum.), vol. 2, Ed. Tehnica, Bucharest, 1972. [219]. D. Jecu, Ph. D. Thesis, Bucharest University, 1987, p. 83. [220]. V.A. Ignatiuk, N.V. Stavnistii, E.N. Minaev, Proc. Intern. Conf. “Amorph. Semic. ‘80”, Chishinau, Moldova, 1, 117 (1980). [221]. W. Bauhofer, R. Kniep, Mat. Res. Bull., 8(8), 989 (1973) [222]. W.R. Blackmore, S.C. Abrahams, J. Kalnajs, Acta Cryst., 9(1), 295 (1956). [223]. Ma Hong Li, Zhang Xiang Hua, Jacques Lucas, J. Non- Cryst.Solids, 135, 49 (1991). [224]. I.M. Rudnev, V.A. Ignatiuk, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 237 (1978). [225]. X. Zhang, J. Lucas, G. Fonteneau, Proc. 4-th Intern Symp. on Halide Glasses, Monterray, Ca.,

Jan. 1987 [226]. V.S. Mineaev, Iu. N. Kuznetov, V.Z. Petrova, A.N. Boguslavski, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 79 (1978)

[227]. S.S. Flaschen, A.D. Pearson, W.R. Northover, J. Appl. Phys., 31, 219 (1960). [228]. S.A. Dembovskii, N.P. Popova, Izv. Akad Nauk SSSR-Ser Neorg. Mat., 6, 138 (1970) and 7, 328 (1971). [229]. V.F. Kokorina, L.G. Aio, E.A. Kislitskaia, V.V. Melnikov, Proc. 5-th Intern. Conf. “Amorph. Liq. Semic, ”, Leningrad, 1975, vol.1, p.39.

[230]. N.T. Gurin, V.V. Himinets, D.G. Semak, I. D Turianitsa, V. V. Fedak, Proc. 5-th Intern. Conf. “Amorph. Liq. Semic.”, Leningrad, 1975, vol. 1, p. 500.

[231]. [232]. [233]. [234].

J.S. Sanghero, J. Heo, J.D. Mackenzie, J. Non-Cryst. Solids, 103, 155 (1988). Guorong Chen, Jijian Cheng, Wei Chen, J. Non-Cryst. Solids, 215, 301, 1997. Z.U. Borisova, Izv. Akad. Nauk SSSR-Ser. Neorg. Mat., 7(10), 1720 (1971). Z.U. Borisova, in Struktura i Svoistva Nekristaliceskih Poluprovodnikov (russ.),

Ed. Nauka, Leningrad, 1976, p 6-12. [235]. Iu. S. Tverianovici, Z.U. Borisova, J. Non-Cryst. Solids, 90, 405 (1987). [236] J. Hautala, P. C. Taylor, J. Non-Cryst. Solids, 103, 155 (1988). [237]. V.A. Bazacuta, Proc. Intern. Conf. “Amorph. Semic. ‘74”, Reinhardsbrunn, Germany, 1, 186 (1974). [238]. S. I. Beruli, V.S. Lazarev, A.V. Sapov, J. Neorg. Him., 16(2), 3363 (1971); 22(5), 1437 (1977). [239]. Z.U. Borisova, Proc. Intern. Conf. “Amorph. Semic. ‘82”, Bucharest, CIP Press, 2, 8,1982. [240]. B. Barrau, M. Ribes, M. Maurin, A. Kone, J. L. Sauquet, J. Non-Cryst. Solids, 1, 37 (1980). [241]. M. Khrishnamurthy, J. Aguayo, J. Amer. Ceram. Soc., 47, 444 (1964) [242]. A.V. Pazin, Proc. Intern Conf. “Amorph. Semic. ‘80”, Khishinau, Moldova, vol. I, p.82, 1980. [243]. M.I. Golovei, E.E. Semrad, E. Iu Peres in Halkoghenidî (russ.), Ed. Naukova Dumka, Kiev, SSSR, V.3, 1974, p. 41.

[244]. M. Ribes, B. Barrau, J.L. Souquet, J. Non-Cryst. Solids, 38/39, 271 (1980). [245]. B.T. Kolomiets, N.A. Goriunova, V.P Shilo, in Steklobraznoe Sostoianie, Ed. Akad. SSSR, Moskva 1960, p. 456. [246]. L. Zhenhua, J. Non-Cryst. Solids, 127, 298 (1991).

100

CHAPTER 1

[247]. Ia. Savan, I.I. Kojina, G.M. Orlova, H. Binder, Izv. Akad Nauk SSSR- Ser. Neorg. Mat., 5(3), 492(1969). [248]. Lukic S.R, Petrovic D.M., J Optoel. Adv. Mater., 1(4), 43, 1999. [249]. S.H. Hunter, A. Bienenstock, T.M. Hayes, Proc. 7-th Intern. Conf. Amorph. Liq. Semic., Edinburg, 1977, ed. W.E. Spear. [250]. V.R Panus, N.A. Alimbarasvili, Z.U. Borisova, Fiz. Him. Stekla, 1(3), 221(1975). [251]. C. Jijian, L. Zhenhua, J. Chim. Ceram Soc., 16, 312 (1988). [252]. V.S. Minaev, V.T. Shipatov, V.N. Kiselev, Izv. Acad. Nauk SSSR-Ser. Neorg. Mat., 16(8), 1481 (1980) [253]. S.I. Radautsan, R.A. Maslianko, R.lu. Liapikova, V.G. Koloshkova, Izv Akad. Nauk SSSR-Ser. Neorg. Mat., 11(8), 1508(1975). [254]. J.K. Kom, J. Flahaut, L. Domange, C.R. Acad. Set. (Paris), 257(25), 3919 (1963). [255]. Y. Kawamoto, M. Agata, S. Tsuchihhashi, Yogyo-Kyokai-shi, 82, 46 (1974). [256]. E.A. Kazenkova, B.E. Kasatkin, T.S. Rikova, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 371 (1978). [257]. M.I. Golovei, Ph. D. Thesis, Kiev, 1974. [258]. I. Bunget, M. Popescu, Physics of Solid Dielectrics, Elsevier, 1984, p. 53. [259]. B.I. Boltaks, T.D. Djafarov, H.Hudoiarova, N,F. Kartenko, R.M. Imamov, A.A. Obraztsov, Fiz. Tehn. Poluprovod. (russ.), 13, 41 (1979). [260]. D. Ruffolo, P Boolchand, Phys. Rev. Lett., 55, 242 (1985). [261]. T. Furukawa, W.B. White, J. Non-Cryst. Solids, 38&39, 87 (1980). [262]. K. Tanaka, J. Non-Cryst. Solids, 170, 27 (1994). [263]. Z.U. Borisova, Proc. Intern. Conf. Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 65 (1978). [264]. M. Mitkova, Z. Boncheva-Mladenova, J. Non-Cryst. Solids, 90, 589 (1987). [265]. Y Kawamoto, M. Nishida, J. Non-Cryst. Solids, 20(3), 393 (1976). [266]. J.K. Kom, J. Flahaut. L. Domange, C.R. Acad Sci (Paris), 255(4), 701 (1962). [267]. N A. Goriunova, B.T. Kolomiets, V. P. Shilo, J. Tehn. Fiz. (russ.), 28(5), 981 (1958). [268]. A. A. Andreev, Z.U. Borisova, E.A. Bicikov, V.G. Vlasov, J. Non-Cryst. Solids, 35&36, 901 (1980). [269]. R. Flasch, M. Izu, K. Sapru, T. Anderson, S.R. Ovshinsky, H. Fritzsche, Proc. Intern. Conf. “Amorph. Liq. Semic.”, Edinburg, 1977, p. 524 [270]. R.R. Chianelli, M.B. Dines, Inorg. Chem., 17, 2758 (1978). [271]. K.S. Liang, J. P. de Neufville, A.J Jacobson, R. R. Chianelli, J. Non-Cryst. Solids, 35&36, 1249 (1980). [272]. M.R. Allazov, P.K. Babaeva, P.G. Rustamov, Izv Akad. Nauk SSSR-Ser. Neorg. Mat., 15, 1177 (1979). [273]. Z.U. Borisova Proc. Intern. Conf. “Amorph. Semic. ‘84”, Gabrovo, Bulgaria, 1, 145 (1984). [274]. A.A. Dunaev, Z.U. Borisova, M.D. Mihailov, A.V. Bratov, Fiz. i Him. Stekla (russ.), 6(2), 174 (1980). [275]. T. Katsuyama, H. Matsumura, J. Non-Cryst. Solids, 139, 177 (1992). [276]. A. Feltz, A. Krautwald, Zeitschr. Chem., 19(2), 78 (1979). [277]. Z. Boncheva-Mladenova, Z.C Ivanova, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 103 (1978). [278]. A.A. Dunaev, Z.U. Borisova, M.D. Mihailov, I. V. Privalova, Fiz. i Him. Stekla (russ.), 4(3), 346 (1978). [279]. Z.U. Borisova, A.I. Bobrov, Solid State Chemistry, Consultants Bureau, New York, 1960, p. 247. [280]. A.A. Dunaev, M.D. Mihailov, Z U. Borisova, Fiz. i Him. Stekla (russ.), 5(3), 370 (1979). [281]. Z. G. Ivanova, Ph. D. Thesis, Sofia, 1980 [282]. T. Matsushita et al., Jap. J. Appl. Phys., 11, 417 (1972). [283]. Z. Boncheva-Mladenova, M. Mitkova, Proc. Intern. Conf. “Amorph. Semic. ‘74”, 2, 244 (1974). [284]. V.S. Minaev, V.T. Shipatov, V.N. Kiselyov, Izv. Akad. Nauk SSSR-Ser. Neorg. Mat., 16(8), 1481 (1980). [285]. N.N. Apihtin, M.D. Mihailov, V.R. Panus, T.N. Salamatova, Fiz. i Him. Stekla (russ.), 6(4), 383 (1980). [286]. M. E. Fleet, Z. Krist., 138, 147 (1973). [287]. F. Pernot, Ph. D. Thesis, Paris, 1971, p. 51

THE CHALCOGENS AND THEIR COMBINATIONS

101

[288]. V.K. Nikitina, Proc. Intern. Conf. “Amorph. Semic.”, Leningrad 1975, Vol. 1, Structure and Properties of Amorph. Semic., p.44. [289]. D. Linke, Proc. 8-th Arbeitstagung “Elektronenmikroscopie”, 1975, p. B 257. [290]. S.A. Dembovskii, J. Neorg. Him., 13(6), 1721(1968) [291]. Fiziceskie Svoistva Halkoghenidnih Redkozemelnih Elementov (russ.), Editor V.P. Juze, Nauka, Leningrad 1973 [292]. W.H. Zachariasen, Acta Cryst., 2, 57 (1949). [293]. S. Barnier, M. Guittard, C.R. Acad. Sci., (Paris), 282, 461 (1976). [294]. E.M. Raspopova, V.A. Mesloboev, L.I. Polejaeva, B.T. Kolomiets, N.N. Kalinina, V.P. Shilo, Proc. Intern. Conf. “Amorph. Semic. ‘78”, Pardubice, Czechoslovakia, 1, 162 (1978).

[295]. A.M. Lozac’h, S. Barnier, M. Guittard et al., Chimie Infrarouge des Solides, Paris, 1974, p. 127. [296]. T.M. Iliasov, P.G. Rustamov, L.A. Mamedova, Proc. Intern Conf. Amorph. Semic. ‘84”, Gabrovo, Bulgaria, 1, 106 (1984) [297]. K. Kadomo, H. Miguchi, M. Takahashi, Y. Kawamoto, H. Tanaka, J. Non-Cryst. Solids, 184, 309-313(1985). [298]. I.D. Olekseyuk, A.N. Borets, T.V. Shopko, I.I. Spak, Proc. Intern. Conf. “Amorph. Semic. ‘89”, Uzhgorod, SSSR, 1, 133 (1989). [299]. Svoistva Sveto-Ciustvitelnih Materialov i ih Primenenie v Golografii (russ.), Ed. Nauka, Editor V.A. Barachevskii, 1987, p. 78.

[300]. A. Feltz, W. Burchardt, Z. anorg. allg. Chem., 461(1), 35 (1980). [301]. A. Feltz, W. Burchardt, J. Non-Cryst. Solids, 41, 301 (1980). [302]. J.I. Petz, R.F. Kruh, G.C. Amstutz, J.Chem. Phys., 34(2), 526 (1961). [303]. Z.U. Borisova, Himia Stekloobraznih Poluprovodnikov (russ.), Leningrad, Ed. LGU, 1972. [304]. A. Feltz, B. Voigt, Z. annorg. allg. Chem., 403, 61 (1974). [305]. A.V. Pazin, Z.U. Borisova, Appl. Chem., SSSR, 32, 1225 (1970). [306]. M. Frumár, H. Tichá, L. Koudelka, J. Faimon, Mat. Res. Bull., 11, 1389 (1976). [307]. N. Tohge, T. Minami, Y. Yamamoto, M. Tanaka, J. Appl. Phys., 51, 1048, (1980). [308]. P. Nagels, L. Tichý, H. Tichá, A. Tríska, in Physics of Disordered Materials, Institute of Amorphous Studies, Plenum Press, 1985, p. 645. [309]. L. Tichý, H. Tichá, L. Bénés, J. Malek, J. Non-Cryst. Solids, 116, 206 (1990). [310]. P.T. Oreskin, S.P. Vihrov, V.N. Ampilogov, A.A. Babaev, V.V. Himinets, Proc. Intern. Conf. “Amorph. Semic. ‘84”, Gabrovo, Bulgaria, 1, 192 (1984). [311]. W. Hillegas, J. Neyhart, J. Non-Cryst. Solids, 27, 347 (1978). [312]. P. Hagenmuller, F. Chopin, C.R. Acad. Sci. (Paris), 255(180), 2259 (1962). [313]. V.A. Boriakova, Ia. H. Grinberg, Z.G. Jukov, Izv. Akad. Nauk SSSR-Ser Neorg. Mat., 5(3), 477(1969). [314]. L.I. Doinikov, Ph. D. Thesis, Leningrad Univ., 1968. [315]. L.I. Doinikov, O.V. Ilinskaia, Z.U. Borisova, J. Prikhlad. Him., 37(6), 1217 (1964). [316]. V.V. Kiripenko, S.A. Dembovski. V.T. Kalinikov, Izv. Akad. Nauk SSSR-Ser. Neorg. Mat., 10(3), 542(1974).

[317]. S.A. Dembovskii, V.V. Kiripenko, Izv. Akad. Nauk SSSR-Ser. Neorg. Mat., 7(3), 510 (1971). [318]. L.I. Doinikov, Z.U. Borisova, in Himia Tverdovo Tela (russ.), Ed. LGU, Leningrad, 1965, p. 93. [319]. A.P. Cernov, S.A. Dembovskii, V.T. Kalinikov, Izv. Akad. Nauk SSSR-Ser. Neorg. Mat., 16(1), 42 (1980). [320]. N.P. Lujnaia, S.A. Dembovskii, V.V. Lavrovskaia, J. Neorg. Himii, under press. [321]. Fizico-Himiceskie Svoistva Poluprovodnikovih Veshcestv (russ.), Moskva, Nauka, 1979. [322]. A.P. Chernov, S.A. Dembovskii, N.P. Lujnaia, Proc. Intern. Conf. “Amorph. Semic. ‘80”, Chishinau,

Moldova, 1, 119(1980). [323]. M. Iovu, S. Shutov, M. Popescu, D. Furniss, L. Kukkonen, A.B. Seddon, J. Optoel. Adv. Mat., 1(2), 15, 1999. [324]. S.R. Ovshinsky, Patent USA No. 3, 271, 591 from Sept. 6, 1966. [325]. S.R. Ovshinsky, Phys. Rev. Lett., 21, 1450 (1968). [326]. S.R. Ovshinsky, J. Non-Cryst. Solids, 2, 99 (1970). [327]. S.V. Phillips, R.E. Booth, P.H. McMillan, J. Non-Cryst. Solids, 4, 510 (1970). [328]. G.V. Bunton, J. Non-Cryst. Solids, 6, 72 (1971)

102

CHAPTER 1

[329]. K.E. Petersen, J. Appl. Phys., 47, 256 (1976). [330]. J. Takashima, Electron Parts and Mat., 8, 59 (1969). [331]. S.C. Moss, J.P. de Neufville, Mat. Res. Bull., 7, 423 (1972). [332]. S.R. Ovshinsky, Patent (Fr) No. 2103895; Patent (USA) Nr. 3530441. [333]. N.T. Gurin, V.V. Himinets, D.G. Semak, I.D. Turianitsa, V.V. Fedak, Proc. Intern.Conf. “Amorph. Liq. Semicond, ”, Leningrad 1975.

[334]. J. Xu, R. Yang, Q. Chen, W. Jiang, H. Ye. J. Non-Cryst. Solids, 184, 302 (1995). [335]. J. Chen, W.Chen. D. Ye, J. Non-Cryst Solids, 184, 124 (1995). [336]. N.A. Heab, M. Fadel, M.M. El-Samanoudy, J. Mat. Sci., 30, 5461 (1995). [337]. Y. Ivanova, A. Yoleva, V. Dimitrov, Y. Dimitriev, D. Lepkova, Mat. Sci. Eng., B26, 197 (1994). [338]. S. Barnier, M. Guittard, C. Julien, Mat. Sci. Eng., B7, 209, (1990). [339]. S. Barnier, M. Guittard, M. Palazzi, M. Massot, C. Julien, Mat. Sci. Eng., B7, 209 (1990). [340]. B. B. Harbison, C.I. Merzbacher, I.D. Aggarwal, J. Non-Cryst. Solids, 213&214, 16 (1997). [341]. S.L. Kuznetsov, M.D. Mikhailov, I. M. Pecheritsyn, E. Yu. Turkina, J. Non-Cryst. Solids, 213&214, 68(1997) [342]. P. Nemec, J. Oswald, M. Frumár, B. Frumárova, J. Optoel. Adv. Mater., 1(4), 33 (1999). [343]. I.D. Olekseyuk, O.V. Parasyuk, V.V Bozhko, I.I. Petrus, V.V Gsalyan, Functional Materials, 6(3), 474(1999). [344]. D.M. Petrovic, F. Skuban, S.R. Lukic, M.M. Garic, Functional Materials, S.J. Skuban, 6(3), 478 (1999). [345]. M. Mitkova, Yu. Wang, P. Boolchand, Phys. Rev. Lett., 83(19), 3848 (1999).

CHAPTER 2

103

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENS AND BINARY CHALCOGENIDE GLASSES

2.1. General Properties 2.1.1. METASTABILITY OF GLASS Among the aggregation states of the matter the non-crystalline state is placed in a special position. Although, usually, this state is related to the solid crystalline state, many noncrystalline substances are either viscous or even approach a liquid-like state, a counterpart of the liquid crystalline state. When the gas phase is condensed into a liquid state, the translation motion of the molecules is hindered and this favours the formation of groups of atoms strongly bonded. The thermal oscillations of the particles with the frequency Hz lead to atomic collisions and, consequently, to small increase of the mean inter-atomic distance. This is important for the mobility specific to the liquid state. The flowing property (viscosity) depends not only on the amplitude of the particle oscillations (atoms or molecules) but also on the character and strength of the chemical bonds between particles. The chemical bond is the fundamental cause that determines the transition from the liquid state either to crystalline or to non-crystalline state. The supersaturated solutions, from which the homogeneous nucleation is possible, represent in fact micro-heterogeneous systems. After Freckle, submicroscopical nuclei (seeds) are present not only in supersaturated but also in saturated solutions. Therefore, the liquid structure can be regarded as built from particulate aggregates of short existence, the so-called kinetical units. By cooling, the liquid structure, defined by the spatial disposal of the kinetical units, is subjected to a permanent modification whose magnitude decreases with the temperature and there is possible that, at a given temperature, the relative configuration of the kinetical units is frozen. The above model can serve to the characterisation of the vitreous structure and the freezing temperature will be the (softening temperature). A similar effect can be brought by the increase of pressure. For a given temperature it exists a critical pressure, that triggers the transformation liquid-glass. The vitreous solidification is related to an appropriate cooling rate. This rate must be enough high to overcome the rate of formation of the crystalline seeds and the rate of diffusion of the kinetical units, which determine their growth. Thus, the vitreous solidification is produced without the formation of new phases. The vitrification occurs in a temperature range centred on The resulting structure at temperatures higher than differs from the equilibrium structure of the liquid

104

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

(obtained during slow cooling) and will freeze definitely somewhat under This is the structural cause of the dependence of on the cooling rate. Below the melting temperature, both seed formation and growth show a maximum rate corresponding to a given temperature. This temperature is situated between and The crystallisation centres are in fact already present in the bulk of the melt but the ampoule walls can also induce the crystallisation. Very small drops can be easier undercooled than the macroscopic systems because, in the absence of the walls, only the homogeneous nucleation in the bulk is possible. The cooling rate between and depends on composition and cooling procedure. For small amount of substance of the order of grams, the air cooling from moderate temperatures ensures an average cooling rate of 1-3 °C/s. In many cases the vitreous solidification is accompanied by the separation, at a lower or at a larger scale, as a function of the cooling rate, of submicroscopical domains of definite compositions, called vitreous microphases. The general properties (mechanical, chemical, thermal, electrical, optical, etc.) are greatly influenced by the type of microphases, their concentration and relative configuration. The separation process consists in the splitting of a homogeneous phase in two or more non-miscible phases. Above a given temperature (co-solvating temperature) the melt is homogeneous and monophasic. The separation of the homogeneous phase in two different phases is produced if the free energy of the diphasic system is lower than that of the monophasic system. The separation process can be avoided only for compositions situated outside the non-miscibility range. In the miscibility range it is possible to avoid the microphase separation by melting the mixture above the co-solvating temperature followed by very high cooling rate. The additives can also reduce the tendency towards micro-separation. Balta [1] has explained the microphase separation in the frame of the theory of polymer solutions. The inorganic polymeric phase (this is the case of the main chalcogenide glasses) splits into two solutions, which differ by the concentration of every polymer, and by their distribution. In the more diluted solution there are polymers with lower molecular weight [2]. The crystalline lattice is characterised by the presence of the chemical and geometrical rules for the atom arrangement while the geometrical order is lacking in the amorphous network. The transition from the liquid to crystalline state implies the spatial rearrangement of the particles. This process becomes very difficult if the particles are built from different kinds of atoms. On the other hand, the amplitude of the thermal oscillations of the particles depends on the nature of the chemical bonds between atoms, and, consequently, on the chemical composition of the material. If the chemical forces between atoms are spherically symmetrical, very polar metallic or are intermolecular weak forces, then the oscillation amplitudes are high. This fact leads to high mobility of the particle in the melt, low viscosity that determines the transformation of the liquid in an ordered, crystalline, solid. For such type of materials, the increase of the distance between the particles leads to a weak and slow decrease of the bonding energy (Fig. 1a). This is the cause of the rapid arrangement of the particles during the crystallisation process of the melt for small energy losses. In the case of

CHAPTER 2

105

short-range directional forces in chemical bonds, e.g. covalent or dipolar bonds, small atom displacements imply a significant loss of chemical bond energy (figure 2.1 b). The

particle arrangement in the melt during cooling is, therefore, hindered and the disordered liquid-like structure will be frozen.

Figure 2.1. The potential energy U as a function of the inter-atomic distance x. a - for polar materials: b - for covalent materials:

As a consequence of the disorder in the spatial distribution of the particles, the amorphous materials will exhibit, essentially, isotropic properties. The particular properties of the crystals (e.g. the presence of cleavage surfaces) will be absent.

As shown in Fig. 2.1, for a given energy E, necessary to shift the point A with in the case of polar materials and in the case of covalent materials, one gets For the shift of the particle A in both types of materials with the same value it is necessary, in the first case, to release a certain amount of energy, significantly lower than the energy while in the second case the ordering is obtained on the account of a small activation energy necessary to shift the particles apart from the equilibrium position with the distance In the case of the ionic bonds the breaking of the bonds

106

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

is not produced as a consequence of the superposition of the potential fields, as opposite to the case of covalent materials, where the particle shift away with the same distance leads to bond breaking by the release of the dissociation energy because the superposition of the potential curves does not occur. The amorphous phase represents an intermediary state between liquid and crystal, and is unstable. The stability is an important property of the disordered material. Dembovski [3] has shown that the glass-forming ability, C, is related to the number of lone-pair electrons, by the relation

where A is the number of atoms and E is the number of structural units of various kind in a glass forming system. The influence of lone-pair electrons on the formation and properties of the chalcogenide glasses was studied by many scientists, beginning with Kastner [4] who called these substances “lone-pair semiconductors”. Kastner has shown that the valence bond is formed by the lone-pair electrons which occupy the middle part between and orbitals. It has been shown that the existence of lone-pair electrons causes a number of peculiarities of structure and chemical bond such as the existence of bond free solid angle,

appearance of atoms with variable valence, charge and co-ordination. These peculiarities are the causes of the specific network defects, which are absent in the substances, which do not contain lone-pair electrons. The correlation between glass forming ability, C, and the microscopic parameter is illustrated in Fig. 2.2. For the glass formation field is I. corresponds to the non-glass formers. corresponds to the hyperbola III which defines the boundary between the domain II and I. It is possible to change simultaneously two parameters, e.g. cooling rate and composition, without leaving the glass formation field.

Figure 2.2. The dependence of the number of lone-pair electrons upon I - domain of glass formers II - domain of non-glass formers III - the hyperbola which corresponds to the equation

parameter.

CHAPTER 2

107

The amorphous state is metastable and the free energy does not reach the absolute minimum value. From the phenomenological point of view, the non-crystalline state is explained on the basis of the step rule of Ostwald [5]. According to this rule the system moves from the initial less stable to final stable state by a series of intermediary states characterised by a gradual increase of their stability. The rule gives the possibility to consider the metastable structures as related to the high-temperature modifications or to high-pressure modifications. In thin amorphous films it is, also, possible to get metastable structures. In some cases, the thickness of the film is important, in other cases the temperature and the nature of the substrate or the pressure of the gaseous traces play the leading role. Under the action of various physical factors (temperature, pressure, mechanical shocks, etc.) the metastable modifications can be transformed into one or more crystalline structures, depending on the composition of the material. In most cases the short-range order of the atomic network, characteristic to the crystalline state, is distorted in the amorphous state and this distortion, according to Joffe and Regel’s rule, strongly influences the properties of the amorphous materials. The driving force of the crystallisation process depends on the relation between internal energy and entropy in the crystalline and amorphous states, as reflected by the Gibbs equation: G = H-TS. Because the amorphous material does not exhibit geometrical order, its internal energy or enthalpy is greater than for the crystalline state of the same composition. In the same time, the entropy of the amorphous material is higher than the entropy of the crystal. As a consequence, for a given composition and temperature it is possible, in principle, to fulfil the condition:

and the amorphous state becomes the thermodynamically stable state. Such amorphous materials, cannot be practically crystallised in the temperature range (softening temperature - melting temperature). The free energy of the non-crystalline (amorphous or vitreous) phase, which is released during crystallisation, plays an important role in the determination of its structural stability and influences significantly the development of the crystallisation process. In several cases, the transformation of the amorphous phase into crystalline phase is accompanied by the release of an amount of heat equal to around half of the melting heat of the bulk phase. The thermo-physical and mechanical properties of the chalcogenide glasses depend on the structural and energetical factors. The formation in the glass of structural units at the atomic scale is correlated to the bond energy and to the preparation conditions of the glass. The bond energy is, nevertheless, the leading factor. Table 2.1 shows the values of the average molar bond energy, E (kJ/mol), for various chalcogenide pairs and pairs of elements involved in the glass structure [6].

108

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The important thermal effects that occur during transition from crystal to glass cannot be explained only by the modification of the long-range order. Because in glasses

no long-range Coulomb forces exist, the bonding between the groups of chemically bonded atoms is realised on the account of the short-range covalent and “residual” forces and this explains the small contribution of the long-range energy and the weak dependence of the energy of such solid materials on the degree of structural ordering. Consequently, the energy of such materials in the crystalline state differs only by a small amount from their energy in the glassy state. The contribution of the long-range energy does not overcome 2-3 % of the total bond energy.

2.1.2. PREPARATION TECHNIQUES Glasses For the preparation of chalcogenide glasses by melt cooling, the best ampoules are made of graphite, glassy carbon or high temperature oxides as e.g. silicon oxide (quartz). Low

network energy of chalcogenide materials requires relatively low melting temperatures, which determine a considerable diminishing of the ampoule wall corrosion if compared to the case of oxide glasses. The high vapour pressure of chalcogenide melts and the tendency, especially at higher temperatures, to react with oxygen, requires the work in closed systems under vacuum conditions. The chalcogenide glasses are, therefore, prepared, as a rule, by melting the corresponding mixture of the elements in quartz ampoules sealed under high vacuum conditions. For the sake of homogeneity, either the ampoules are rotated in the oven, or the oven itself is subjected to periodical change of position or even to large periodical oscillations. The melt quenching is realised by throwing the ampoules in a cold liquid (usually water) by decreasing the electric power applied to the oven, following a time schedule, or by simply letting the oven to cool in normal atmosphere after turning off the electrical power.

CHAPTER 2

109

Because the thermal expansion coefficients of the chalcogenide glasses are enough large, the samples exhibit a marked tendency towards the formation of internal stresses, so that it is necessary to perform a mechanical processing of the glass consisting of thermal annealing in the range situated around followed by slow cooling down to room temperature. Amorphous precipitates In some cases it is possible to transform the dissolved substances, by specific

decomposition reactions, into amorphous or insoluble powder. In the chemistry of water solutions the problem is to establish the adequate conditions for getting stoichiometric, filterable, crystalline substances. Several chalcogenide compounds with stoichiometric composition were obtained in amorphous form by decomposition in water solutions with and and also are known only as amorphous compounds. and were obtained by decomposition of thio- and seleno-hypodigermanates, respectively. They can be also prepared as glassy compounds by melt quenching. Other compounds are obtained as amorphous precipitates: (from (from ) and (from powder was obtained from and in 2-methoxyethyl-ether. Thin films

a) Thermal deposition by vacuum evaporation. The thin films obtained by this procedure are porous, show pinholes and in normal atmosphere rapidly crystallise. The softening temperature and vapour pressure of a given material must be known in order to use the highest substrate temperature and the as slow as possible deposition. A deposition rate of 0.1 - 0.5 nm/s at the substrate temperature of 298 K must be used for the case of in order to get an enough large surface mobility so that thick films be obtained. The properties of these Ge-Se-Te films are not far from the bulk glasses. b) Electron and laser ablation. In the ablation method it is possible to get thin amorphous films. Electron or laser beams of high energy are used in order to ablate the material from a target. In the process the high-energy particles are transported as a plasma product and deposited on a substrate. High local temperature in targets can be monitored without the danger of the influence of the walls of the vessel or of the material of the evaporation source. c) Flash evaporation. In this method free falling within an oven melts small size particles. Wide range of fragments deposited on the substrate gives rise to a very disordered amorphous film.

d) Plasma-jet evaporation. In this method high temperature plasma is used. A complete evaporation and a high rate of the vapour beam are obtained. The directional beam which determines the vaporisation is characterised by particle speed of ~ 20 km/s.

110

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

e) Cathode sputtering. In order to get amorphous films one starts from an inert gas (usually argon) under low pressure wherein a gas discharge is produced. The ions are accelerated toward a target. On the target surface, by thermal evaporation and, also, by direct impulse transfer, atoms and molecules are sputtered. When the electric field is switched on, the ions receive energies from several hundreds up to several thousands of electron-volts and this overcomes the bonding energy in solids. The effect of fractionation, common to the case of thermal evaporation of many compound systems, is much lower. For the deposition of high electrical resistance materials it is recommended to use a periodic change of the sign of the electrical charging of the target. In radio frequency sputtering, a 10 MHz frequency is used. For chalcogenides, this method applies with success and complex compositions as e.g. can be obtained in a reproducible way and with accurate stoichiometry. This preparation method stimulated the study of the electrical properties of the switching memory compositions. The shortcoming of the method is the impurification of the films by argon. f) Chemical vapour deposition. Very pure films can be obtained by deposition from vapours. The decomposition reactions, which take place during heating of a particular

chalcogenide compound, are exploited in this method. Amorphous surfaces and very disordered materials based on crystalline lattice transformations.

i. mechanical processing. The crystalline material can be transformed into amorphous powder by long time grinding. ii. irradiation. Very energetic neutron beams or ion beams destroy the lattice periodicity, introduce defects and finally transform the crystal surface or bulk into amorphous material.

iii. shock waves. Mechanical shocks by pressure waves of several megabars, accompanied by high temperatures, can induce the formation of the amorphous state. 2.1.3. IMPURITIES AND PURIFICATION

The impurities strongly influence the stability of the vitreous state. They catalyse the

recrystallisation of the glass. This effect greatly influences the quality of the manufactured devices such as their reproducibility, stability and durability. On the other hand, small additions of some metals such as Ag, Cu, Tl, Mn, do cause a large increase in conductivity accompanied with a decrease of the activation energy. Therefore, the electrical properties is important in the control of the preparation of very high purity glasses, comparable with the purity of the crystalline, device quality, semiconductors. Suitable purification of amorphous semiconducting glasses is more difficult than for crystalline materials because of different physical behaviour of vitreous and crystalline states.Conventional purification, such as crystallisation, zone melting, etc., is not possible.

CHAPTER 2

111

A purification method was developed [19] in order to remove oxides and some metal impurities. The method is based on the utilisation of the purifying effect of products, which arises during the pyrolithic decomposition of the urea. In practice the method is applied in such a way that either the components in the stoichiometric ratio or the already prepared raw glasses are heated together with the urea in the sealed evacuated ampoules to temperatures of 700÷750 °C. During heating the mixture, urea decomposes as follows: at temperatures of 700 ÷ 800 °C: at temperatures of 800 ÷ 1000 °C: The condition for the removal of the impurities is the presence of CO and because CO binds oxygen and reacts with metal impurities giving rise to nitrides. The gaseous evaporates from the melt. The nitrides separate from the melt due to their different specific weights. The method can be used for the purification of those glasses, whose elements do not form nitrides with ammonia or nitrogen within the temperature range of 700 -1000 °C. The suitable systems are: S-As, Se-As, As-Te, Ge-As-Te, As-Se-Te. 2.1.4. MECHANICAL PROPERTIES The expansion coefficient results from the anharmonicity of the lattice oscillations. The

potential energy relation can be written as: where b gives the elastic part of the returning force for the harmonic oscillator, and c is a measure of the anharmonicity. The temperature dependence of the average backshift “u”, a parameter proportional to the thermal expansion coefficient, is: The weakening of the elastic force of the lattice, e.g. by the increase of the network dimensionality, leads to the increase of the relative contribution of c and, correspondingly, to the increase of

The elastic force of the network, which determines the strength of the material, will increase as a consequence of the growth of the mean number of bonds, expressed by the

increase of the parameter b. Then c will be equally weakened and

will decrease. After

reaching a minimum, will come out of the new structure and new bonding relations. By analysing the expansion coefficient and density curves it is possible to evidence

the structural transformations that occur in melts at various temperatures and to investigate in details the crystallisation process. The hardness of the glass is measured by evaluating the mechanical resistance against the penetration of a hard body. The hardness is a function of composition, temperature and of the state of the glass surface. According to Rebinder, the hardness, H, of the single crystal faces are in the same relation as the specific surface energy of their crystallographic planes (hkl).

112

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The value of is equal to the product of the energy of a single covalent bond by a number of bonds passing through of a crystal face. The relation can be applied to the isotropic glasses with continuous framework structure [7]. Suppose the glass having different structural units contains 1 mole of atoms. Mean volume concentration of atoms is equal to where is Avogadro’s number, is the mean atomic volume of glass. Then, we find the mean surface concentration of atoms and covalent bonds where is the mean number of covalent bonds in atom. The specific surface energy of isotropic polymeric glass with framework structure is being determined by the relation (2.6):

where is the mean energy, taken over one covalent bond. Since then, from the above equations follows:

where is the mean enthalpy of crystallisation (kJ/at.g). Then we shall have:

The calculation of hardness for glasses in the series : gives the results: 1 :

1.0

:

2.2

:

3.0

which compares rather well with the experimental values of microhardness:

In glasses with polymeric layer structure the role of the weak forces of interaction between the layers during the change of microhardness is noticeable, the agreement between calculation and experiment being worse. In particular for the series of glasses the calculation gives: 1: 0.83: 0.81: 0.65 while the experimental scale is: 1: 0.73: 1.13: 1.18. The experimental data are correlated to the layer distortion and inter-layer interaction. The theoretical scales of glass transition temperatures, can be derived by considering the relation:

between the shear module G, the free activation energy E* and the activated volume V* at viscous flow of glass as well as the relation

CHAPTER 2

113

obtained from Eyring equation at Then, for polymeric glasses with one kind bridge atoms the scale will be in agreement with the experiment. For glasses of different grades the following equation may be recommended for the calculation of

where f(n) is the structural factor in the equation:

The calculation of by the above equation shows that for polymeric glasses of different nature In particular, for selenide glasses of the systems Se-As(Ge) and Se-As-Ge (I, Tl) where For polymeric glasses with only one kind of bridge atoms one gets the scale:

which agrees well with the experimental data. At the equal mean enthalpy of atomisation, the polymeric glasses with higher covalent linkage have higher softening temperatures. The softening temperature is linearly related to microhardness: where K and B are constants for the given systems ( for the chalcogenide systems). In this case the properties in question are not only structure-sensitive but also bond-sensitive. The molecular-kinetical processes in glasses are, to a large extent, defined by the local fluctuation disorder in the structure at the expense of the deformation of the glass network, without breaking the valence bonds. These processes can be considered within the free-volume theory as formation and migration of the fluctuation microvolumes. This gives reason to correlate the microhardness, glass transition temperature and elastic constants. Based on the free volume concept, the microhardness, H, can be explained by the following formula:

where is equal to the microvoid formation energy with a volume by the relation: where

is the Boltzmann constant.

which is defined

114

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

From the above equations one obtains:

Based on the equation for the energy of molecular interaction and on the microhardness definition as the pressure necessary to overcome the intermolecular forces, Sanditov [8] wrote the following relationship:

where E is the elasticity modulus at single axis expansion/shrinking and coefficient, which can be defined by:

is the Poisson’s

where is the free volume part, which is maintained below and, according to Simha’s and Boyer’s rule [9], for a number of glasses with nearly equal is around 0.1. A correlaton between the Vickers hardness number (VHN) and the average atomic co-ordination number ( ) of the glass has been hinted at by the work of Shkol’nikov

[7]. Within the binary systems, Shkol’nikov noted that when the hardness is plotted versus the average enthalpy of atomization, straight lines are obtained following the general relationship: where k and B are constants. The slopes of these lines are discontinuous where stoichiometric compositions are reached. Shkol'nikov has calculated using the formula: where is, as shown before, the average energy of the covalent bond. The combination of the above equations hints at the linear relationship between VHN or H, and A replot of the VHN data versus for Ge-Sb-Se glasses gives an excellent linear fit to the equation: A steep rise in the bulk modulus of several glasses (as e.g. Ge-As-S system) has been observed at and not at The transition point at was ascribed to the maximum stability of the layered structures, still staying within the broader vector percolation principles advanced by Phillips and Thorpe (see [248]). 2.1.5. THERMAL AND THERMO-ELECTRICAL PROPERTIES The heat capacity of chalcogenide glasses is related to its composition. According to the Neumann-Kopp rule, the molar heat of a solid equals the sum of the atomic heats of

the elements of which the solid consists and it is well-known, for instance, that the molar heats of many alloys are well-expressed by this rule. In other words, the rule states that the gram-atomic heat capacity of a solid is equal to the weighted sum of the atomic heats of the elements that formthe solid. According to the theory of heat capacity in hetero-dynamical structures, the atomgram heat capacity of a chain like structure with a lateral interaction between the chains is expressed by the following equation [10]:

CHAPTER 2

115

where and and are the characteristic temperatures for intra-chain and inter-chain vibrations, respectively. A vitreous solid shows an anomalous increase of the specific heat at its glass transition temperature, The significant thermal effect produced by crystallisation was ascribed to the change of both long range order and short range order. The heat capacity is proportional to T for chain structures in the temperature range and this was proved for the case of glasses [11] The thermoelectric power has been measured in many chalcogenide glasses and with few exceptions has been found to be positive. It has a magnitude typical for semiconductors, of several milivolts per Kelvin and decreases with the temperature. The basical equation for the thermoelectric power, S, was appropriately modified so as to take into account the contribution of both electrons and holes. Thus

where the ratio of electron to hole mobility. Using the equation (2.18) and obtained from the conductivity, reasonable values for b in the range 0.1 - 0.3 are obtained [12]. Once again this is consistent with intrinsic conduction and since the holes are more mobile the conduction will be of p-type. The predominance of hole transport is confirmed by measurements of carrier mobility using a direct “time of flight” technique. 2.1.6. ELECTRICAL CONDUCTION

In many non-crystalline materials and in a large temperature range the d.c. conduction, as a function of temperature, can be expressed by the relation:: with constant activation energy over the whole temperature range of the measurements. The activation energy is usually referred to as half of the energy gap. The activation energy is usually situated in the range 0.5 - 1.0 eV although values as low as 0.2 eV and as high as 1.5 eV have been observed. The pre-exponential factor is often situated in the range but it can be as low as and as high as The annealing of the chalcogenide glasses leads, generally, to a decrease of the electrical conductivity and to a small increase of the activation energy. As opposite to the case of non-crystalline semiconductors prepared by quenching from the melt, the conductivity in amorphous thin films are much more sensitive to heat treatment.

116

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES As in the crystalline semiconductors, the carriers will be thermally or optically

activated into the bands of non-localised states, above some critical energy, where they can drift in an applied electrical field. In these states, the carriers in a glassy semiconductor will be subjected to the same collision (scattering) and trapping processes as in the crystal, but there will be at least one important additional scattering mechanism resulting from the disappearance of the strict periodicity in the lattice. This will decrease the mean free path and mobility of the carrier and hence increase the resistance. With the model of density of states and mobilities in an amorphous semiconductor, as discussed by Mott and Davis [13], three mechanisms of electrical conduction should be distinguished:

a. conduction due to carriers excited beyond the mobility edges into the nonlocalised states:

(for the case when the main current is carried by holes) where is the Fermi level and is the energy which separates the localised from non-localised states in the valence band. b. conduction due to carriers excited into localised states at the band edges If the main current is carried by holes and conduction is by hopping we have:

where

is the activation energy for carrier hopping,

c. conduction due to carrier hopping (tunnelling) between localised states near the Fermi energy. This process is analogous to impurity conduction in heavily doped semiconductors:

where is the hopping energy of the order of half the width of the defect band. A straight line in a plot Incr against 1/T is expected only if carrier hopping takes place between nearest neighbours. As the temperature is lowered, it may become favourable for the carriers to tunnel to more distant sites, will behave after the Mott law:

will fall and the conductivity

In 1962 R.L. Myuller developed the chemical theory of the electrical conduction in glasses (dielectrics and chalcogenides). He treated the glass as consisting from atoms linked by covalent bonds with localised pair electrons. During the absorption of a quantum of energy it occurs the ionisation of the covalent bond with the transition of the pair electron bond “v” in one electron bond “h”. This process is an equilibrium process and can be described as: The equilibrium is related to the interac-tion of the expulsed electron e with the other atoms by means of the polarisation of the bond. By

CHAPTER 2

117

applying the electric field, the ionised electron and the one-electron bond (hole ) shift with small activation energy in different directions, thus conditioning the transport of the electrical current. Taking into account the equilibrium and the activated shift of the carriers of the order of inter-atomic distances along the vacancies and along the dangling bonds we are in the situation to find the theoretical expression for temperature dependence of the electrical conduction on the energy of electrical conduction, which is

composed from the ionisation energy of the covalent bond activation energy of the electron expulsion and of the vacancy:

and the double of the

In the frame of this model, the statistical calculation of the pre-exponential factor in the equation

shows that this parameter is proportional to the volume concentration [v] of the valence bonds, which belong to the predominant structural units in the glass. According to this theory, the logarithm of the ratio of the pre-exponential factor to the concentration of the covalent bonds, the so-called electro-conduction modulus, of the solids with three-dimensional continuous network is constant:

The incertitude can be essentially diminished if the frequencies of the oscillations of the valence bonds are known. The deviation of the experimental modulus from the theoretical one is called steric factor.

The steric factor is a measure of the deviation of the nature of the given solid from the standard one with maximum discrete pair electron bonds. In the case of the destruction of the skeleton of the valence bonds and of the appearance of radicals, then

2.1.7. OPTICAL ABSORPTION

The investigation of the infrared absorption in various chalcogenide glassy systems is important from two points of view. The first is the discovery of systems with very good optical transmission and high temperatures for practical applications and the second is the fundamental information regarding the structure and bonding in these non-crystalline materials.

118

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

In the region of the spectrum where the photon energy there should be a rapid increase in optical absorption as a function of the frequency because the electrons are excited across the crystal and photons of the corresponding frequency are absorbed. The spectral region where this occurs is called “optical absorption edge” and the

corresponding photon energy should give a measure of The optical absorption coefficient, show an exponential dependence on photon energy, and obey Urbach rule:

where is a constant, is the angular frequency of the incident photon, (h is the Planck’s constant) and is the band tails’ width of the localised states in the band gap. In general it represents the degree of disorder in an amorphous semiconductor. For the case of high absorption region, was discussed by Davis and Mott [13] who found the equation:

where b is a constant, M is a number that characterises the transition process and is the optical gap of the material. It is frequently found in the experiments that

so that the equation (2.29) has the same form as that predicted for the optical absorption by indirect electronic transitions in a crystalline semiconductor. Usually, was found to be approximately The optical absorption at energies above the edge is best fitted by the equation:

This equation represents the normal expression for direct allowed transitions between parabolic energy bands. The absorption coefficient, can be calculated from the optical absorption spectra using the relation:

where t is the film thickness and A is the optical absorbency of the film. The degree of disorder and defects present in the amorphous structure changes due to heat treatment. Thus a study of the variation of as a function of temperature and time

and of heat treatment may provide a deeper insight into the mechanism of disorder and defect formation in the amorphous chalcogenides.

CHAPTER 2

119

2.1.8. DIELECTRIC PROPERTIES

The a.c. conductivity and dielectric properties of glasses are caused by the delayed response of carrier “hopping” motion in an applied field [14]. The simplest case occurs when there are pairs of isolated centres at a distance R from each other and two situations are usually envisaged: (1) the two centres differ in energy by an amount In the absence of a field the population of the upper level is exp(-W/kT) and the application of a field F changes the populations by the factor The carrier then moves between the centres by phonon-assisted quantum-mechanical tunnelling when the site energies are brought into equivalence by a suitable phonon. (2) the potential minima are separated by an energy barrier of height W. In the absence of the field there is equal probability of being occupied but with the application of a field the relative energies of the minima change by and their populations are altered. In this case the carrier is thermally activated and crosses over the barrier. In both cases the modification of the distribution between pairs of sites is experimentally manifested by a dielectric polarisation and by the increase of the conductivity with the frequency of the applied field. A dielectric loss which is nearly independent of frequency (or, in terms of conductivity,

over a substantial range of frequency is a common

occurrence in many amorphous dielectrics. In the context of semiconducting glasses the a.c. conductivity representation is normally used and the results are expressed in the form:

where A is a constant which increases slightly with the temperature. The constant n is slightly different in different materials but is situated in many cases in the region 0.70 - 1.10. It exhibits the tendency to decrease slightly when the temperature increases. The data for and collected by Lakatos and Abkowitz [15], show that the power law (eq. 2.32) holds over the frequency range Hz to at least Hz with n ~ 0.85 at room temperature. This type of frequency dependence is commonly observed in the chalcogenide glasses. Mott and Davis [13] followed the method of Pollak and Geballe [16] and interpreted the power law a.c. conductivity in terms of hopping between pairs of isolated sites close to the Fermi level and with a random distribution of separation distances R. Thus,

where

and is given by the equation:

is the density of states at the Fermi level. The factor

120

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

with is the static and is the high frequency dielectric constant. is the relaxation time, which characterises the delayed response. It is assumed that the main contribution to the integral comes from values of R such that For phonon assisted quantum-mechanical tunnelling the result is:

The frequency factor In

varies as

over a wide frequency range so that

For thermally assisted hopping over a barrier the same procedure leads to:

Pollak and Geballe [16] has developed the analysis more rigorously by transforming the microscopic random variables of the physical problem into a random impedance network of capacitors and resistors. He found for tunnelling for hopping over a barrier for a mixture of tunnelling and hopping

The equations assume that the pairs of centres are isolated. If this condition is relaxed, one must consider the possible correlated occupancy of electron states and Pollak has shown that this correlation has the effect of lowering the exponent of the temperature dependence by 1. 2.1.9. MAGNETIC SUSCEPTIBILITY

The diamagnetic susceptibility of the glasses increases with temperature and approaches a constant value at room temperature. There was possible to describe these curves by an empirical equation:

CHAPTER 2

showing that the measured susceptibility is given as a sum of a diamagnetic term paramagnetic term A/T. If one writes

121

and a

it is possible to calculate the number of free spins The existence of the weak paramagnetism can be explained in two ways. Because of the disorder, it is possible to assume the existence of dangling bonds. A dangling bond is represented by an orbital in which one electron is trapped. This configuration has in its first approximation a magnetic moment of 1.73 The second explanation uses the theoretical considerations concerning the localised states [17]. According to this theory, the localised valence states are occupied by two electrons and the repulsive energy between these electrons in the states lying above the Fermi level is so weak that these states can be ionised. In this way, states with one electron are created and have, like in the preceding case, a magnetic moment of 1.73 is the diamagnetic term and is temperature independent. It can be easily measured at room temperature when the weak paramagnetism is practically zero. The diamagnetic term is given by the contribution of atoms and bonds between them. Hudgens et al. [18] proposed a model which enables one to separate the magnetic susceptibility into a diamagnetic part (Langevin susceptibility of the “core” and valence electrons) and a paramagnetic part (VanVleck paramagnetic susceptibility originating from virtual interband transitions). In chalcogenide glasses the ratio between in crystalline and that in glassy materials is ~1 while in glasses with tetrahedral co-ordination of the atoms is considerable greater than unity. Cimpl et al. [107] explained the diamagnetic term using van Vleck formula. A diamagnetic glass can be considered as a giant molecule having zero magnetic moment in the ground state. Then, the susceptibility of one mole of material is given by:

where is Avogadro number, is the matrix element of the Z component of the orbital moment connecting the ground state 0 with the excited state k and is the energy separation of the two states. First term in the equation is the Langevin diamagnetism of atoms forming the glass and requires that the electron density distribution be spherical or cylindrical. If there are deviation from these types of distributions, caused for example by the existence of covalent bonds in the glass, the second term in equation, called van Vleck paramagnetism, will be not zero but always smaller than the Langevin term if one compares absolute values. The Langevin diamagnetic term can be easily calculated using the Kirkwood formula [20]:

122

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

In the above expression is the index of refraction at Z is the number of electrons per molecule and is the molar volume of the glass. The combination of the two equations permits to point out the dependence of the diamagnetic term upon the chemical composition of the glass [21].

The enhancement of the diamagnetism in glasses was explained by White and Anderson [22]. They considered two mechanisms. One mechanism is connected with the

valence states created by disorder and lying near the mobility edge. These states are spatially distributed and enhance the diamagnetism of the glass. Their contribution can be calculated using the classical Langevin expression. The second mechanism has its origin in the diminishing of paramagnetic interband contribution of the Van Vleck type. This contribution can be written as:

where indices 0 and 1 correspond to the valence and conduction band. is the orbital moment and the matrix element connect the ground state with the excited state is the Bohr magneton. The denominator represents the energy separation of the two states. The authors assume that the matrix elements do not change: the paramagnetism changes because of the summation, which is different in crystalline and glassy state. They argue that the sum is smaller for the glassy state and, therefore, the effect of this second mechanism is also the enhancement of the diamagnetism of the glass. The fact that diamagnetism does not change in sulphur, selenium and in chalcogenides was explained by Hudgens [23]. He supposed that disorder produces significant mixing of valence band states with higher lying states and that the value of the overall valence band diamagnetic contribution will be unchanged. Because the paramagnetic term in last equation is very small, then, according to Hudgens, it is possible to understand why the diamagnetic term does not change during the transition from crystalline to glassy state. In conclusion, the paramagnetic term in susceptibility is related to the states in the gap of amorphous semiconductors and to the presence of paramagnetic or ferromagnetic impurities, while the diamagnetic term is related mainly to the structure and to the characteristics of the bonds between the atoms. 2.1.10. THE INFLUENCE OF WATER ADSORPTION

The chalcogenide glasses can modify their electro-physical properties by water adsorption. The chemosorption and physisorption are related to the change of the surface electrical charge on local centres and this leads to the change of the surface potential, of the thickness of the spatial charge in the near-surface layer, of the recombination processes and of the parameters of relaxation and transport (mobility). The water adsorption gives rise to a n-type surface layer on the p-type semiconductor (chalcogenide) and this fact explains the behaviour of the relaxation process (this process occurs by surface recombination). The efficiency of the modifications increases when the destruction of the surface increases. The process of water absorption is stimulated by the surface destruction. The destruction amplifies the number of adsorption centres. The adsorption is

CHAPTER 2

123

intensified not only as a consequence of the increase of the effective surface during destruction, but also as a consequence of some electron processes (or chemical processes) The participation of the structural defects is a favourable factor for water adsorption. The increase of surface destruction leads to the increase of the amount of effective centres for water adsorption, that is the polishing of the surface gives rise to a high number of such centres. The structural defects are very important for the chemical stability of the glass.

2.1.11. THE ADDITIVITY METHOD FOR THE DETERMINATION OF THE PROPERTIES OF COMPLEX GLASSES The additivity method can be used for the successful determination of some properties in complex glasses. The main properties of the ternary and more complex chalcogenide systems can be explained by the weighted contribution of specific structure-chemical units as shown for example in [24]. The dependence of the property f on the concentration expressed in molar parts of the component can be defined by:

where The constants characterize the additive contributions of the components a, b, c. By generalisation one gets:

A physical property f can be represented differently by means of this formula. Nevertheless, for a group of compositions {N} situated in a glass formation domain, it is possible to find an optimum polynomial. Let see the properties of the optimum polynomials, which are symmetrical for all the compositions of the system. For every k component of the system one can write, more conveniently:

where

corresponds to the interaction of order “i”:

124

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Here is the rest term. The orthogonality relations are valid: {N}

and they are analogous to the orthogonality relations for optimum decomposition [25]. Summation in (2.51) is made over all the compositions of the system {N}, for which the polynomial is optimum. The relation (2.51) is an additional condition, which circumscribes the above non-univocity in a narrow range. The domain of application of the optimal decomposition is wider than the domain of application of the optimum polynomials. Therefore, either the results of measurement of the parameter f for disordered systems of the same composition, obtained in the frame of a given experiment, coincide in the error limit or do not coincide.

In the first case one can use the optimum polynomials (2.47) because f is, in fact, an unique function of the component concentration while in the second case it is not possible to use them. For optimum decomposition such limitation does not exist. This is important because to a given chemical composition corresponds, a number of disordered systems with structural properties which differ one from another. The coefficients of the optimum polynomial (2.47) can be found from the condition of minimum for every value l of the expression:

which characterises the rms deviation from mean (error) on the multitude of compositions {N} of the parameter f. In other words, firstly for one finds from the condition of minimum (2.52) the coefficients corresponding to the additive contribution of the components, then, analogously, step by step, one finds for the coefficients that correspond to the second order interactions, third order interactions and so on. As regarding the dependency on {N} it is necessary to take into consideration the following alternative: If f depends on following an additive rule, then the coefficients are the same for all assemblies {N}. For this equal to are the partial contributions of the components. In the opposite case, for different {N} and{N} corresponding to and are different and can be considered as partial contributions averaged on {N}, to which corresponds the minimum (see eq. 2.52). Let us discuss how it is possible to estimate the coefficients of the polynomial which is optimum for the compositions {N} if it is known the optimum polynomial for these compositions. The values of the polynomial in eq. 2.47 are uniquely determined

CHAPTER 2

125

for all and it is not difficult to find the contribution in the unknown coefficients [26]. In the term is enclosed the unknown supplementary information necessary for precise determination of the coefficients Because the total information is lacking, it is not possible to find out precise values It is only possible to determine the intervals within which the values are circumscribed. The following alternative is valid: the

determination is made according to (2.47), (2.52) for {N} and because the remaining member is small for {N} (i.e. its contribution in is significantly smaller than the contributions then the information on is enough for the determination of with the precision of interest for us. In the opposite case it is necessary to bring supplementary information. The method of optimum polynomial has been used for the calculation of the refraction by the atomic components in the chalcogenide glasses [27]. For the case of the optical refraction property it is used the Lorentz-Lorenz formula:

based on the measurement of the n and density d and the value of the mean atomic mass, A. For several glassy systems, the measurements of the refractive index is difficult because of the high crystallisation ability or due to non-transparency in the near infrared (IR) spectra.

In the optimum polynomial method the physical parameter R can be represented with the eq. 2.47. In the decomposition we retain the terms corresponding to the additive contribution of the components:

This is a polynomial of the first degree and the coefficients are the refraction indices of the elements. The remaining member takes into consideration the possible

deviation of the parameter R from the additivity when the atomic parts are modified. The precision of the calculated values by the method of optimum polynomials depends only on the experimental errors of determination of the refractive index and of the density. In the case of the condensed systems the method of optimum polynomials must be regarded as a possibility to determine uniquely a number of characteristics of atoms and of structural fragments. The difficulties in the determination of the values are naturally resolved in the frame of the method of optimum polynomials. This method, in principle, can lead to the same unique value for the refraction both in the case of the quantum-mechanical calculations and, in the case of calculation with the use of the experimental data.

126

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

2.1.12. SURVEY OF THE GENERAL PROPERTIES OF THE NON-CRYSTALLINE CHALCOGENIDES

The amorphous chalcogenides are chemically stable materials in normal conditions. They are not greatly affected by the aggressive media, are not soluble in water and do not dissolve in organic solvents. The characteristics of the devices based on chalcogenide materials do not change under the action of neutron irradiation with the fluency lower than or of ionising radiation up to rad/s. Nevertheless, there was shown that various external factors (electromagnetic radiation, particle beams, electric and magnetic fields etc.) determines particular modifications of interest in various applications (see chapters 3 and 4). The chalcogenides are less stable in alkali solutions. The investigations of the kinetics of dissolution of such solutions have shown [28] that the dissolution rate is low: The chemical stability of the chalcogenide glasses increases with the increase of the

degree of metallization of the bonds in the series [29,30]. This is proved by the increase of the value of the activation energy of dissolution and of the alkali concentration wherein the glass dissolution takes place. The introduction of the halogens in As-Ch glasses leads to the destruction of the compact character of the glass network and, correspondingly, to the decrease of the chemical stability of the glass. The elements of the group IV of the Periodic Table (the tetrahedral elements as e.g. Si, Ge) increase the stability of the glass by the formation of the tetrahedral structural units as e.g. etc.[30]. The density of the chalcogenide glasses is higher than that of the oxide glasses and the microhardness is essentially lower. The highest hardness values were found in the sulphur systems (Ge-S, Ge-As-S, Ge-As-S-I) and the lowest values were found in glasses with atomic chains (i.e. S-Se). The microhardness and the density of the semiconducting glasses decrease linearly with the increase of the temperature. As a function of composition, of the chalcogenide glasses changes in a large range The introduction of halogens in As-Ch glass leads to the abrupt decrease of while the introductions of tetrahedral elements determine the increase of The glassy As-Ch chalcogenides exhibit enough high expansion coefficients The introduction of halogens in As-Ch leads to the increase of up to . The halogenated glasses exhibit high thermo-plasticity in a large range of temperatures. For glassy materials with high the coefficient is situated in the range We shall discuss in the next section the main physico-chemical properties of the chalcogens and of the binary chalcogenide glasses.

CHAPTER 2

127

2.2. Properties of Non-crystalline Chalcogens 2.2.1 SULPHUR Mechanical and thermal properties The plastic sulphur is prepared by stretching the liquid sulphur from the boiling point, or

by drawing filaments from hot liquid sulphur. After several hours of storage at room temperature the material becomes hard and brittle and this effect was explained by the transition from plastic sulphur to micro-crystalline -sulphur. The S-S average bond energy is about 265 kJ/mol (280 kJ/mol after [6]) and the dissociation energy is about 138 kJ/mol. The density of the plastic sulphur is 2.01 [32], the same as that of fibrous sulphur [33]. depends on the cooling rate of the melt. The values are in the range Usually, a value of -27 °C is considered. In the range 40 - 60 °C the expansion coefficient is Polymeric sulphur (plastic, fibrous) can be quenched as a thin film. Slowly quenched polymeric sulphur is yellow. The polymer quickly quenched to 76 K is red, because it contains small molecules, which recombine at -100 °C [33]. The thermal conductivity of sulphur decreases from 11 W/m-K at 4.2 K to 0.29 W/m-K at 0 °C. At 100 °C it is 0.15 W/m-K. [33]. Sulphur ranks with mica and wood among the best thermal insulators.

Electrical and dielectric properties The plastic sulphur is a good insulator. Under high pressure it transforms into the metallic form, which at temperatures below 10 K becomes superconducting [31]. The thin films of sulphur deposited in vacuum at room temperature show resistivity values of the order of (as compared to rhombic sulphur: The dielectric constant is

Optical properties The elastic sulphur is yellow and translucent. By transformation in plastic sulphur the colour is changed in orange-red till brownish and the transparency is lost. is lemon yellow, is light yellow and is orange-yellow. Under the action of ultraviolet rays the plastic sulphur changes its colour. The absorption edge is situated in the ultraviolet domain of the spectrum. Liquid polymeric sulphur is dark yellow and exhibits an optical absorption edge situated at 350 nm [33]. The optical absorption edge of the bulk samples prepared by heating sulphur at 600 °C for several days and, then, slowly cooled at room temperature, is situated at 0.42 The optical gap is eV [34]. For crystalline sulphur an optical gap of 3.82 eV was reported [35]. Figure 2.3 shows the transmission spectra of sulphur compared to those of Se and Te [36]. The refractive index for melted sulphur is n = 2.068 for nm and n = 1.927 for nm [37].

128

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.3. The transmission spectra of sulphur, selenium and tellurium (after [36]).

Magnetic properties. The diamagnetic susceptibility of amorphous sulphur is [38] and this value seems to be identical to that of the polycrystalline sulphur (rhombic sulphur exhibits a susceptibility value of [39].

2.2.2.SELENIUM

Mechanical and thermal properties. The vitreous selenium obtained by cooling the melt is composed from rings and long selenium chains. By storage selenium crystallise at the room temperature. Gupta et al. [40] have shown that, after long time, both vitreous selenium and the undercooled liquid selenium transform into monoclinic and trigonal structural modification. Crystallisation of the vitreous selenium has been investigated by Kawarada and Nishina [41] with particular attention to the density of the nucleus for crystallisation as a function of the quenching temperature. Nucleation of crystallisation has been found to be heterogeneous because of the fact that the distribution of the crystallisation nuclei hardly changes with the heating time at a given temperature situated above ~60 °C. The distribution curve of the nucleus density with respect to the quenching temperature shows a minimum near 500 °C. The amorphous selenium films are different from the melt-quenched selenium in its weight fraction of rings. Thus, amorphous selenium films are composed predominantly of rings. The structure of amorphous selenium films depends on the

CHAPTER 2

129

deposition conditions, i.e. substrate temperature, deposition rate, etc. Montrimas and Petretis [42] and Andrievskii and Nabitovitch [43] have observed the effect of the substrate temperatures and have found that if the substrate is kept at room temperature, the evaporated film consist of Se8 rings, but if the substrate is kept at a temperature around 70°C, the films consist mainly of the polymeric chains. By heating the film in the temperature range 30 - 100 °C, the crystallisation is produced under the form of spherulites developed at some nuclei [44], which are generated at the substrate-film interface as well as on the free surface. The radius r of the spherulites increases linearly with the heating time, The crystallisation velocity dr/dt is related to the heating temperature, T, in terms of the following phenomenological expression:

where cm/s is a constant and E is the activation energy for crystallisation. E = 1.4 eV at the interface to the substrate [45] and 0.85 eV on the free surface [46]. Kawarada and Nishina [47] have shown that the velocity of the advancement of the crystallisation process in amorphous selenium films is anisotropic and depends on the parallel or perpendicular direction to the interface. The velocity in the direction of film thickness is found to be one or two order of magnitude lower than in the direction parallel to the substrate. The parallel component of the velocity depends on the deposition rate. The crystallisation is found to originate from the heterogeneous nuclei mostly distributed on the interface between the evaporated film and the substrate. The total heat of the transformation evolved in the process of crystallisation is influenced by the deposition rate. The larger is the deposition rate, the smaller is the heat evolved. The Se-Se bond strength is 184.3 kJ/at.g [48]. Vitreous selenium is soluble in sulphuric acid and slightly soluble in carbon disulphide. It is insoluble in water and alcohol. The organic solvents as alcohol, benzene, toluene, chloroforms, etc. produce the transformation of a-Se into “red” crystalline form. The solvents as quinoline, aniline, pyridine, etc. leads to the transformation a-Se metallic Se. The density of vitreous Se is 4.27 and increases to 4.79 when amorphous selenium is converted into the metallic form. In liquid the density is 3.99 at 220 °C. The Young modulus is 4.89 × Amorphous selenium is rather a soft material. Its hardness is 2.0 on the Moss scale (as compared to Pb: 1.5 and Cu: 2.5-3.0). The Vickers hardness, of vitreous selenium at 23 °C is 37.5 [49]. The thermal expansion coefficient is 5.6 × while that of hexagonal (metallic) selenium is 3.2 × [50], The thermal conductivity at 20 °C is 2 × W/cmK for both a-Se and metallic Se [51].

130

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The heat capacity is 6.13 cal.g/mol K for a-Se and 5.95 cal.g/mol.K for hexagonal selenium [52] Amorphous selenium films are prepared by conventional vacuum evaporation of vitreous pellets onto pre-oxidised aluminium substrate [53] Following deposition the films are aged in the dark for about three months at 24 °C in order to stabilise their physical properties. During the ageing process the structure relaxes towards the stable equilibrium state. Figure 2.4 shows typical results from differential scanning calorimetry (DSC) heating and cooling scans on well-aged a-Se films at heating and cooling rates of 10 °C/min. It can be seen that the heating scan exhibits an endothermic peak, which is a typical manifestation of a relaxation process.

Figure 2.4 Typical DSC heating and cooling scans on amorphous selenium films, r - heating rate q - cooling rate

Figure 2.5 shows the microhardness variation versus temperature in amorphous selenium films when the temperature is raised with the rate of 0.1 °C/min. Apparently, at a

temperature there is the onset of a sharp decrease in the microhardness that occurs in the glass transformation region. was observed to depend on the heating rate in a similar fashion to the in DSC measurements, although the two experiments probe different properties.

CHAPTER 2

131

Figure 2.5. The dependence on temperature of the Vickers microhardness, through the glass transformation region [50,51 ].

In selenium (and in tellurium, too) the transition glass-melt or crystal-melt is accompanied by a significant decrease of density. This diminishing of density proves the existence of a significant change in the local structure during the transition The change of density (decrease) in glassy selenium at the transition in the melted state is 6 % [54] (Figure 2.6).

Figure 2.6. The dependence of the density of the glassy and liquid Te on temperature:

• [54]

× [55].

The thermal expansion coefficients of amorphous selenium are compared in table 2.2 with those of tellurium and of other chalcogenide materials in glassy state [56].

132

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Electrical and dielectric properties The resistivity of glassy selenium at room temperature is but values in the range were reported. For hexagonal (grey) selenium: Analysis of the potential decay curves indicates values of for the dark resistivity of vacuum-evaporated thin selenium films. The thermal activation energy obtained from resistance-temperature measurements is 1.7 eV but values of 1.6 eV and 1.95 eV were, also, reported. In thin evaporated films of a-Se the effective mobility of electrons is and or even for holes [57]. The dielectric constant ranges from 5.97 to 6.60. The value 6.3 seems to be the most correct. It decreases somewhat at high frequencies and takes the value 5.97 at the frequency of Hz. For liquid selenium Ilyás et al. [58] reported for glassy selenium obtained by quenching method (quenching in water + ice from the melt heated at 800 °C for 8 hours) at the temperature of 346 K and frequency f = 200 Hz, the dielectric constant and dielectric loss

Optical properties The optical bandgap of selenium is 2.1 eV [59]. The absorption edge

is situated at 540 nm for a-Se and 620 nm for hexagonal Se and for liquid selenium at 220 °C. The absorption spectrum in selenium is shown in figure 2.7 [60]. The typical value of the optical gap in thin amorphous films is 2.05 eV [61]. The reflectivity of amorphous selenium decreases from about 27 % at 300 nm to 25% at about 550 nm wavelength. At this point there is a rather sharp drop to about 20 % at 700 nm. The activation energy found from photoconductivity measurements is 2.5 eV. This value agrees approximately with the wavelength at which the absorption constant begins to fall down rapidly. This value agrees, also, with the energy associated with the Se-Se

CHAPTER 2

133

bond, which is 225kJ/mol. The hexagonal and monoclinic selenium show activation energies in the range:

Figure 2.7.

Absorption in selenium:

amorphous ------- crystalline, metallic Se.

The refractive index has a maximum value of n = 3.13 at about 500 nm. In the infrared range n = 2.46. Figure 2.8 shows the refractive index of selenium as a function of wavelength.

Figure 2.8. The refractive index for selenium.

134

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Magnetic properties. The diamagnetic susceptibility of a-Se is selenium is

while for polycrystalline

2.2.3.TELLURIUM

Mechanical and thermal properties In the case of tellurium, the transition crystal-glass around melting temperature corresponds to a density modification (decrease) of 6.5 %. Pure tellurium cannot be produced in the glassy state by quenching from the melt. The viscosity of the tellurium melt is, on the basis of the partial delocalisation of the volume electrons, so much reduced, that the transition in the glassy state can be reached only through very high cooling rates. Rates as K/s are necessary to get thin amorphous film [62]. Amorphous films can be also obtained in the amorphous form, by condensation from vapours on substrates cooled at 77 K. The

value for a-Te film is

285 K [63]. The density of bulk (polycrystalline) sample is

The crystallisation ability of tellurium is influenced by the rate of deposition and by the subsequent annealing cycles. The melting entropy, of tellurium is 24.3 J/mol·K while for selenium is 12.6 J/mol·K [64]. The atomisation enthalpy is 197 kJ/mol while for selenium is 22 kJ/mol [65]. In the case of tellurium and of solution of selenium into tellurium, the volume during melting increases but above the melting point it exists a domain of increased density (Figure 2.9) [66]. The relative increase of density by melting is of -15.2 % for Se and -5.3 % for Te, while for is -3.2 %, for is -4.1 % for Si is +10 % and for Ge is +4.15%) The behaviour of the viscosity, of tellurium illustrates the relation with the formation of the chain molecules. The density shows that melting destroys the ordered chain structure of the crystalline tellurium only partially. Thus one gets a mixture of chain molecules, whose average length decreases when temperature is raised. The trend of the correspond to this model. The electrical conductivity increases when the viscosity decreases and this behaviour indicate that the metallic character is related to the increase of the density of broken bonds. The crystalline selenium has also an ordered chain structure and the behaviour of during melting is qualitatively similar to that of tellurium. In Te-Se alloys the viscosity is much higher and decreases when tellurium concentration increases. The enthalpy of formation for pure tellurium is 193 kJ/mol at 298 K and the standard entropy at 298 K is 49.61 J/K. Vapour pressure at 20 °C is 525.5 Torr. The dissociation energy of tellurium is The thermal conductivity at 20 °C is 0.014 cal/cm·s·K.

CHAPTER 2

135

Figure 2.9. The temperature dependence on the specific volume of the electrical resistivity and of the viscosity of tellurium (after [59]).

Electrical properties The electrical conductivity at room temperature is The thermal activation energy is 0.44 eV. [67]. Between 77 and 170 K the Mott law is valid and the activation energy is 0.046 eV [68]. The electrical conductivity increases during tellurium melting and for further heating. This effect is related to the increase of the co-ordination number of tellurium. Optical properties

The absorption coefficients for amorphous and crystalline tellurium versus wavelength are given in figure 2.10. By amorphization the absorption edge shifts toward short wavelengths. The shift of the absorption edge is accompanied by the decrease of the refractive index from 5.3 (for polycrystalline tellurium) to 3.4 for amorphous tellurium [69]. The essential decrease of the refractive index of the amorphous tellurium as compared to polycrystalline tellurium was explained by the change of the effective number of nearest neighbours in the disordered material on the account of the disorder between atom chains. The optical gap of amorphous tellurium is 0.9 eV.

136

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.10. The dependence of the absorption coefficients on the photon energy for tellurium, 1 . crystalline state 2. amorphous state

2.2.4 THE CHALCOGEN-CHALCOGEN SYSTEMS

Sulphur – Selenium Mechanical and thermal properties. The transition temperature for vitreous S0.11Se is 38 °C. The films of thickness µm, with at.% S, exhibit The complete melting occurs at 140 °C. The crystallisation from melt leads to the separation of solid solutions with various structures [70]. The composition with at.% Se gives rise to crystalline compound with rhombic symmetry of the type For compositions with at.% Se one gets crystals with monoclinic structure of the type In the composition range at.% Se one obtains by crystallisation crystals with hexagonal structure. The density of is 4.20 and increases by crystallisation up to [71].

The specific heat at the melting temperature for and are, respectively: 8.3 cal/mol·K and 8.2 cal/mol·K, while for tellurium at 500 K. C= 9 .33 cal/mol·K and cal/mol·K [72].

CHAPTER 2

137

Electrical properties. The alloy with the composition exhibits a conductivity of at room temperature. The conductivity increases by crystallisation The activation energy of this material is 1.7 eV and, after crystallisation, it decreases down to 0.7 eV [73]. The main electrical data are collected in Table 2.3.

Optical properties. The refractive indices for the typical compositions of the system are:

The photoconductivity and the optical absorption curves of S-Se amorphous thin films, as a function of the photon energy are shown in figure 2.11.

Figure 2.11. The photoconductivity and the optical absorption curves for amorphous S-Se films. The two

curves of the photoconductivity data are for illumination incident on the positive and negative electrodes.

138

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Magnetic properties. The magnetic susceptibility for

and for

is:

is:

Sulphur-Tellurium Mechanical and thermal properties. Thick amorphous films were prepared by splat-quenching [74]. The softening temperatures for and are 5 °C and respectively. Therefore the stability of the amorphous phase is very low. At room temperature is observed a rapid crystallisation with the separation of crystalline tellurium and an unknown additional crystalline phase. Optical properties. The optical absorption edge is situated at composition [74].

wavelength for

Selenium - Tellurium

Mechanical and thermal properties. In the Se-Te alloys are formed mixed chains consisting of different length of selenium and tellurium chains. 8-member rings and long polymer chains of selenium prevent the crystallisation of the melt. The glassy phases are rich in selenium and do not have more than 20 at.% Te. The X-ray diffraction of the crystalline glasses has shown that selenium phases and Se-Te solid solutions are formed. Isothermal annealing of (x= 10 at.%) leads firstly to the separation of selenium and later, by increasing the temperature, the solid solution of Te in Se is separated. Afify [75] reported the structure of the crystalline hexagonal form of The crystallisation mechanism of this glass corresponds to surface and one-dimensional growth. The effective activation energy for crystallisation is Some data on are given in the table below [76].

CHAPTER 2

139

The crystallisation temperatures, and the glass transition temperatures, for glasses were measured at heating rates from and exhibit weak rate dependencies at low heating rates, but the slopes of their heating rate dependencies become steep with increasing heating rate. The evolution of and as function of the heating rate was interpreted in terms of nucleation and fragmentation processes in random networks. Figure 2.12 shows the and obtained from the dependence of the electrical conductance on temperature [77]. Yang et al. [78] reported the thermal data in the amorphous system (see Table 2.4).

The density, microhardness and heat capacity as a function of annealing time at room temperature for glassy Se - 10 at.% Te are given in Figure 2.12 [79]. The heat capacity versus temperature is illustrated in figure 2.13. Carini et al. [80] reported more complete data on several compositions (Table 2.5).

Illeková et al. [81] reported for the glassy composition crystallisation temperature and melting temperature specific heat was found

while the

140

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.12. The effect of annealing time at room temperature on microhardness, density and heat capacity at the glass transition peak in the amorphous Se-Te 10 at.% alloy.

The mechanical properties of several compositions in the system were reported by

Carini et al. [80] (Table 2.6).

Sreeram [82] reported several mechanical data in high purity glasses: Young modulus, E = 14.15 GPa, Poisson ratio, Young modulus, E = 12.18 GPa, Poisson ratio,

CHAPTER 2

141

Figure 2.13. Heat capacity versus temperature in the region of the glass transition of amorphous Se-Te alloys after annealing for various times at room temperature.

Figure 2.14 shows the crystallisation and the transition temperature for some S-Te alloys.

142

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.14. The crystallisation and transition temperature obtained from the variation of the electrical conduction with the temperature in a alloy. a. temperature dependence of conductance and its derivative b. Heating rate dependencies of crystallisation, and transition temperature,

Electrical properties. The main electrical parameters in the system are given in table 2.7.

Hagen [83] reported electrical data in a large range of compositions (Table 2.8).

CHAPTER 2

143

The dielectric constants of some amorphous films are given in Table 2.9.

Optical properties. The optical gap decreases with the increase of tellurium in the amorphous alloy. Figure 2.15 shows the evolution of the optical gap with the Te content in Se-Te samples [83].

Figure 2.15. The optical gap in Se-Te vitreous alloys.

144

and

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The refractive index near the optical absorption edge is for

for

Magnetic properties. The typical values of the magnetic susceptibility in the system are:

Sulphur - Selenium - Tellurium Mechanical and thermal properties. During the crystallisation of the glasses in the system firstly are separated solid solution Se-Te. Long annealing enhances the separation and improves the crystallization. The density of the glasses changes from 3.7 up to and increases linearly with the increase of the tellurium content. Relative low microhardness and are characteristic. Hardness changes within the range It decreases when the sulphur content raises and increases by raising the tellurium content. The increase of Te concentration leads to the increase of the thermal stability, when sulphur is increased, then decreases (see table 2.10.).

CHAPTER 2

145

The decrease of the mechanical resistance and of the thermal stability with the increase of sulphur content are caused by the appearance of weak Van der Waals forces between sulphur chains and rings. The effect of improving mechanical strength when Te is added, which is reflected in the increase of hardness and can be explained by the densification of the Se-S glass and by the destruction of the rings and introduction in them of short Te chains and, also, on the account of the raise of the co-ordination number of

tellurium. Electrical properties. The main electrical parameters for a broad range of compositions are given in Table 2.11.

Optical properties. The index of refraction in several S-Se-Te glasses is given in table 2.12. Magnetic properties. The magnetic susceptibility data in the vitreous system S-Se-Te are shown in table 2.12

146

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

2.3. Properties of Non-crystalline Binary Chalcogenides

2.3.1. PHOSPHORUS - CHALCOGEN

Phosphorus - Sulphur (P-S) Mechanical and thermal properties. The glasses with the composition in the concentration range show yellow colour and low thermal stability against crystallisation [84]. In normal atmosphere they rapidly hydrolyse and eliminate gas. For low phosphorus content, the pnictide atoms build bridges between sulphur chains. For high phosphorus content it is formed a polymeric structure based on molecules. The melted compositions transform into polycrystalline solids by quenching in water. The dependencies of the softening temperature and of the molar volume on composition are given in figure 2.16.

Figure 2.16. The properties of the glasses in the systems P-S and P-Se

P-S: P-Se:

softening temperature softening temperature

volume • molar molar volume

CHAPTER 2

147

While increases rapidly with the phosphorus content, the molar volume slowly decreases (the density increases). Phosphorus - Selenium (P-Se) When sulphur is substituted for selenium the glass ability formation increases. In the system P-Se the glasses are hygroscopic. By interaction with water vapour the

system eliminates

The chemical stability of

glasses extends up to

The glasses and also the glass are stable in air for more than one year. When compositions in the range are stored in air the glass surface covers with a waved film consisting of hydrolisis products. The most stable is the glassy composition PSe. The colour of the glasses with x>0.4 is deep red. Mechanical and thermal properties. The density, microhardness and glass transition temperature of some P-Se glasses are given in Table 2.13 [85].

The density of the glasses in the first domain of glass formation decreases monotonously from with the increase of the P content.

148

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

In the second domain, the decrease of the density becomes slower for the compositions with 60 - 70 at.% P, i.e. in the vicinity of glass. The microhardness, H, and the softening temperature, are parameters sensitive to

glass structure. For the alloys with 50-60 at.% phosphorus one observes an abrupt decrease of H and This is related to the advanced inhomogeneities of these glasses. The transition towards more homogeneous glass is accompanied by the increase of The glass transition temperature is of the same order of magnitude as that of the PSe glass. By DTA one observes an endothermal effect during softening. The introduction of P in Se glass leads to ther increase of Selenium with 2.5 at.% P (or As) and, also, the pure selenium glass crystallise when measured in DTA apparatus and one gets hexagonal selenium. When no more than 5 at.% P or As is added to Se, then no crystallisation is observed by DTA. Therefore, the introduction of small amount of P (or As) increases the stability of selenium against crystallisation. For 2.5 at.% amount of P in alloys, the crystallisation takes place very slowly and can be controlled easier than the crystallisation of pure selenium. For glass it appears in DTA an exothermic peak due to melting of crystalline selenium.

The thermal expansion coefficient, coefficient, Table 2.14.

the ultrasound speed,

the Poisson

and the elasticity modulus, E, for several P-Se glasses are given in

Electrical properties. The electrical conductivities at room temperature for the P-Se glasses are situated in the range The logarithm of the conductivity, at room temperature, the activation energy, the logarithm of the preexponential factor, and the steric factor, together with the permittivity, [86] are given in table 2.15. The dielectric constant was measured by the immersion method. The electrical parameters speak in favour of the dielectric character of these glasses. By introducing around 5 at.% P in the selenium glass the

CHAPTER 2

149

conductivity decreases by five orders of magnitude and the activation energy increases more than 0.5 eV. The steric factor indicates an intrinsic semiconductor.

Significant photoconductivity was discovered in three vitreous compositions: and PSe [87]

Optical properties. The refractive index, n, was determined interferometrically in the spectral range: The results are given in table 2.16. Glassy films with polymeric units show a photo-contraction effect. The addition of Sn and Pb diminishes rapidly the effect [88].

Magnetic properties. The magnetic properties of the P-Se glasses are determined, in the first approximation, by the electronic configuration of phosphorus atoms. The most important contribution to the magnetic susceptibility is given by the distortion of the

150

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

electron distribution of P and Se atoms as a result of the formation of P-Se chemical bonds and of the character of the distribution of the phosphorus selenides in the matrix of glassy selenium. Significant change of the magnetic susceptibility was observed in the glass for long storage in air. The dependencies of the magnetic susceptibility

and of the magnetic properties in

P-Se vitreous system on phosphorus content are given in figure 2.17 [90]. The atom-gram susceptibility is related to the ionicity of the chemical bonds.

Figure 2.17. The dependence of the specific magnetic susceptibility,

susceptibility

and Van Vleck paramagnetism

atom-gram magnetic

on the P content in the P-Se alloys.

Phosphorus-Tellurium (P-Te) The only known compound in this system is Glassy compositions P-Te were obtained in high speed cooling regime, by quenching the melts. No data on properties are available after our knowledge. Phosphorus - Selenium - Tellurium (P-Se-Te) A large range of homogeneous glasses can be prepared in this system. In the compositions of the type one observes the stratification in two glassy phases. The volume ratio of the two phases changes with the Te content. The two phases differ by hardness and softening point. For the properties are for the upper layer and while for the bottom layer and

Mechanical and thermal properties. When the tellurium content increases, the density increases. A very rapid increase of density takes place for the first 10 at.% Te. Thereafter the density increases close to linear. The deviation from linearity is observed for partially

CHAPTER 2

151

crystallised glasses. When the contentration of phosphorus increases the density of the

glasses decreases. The glasses show in the range 75-100 °C (table 2.17). The decrease of for high tellurium content is probably due to submicroscopic inhomogeneities, to the formation of a highly dispersed phase and to the change of the glass composition. For larger concen-trations of P or Te the crystallisation ability strongly increases. In the DTA curves one observes significant endothermic effects (one - two peaks) in the range 160 - 240 °C due to crystallisation.

Electrical properties. The conductivity data for the P-Se-Te glasses are given in the table 2.17. The conductivity increases and the activation energy decreases with the increase of Te content. It is easy to observe in these glasses the transition from dielectric to semiconducting properties. Because the steric factor negative, no intrinsic conductivity exists. Low

low chalcogen content.

values prove the existence of blocking charge current for

152

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Magnetic properties. The magnetic susceptibility as a function of tellurium content is given in figure 2.18. For the glasses in the section the change of with Te concentration is non-linear and non-monotonous. There are two ranges of concentration with different magnetic properties, probably related to the different character of the bonds in the local structure. The increase of is accompanied by the decrease of the activation energy,

Figure 2.18. The dependence of the component of the paramagnetic susceptibility on the tellurium concentration in two glassy P-Se-Te samples:

2.3.2. ARSENIC - CHALCOGEN Arsenic - Sulphur (As-S)

For increasing concentration of arsenic in the system As-S, the stability against crystallisation increases. The glass with the content of 6 at.% As crystallises at room temperature in a day with the formation of rhombic sulphur, while the glass of composition cannot be crystallised by thermal annealing. The glasses situated in the range As usually crystallise by long time annealing at 60 °C with the formation of rhombic sulphur. crystallises completely by annealing at 280 °C in 30 days. The glasses in all the concentration range crystallise under the action of high pressure and temperature As a result, new crystalline modifications are produced. The glasses with high sulphur content crystallise under light irradiation (see chapter 3). In the composition range As the crystallisation ability is different. The glass stability increases when the As content increases and starting from 56 at.% As the glasses do not crystallise at room temperature. The glasses with As crystallise at room temperature in several hours. These glasses are plastic at room temperature: is situated under room temperature.

Mechanical and thermal properties. The expansion coefficient in the As-S system as well as are given in Fig. 2.19 [91]. The density of glass is

CHAPTER 2

[93]), while in the crystal (orpiment) is The thermal diffusivity for

153

Hardness is

and

bulk glass is

Figure 2.19. The expansion coefficient (a) and the softening temperature (b) in As-S glasses.

Electrical and dielectric properties. The temperature dependence of the electrical conductivity (d.c.) was investigated. For the most part of the amorphous semiconductors, and for a very large temperature range, the following relation is valid: (2.55)

where is a proportionality coefficient and is the activation energy. For relatively high temperatures the preexponential factor, is considerably lower than in other amorphous semiconductors and takes the values in the interval and even The results obtained on samples of thickness with two types of electrodes (Al and Pt) are given in figure 2.20. For aluminium electrodes the plot exhibits two slopes. The activation energy in the low temperature range depends on the time of storage of the sample in dark and for most samples is situated in the range With the raise of the temperature a new linear dependency characterised by higher activation energy appears and this is much higher than the half width of the forbidden gap obtained from optical measurements or than the activation energy of the bulk In the case of Pt electrodes, one observes three slopes on the conductivity plot corresponding to three activation energies: and the last one, which corresponds to bulk glasses. The dark conductivity at room temperature is while during light irradiation

154

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.20. The dependency of the electrical conductivity on temperature. a. 1) during heating 2) during cooling b.

In the system As-S the dielectric constant for each composition is independent of frequency over the range (Fig.2.21a). The dependency on frequency of the dissipation factor for the compositions investigated is shown in figure 2.21 b.

Figure 2.21. a. The dielectric constant vs frequency b. the dissipation factor vs frequency for various As-S glasses.

CHAPTER 2

155

Stevels [98] has shown that the decrease of with the frequency in the range is due to the summation of the conduction losses and the relaxation losses. Therefore, the increase of at low frequencies indicates that ionic contributions are present. This effect is clearly evidenced for the case of glasses with high arsenic content because the conduction losses in As-Se decrease rapidly with increasing frequency. On the other hand, since the viscosity of the glasses decrease with the increase of sulphur, the relaxation time, may decrease. The relation between the relaxation loss and the frequency is expressed by the formula:

The relaxation loss takes a maximum value for the frequencies at the condition Therefore, the maximum moves to the high frequencies. In practice the maximum of the dissipation factor moved to the region of high frequencies for the glasses with high sulphur content (see sample 4). The dissipation factor increases with the sulphur content in the glassy matrix, which is presumably the result of the increase of the number of dipoles appeared in the bound sulphur fraction. The dielectric constants of the glasses in the systems As-S and As-Se have been measured by Gutenev [99] using the immersion method in the frequency range 300 - 700 Hz. The polarisation processes produced in bulk sample condition the results on permittivity. is determined mainly by the contribution of the elastic part of the polarisation. The permittivities, of the investigated glasses are given in table 2.18.

Optical properties. The optical absorption edge for crystalline and glassy temperature range is shown in figure 2.22 [101].

in the

156

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure2.22. Dependence on temperature of the absorption edge of crystalline and glassy

Zakis and Fritzsche [101] have shown that from 4 to 200 K the position of the absorption edge does not depend on temperature. In the range 300-400 K the absorption edge shifts towards higher energies with the rate and this can be explained by the electron-phonon interaction. The calculated refractive indices of several glasses in the system [99] are given in Table 2.19 together with the dielectric constant and the density d.

CHAPTER 2

157

Heo and Mackenzie [102] reported the value n = 2.66 for glass at For n = 2.395 and the temperature coefficient of the refractive index is

Magnetic properties. The magnetic susceptibility of the glassy As-S system at room temperature was studied by Bajdakov [105]. The susceptibility is negative and slightly dependent upon the adsorbed oxygen. Tauc and Menth [106] studied the susceptibility of glasses within the temperature range for x situated from 2.9 to 26.5. They found that the susceptibility is negative in the whole temperature range. It consists of a temperature independent diamagnetic term and a superimposed paramagnetic Curie term. One finds an approximate value of paramagnetic centres Cimpl et al. [107] have measured the specific magnetic susceptibility on powdered samples and obtained the results shown in Table 2.20.

Arsenic - Selenium (As-Se) The chalcogenides show high stability against aggressive media. They are stable in humid atmosphere but less stable (with the exception of in alkali solutions. The stability in alkali solutions was studied for NaOH and in the temperature range

In the dissolution rate in alkali solutions is determined by the reaction, which takes place at the surface and does not depend on the solution modification. The stability of the glasses depends on the character of the chemical bonds and increases in the series Table 2.21 shows the parameters of the chemical stability in the glasses together with the average bond energy per mole of atoms of compounds [108].

158

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The air - vapour interaction with

fibre surface does not degrade the fibre

surface in the absence of ultraviolet light [151]. The crystalline was prepared by annealing glassy materials at 688 K for ten

days [110]. In the system As-Se the glasses crystallise under pressure and high temperatures in all the composition range. At 70 kbar and 400 °C the following phases are formed: which transforms into at 220 °C. Although and are isostructural, the first one is more stable against crystallisation and the second one cannot form glass even by rapid quenching and it is necessary to use high pressure. The microhardness of films is around 110 [112]. The softening temperature of bulk deduced from DTA is 180 °C and the melting temperature is 202 °C. From X-ray diffraction, the crystallisation temperature was found 180 5 °C [113]. Mechanical and thermal properties. A typical thermo-differential curve for a fresh glassy sample heated at a rate of 2 °C/min is shown in figure 2.23.

Figure 2.23. Differential scanning calorimeter curve of bulk recorded at a rate of 2 °C/min showing the glass transition, crystallisation and melting phenomena.

One observes the glass transition at the crystallisation exothermal peak (with maximum crystallisation rate occurring at the temperature and the melting endothermal minimum (with melting point While shows no dependency on the heating rate and the dependency of on heating rate is slight, the crystallisation exotherm location changes markedly as the heating rate or sample age is varied [114]. The thermoelectric power of measured at 60 °C is

The main mechanical and thermal properties of several As-Se glassy compositions are given in the table 2.22 [115]. Seddon and Laine [116] reported a density of for glass.

CHAPTER 2

159

For glass The linear coefficient of expansion for glassy is : for and for [117]. The Vickers hardness in As-Se glass fibers has been investigated by Hach et al. [118]. Hardness increases linearly from GPa for 5 at.% As up to GPa for 40 at.% As and then decreases at GPa for 50 at.% As. The graph of the heat capacity, for the glassy and crystalline can be seen in figure 2.24 [119]. The

values for glass and crystal practically coincide in a very

large range of temperatures

The mean atomisation heat, of the glass is a measure of the strength of the chemical bond, as proposed by Sadagopan and Gatos [120] as basis for the correlation of the physical properties as e.g.: n and (resistivity). The As-chalcogenides exhibit also a direct correlation between and the forbidden gap,

Figure 2.24.

The experimental values of the heat capacity of 1 - crystal 2. glass

160

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The increase of the melting entropy of the compound in the series is related to the weakening of the bond energy during melting. The melts must be characterised by high dissociation and disorder. This is in agreement with the observation of the diminishing of the glass formation ability with the decrease of the melt viscosity in the series The analysis of the superheated vapour above the melt has shown that is stable and up to and in vapour phase exists as monomer and polymer dimmer) molecules. and are in these conditions significantly dissociated because the dissociation ability increases from Se to Te. Figure 2.25 shows the distribution of the cluster composition in the vapour phase above target subjected to laser irradiation [121] and to explosive vaporisation. It is remarkable that if the sample is single crystal or glass the vapour composition is nearly identical. The tendency to form mainly and molecular units in vapour phase reflects the peculiarity of the short-range-order in the solid.

Figure 2.25. The distribution of the molecular composition in the vapour phase above the sample: a - evaporated by laser radiation b- explosive vaporisation (T=1000 °C).

Electrical properties. The experimental values of the electrical conductivity in bulk glass of are shown in figure 2.26 [122].

CHAPTER 2

Figure 2.26. Electrical conductivity of amorphous

161

as a function of reciprocal temperature.

The activation energy, is 1.05 eV and 0.9 eV at the temperature range above and below 200 °C, respectively. One of these activation energies may correspond to half of the width of the mobility gap. In the crystal is 0.8 eV [123]. Iovu et al. [124] reported an activation energy of 0.85 eV in bulk glassy At room temperature the conductivity in is Only p-conduction was revealed. The mobility of the charge carriers at 298 K is of the order of The hole band increases when arsenic is added to selenium up to the value of 1.73 eV (at 300 K) for glassy Further increase of the arsenic leads again to an increase of the gap [125]. The crystalline exhibits a forbidden gap of width 2.0 eV [126]. The value of the dielectric constant obtained from parallel capacitance measurements up to 5 MHz is and is nearly independent of both frequency and temperature. At microwave frequency it is The dielectric constants for some compositions in the system are given in the table 2.23. Feltz et al. [127] reported the value at room temperature and 50 kHz.

Optical properties. The optical gap of is 1.70 eV while in is 1.76 eV [128]. The refractive index together with the permittivity in the system are given in table 2.23. The refractive index of films, reported by Lyubin et al. [129] for 600 nm wavelength is 2.27. Skordeva [130] reported optical band gap values of 1.75 eV for 1.83 eV for AsSe and 1.89 eV for The photoconductive gaps in arsenic chalcogenides, reported by Tanaka and Nakayama [131] are:

162

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The refractive index of AsSe amorphous films (2 µm thickness) as function of the frequency of the light is reported in [132] (Fig. 2.27).

Figure 2.27. Indices of refraction in the AsSe film before

index in

and after

photodarkening.

Voigt and Linke [133] reported recently high precision values of the refractive at room temperature for light wavelength range (Table 2.24.).

CHAPTER 2

163

Magnetic properties.The diamagnetic susceptibility for is while for is The change of with composition depends on the saturation of the unpaired electrons at the ends of the chalcogenide chains (for high chalcogen concentration in the system As-Se) and on the statistical distribution of chalcogen in glassy matrix for low chalcogen concentration in the same As-Se system. Tellurium (As-Te) Mechanical and thermal properties. The densities of the glasses are shown in figure 2.28. [134]. It can be pointed out that the density is a linear function of the mass concentration of arsenic. The extrapolated value for pure arsenic is which is just the density of black vitreous arsenic. The extrapolated value for pure tellurium is

Figure 2.28. Densities in the As-Te system.

164

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Amorphous tellurium cannot be prepared in bulk state and, therefore, this value can be considered only as an estimation of the density of vitreous tellurium. The density in the amorphous state is about 10% lower than that in the crystalline state. For glassy the density is while for crystalline is This difference is surprisingly high, if we refer to other typical vitreous chalcogenides as e.g. As-S and AsSe [135]. It cannot be attributed to the presence of cavities which contribute less than to the total volume. In fact this indicates that differences exist between the molecular arrangements in vitreous and crystalline state due to the partial ionic character of the AsTe bonds. In the case of homogeneous glasses, microhardness is a property which depends directly on the molecular binding forces. Figure 2.29 shows the variation of the microhardness versus composition for both vitreous and crystalline compounds.

Figure 2.29 Vickers microhardness versus composition in As-Te system.

It appears that i) microhardness of the glasses (due to the rigidity of the bonds) increases continuously with the arsenic concentration, ii) as far as the vitreous state is concerned, the experimental curve exhibits a kink just at the composition which suggests the existence of more rigid bonds (As bonds) on the As-rich side of the

CHAPTER 2

165

composition, iii) glass is 3 times harder than the polycrystalline and twice harder than the (100) planes of the single crystal These planes are parallel to the [010] axis of the molecular chains of the crystal is made of. Between these chains the molecular forces are partly of Van der Waals type. Thus the strong hardening of the glasses, as compared to the crystalline stoichiometric compound, indicates that the molecular arrangement in the disordered state is deeply modified with the purpose to eliminate the non-covalent forces. iv) the two maxima of microhardness in the polycrystalline state occur at the two eutectic compositions of the system. The phases appear as dense aggregation of minute crystallites. The contribution of the interfacial energy of grain boundaries is responsible for the extrinsic hardening of these eutectic alloys. The thermograms of the compositions and are given in figure 2.30. Tr1 and Tr2 are the onsets of crystallisation in these compositions.

Figure 2.30. Thermograms in As-Te system.

Electrical properties. The crystalline exhibits at room temperature a relatively high conductivity with activation energy of 0.03 eV. This has been interpreted in the frame of conduction by defects. Only above 500 K is obtained from the intrinsic conduction a definite width of the gap of 0.45 eV [136]. The electrical conduction of the glass at r.t. is with the activation energy which decreases slowly with the increase of the temperature. The conduction is p-type. The

166

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

activation energy for crystallisation of is 0.40 eV [137]. An essential diminishing of the activation energy is observed in the glass AsSe, at low temperatures. Optical properties. Glassy exhibits a band gap of 0.87 eV [138]. Baidakov et al. [139] reports a band gap eV for glassy and 0.48 eV for polycrystalline samples. The amorphous films show nearly identical band gap value [136]. Magnetic properties.

[137].

Arsenic - (Sulphur, Selenium, Tellurium) Mechanical and thermal properties. In the system

a continuous range of

solid solutions is formed, as proved by the IR spectra, phase diagram and X-ray analysis

[95]. The softening temperature is constant for all the concentration range and is (Fig.2.31). The microhardness increases monotonously with the raise of the content in the alloy. The crystallisation is possible for long annealing (30 days) at temperatures of 250-280 °C.

Figure 2.31. State diagram (a), glass temperature (b) and microhardness (c) in the section

The densities in the system have been measured by the pycnometric method at room temperature [140]. The results, summarised in Table 2.25, show that the glasses are 5-8 % less dense than the crystalline samples

CHAPTER 2

167

The viscosity of the glasses has been determined by the method of Nemilov [141]. There was measured the velocity of a needle driven through the material investi-

gated, by applying a known force. The viscosity data as function of concentration at two temperatures is shown in the Table 2.26.

The glass transition temperature and the crystallisation temperature measured at a heating rate of 10 K/min are shown in figure 2.32. is represented by the midpoint of the heat capacity curve and the peak temperature of the crystallisation exothermal effect is taken at decreases with increasing tellurium content and shows negative deviation from linearity. decreases more steeply with the Te content than i.e. that measures the thermal stability of the glasses, also decreases with increasing concentration of Te. The markedly non-linear molar volume dependency calculated from the density data suggests that the atomic volume of tellurium changes with concentration. There was observed a power law dependence of on the heating rate, which is in fair agreement with the free volume concept of the glass transition [142]. The specific heat below agrees within 5% with the Dulong-Petit limit, i.e. 3R/mole, and increases with at where new translation and rotation degrees of

freedom become accessible for the glass (figure 2.33 ).

168

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.32. Transition temperatures and molar volumes of

Figure 2.33. Heat capacity of

glasses.

glass versus the heating rate and temperature.

CHAPTER 2

169

The theory would predict a stepwise change in the heat capacity which is smeared out in real measurements, where, at finite heating rate the experimental time scale becomes compatible with the atomic reorganisation time scale at By reducing the heating rate, the glass transition should become sharper and the measured softening temperatures should converge, which fact was observed in measurements.

Electrical properties. The main electro-physical parameters are given for several As-S-Se-Te compositions in table 2.27.

Optical properties. In amorphous films of for x = 0.66 the refractive index is The optical gap at 0 K is 1.1 eV, somewhat larger than the conductivity gap (1.0 eV). Its coefficient of temperature is [144]. The position of the absorption edge for bulk glasses in the system varies between 2.3 eV for x = 0 and 0.75 eV for x = 1. The spectral dependence of the index of refraction n for several compositions in this pseudo-binary system is shown in figure 2.34. The index n varies continuously from 2.4 to n = 3.7 [145].

170

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.34. Spectral dependence of the index of refraction n for pseudobinary arsenic chalcogenide glasses.

Properties of

glasses compared to crystals

The tables 2.28 and 2.29 show the main mechanical and electrical parameters of the glasses and of the corresponding crystalline phases with the same composition.

CHAPTER 2

171

2.3.3. ANTIMONY - CHALCOGEN (Sb - Ch)

Antimony - Sulphur (Sb-S) Mechanical and thermal properties. Vitreous has been prepared. The density is [152] while single crystals exhibit The crystallisation temperature is 250 °C. Glassy was obtained by heating to 650 °C commercial non-stoichiometric powder. The glass exhibits a dark red colour, has a density of and is 277 °C. [153]. The thin films of and prepared by thermal evaporation show densities of and respectively.

172

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Amorphous films prepared by ion sputtering method [154] show : and for composition and and (Sb crystallites); 533 K ( crystallites) for composition.

Electrical properties. The electrical conductivity at room temperature is for vitreous samples of and for single crystals. The variation of the permittivity, of thin films, as a function of thickness is given in figure 2.35 [1551.

Figure 2.35. The dielectric constant as a function of the light wavelength in a thin film of thickness calculated from the optical data.

amorphous

An electrical switching effect was observed both in single crystal and vitreous The switching in vitreous is a threshold phenomenon. The threshold field is This value is times higher than that for the single crystal. The mechanism of switching is electronic. The switching parameters in vitreous differ sharply from those in typical chalcogenide glasses based on arsenic. Optical properties. The dispersion of the refractive coefficient for amorphous films of with various thicknesses is shown in figure 2.36 [156].

Figure 2.36. The dispersion curves of the refractive index in thin amorphous films of thicknesses:

of various

Antimony - Selenium (Sb - Se)

Mechanical and thermal properties. The density of vitreous is the microhardness is and the softening temperature is 190 °C. The crystallisation activation energy is and the glass transition activation energy is The molar volume in several glasses is given in table 2.30 [158].

CHAPTER 2

173

Electrical properties. The activation energy in thin amorphous films of composition is 1.4 eV while for single crystal is 1.1 eV [159]. The electrical conductivity at 20 °C is and the electrical gap is 1.8 eV [160]. The single crystal shows typical conductivity of and in several amorphous films has been measured a conductivity of Splat cooled glass has the characteristics of a memory glass at room temperature. It will switch from the high resistance or off-state to the low-resistance or memory state (on-state) upon the application of fields of the order of V/cm. The resistance in the off-state is of the order of times greater than that of the memory state [161]. Optical properties. Thin films were investigated using optical absorption and transmission experiments [163]. The optical transmission coefficient as a function of temperature changes abruptly at 160 °C and this effect was ascribed to the crystallisation (figure 2.37).

Figure 2.37. The dependence of the transmission coefficient of amorphous thin films of and on the annealing temperature.

The optical and thermal activation energies for typical Sb-Se amorphous films are illustrated in figure 2. 38.

174

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.38. Optical (o) and thermal (×) activation energies as a function of composition in Sb-Se amorphous films.

Antimony - Tellurium (Sb-Te)

Mechanical and thermal properties. The crystallisation temperature in these glasses is situated in the range 100-150 °C. The glass transition temperature could be below room temperature. Electrical properties. Electrical conductivity vs 1/T for Sb-Te glass is shown in Fig. 2.39.

Figure 2.39. The Arrhenius plot of the electrical conductivity in Sb-Te glasses.

CHAPTER 2

175

The activation energy for is eV while for the crystalline compound eV (probably eV) [164]. All Sb-Te compositions are p -type semiconductors with the Fermi level close to the middle of the gap [164]. The Seebeck coefficient calculated on the basis of one-carrier system, is at least a factor of two higher than the experimental values. This suggests that both electrons and holes are contributing to the coefficient and that, in the range 60-70 at.% Te, the electron contribution increases (perhaps due to a mobility increase) although the Fermi level remains approximately fixed. The room temperature Seebeck coefficient versus film composition is shown in figure 2.40 [162].

Figure 2.40. The Seebeck coefficient in thin amorphous Sb-Se films.

Optical properties. Thin amorphous films of were investigated by optical transmission measurements for various annealing temperatures (figure 2.41). A crystallisation temperature of 60 °C has been determined. The optical activation energy is also shown in figure 2.41.

176

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.41. Optical activation energy at room temperature optical activation energy extrapolated to 0 K and thermal activation energy extrapolated to 0 K versus film composition in amorphous Sb-Te films.

2.3.4. SILICON-CHALCOGEN (Si-Ch) Silicon-Sulphur (Si-S)

The Si-S bond energy is lower with than the Si-O bond energy. Consequently a strong tendency toward hydrolysis is exhibited by the compounds and alloys [165].

Mechanical and thermal properties. The alloys are liquid in a narrow temperature range: The glasses with the composition obtained by melt cooling under pressure are more stable in water than SiS and obtained in normal melt cooling conditions. SiS sublimates at 940 °C. SiS and decompose easily in humid air. Silicon-Selenium (Si-Se) The glass shows yellow colour and crystallises easily by slowly cooling the melt. For silicon content exceeding 20 at.% the glass rapidly decomposes in air. The glasses hydrolyse in air [166]. melts at 1060 °C. Silicon Tellurium (Si-Te)

Mechanical and thermal properties. Amorphous silicon-tellurides were prepared in both bulk (by melting of the components) and thin film (by sputtering from polycrystalline targets). The amorphous material seems to be very stable against the decomposition in air. The crystal is hygroscope but is more stable than Si-S and Si-Se compounds. The deep red crystal quickly turns black upon exposure to ordinary air (the black appearance being due to free tellurium identified on its surface). The glass transition temperatures for several glasses are given in figure 2.42 [167].

CHAPTER 2

177

Figure 2.42. The glass transition temperature for bulk amorphous Si-Te samples and crystallization temperature for

and Te.

The increases with the increase of silicon concentration due to the hardness of silicon and softness of tellurium. In the DTA curves [168] one observes an exothermal peak with the maximum situated at For glasses with high Te content a second peak appears, which shifts to lower temperatures with the increase of tellurium content (Fig. 2.43)

Figure 2.43. DTA curves in Si-Te glasses. 1 45-

2-

3-

178

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The crystallisation of the glasses with less than 80% Te occurs in one stage at the temperatures 290-300 °C. For higher tellurium content the glasses crystallise in two stages. The X-ray analysis shows the formation of a new crystalline Te phase embedded into the amorphous matrix. The high temperature exothermal effect for is related to the crystallisation: crystalline tellurium and other crystalline phases related to the decomposition of the eutectic and possible crystals.

Electrical properties. The resistivity versus temperature in bulk non-crystalline (vitreous) is given in figure 2.44.

Figure 2.44. Electrical resistivity vs inverse temperature for amorphous The activation energy varies from 0.55 to about 0.44 eV at 400 K and 200 K, respectively.

Room temperature resistance versus composition for bulk amorphous samples is given in figure 2.45. The electrical activation energy vs. composition in bulk amorphous samples is given in figure 2.46.

CHAPTER 2

179

Figure 2.45. Room-temperature resistance vs. composition for bulk amorphous Si-Te samples.

Figure 2.46. Electrical activation energy vs. composition for Si-Te bulk samples.

The amorphous thin films show switching properties. The switching field as a function of the thickness of the film has been investigated by Baliavicius et al. [170].

180

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Optical properties. The infrared transmission spectra at room temperature in crystalline and bulk amorphous are given in figure 2.47 [167]. The optical energy gap in the Si-Te system can be followed in figure 2.48 [167]. The band gap of crystal is 2.0 eV.

Figure 2.47 Infrared transmission for crystalline

(normal to c axis) and bulk amorphous

Figure 2.48. Optical energy gap vs. composition for various amorphous and crystalline

samples.

CHAPTER 2

181

Madhosoodnan et al. [171] reported the following optical gaps in Si-Te glasses:

2.3.5. GERMANIUM-CHALCOGEN (Ge-Ch)

Germanium-Sulphur

(Ge-S)

Mechanical and thermal properties. The glasses in the system Ge-S are stable in air. The glass transition temperature increases with the germanium content from -40 °C (pure sulphur) to 300 °C for the composition with 30 at.% Ge [173]. The glassy GeS was prepared at unusually high cooling rates The DTA curve exhibits a notable

figure 2.49 [173]. For amorphous films Feltz [174] reported the density of Heo and Mackenzie [102] reported the softening temperature for glass: and the molar volume of For glass were reported the following data [175]:

and

For glass were reported room temperature

and the thermal expansion coefficient at

[103].

Figure 2.49. (a) The glass transition temperature of Ge-S glasses [176,177].

and of Ge-Se glasses

182

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The densities of the glasses are shown in figure 2.50.

Figure 2.50. The densities in the glassy systems Ge-S and Ge-Se.

Electrical properties. The resistivity of glass is is an intrinsic semiconductor with the activation energy = 1.76 eV and The mobility gap of glassy has been evaluated to be at 4.2 K. Electrical data on films compared to data on bulk materials were reported by Watanabe et al. [178] (see table 2. 31).

CHAPTER 2

183

The temperature dependence of the conductivity for crystalline and amorphous GeS films of thickness is given in figure 2.51 [179].

Figure 2.51. The temperature dependence of conductivity for crystalline and amorphous GeS layers. (deposition temperatures:

For films deposited at the substrate temperature of 220 °C, while the crystal exhibits a value of 1.58 eV. The absolute photoconductive sensitivity for thin amorphous GeS films is at 2 V for the incident radiation intensity of (radiation with The photoelectric gap for thick amorphous GeS films is For bulk glass was reported [175] at room temperature the conductivity Optical properties. The optical gap in the glass is 3.10 eV and in is 2.14 eV [180]. In crystal eV [35]. The optical gap of amorphous GeS films was found 1.5 eV while in GeS crystal the in composition are shown in Table 2.31. As in almost all the semiconductors the temperature coefficient of the absorption edge in both forms of GeS is negative: for amorphous and for GeS crystal. The increase of sulphur concentration results in an almost parallel shift of the absorption edge also towards higher energies [181]. The refractive index is n = 2.3 [102]. The stress induced birefringence effect, for 106 Pa [182] in Ge-S alloys and in selenium is given below [176]:

184

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Germanium-Selenium (Ge-Se) Mechanical and thermal properties. The glasses are stable in ambient conditions. increases from 42 °C for pure selenium till 363 °C for and then decreases by adding GeSe to down to 299 °C for the glass: 57.6 mol% mol% The annealing at temperatures higher than leads to the crystallisation of The density, microhardness, and expansion coefficient are given in Table 2.32 [182]. De Neufville [183] reported for for bulk samples and for sputtered thin films. The glass transition activation energy for the glass is 271.4 3.8 kJ/mol and the crystallisation energy is 245.2 18.9 kJ/mol [157]. For Ge2Se3 glass we determined and for glass, (Mihai Popescu, unpublished data). For vacuum evaporated amorphous films the crystallisation temperature is only 360 °C [184]. For glass Ivanov et al. [175] reported and

The thermal capacity ,

is given in figure 2.52.

Figure 2.52. The thermal capacity versus Ge content in Ge-Se glasses.

CHAPTER 2

185

varies linearly with the germanium concentration in two domains. This fact is related to the content of tetrahedral structural units in the mixture of two configurations: and For is an additive function of the thermal capacity corresponding to every type of structural unit, according to their concentration. Electrical and dielectric properties. The permittivities in the glassy system Ge-Se are given in Table 2.33:

The electrical conductivity at room temperature and the activation energy for some glasses are given in Table 2.34.

For the composition two linear dependencies are defined. The low temperature part corresponds to and the high temperature part has The conductivity at 300 K is greater as compared to the conductivity in and reaches [187]. For bulk chalcogenide glass the value varies in the range depending on composition [13]. Ivanov et al. [175] reported for bulk glass at 300 K and, after crystallisation Olekseyuk et al. [188] reported the value of electro-conductivity in glass:

186

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Optical properties. The optical absorption edges of as-deposited evaporated (x=0; 0.20; 0.33; 0.43 and 0.50) have been determined by Kotkata et al. [189]. In the high-energy range of the spectra the data obey the Tauc law and the

extrapolated optical gap has been thus determined. Table 2.35 shows the results.

The optical gap

for

films prepared by PECVD and evaporation

are given in figure 2.53 [190].

Figure 2.53. The optical gap in

× PECVD films, virgin

annealed films

films.

evaporated films, virgin

In the case of virgin films, the PECVD films exhibit while the evaporated films exhibit The plasma enhanced chemical vapour deposited films, annealed at 175 °C, exhibits the gap and those annealed at 250 °C show The mobility gap in glassy is 2.3 eV [191]. The refractive index of the sputtered GeSe films (with the thickness in the range is almost constant in the wavelength range and has the

value of 3.05 [192]. The refractive index, n, in the Ge-Se system is shown in figure 2.54. [193].

CHAPTER 2

Figure 2.54. The refractive index as a function of atomic concentration, x, for the The solid curve indicates the calculated value.

187

system.

Magnetic properties. The magnetic susceptibility as a function of germanium content is given in figure 2.55. For amorphous

-0.250 ×

while for crystalline

the experimental magnetic susceptibility, is -0.246 ×

is

Figure 2.55. The dependence of the magnetic susceptibility on the germanium content in Ge-Se glasses.

188

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

The experimental (total) magnetic susceptibility, or and the paramagnetic component show that the structure of the glasses in Ge-Se system exhibits an essential change in a narrow range of concentration. Germanium-Tellurium (Ge-Te) Mechanical and thermal properties. The softening temperature increases monotonously

with the raise of the germanium content from 60 °C for the glass GeTe9 up to 220 °C for the glass The variation of the softening temperature is shown in figure 2.56.

Figure 2.56. The evolution of the in Ge-Te glassy alloys. The samples were prepared [194] by: 1 - spray cooling 3- splat cooling 2- cathode sputtering 4-cooling in NaCl solution [195].

The maximum is for the composition although a compound with this formula does not exist. The results are consistent with the view that each Ge atom creates a point of Te chain cross-linking, and, further, that this process saturates at the composition for which every pair of Te atoms has been cross-linked to another pair of Te atoms via a Ge atom. The temperature where the crystallisation occurs during heating is a highly sensitive and reproducible parameter of the Ge-Te glasses [196]. The crystallisation temperature reaches a maximum at ~20 at.%Ge (220 °C) and decreases with more than 60 °C for 10 at.% increase of tellurium content in the splat-cooled glasses. The thin films crystallise around 145 ± 3 °C. The crystallisation energy (1 cal/at.g) is about one half the crystallisation heat of GeTe at the melting point. This value is higher than that in the case of the crystallisation of thin Ge films. The crystallisation of GeTe films is produced with high rate and this makes difficult the research of the kinetics and determines a more evident change of the physical properties of the films [197]. An anomalous high value of the crystallisation hat and also other particularities of the phase transition leads to the conclusion that in a-GeTe films are produced a short range order different from that of the crystal. Table 2.36 shows the main electro-physical parameters of some amorphous Ge-Te alloys [198].

CHAPTER 2

189

Electrical properties. The conductivity σ and the activation energy, are shown in Table 2.36 for some amorphous materials. The Arrhenius graph of the conductivity is given in figure 2.57.

Figure 2.57. Electrical conductivity of sputtered films versus reciprocal temperature. Film thickness: Activation energy at 25 °C is 0.51 eV on heating and 0.59 eV on cooling.

For following data:

bulk glass and thin amorphous films, Vihrov et al. [199] give the

Bulk glass

Optical properties. The optical gap of the GeTe glass is parameter B in the absorption law is

Thin film

and the

190

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Magnetic properties. The atom-gram magnetic susceptibility and the paramagnetic component of the susceptibility, conditioned by the amplification of the asymmetry of the domains of superposition of the electron orbitals are given below [200] for two Ge-Te glassy compositions.

Germanium-(Sulphur-Selenium)

Mechanical and thermal properties. The density, hardness and glass transition temperature for some glasses are given in Table 2.37.

The glasses in the section have been prepared by heating the corresponding mixtures of the elements in quartz ampoules. Synthesis was carried out in vibrating furnaces at 900 °C for ten or 20 hours. The cooling down to 700 °C was carried out by quenching in water. The amount of material was 40 g excepting which exhibits high tendency toward crystallisation, and therefore a smaller mass was used in preparation: 10g. The glasses are stable in water, ethanol and exhibit a well developed softening range. Table 2.38 shows the mechanical and thermal properties of some glasses in this composition section.

CHAPTER 2

191

Electrical and dielectric properties. The dielectric constant for several glasses is given in Table 2.37. Optical properties. The refractive index, n, for some glasses is given in Table 2.37. Germanium - (Se, Te); Germanium-(S, Te) Mechanical and thermal properties. In the system Ge-Se-Te, varies from 89 °C to 336 °C. The softening temperature increases when the germanium content increases. In the system Ge-S-Te the hardness and the softening temperature increase when sulphur is substituted by tellurium. The density, microhardness and glass transition temperatures are given in Table 2.39.

The

glass exhibits

and

192

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

Figure 2.58 shows the composition dependence of glass transition temperature and density in glassy system.

Figure 2.58. Composition dependence of and density in glassy system. data are for bulk glasses between 40-1000% all other data and all density data are for sputtered glass films.

Electrical properties. The conductivity and the activation energy data for several glassy compositions Ge-S-Te and Ge-Se-Te are shown in Table 2.39. The accuracy in is ±0.3. In the glassy alloys at.% Te the d.c. conductivity measurements did not revealed any systematic influence of the tellurium content up to ~10 at.% Te. The Arrhenius plot for is shown in figure 2.59 for several compositions.

Figure 2.59. Direct conductivity of

Te glasses. The activation energy is

CHAPTER 2

193

2.3.6.TIN-CHALCOGEN (Sn-Ch)

Tin-Sulphur (Sn-S) Thin films were obtained by vacuum deposition. The colour of the amorphous films is deep black while that of the single crystal SnS is silver-like. The enthalpy of atomisation of SnS and of is 18.6 kcal/mol and 22.5 kcal/mol, respectively, The cleavage energy for Sn-S bond in SnS is 60 kcal/mol and for Sn-S bond in is 56 kcal/mol. Tin-Selenium (Sn-Se) The colour of thin amorphous films is deep black while SnSe crystals exhibit silver-like colour [201]. The mass spectrometric analysis has shown that at temperatures of 8001400 °C in vapour phase the SnSe phase is constituted from two-atom molecules [202]. After cooling from 900 °C in 10 minutes at 20 °C, one gets a partially crystallised glass with the physical constants: and After annealing for 25 h at 270 °C one obtains and Films with unstable composition have been reported by Palatnik and Levitin [203]. Tin-Tellurium (Sn-Te) No Sn-Te glasses been reported. The Sn-Te crystal is a semi-metal and a true p-type

semiconductor with

Many properties approach those of the metallic ones.

2.3.7.THALLIUM - CHALCOGEN

Thallium - Sulphur (Tl-S) Mechanical and thermal properties. Amorphous

is plastic, liquid-like at room

temperature and brittle solid-like at temperatures under 0 °C. The glasses show low glass transition temperatures [204]. crystallises at room temperatures in several days. melts congruently at 450 °C. The glass transition temperature, crystallisation and melting temperatures for several glasses in this system are given in Table 2.40.

194

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

For

and [155] gives the heat of melting the entropy of melting This is an extraordinary low value which may support the idea tyhat the local structure the type of bonds changes only little during melting. The density of (pycnometric value) is The crystalline has the density of Zavetova and [207] reported the density of After long exposure at room temperature the bulk glass crystallises.

Optical properties. In the rate of variation of the absorption edge is at hv=2 eV [208]. The index of refraction of glass as a function of wavelength of the electromagnetic radiation is shown in figure 2.60 [207].

Figure 2.60. Index of refraction of glassy

Thallium -Selenium (Tl-Se)

Definite compounds are and Tl2Se with chain structure where tetrahedral distorted units are linked by edges in a spiral (elicoidal) form analogous to and TlSe melt congruently at 390°C and respectively at 330°C. Mechanical and thermal properties. The softening temperature, the crystallisation temperature and the melting temperatures are given in Table 2.41.

CHAPTER 2

Electrical properties. For

195

glass, the conductivity is

Thallium - Tellurium (Tl-Te)

Amorphous thin films can be prepared by vacuum evaporation in a large range of concentration. The system Tl-Te is one in which charge transfer localises electrons on the tellurium and leads to semiconducting behaviour of the stoichiometric composition Glassy was obtained. This glass rapidly crystallises at 20 °C. melts congruently at 453 °C. An eutectic with the composition at.%Te was reported [210].

2.3.8. ALKALI-CHALCOGENIDES The first data on an alkali glass (Cs-Se) were published in 1982 by Chuntonov et al. [211] The domain of glass formation is situated from up to 76 % for Na-S, K-S, Rb-S and Cs-S systems,

for Na-Se, K-Se and Rb-Se systems and

for Na-Te, K-Te, Rb-Te and Cs-Te systems [212].

2.3.9. BISMUTH-CHALCOGEN (Bi-Ch)

Bismuth-Sulphur (Bi-S) The electrical gap for crystalline is 1.2 eV [35]. No reliable data for amorphous materials of the same composition are known. Bismuth-Selenium (Bi-Se) The first n-type amorphous semiconducting chalcogenides were described in the system [213]. The elements Bi and Se were co-evaporated onto a substrate at a temperature of 50 °C. X-ray diffraction measurements have shown that the films are amorphous up to 30 at.% Bi. At higher concentration, microcrystallites were detected. The resistivity shows a drop of twelve orders of magnitude for a composition change from 0 to 30 at.% Bi. The band gap shows a large increase from 0 to 30 at.% Bi and remains unchanged for higher bismuth content. Comparison of the activation energy of the conductivity with the optical gap points to an intrinsic-like behaviour. The thermopower indicates n-type conductivity for x > 0.03. Crystalline

is a n-type semiconductor. Takahashi [214] measured the drift

mobility of both type of charge carriers in amorphous films by means of the time-of-flight technique. He observed a drastic diminishing of the hole drift

mobility upon addition of 2 at.% Bi. A similar behaviour, although much less pronounced, was also found for the electron drift mobility. The bismuth doped selenium films are n-type semiconductors.

196

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

2.3.10. HALOGEN-CHALCOGEN Bromine-Selenium (Br-Se) In the system there were found two compounds:

which melts at 123 °C distectically and which melts at ~0 °C peritectically. In the range 4 - 14 at.% Se one observes a stratification effect with the monotectic temperature of 110 - 115 °C. Iodine-Selenium (I-Se)

The selenium and iodine form a simple eutectic for 52 at% Se. The melting temperature of the eutectic is 58 ± 1 °C. The solubility of iodine in selenium does not exceed 2 at.% [215]. The crystallisation of I-Se melts occurs in several days. The activation energy of crystallisation for Se + 2 at.% I is: Iodine-Tellurium (I-Te) The stable compounds are ITe and on the phase diagram. Glasses of composition have been prepared [217] (Cl,Br)-Tellurium (Selenium)

and glasses have been prepared [218]. deduced from the differential scan (fig. 2.61).

and

Figure 2.61. Differential calorimetric scan of glassy Te-Br, Te-Cl, and Te-Se-Br. (Tx is the crystallisation temperature)

have been

CHAPTER 2

197

Boron - Chalcogen (B - Ch)

Boron - Sulphur (B - S) was obtained in glassy state by vacuum synthesis from elements (at 950 °C) and in amorphous state by the reaction of (in hydrogen flow) with boron pellets at 600 - 900 °C. White powder of composition is thus obtained. This powder is thereafter decomposed at 100 °C and the final product is [219]. The glasses in this system are hygroscopic. melts at 417 °C. melts at 567 ± 10 °C. Boron - Selenium (B – Se) The glasses are obtained by vacuum synthesis. The glass temperature, decreases from 325 °C for down to 40 °C for glassy selenium. The glasses are not stable and hydrolyse easily in normal atmosphere. The eutectic in this system exhibits a melting temperature of 90 °C [220]. The reaction of boron with gives fine grain powder. 2.3.11. HEAVY METAL - CHALCOGEN

Heavy metals do not form glasses with the chalcogens. By special procedures one obtains very fine amorphous powders. Several amorphous metal-chalcogen compounds were obtained and studied. The amorphous heavy metals-chalcogen materials were prepared by precipitation with from thio-, seleno- and telluro- salt solutions or by chemical or thermal decomposition of the corresponding ammonium salts [221]. The amorphous has been prepared from and by Diemann [222]. was prepared and its structure was investigated [223]. has been prepared from and in 2-methoxyethyl-ether. Ruthenium is divalent. The amorphous crystallises above 623 K in a pyrite structure [224]. and were prepared from with They are black and diamagnetic powders with semiconducting properties and relatively high-density [221]. The amorphous structure for the chalcogenides results from the random distribution of disulphide groupings in the co-ordination polyhedra Therefore the compounds show very high specific surfaces. Many transition metal chalcogenides can be obtained in the amorphous state. Antimony compounds and are included.. and were prepared by thermal decomposition of the ammonium tetrathio compounds, and at 200 °C under dry flow. The low conductivity at room temperature of for suggests [226] that the metal-metal bonds are localised. Amorphous is diamagnetic. The only natural amorphous material is the mineral jordisite It is a black-brown material and its structure was studied by Diemann [227]. Amorphous is obtained by the decomposition of ammonium thio-molybdate is stable between 280 °C and 335 °C and at higher temperatures losses

198

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

sulphur and gives rise to

The material is porous with the surface area

[228].

The amorphous materials based on tungstene – sulphur show the following dielectric constants:

The d.c. conductivity of amorphous W-S versus 1/T is given in figure 2.62 [229].

Figure 2.62. DC conductivity as function of

for several W-S amorphous samples.

CHAPTER 2

199

2.3.12. SILVER-CHALCOGEN (Ag-Ch)

Silver-Sulphur (Ag-S) is a superionic conductor with a large amorphous component whose structure was explained in the frame of a model with liquid-like distribution of Ag [230]. 2.3.13. SILVER-SELENIUM (Ag-Se)

The introduction of silver leads to the increase of For 10 at.% Ag, By deposition of Ag on Se and illumination one gets amorphous Ag-Se layers. The compound was proved to exist be electronography [231 ] The ionic conductivity of Ag ions in is about This value is times larger than that of (low temperature phase) [232,233]. Miyatani [232] has found the activation energy for the diffusion of silver diffusion in 0.1 eV. The ratio of the diffusion constant to the mobility deviates remarkably from the Einstein relation. These facts suggest that the Ag atoms in are in highly disordered state or rather in liquid-like distribution. 2.3.14. SILVER - (S, Se) The glasses show colours from deep red to black. They are stable in air. slightly depends on composition and varies from 70 °C up to 111 °C The system forms solid solutions with a minimum [234]. This minimum corresponds to which has the same composition with the mineral aguillarite. Amorphous films of Ag-Se-S and Ag-Se-Te systems were investigated as regarding the kinetics of the selective solubility [235]. The illumination with UV light caused structural changes and photodarkening. The Ag-Se-S dissolves better in KOH and best results are obtained by a dissolution in a mixture of for 2.3.15. SILVER - (S, Te)

Iida et al. [236] investigated the structure of and found that the amorphous component is based on a liquid-like distribution of Ag atoms into the Se network.

2.3.16. OXYGEN - CHALCOGEN (O-Ch)

Tellurium-Oxygen (Te-O) Mechanical and thermal properties. Tellurium oxide glass was prepared [237]. The of this glass is ~273 K. The glass exhibits low melting point, and is not a hygroscopic

200

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

material. The density is [238]. The molar volume is The Debye temperature is 250 K. Bulk modulus is 32 GPa, Elastic modulus is 59 GPa (longitudinal) and 21 GPa (shear). Ultrasonic velocity is 3400 m/s (longitudinal) and 1980 m/s (shear). Electrical properties. The permittivity of glass is conductivity as a function of 1/T is given in figure 2.63[239].

The plot of the

Figure 2.63. The Arrhenius plot for the conductivity of the Te-O glass.

The activation energy, is 0.02 - 0.03 eV in the low temperature range.

eV in the high temperature range and

Optical properties. Thin amorphous films show at indices, n, and the extinction coefficients, k, are given in Table 2.42.

the refractive

CHAPTER 2

201

2.3.17. OTHER BINARY CHALCOGENIDES Aluminium-Tellurium (Al-Te) The glassy samples were prepared by heating the Al-Te mixture at 1000 °C in evacuated ampoules, followed by water quenching. was studied by gamma resonance spectroscopy (Mössbauer effect). The spectra of the glasses speak in favour of the unique

configuration of tellurium in the glass network. By very rapid quenching of the melt were obtained several glasses whose composition and transition temperature are given in the table 2.43 [242]:

Gallium-Selenium (Ga-Se) In amorphous and dielectric parameters: the electrical conductivity

et al. [58] have determined the following and respectively, and and respectively.

Gallium-Tellurium (Ga-Te) Only one configuration of tellurium has been confirmed by Mössbauer spectroscopy in glasses films with have been prepared. Two conduction state have been revealed: low resistivity state and high resistivity state. Independently of the structural state (amorphous or crystalline) and of the switching regime, the conduction in the low resistivity state is quasi-metallic. The most probable cause of the appearance of this state is the electron-phase transition [243]. The dielectric breakdown was shown to be of electro-thermal nature. Manganese-Chalcogen (Mn-Ch) Manganese chalcogenides as MnTe [244] and MnSe [245] show photoconductive properties. Films of Mn-Ch prepared by deposition on substrates held at temperatures less than 373 K are amorphous as shown by X-ray diffraction [246]. Amorphous Mn-Se films deposited on substrates held at 343 K exhibit sheet resistance of [247]. The amorphous film photoresponse was studied in the wavelength range 320 - 900 nm.

202

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES

REFERENCES [1]

Verres et Refractaires, 28(2), 49 (1974).

[2] W. Vogel, K. Gerth, Glastech. Res., 31, 15 (1958). [3] S.A. Dembovski, J. Neorg. Him., 22, 3187 (1977) [4] M. Kastner, Phys. Rev. Lett., 28, 6 (1978). [5] W. Ostwald, Z. Phys. Chem., 22, 289 (1897). [6] D. Linke, Proc. Intern. Conf. “Amorph. Semic. ’78”, Pardubice, Czechoslovakia, Vol. 1, p. 78, 1978. [7] E.V Shkolnikov, Proc. Intern. Conf. “Amorph. Semic. ’84”, Gabrovo, Bulgaria, Vol. 1, p. 36, 1982 ; E. V. Shkolnikov, Sov. J. Glass. Phys. Chem., 37, 1003 (1962). [8] D.S. Sanditov, Phys. Chem. Glass. (russ.) 3(1), 14 (1977). [9] R. Simha, R. Boyer, J. Chem. Phys., 37, 1003 (1962). [10] K. Kageyama, M. Imaoka, H. Suga, S. Seki, Fourth Jap. Calorimetry Meeting, Tokyo, 1968.

[11] M. Hattori, K. Nagaya, S. Umebachi, M Tanaka, J. Non-Cryst. Solids, 3, 195 (1970). [12] A. E. Owen, J.M. Robertson, J. Non-Cryst. Solids, 2, 40 (1970). [13] N.F. Mott, E.A. Davis, Electronic Processes in Non-Crystalline Materials,

Clarendon Press, Oxford, 1971. [14] A.E. Owen, J. Non-Cryst. Solids, 25(1-3), 372 (1977)

[15] A.L. Lakatos, M. Abkowitz, Phys. Rev., B3, 1791 (1971). [16] M. Pollack, T.H. Geballe, Phys. Rev., 122, 1742 (1961). [17]

phys. stat. sol. (b) 23, 473 (1973).

[18] S. Hudgens, M Kastner, H Fritzsche, Phys. Rev. Lett., 33, 1552 (1974). [19] V. Trkal, I. Srb, S. Dokoupil, V Rosická, Proc. Intern. Conf. “Amorph. Semic ’72, Sofia, Bulgaria, Vol 1, p. 126, 1972. [20] J. G. Kirkwood, Phys. Z., 33, 57 (1932). [21]

Proc. Intern. Conf. “Amorph. and Liq. Semic.”, Leningrad, 1976, Vol. 1, Structure of Amorphous Semiconductors, p. 13.

[22] R.M. White, P. W. Anderson, Phil. Mag., 25, 737 (1932).

[23] S.J. Hudgens, Phys. Rev. B7, 2481 (1973). [24] R. Andreichin, M. Nikiforova, E. Skordeva, L. Yurukova, R. Grigorovici, R. Manaila, M. Popescu, A. Vancu, J. Non-Cryst. Solids, 20, 101 (1976). [25] M.D. Balmakov, Z. U. Borisova, Jurn. Fiz. Him. Stekla (russ.), 2(3), 234 (1976).

[26] M.D. Balmakov, Jurn Fiz. Him, Stekla (russ), 3(3), 255 (1977). [27] M.D. Balmakov, M. S. Gutenev, L. A. Baidakov, J. Fiz. Him. Stekla (russ.) 3(5), 537 (1977). [28] Z.U. Borisova in Kalkoghenidnîie poluprovodnikovîe stekla (russ.), Leningrad Univ. 1983. [29] D. Turianitsa, V V. Himinets, T. V. Timko, Jurnal Prikhladnoi Himii (russ.), 48, 1360 (1975). [30] V.V. Himinets, L.B. Baranova, J. Prikhladnoi. Himii (russ.), 56(7), 1514 (1983). [31] E.N Jakovlev, Pisma v JETF (russ.), 28, 369, 390 (1978). [32] K. Shimakawa, J. Non-Cryst. Solids, 43, 229 (1981) [33] B. Meyer, Chemical Reviews, 76(3), 367 (1976). [34] G. Zegbe, J. Olivier-Fourcade, J.C. Jumas, I. Deszi, G. Langouche, European J. Solid State and Inorg. Chem. 30(1-2), 165 (1993)

[35] W. H. Strehlow, E.L. Cook, J. Phys. Chem. Ref. Data, Vol. 2, No. 1, 1973. [36] A.R. Hilton, J. Non-Cryst. Solids, 2, 28 (1970).

[37] Gmelins Handbuch, Verlag Chemie, Weinheim, 8. Auflage, Teil A, Lieferung 3, 1953, p. 641. [38] H. Fritzsche, S.J. Hudgens, Proc. Intern. Conf. on Amorph. and Liq. Semicond., Leningrad, Vol. 2, Electronic properties in non-crystalline semiconductors, 1976, p. 6.

[39] P.S. Varadachari, J. Annamalai Univ., 4, 73/81, 7-9 (1935). [40] K. das Gupta, S.R. Das, B. B. Ray, Indian J. Phys., 15, 389 (1941).

[41] M. Kawarada, Y. Nishina, Jap. J. Appl. Phys., 16(9), 1525 (1977).

CHAPTER 2

203

[42] E. Montrimas, B. Petretis, phys. stat. sol. (a), 15, 361 (1973). [43] A.I. Andrievskii, I.D. Nabitovitch, Sov. Jurn, Krist., 5, 442 (1960).

[44] O. Renner, J. Zemek, Czech. J. Phys., B22, (1972). [45] K.S. Kim, D. Turnbull, J. Appl. Phys., 44, 5237 (1968). [46] T. Uena, S. Muraoka, T. Shimada, A. Odajima, Bull. Fac. Eng., Hokkaido University, No. 74, 1 (1974). [47] M. Kawarada, Y. Nishina, Jap. J. Appl. Phys., 16(9), 1531 (1977) [48] N. Clavaguera, M.T. Clavaguera-Mora, S. Surinach, J. Non-Cryst. Solids, 104, 283 (1988). [49] S.O. Kasap, S. Yannapoulos, P. Gundappa, J. Non-Cryst. Solids, 111, 82 (1989). [50] T. Sato, H. Kaneko, Tech. Reports, Tohoku Univ., 14, 45 (1950). [51] H. Stegmann, Naturwiss., 44, 108 (1957). [52] E.M. Pell, Properties of Amorphous Photoconductors in Xerography and Related Processes, Ed. J. H. Dessauer and Harold E. Clark, 1977 [53] S.O. Kasap, B. Polischuk, V. Aiyah, S. Yannacopoulos, J. Non-Cryst. Solids, 114, 49 (1989). [54] V.A. Ananicev, L.A. Bajdakov, L.N Blinov, Proc. Intern. Conf. “Amorphous Semicond. ’80”, Chishinau, Moldova, Vol. 1, p. 79, 1980

[55] V.M. Glazov, S.N Chijevskaia, N. N. Glagoleva, Liquid Semiconductors (russ), Nauka, Moskva, 1967. [56] V.A. Ananicev, L.A. Bajdakov, L.N. Blinov, Proc. Intern. Conf. “Amorph. Semicond. ’80”, Chishinau, Vol. 1, p. 80, 1980. [57] W.E. Spear, Proc. Phys. Soc. B70, 669 (1957); B76, 826 (1960). [58] M. Ilyás, M. Zulfequar, Z.H. Khan, M. Hussain, Physica, B254, 57 (1998). [59] E.A. Davis, Electronic and Structural Properties of Amorphous Semiconductors,

Eds. P.G.L. LeComber, J Mort, Academic Press, N. Y., London, 1973, p. 444. [60] T.S. Moss, Photoconductivity in the Elements, Academic Press, N Y., 1952, p. 190.

[61] J.P. Audière, C. Mazières, J.C. Carballes, J. Non-Cryst. Solids, 27, 411 (1978). [62] H.A. Davies, J.B. Hull, J. Mat. Sci., 9, 707 (1974). [63] I.W. Donald, H.A. Davies, J. Non-Cryst. Solids, 30, 77 (1978).

[64] G. Tourand, B. Cabane, M. Breuil, J. Non-Cryst. Solids, 8-10, 676 (1972). [65] A. Feltz, Amorphe und Glasartige anorganische Festkörper, Akademie Verlag, Berlin, 1983.

[66] A.F. Joffe, A.R. Regel, Progress in Semicond., 4, 238 (1960). [67] J. Stuke, Festkörperprobleme, IX, 46, (1969).

[68] A.M. Phahle, Thin Solid Films, 61, L21 (1979). [69] J. Mossinski, A. Renninger, B. L. Averbach, Phys. Lett. A42, 453 (1973).

[70] F. Halla, E. Mehl, F X. Bosch, Z. Phys. Chem., B12, 377 (1931). [71] M.K. El-Mously, M.M. El-Zaidia, J. Non-Cryst. Solids, 11, 519 (1973). [72] F. Kakinuma, S. Ohno, K Suzuki, J. Non-Cryst. Solids, 156-158, 691 (1993). [73] W.D. Gill, G. B. Street, J. Non-Cryst. Solids, 13, 120 (1973/1974). [74] D. Jecu, J. Jaklovszky, A. Trutia, I. Apostol, M. Dinescu, I.N. Mihailescu, G. Aldica, M. Popescu, N. Vlahovici, S. Zamfira, E. Indrea, J. Non-Cryst. Solids, 90, 319 (1987). [75] N. Afify, J. Non-Cryst. Solids, 14, 247 (1992).

[76] L.N. Suvorov, Z.U. Borisova, G.M. Orlova, Izv. Akad. Nauk SSSR, Ser. Neorg. Mat. (russ.), 10(3), 441 (1974). [77] K. Nakayama, K. Kojima, I. Tamaru, Y. Masaki, A. Kitagawa, M. Suzuki, J. Non-Cryst. Solids, 198&200, 758 (1996). [78] H. Yang, W. Wang, S. Min, J. Non-Cryst. Solids, 80, 503 (1986). [79] G.C. Das, M.B. Bever, D.R. Uhlmann, J. Non-Cryst. Solids, 7, 251 (1972). [80] G. Carini, M. Cutroni, M. Federico, G. Galli, J. Non-Cryst. Solids, 64, 317 (1984). [81] E. Illékova, M.T. Klavaguera-Mora, M.D. Baro, S. Surinach, Mat. Sci. Eng., B22, 181 (1994). [82] A.N. Sreeram, A.K. Vashnyia, D.R. Swiler, J. Non-Cryst. Solids, 128, 294 (1991).

204

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES [83] S.H. Hagen, P.J.A. Derks, J. Non-Cryst. Solids, 65, 241 (1984). [84] F. Heyder, D. Linke, Z . Chem., 13, 480 (1973). [85] T.S. Rikova, Z.U. Borisova, Vestnik Leningrad Univ., No. 10, 153 (1977).

[86] E.I. Kim, M.S. Gutenev, I.V. Viktorovski, L.A. Bajdakov, J. Prikhladnoi Himii (russ.), 2, 281 (1975). [87] V.M. Lyubin, G.A. Feodorov, Neorg. Mat. (russ.), 8(9), 1559 (1972). [88] K.L. Chopra, L.K. Malhotra in Physics ofDisordered Materials, IAS Series,

Eds. D. Adler, H. Fritzsche, S. R. Ovshinsky, Plenum Press, 1985, p. 215. [89] Iu. S. Tverianovici, L.A. Baidakov, T.S. Rikova, V.N. Jurba, in Amorphous and Vitreous Semiconductors (russ.), Vol.2, p. 36, Kaliningrad, 1977.

[90] N.V. Timofeeva, G.Z. Vinogradova, E.M. Feklicev et al., Dokl. Akad. Nauk SSSR, 190(4), 902 (1970). [91] M. Tanaka, T. Minami, M. Hattori, Jap. J. Appl. Phys., 5, 185 (1966).

[92] J. Tauc in Amorphous and Liquid Semiconductors, ed. J. Tauc, Plenum Press, London and New York, 1974, p. 209. [93] M. Popescu, A, Andriesh, V. Chumash, M . Iovu, S. Shutov, D. Tsyuleanu, Fizica Sticlelor Calcogenice (romanian), Ed. Bucharest-Chishinau, 1996. [94] R. Andreichin, M. Nikiforova, E. Skordeva, Proc. Intern. Conf. “Amorph. Semic. ’74”, Reinhardsbrunn, Germany, p. 224, 1974

[95] T. Velinov, M. Gateshki, D. Arsova, E. Vateva, Phys. Rev., B55(18), 11014 (1997). [96] P.K. Bhat, K.L. Bathia, S.C. Katyal, J. Non-Cryst. Solids, 27, 399 (1978). [97] A.M. Andriesh, et al., Kvantovaia Elektronika, 4, 629 (1977). [98] J.M. Stevels, J. Ceram. Assoc. Japan, 72, 176 (1964). [99] M.S. Gutenev, Fiz. Him. Stekla (russ.), 11(3), 311 (1985). [100] A. Feltz, H. Aust, A. Blayer, Proc. Intern. Conf. “Amorph. Semicond’ 82”, Bucharest 1982, Vol. R, p. 126.

[101] J.K. Zakis, H. Fritzsche, phys. stat. sol. (b), 64, 123 (1974). [102] J. Heo, J.D Mackenzie, J. Non-Cryst. Solids, 111, 29 (1989).

[103] A.B. Seddon, J. Non-Cryst. Solids, 184, 44 (1995). [104] J. A. Savage, Infrared Optical Materials and their Antireflection Coatings, Adam Hilger, Bristol, 1985. [105] L.A. Bajdakov, N.A. Novoselova, L.P. Strachov, Izv. Akad. Nauk SSSR, Neorg. Mat., 4, 193 (1968). [106] J. Tauc, A. Menth, Rep. Washington Meeting of the APS, April, 1970. [107] Z. Cimpl, F. Kosek, M. Matyáš, phys. stat. sol., 41, 535 (1970). [108] A. Feltz, W. Burchardt, B. Voigt, D. Linke, J. Non-Cryst. Solids, 129, 31 (1991). [109] Z.U. Borisova, Himia Poluprovodnikov, (russ.) Leningrad Univ. 1972. [110] G.C. Das, N. S. Platakis, M.B. Bever, J. Non-Cryst. Solids, 15, 30 (1974). [111] Sovremennie problemi Fiziceskoi Himii (russ.), Moskva, Izd. MGU, 6, 234 (1972).

[112] C.A. Maijd, P. R. Prager, N. H. Fletcher, J. M. Bretell, J. Non-Cryst. Solids, 16, 365 (1974). [113] S. Yannacopoulos, Wiswanath Aijah, S. O. Kasap, J. Non-Cryst. Solids, 115, 54 (1989). [114] D.D. Thornberg, R.I. Johnson, J. Non-Cryst. Solids, 17(1), 2 (1975).

[115] A. P. Chernov, Ph. D. Thesis, Moscow 1970. [116] A.B. Seddon, M. J. Laine, J. Non-Cryst. Solids, 213&214, 168 (1997). [117] D. W. Henderson, D. G. Ast, J. Non-Cryst. Solids, 64, 43 (1984).

[118] C.T. Hach, K. Cerqua-Richardson, J. R . Varner, W C. LaCourse, J. Non-Cryst. Solids, 209, 159 (1997). [119] V.M. Jdanov, A.K. Maltsev, J. Fiz. Him (russ.), 42(2), 2051 (1968).

[120] V. Sadagopan, H.C. Gatos, J. Solid State Electronics, 9(1), 17 (1966). [121] B.E. Knox, V.S. Ban, Mat. Res. Bull., 3(11), 885 (1968). [122] M. Kitao, Jap. J. Appl. Phys. 11(10), 1472 (1972).

CHAPTER 2

205

[123] N. S. Platakis, H. C. Gatos, J. Electrochem. Soc., 119, 914 (1972). [124] M. Iovu, S. Shutov, M. Popescu, D. Furniss, L. Kukkonen, A.B. Seddon, J. Optoel. Adv. Mat., 1(2), 15 (1999). [125] F.D. Fisher, J.M. Marshall, A.E. Owen, Phil. Mag., 33, 261 (1976). [126] H.L. Haltaus, G. Weiser, S. Nagel, phys. stat. sol. (b), 87, 117 (1978). [127] A. Feltz, H. Aust, A. Blayer, J. Non-Cryst. Solids, 55, 179 (1983). [128] K. Tanaka, J. Non-Cryst. Solids, 35&36, 1023 (1980). [129] V. Lyubin, M. Klebanov, L. Shapiro, M. Lisiansky, B. Spektor, J. Shamir, J. Optoel. Adv. Mat., 1(3), 31 (1999). [130] E. Skordeva, J. Optoel. Adv. Mat., 1(1), 43 (1999). [131] K. Tanaka, S. Nakayama, J. Optoel. Adv. Mat., 2(1), 5 (2000). [132] R. Grigorovici, T. Stoica, A. Vancu, Proc. Intern. Conf. “Amorph. Semicond. ’80”, Khishinew, 1980, part B, p. 201. [133] B. Voigt, D. Linke, Physics and Applications of Non-Crystalline Semiconductors in

Optoelectronics, Ed. A. Andriesh and M. Bertolotti, NATO ASI Series, 3 High Technology, Vol. 36, 1996, p. 155 [134] J. Cornet, J. Non-Cryst. Solids, 12(1), 61 (1973).

[135] B.T. Kolomiets, O. V. Pavlov, Fiz. Tehn. Poluprovod. (russ.), 1, 350 (1967). [136] N.S. Platakis, IBM Technical Report 02.769, 1977.

[137]

I. Kubelik, A . Triska, Proc. Intern. Conf. “Amorph. Semic. ’74”, Reinhardsbrunn, Vol. 2, 1974, p. 421.

[138] A. Abraham, A. Hrubý, J. Non-Cryst. Solids, 8-10, 353 (1972). [139] L.A. Bajdakov, L. P. Strachov, Proc. Intern. Conf., ICALS’6, Leningrad, 1974, Vol. 1, p. 92. [140] J. Hajtó, T. Kemény, Proc. Intern. Conf. “Amorph, Semic,’ 76”, Balatonfüred, 1976, Ed. I. Kósa Somogyi, p. 493. [141] A. Nemilov, Y. Petrovsky, Fiz. Tverd. Tela, Solid State Phys. (russ.), 36, 222 (1963). [142] D. Turnbull, M. H. Cohen, J. Chem. Phys., 52, 3038 (1970).

[143] Z.U. Borisova, Chalcogenide Semiconducting Glasses (russ.), Leningrad, Univ. Press, 1983, Chap. 1. [144] K. Weiser, R. Fischer, M.H. Brodsky, Proc. X-th Intern. Semic. Conf. Cambridge Mass., 1979. [145] W. Henrion, Proc. Intern. Conf. ‘Amorph. Semicond, ‘76”, Balatonfüred, 1976, Ed. I. Kósa Somogyi, p. 255 [146] A.R. Hilton, C.E. Jones, M. Brau, Phys. Chem. Glasses, 7(4), 105 (1966) (part. I).

[147] S.A. Dembovskii, Izv. Akad. Nauk SSSR, Neorg. Mat. 4(11), 1920 (1968). [148] Stekloobraznii sulfid mishiaka (russ.), ed. B.T. Kolomiets, Stiintza, Khishinau, 1981 [149] M. Popescu, A. Andriesh, V. Chumash, M. Iovu, S. Sutov, D. Tsiulyanu, Fizica Sticlelor

Calcogenice (romanian), Chapter 1, Ed. Stiintzifica-Stiintza, Bucharest-Chishinau, 1996. [150] M. Popescu, F. Tudorica, A. Andriesh, M. Iovu, S. Shutov, M. Bulgaru, E. Colomeiko, S. Malkov, V. Verlan, M. Leonovici, V. Mihai, M. Steflea, Bul. Acad. St. Rep. Moldova, Fiz. i Tehn. (romanian), Nr. 3, 3 (1995). [151] F. Hulderman, J. S. Sanghera, J. D. Mackenzie, J. Non-Cryst. Solids, 127(3), 312 (1991). [152] M.S. Ablova, A.A. Andreyev, B.T. Melekh, A.B. Pevtsov, Proc. Intern. Conf. “Amorph. Semicond. ’76”, Balatonszéplak, 1976, p. 423. [153] G. Dalba, P. Fornasini, G. Giunto, J. Non-Cryst. Solids, 107, 261, 1989. [154] K. Morii, S. Wanaka, Y. Nakayama, Mat. Sci. Eng., B15, 126 (1992). [155] V.A. Bazacuta, V.D. Kulibaba, V.A. Moghileski, Voprosi fiziki poluprovodnikov(russ.), Kaliningrad University, VIP.1, 126 (1975). [156] M. A. Iovu, M. S. Iovu, S. D. Shutov, Fizika Poluprovodnikov, (russ.),

Ed. Stiintza, Chishinau, 1979, p. 11. [157] N. Afify, J. Non-Cryst. Solids, 126, 130 (1990). [158] A.N. Sreeram, A. K. Varshneya, D. R. Swiler, J. Non-Cryst. Solids, 128, 294 (1991).

206

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES [159] Z. Hurych, R. Mueller, C.C. Wang, C. Wood, J. Non-Cryst. Solids, 11, 153,(1972).

[160] V.V. Himinets, Ia. P. Kupenko, V.P. Pinzenik, M.V. Dobosh, I.D. Turianitsa, Proc. Intern. Conf. “Amorph. Semicond. ‘80”, Chishinau 1980, p. 120 (Vol. 1). [161] D. Brasen, J. Non-Cryst. Solids, 15, 395 (1974). [162] C. Wood, R. Mueller, L.R. Gilbert, J. Non-Cryst. Solids, 12, 295 (1973).

[163] K. Watanabe, N. Sato, S. Myaoka, J. Appl. Phys., 54, 1256 (1983). [164] C.M. Garner, L.R. Gilbert, C. Wood, J. Non-Cryst. Solids, 15, 63 (1974). [165] G.Z. Vinogradova, Glass Formation and Phase Equilibrium in Chalcogenide Systems, Nauka, Moskva, 1984, p. 16. [166] H. Gabriel, C. Alvarez, Tostado, J. Amer. Chem. Soc., 74(1), 262 (1952). [167] K. E. Petersen, U. Birkholz, D. Adler, Phys. Rev. B, 8(4), 1453 (1973). [168] S.A. Altunian, G.V. Greciushkin, V.S. Minaev, B.K. Skacikov, IU. M. Ukrainskii, Proc. Intern. Conf. “Amorph. Semicond. ’72”, Sofia 1972, Vol 2, p. 43.

[169] M.K. Gauer, I. Dezsi, U. Gonser, G. Langouche, H. Ruppersberg, J. Non-Cryst. Solids, 109, 247 (1989). [170] S. Baliavicius, A. Deksnish, A. Poshkus, N. Shiktorov, Proc. Intern. Conf. “Amorph. Semicond. ’82”, Bucharest 1982, Vol. R, p. 235. [171] K.N. Madhudsoodanan, J. Philip, G. Parthasarathy, E. S. R. Gopal, J. Non-Cryst. Solids, 109, 255 (1989).

[172] Y. Kawamoto, S. Tsuchihashi, J. Amer. Ceram. Soc., 52(11), 626 (1969). [173] A. Feltz, W. Burckhardt, L. Senf, K. Zickmüller, Z. anorg. allg. Chem., 435, 172 (1977). [174] A. Feltz, J. Non-Cryst. Solids, 69, 271 (1985).

[175] G.H. Ivanov, B.T. Kolomiets, V.M. Lyubin, V.P. Shilo, Proc. Intern. Conf. “Amorph. Semic. ’72”, Sofia, 1972, Vol. 1, p. 88. [176] A. Feltz, W. Burchardt, B. Voigt, D. Linke, J. Non-Cryst. Solids, 129, 31 (1991). [177] Xingwei Feng, W. J. Bresser, P. Boolchand, Phys. Rev. Lett., 78(23), 422 (1997).

[178] J. Watanabe, T. Maeda, T. Shimizu, J. Non-Cryst. Solids, 37, 335 (1980). [179] A. Abraham, A. Hrubý, Proc. Intern. Conf. “Amorph. Semic. ’74”, Reinhardsbrunn 1974, part. II, p.358. [180] A. Abraham, Czech. J. Phys., B24, 1406 (1974). [181] Proc. Intern. Conf. “Amorph. Semic.’ ’74”, Reinhardsbrunn 1974, part. II, p.288. [182] S. Surinach, M. D. Baro, M. T. Clavaguera-Mora, J. Non-Cryst. Solids, 111, 113 (1989). [183] J.P. deNeufville, J. Non-Cryst. Solids, 8-10, 85 (1972).

[184] H. Takeuchi, O. Matsuda, K. Murase, J. Non-Cryst. Solids, 238, 91 (1998). [185] S.S. Fouad, S.A. Fayek, M.H. Ali, Vacuum. 49(1), 25 (1998).

[186] F. Smektala, C. Quémard, L. Leneindre, J. Lucas, A Barthélémy, C. De Angelis, J. Non-Cryst. Solids, 239, 139 (1998). [187] G.H. Ivanov, V.M. Lyubin, Proc. Intern. Conf. “Amorph. Semic. ’72”, Sofia 1972, part I, p. 88. [188] I.D. Olekseyuk, V.V. Bozhko, O.V. Parasyuk, V.V. Galyan, I.I. Petrus, Functional Materials,

6(3), 550(1999). [189] M.F. Kotkata, K.M. Kandil, M.L. Thèye, J. Non-Cryst. Solids, 164&166, 1259 (1993). [190] E. Sleeckx, L. Tichý, P. Nagels, R. Callaerts, J. Non-Cryst. Solids, 198&200, 723 (1996). [191] K. Murayama, J. Non-Cryst. Solids, 59&60, 983 (1983)

[192] V. Petrova, T. Stoica, H. Stötzel, A. Vancu, L. Vescan, Proc. Intern. Conf. “Amorph. Semic. ’78”, Pardubice 1978, p.472. [193] T.T. Nang, M. Okuda, J. Non-Cryst. Solids, 33(3), 311 (1979). [194] J.P. deNeufville, J. Non-Cryst. Solids, 8-10, 85 (1972). [195] J. Cornet, Ann. Chim. (Fr.) 10(4-5), 239 (1975). [196] T. Takamori, R. Roy, G.J. McCarthy, Mat. Res. Bull., 5, 529 (1970). [197] K.L. Chopra, S.K.. Bahl, J. Appl. Phys., 40, 4171,1971.

CHAPTER 2

207

[198] Z.U. Borisova, Khalkoghenidnîe Poluprovodnikovîe Stekla (russ.), Leningrad Univ., 1983, p. 112-113. [199] S.P. Vihrov, P.T. Oreshkin, V.N. Ampilogov, A.S. Glebov, B.N. Sajin, Proc. Intern. Conf. “Amorph. Semic. ‘82”, Bucharest 1982, Vol. 2 in russian, 1982, p. 77. [200] L.A. Bajdakov, V.R. Panus, V.G. Somova, Vestnik. Leningrad Univ., 10, 119 (1974). [201] Ia. M. Nesterova, A.S. Pashinkin, A.V. Novoselova, J. Neorg. Him. (russ.), 6, 2014 (1961). [202] J.M. Walker, J.W. Straley, A.W. Smith, Phys. Rev. 53, 140 (1938).

[203] M.S. Palatnik, V.V. Levitin, Dokl. Akad. Nauk SSSR (russ.), 96, 975 (1954). [204]

A. Hrubý, J. Non-Cryst. Solids, 30, 191 (1978).

[205]

A. Hrubý, Proc. Intern. Conf, “Amorph. Semic. ’76”, Balatonfüred, 1976, p. 481.

[206] H. Hahn, W. Klinger, Z. fur Anorg. Chemie, 260, 110 (1949).

[207]

A. Hrubý, Proc. Intern. Conf. “Amorph Semic. ‘76”, Balatonfüred, Hungary, Akadémiai Kiadó, Budapest 1976, p. 261.

[208] F. Kósek, Z. Cimpl, J. Chlebný, Proc. Intern. Conf. “Amorph. Semic. ’76”,

Balatonfüred, 1976, p. 277. [209] N.A. Goryunova, B.T. Kolomiets, Akad. Nauk SSSR, Voprosi metallurghii i fiziki poluprovodnikov (russ.), Moscow, 1957, p. 110. [210] G.Z. Vinogradova, Glass formation and phase transition in chalcogenide glasses, Moscow, 1984, p. 15 [211] K.A. Chuntonov, A.N. Kuznetsov, V.M. Feodorov, S.P. Iatsenko, Izv. Akad. Nauk SSSR, Ser. Neorg. Mat., 18(7), 1108 (1982). [212] E.N. Minaev, Elektronnaia Tekhnika (russ.), Ser. Mat. 9(208), 44 (1982). [213] J.C. Schotmiller, D.L. Bowman, C. Wood, J. Appl. Phys., 39, 1663 (1968). [214] T. Takahashi, “J. Non-Cryst. Solids”, 44, 239 (1981). [215] V.G. Golubkova, E.S. Petrov, Izv. Sibir. Otdel. Ak. Nauk SSSR, Ser. Him. Nauk, 2, 114 (1975). [216] W. Swiatkowsky, J. Phys. Chem. Solids, 41, 665 (1980).

[217] V.A. Ignatiuk, N.N. Stavnistii, E.N. Minaev, Proc. Intern. Conf. “Amorph. Semic. ’80”, Chishinau 1980, p. 117.

[218] W.J. Bresser, J. Wells, M. Zhang, P. Boolchand, Z. Naturforsch., 51a, 373 (1996). [219] P. Hagenmuller, F. Chopin, C.R. Acad. Sci. (Fr.) 255(18), 2259 (1962). [220] R. Hillel, J. Cueilleron, Bull. Soc. Chim. (Fr.), 1972, p. 98. [221] E. Diemann, Z. anorg. allg. Chemie, 432, 127 (1977).

[222] E. Diemann , Z. anorg. allg. Chem., 461, 201 (1980). [223] E. Diemann, A. Müller, Z. anorg. allg. Chem., 444, 181 (1978).

[224] J.D. Passaretti, R.C. Collins, A. Wold, Mat. Res. Bull., 14, 1167 (1979). [225] K.S. Liang, S.P. Cramer, D.C. Johnston, C.H. Chang, A.J. Jacobson, J.P. de Neufville, R.R. Chianelli, J. Non-Cryst. Solids, 42, 345 (1980). [226] K.S. Liang J.P. deNeufville, A.J. Jacobson, R.R. Chiavelli, J. Non-Cryst. Solids, 35&36, 1249(1980). [227] E. Diemann, Die Naturwiss., 63(8), 385 (1976). [228] A.J. Jacobson, R.R. Chiannelli, S.M. Rich, M.S. Whittingham, Mat. Res. Bull., 14, 1437 (1979). [229] R. Deroide, P. Belougne, J.C. Giuntini, J. V. Zanchetta, J. Non-Cryst. Solids, 85, 79 (1986). [230] T. Sakuma, K. Iida, K. Honma, H. Okazaki, J. Phys. Soc. Japan, 43(2), 538 (1977). [231] M.I. Korsunsky, S. Ia. Maksimova, Sh.Sh. Sarsembinov, Proc. Intern. Conf. “Amorph. Semic. ’76”, Balatonfüred, 1976 Akademiai Kiadó p. 111. [232] S. Miyatani, J. Phys. Soc. Jap., 15, 1586 (1960). [233] H. Okazaki, J. Phys. Soc. Jap., 23, 355 (1967). [234] Z. Bontscheva-Mladenova, K. Zaneva, Z. anorg. allg. Chem., 437, 253 (1978). [235] M. Mitkova, Z. Boncheva-Mladenova, J. Non-Cryst. Solids, 90, 589 (1987). [236] K. Iida, K. Honma, H. Okazaki, Nippon Kinzoku Gakkaishi (jap.) 38, 682 (1974). [237] E.F. Lambson, G.A. Saunders, B. Bridge, R.A. El-Mallawany, J. Non-Cryst. Solids,

208

PHYSICO-CHEMICAL PROPERTIES OF CHALCOGENIDES 69, 117 (1984). [238] M.A. Sidkey, R. El Mallawany, R.I. Nakhla, A. Abd El-Moneim, J. Non-Cryst. Solids, 215, 75 (1997). [239] R.N. Hampton, W. Hong, G.A. Saunders, R.A. El-Mallawany, J. Non-Cryst. Solids. 94, 307 (1987). [240] H. Seki, Appl. Phys. Lett., 43, 1000 (1983). [241] V. Kozhukharov, D. Dimitrov, M. Marinov, J. Non-Cryst. Solids, 129, 117 (1991). [242] J.A. Savage, J. Non-Cryst. Solids, 11, 121 (1972). [243] A. Chesnis, A. Oghinskas, A. Lisauskas, Proc. Intern. Conf. “Non-Cryst. Semic. ’89”, Uzhgorod, 1989, Vol. 2, p. 34. [244] P. Singh, Indian J. Pure and Appl. Phys., 18, 950 (1980). [245] M.A. Angachi, V. Thanigaimani, phys. stat. sol. (a), 135, 183 (1993). [246] C. Junien, M. Eddrieff, K. Kambas, M. Balkanski, Thin Solid Films, 137, 27 (1986). [247] V. Thanigaimani, M.A. Angadi, Mat. Sci. Eng., B18, L7 (1993). [248] K. Tanaka, Solid State Comm., 60, 295 (1986).

CHAPTER 3

209

MODIFICATIONS INDUCED IN NON-CRYSTALLINE CHALCOGENIDES

The non-crystalline chalcogenides are solids without long-range order and, as a consequence, they are intrinsically metastable. They exhibit properties different from their crystalline counterparts. The structure and bond configuration of these disordered materials can be changed, sometimes reversibly, either by thermal treatment or by various external factors: light and other electromagnetic radiation, particle beams (electrons, neutrons...), electrical and magnetic fields, pressure, etc. The structure and properties of the glassy or amorphous chalcogenides can be, thus, modified. 3.1. Modifications Induced by Light The chalcogenide glasses are susceptible to light-induced changes because they are characterised by intrinsic structural flexibility. The chalcogen elements are two-fold coordinated. The atoms possess lone pair electrons, which are normally non-bonding, but they undergo light induced reactions and give rise to structural defects of three-fold and single co-ordinated chalcogen. The states associated with the non-bonding electrons lie at the top of the valence band and hence are preferentially excited by illumination. Significant changes in the physical properties and structural modifications induced by light have been evidenced in many chalcogenide glasses and amorphous chalcogenide films. The first reports on the changes induced by light in chalcogenide glasses were published by Keneman [1], Berkes, Ing and Hillegas [2] and by Pearson and Bagley [3] in 1971. The photo-induced modifications in chalcogenide glasses include changes in density [4], hardness [5], rheological properties [2], chemical reactivity [6], dissolution rate [7,8], electrical [9-11] and optical properties [4,6,12,13] as well as decomposition [4,8,14] and crystallisation [15]. As a function of the experimental conditions and of the material composition the modifications induced by light can be reversible, partially reversible and irreversible. The photo-induced transformations were observed in elemental chalcogens (Ch) sulphur and selenium, in binary systems (As-Ch, Ge-Ch) in ternary systems (As-Ch-Ge) and in more complicated compositions. Irreversible changes are possible both in crystalline and amorphous (or glassy) chalcogenides. Irreversible photo-induced modifications were evidenced in unannealed films prepared by vacuum evaporation or sputtering [13]. The reversible modifications are usually observed in well-annealed amorphous films and in bulk glasses. The irreversible component can be eliminated by annealing the material at temperatures near The reversible changes are proper to non-crystalline chalcogenide materials and are

210

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

lacking in crystalline chalcogenides. They are not present in pure arsenic. The crucial point of the phenomenon is the presence of the chalcogen element in the non-crystalline alloy [16]. The optical irreversible modifications induced by light with near band gap energy were evidenced in the systems As-S and As-Se [2,13] and called photodarkening. The photodarkening is accompanied by a remarkable structural change (related to a complex polymerisation process), which determines the red shift of the optical absorption edge [8,13]. The structural basis of the photodarkening process is not yet fully understood. Other types of irreversible changes induced by light are also possible. Usually, the modification of the refractive index during illumination of an amorphous chalcogenide film is not high (for As2S3 the refractive index changes from 2.447 to 2.569 [17]). Nevertheless, in As-S and As-Se films were observed catastrophic alterations of the refractive index (down to accompanied by the increase, with a factor of of the ratio of the dissolution rates in alkaline solvents of the irradiated and non-irradiated parts of the film [18]. The irreversible processes are at variance with the reversible ones not only qualitatively but also quantitatively.

In many cases the reversible changes are characterised by the appearance of photodarkening while the irreversible ones give rise to the so-called photobleaching effect (the blue shift of the optical absorption edge). The direction of the shift of the optical absorption edge depends on the initial irradiation of the sample, on the temperature, and on the light intensity. The irreversible component of the photo-modification is strongly influenced by the

preparation conditions. Thus, the irreversible changes induced in Ge-Se amorphous films are enhanced by the oblique deposition of the material [19] while in As-Se films the ultrahigh deposition rate is favourable [20]. Irreversible photo-induced modifications of the thickness of the amorphous films (contraction) have been also observed [21-23]. In chalcogenide alloys it is possible to have photolysis [2], photo-evaporation [24], photo-crystallisation [25] and various photochemical processes. Usually, several processes act concomitantly and this fact makes more difficult the understanding of the atomic-scale phenomena. Photo-induced losses (absorption and scattering) are produced in chalcogenide glasses under exposure to high-power laser radiation, when the photon energy is higher than the energy gap of the material Such irradiation results in both reversible and irreversible effects [26]. The refractive index variation and darkening under irradiation with Ar and He- Ne lasers, in the green and red spectral regions, was investigated [27-29]. It was found that in the range of irradiance of tens of watt/cm2, optical bistability, transparency oscillations and complicated diffraction patterns do occur. Structural relaxation of different types, connected with weakening of the chemical bonds, is supposed to cause these phenomena. Tikhomirov and Elliott [29] observed scattering from the irradiated glass volume. This scattering was attributed to the interaction of the laser beam with imperfections situated in the glass volume and at the surface. The glass imperfections at larger scale can play an important role in the optical modifications induced by laser beams.

CHAPTER 3

211

A recent review of the photo-induced metastability in amorphous chalcogenides has been done by Kolobov and Tanaka [30]. 3.1.1. IRREVERSIBLE MODIFICATIONS Photo-physical transformations

a) Photo-vaporisation. Firstly, the photo-vaporisation was observed in by Janai and Rudman [31,32]. They concluded that this phenomenon is controlled by the photooxidation reaction followed by thermal evaporation of the volatile product. Under the action of high power laser pulses the amorphous chalcogenides exhibit local evaporations. Feinleib et al. [33] have shown that the modifications of the optical properties of the chalcogenide semiconductors are closely related to the formation of bubbles induced by rapid vaporisation during the impact of the light pulse. In the chalcogenide films the softening temperature is several tens of degrees above the ambient temperature and the evaporation temperature is situated above the softening temperature by about several hundreds of degrees. As a consequence, the intense illumination of the amorphous material can induce melting and vaporisation. The vapour pressure increases during light absorption and, gradually, will appear some cavities (bubbles) filled by vapours. If the light is switched off, then a rapid cooling occurs and the bubble wall strengthens while the internal space becomes empty due to the vapour condensation. The bubbles migrate from the film surface towards the film centre. If the substrate is transparent then it will preserve its temperature and the amorphous layer in contact will be cooled. As a consequence, the bubbles will be formed within the film at some distance from the interface. If the sample is heated up to then the material will flow and the bubbles will disappear. Such behaviour is specific to amorphous alloys, as e.g. the selenium-based alloys, which soften by heating, without decomposition. The selenium alloys exhibit down to ~50 °C In order to facilitate the formation of the bubbles it is important to choose an amorphous film with a narrow temperature interval for the transition in the glassy state. As a function of the rheological properties of the amorphous film the bubbles form clusters, disappear or remain distinct spheres. In the case of the amorphous selenium films which exhibit high plasticity at room temperature, the bubbles takes a diskoidal shape due to the decrease of the vapour pressure as a consequence of the condensation of selenium vapours on the bubble walls. In the thin film of alloy have been inscribed bubbles of in diameter by irradiation with pulses of 200 mW and duration (800 nJ per pulse) emitted by a krypton laser. Other alloys based on Ge, As, Te, show higher softening temperatures and, as a consequence, the bubble shell will be more rigid and the bubbles will be more stable at room temperature.

b) Photo-crystallisation and photo-amorphisation. The crystallisation of the amorphous films under illumination is a challenging physical effect. The photo-crystallisation was observed firstly in amorphous selenium by Dresner and Stringfellow [34]. Some scientists

212

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

suggest that the crystallisation is the result of the direct action of the light and not a thermal effect. Other scientists believe that the thermal effect plays an essential role. Chaudhari et al. [35] have shown that, by irradiation with He-Ne or Ar laser beam, in the amorphous films of composition occur structural modifications. In the illuminated regions are formed crystallites, which disappear after further illumination in different regime of irradiation. The illuminated zones show optical properties at variance

with those from the virgin ones. This fact can be explained both by the appearance of small crystallites and, by the morphological change of the amorphous film subjected to the laser beam. After Feinleib et al. [36-37] during the absorption of light of given wavelength takes place a stepwise increase of the free carrier concentration, related to the breaking of the covalent bonds. The breaking of the bonds weakens the metastable amorphous state and, therefore, it results the increase of the rate of crystallisation. The transition from the amorphous to the crystalline state under the action of the light is the direct consequence of the electron-hole formation by the absorption of light quanta. The reversible crystallisation takes place during illumination. The formation of the electron-hole pairs is a very efficient process in the case of the absorption of the high-energy photons. As a consequence the following effects must be considered: a). The recombination of the charge carriers thus formed followed by the increase of the temperature and of the mobility of the atoms and b). The weakening of the bonds between the atoms as a consequence of the formation of the charge carriers. Paribok-Alexandrovitch [38] revealed that the crystallisation rate of the amorphous selenium increases under light irradiation. A significant effect on the crystallisation rate was observed for the light of wavelength less than 560 nm. The energy of these quanta determines the breaking of the selenium chains, thus facilitating the reordering of the shorter chains. A crystallisation under the light irradiation was also observed in Se-Te [36]. The crystallisation rate of a film with the composition under the laser pulses film thickness: has been analysed. The rate of advancement of the crystallisation front was found The crystallisation is accompanied by a strong change of the reflection and transmission of the film. It has been concluded that the mechanism of crystallisation is optical in nature and not thermal. In favour of the optical mechanism of crystallisation speaks the peculiarities of the re-amorphisation process. In the photo-crystallised and then amorphised zones under the action of an other different light pulse, the material becomes completely amorphous. Therefore, the heat dissipation is not sufficient for the crystallisation of the surrounding amorphous material. This conclusion is based on the fact that the thermal crystallisation cannot be produced in very short times (less than a few milliseconds). Nevertheless, no definite proof for the purely optical character of the photo-crystallisation effect seems to exist [39,40]. Haro et al. [41] studied the photo-crystallisation induced by a laser beam in amorphous They demonstrated that the initiation of the process needs a minimum density of energy. Above this threshold value the energy transferred to the system is spent either for electronic transitions (with the consequence of bond breaking) or for

CHAPTER 3

213

mechanical transitions (vibrations and rotations of the clusters). This energy transfer to the system leads to structural changes in the material. Three stages in the phototransformations as a function of light intensity have been observed. The first one consists in the formation of nuclei and their growth due to the appearance around them of free volumes. The growth continues up to the limit of a void-free system with crystallites embedded in the glassy matrix. In the second stage, when the irradiation power increases, the crystallites coalesce and a new dense crystalline material is formed. This transformation is partially reversible. If the laser power is diminished the system will relax towards an equilibrium state between microcrystallites and crystallites. The third stage appears for long time irradiation with light beam of very high energy. The system becomes completely fixed in the crystalline state (the transformation is irreversible). Dresner [34] observed an effect of amplification with a factor of 20 of the crystallisation rate of selenium during illumination, this corresponding to the creation of electron-hole pairs/ µm2s. The crystallisation rate is cm/s. From the spectral dependency of the crystallisation rate was concluded that the process is controlled by the concentration of the charge carriers and not by the absorbed energy. Jecu et al. [42,43] observed the crystallisation effects in glassy compositions from the system Se-S subjected to ruby laser pulses. The glass corresponds in the binary diagram to the eutectic with minimum melting temperature of 105 °C. The absorption edge of this material is situated in the neighbourhood of the photon energy of the laser pulse (the photon wavelength in the laser pulse is and the wavelength corresponding to band gap in the material is Darkening of the material appears immediately after irradiation if the energy of the pulse is more than 980 mJ or after one hour (for energies of or after 24 hours at energies of or does not appear anymore for energies less than 100 mJ. The focal spot was maintained constant with in diameter. There was concluded that, as a function of the power of the electromagnetic radiation, soon or later ordered nuclei do appear and they induce the crystallisation. It is quite remarkable that in all the samples the crystallisation induced by light appears after some weeks of storage. The crystalline phase was identified as stable phase. Oriented crystallisation of amorphous selenium induced by linearly polarised light was observed by Tykhomirov et al. [44]. Illumination with polarised (laser) light above causes the formation of crystalline nuclei with a specific preferred orientation. Prolonged illumination causes anisotropic crystal growth from those nuclei. In general, the crystallisation of glassy and amorphous chalcogenide semiconductors is induced by near or above band-gap light. The photo-induced crystallisation process occurs at lower temperature than the thermal crystallisation temperature in the dark. This implies that the most expensive part of the thermal crystallisation process is substituted by the photo-process. The photo-induced crystallisation implies the long-range ordering as well as the local or mesoscopic ordering. In chalcogenide semiconductors, the electronic excitation by photon absorption is supposed to be able of achieving the bond switching which is the origin of the other photo-induced structural changes such as the photoinduced anisotropy [45]. The same kind of bond switching mechanism should also be involved in the photo-induced crystallisation process. On the other hand, there should be

214

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

is achieved by thermal process and requires lower temperature than the thermal crystallisation temperature in the dark. During the photo-induced crystallisation process

the high temperature crystalline phase seems to grow much more favourable than the low temperature crystalline phase does, though the crystallisation temperature is well below thermal crystallisation temperature in the dark, as shown by Matsuda et al. [46] on the case of amorphous evaporated films. Gazs6 et al. [47] observed that the illumination of the amorphous chalcogenide by high intensity light can produce a transformation to a new amorphous state after melting and quenching of the material. Due to high light intensity involved in this process the mechanism is probably of thermal nature. The illumination causes the film to melt and the subsequent rapid quenching in air gives rise to an amorphous film of somewhat different structure. If the photo-crystallisation is the intermediary step then this effect can be considered as photo-amorphisation. The transition from the amorphous material to the ordered crystalline phase is a long time process. Much less time is necessary for the amorphisation of the crystalline phase and this feature is exploited in the system for optical recording of information. A new type of photo-induced structural change in amorphous chalcogenide materials

has been discovered in AsSe molecular films by Kolobov et al. [48]. This is the athermal photo-induced transformation in the amorphous state, i.e. a process that is not caused by local melting/quenching. The AsSe films were firstly crystallised by annealing at for time intervals up to 24 h. Thereafter, the crystallised films were exposed to broad band white light (2 W/cm2) for at room temperature and complete amorphous films were obtained. Although the mechanism of photo-induced transformation to the amorphous state is not known, two possibilities were suggested: either photon-induced intramolecular bond breaking which leads by cross linking to a continuous random network (CRN) or intermolecular bond-breaking resulting in an orientationally disordered molecular glass. It was suggested in the case of photo-induced amorphisation of crystallised glass, that the presence of the amorphous phase and strain were the driving force for the amorphisation [49]. Photo-induced amorphisation was also observed in The starting material was natural orpiment crystal. Investigation of the temperature dependence of the process has lead the authors to the conclusions that amorphisation has essentially an electronic and not thermal origin. Roy et al. [51] have discovered the laser induced suppression of photo-crystallisation in amorphous selenium films. A film of amorphous selenium exposed to simultaneous action of two different (Kr+ and Ar+) lasers whose photon energies are on different sides of the optical band gap, crystallises more slowly than the one exposed to only one of the above laser beams. A decisive role of the polarisation of the two laser beams has been demonstrated, namely, the suppression of the crystallisation rate is observed only for the polarisation of the two light sources being parallel to each other. The crystallisation suppression is due probably to the fact that, while the sub-band gap light creates nuclei with a certain optical axis, the cross-band gap light breaks them, and vice versa. The role played by the silicon substrate in the light induced vitrification of thin films was investigated by Prieto-Alcón et al. [119]. It was shown that the films

CHAPTER 3

215

crystallise in different structures, depending on the substrate they are attached to. Silicon plays an active role during illumination. The spectral composition of the radiation emitted by the light source influences the photo-amorphisation phenomenon, because the existence of a larger proportion of photons with higher energy will probably cause a larger degree of disorder in the films. c). Photo-contraction and photo-expansion. Some amorphous chalcogenide films show significant modifications of thickness (contraction) when exposed to light. The magnitude of the effect depends on the film composition. Bhanwar Singh et al. [52] have studied this effect in Ge-Se films. They found that the photo-contraction is produced only in obliquely deposited films and increases appreciably for depositions at incident angles of more than

40°. No photo-contraction was found in amorphous selenium films, nor in amorphous germanium films, nor for normal incidence in Ge-Se films. Exposure to band-gap illumination results in a maximum photo-contraction of 12 % in GeSe2 and 19 % in [19]. The analysis of the experimental data allows concluding that the photo-contraction determines the modification of the topography and the inclination of the columns of material specific to the oblique depositions. After exposure to light the inclination of the columns to substrate decreases. Therefore, the columnar structure seems to be essential in the photo-contraction process. The light determines finally the collapse of the columns. If the amorphous film is annealed before exposure to light or if the film is deposited at high temperatures, the photo-contraction phenomenon is strongly diminished and can be even inhibited. The phenomenon seems to be more complex because in the above conditions the columns do not disappear. Sing et al. [52] explained the photo-contraction by a volume change induced electronically. In this process the main role is played by the volume density of the dangling bonds, which controls the magnitude of the effect. If the reordering of the network takes place at large scale, then, even the collapse of the columns can be induced. There was demonstrated that glassy films expand by when illuminated with band gap light [53-55] that is, with radiation of where is the optical band-gap energy. The expansion can be recovered by annealing at the glass-transition temperature and the phenomena can be repeated by exposure and annealing. However, X-ray structural studies have not been able to provide reliable results accounting for the macroscopic expansion phenomena [56-57] and the mechanism of the photo-expansion is still speculative [58-59]. It is also known that stress accumulated in chalcogenide glasses can be released with illumination [60-63]. Hisakuni and Tanaka [64] have shown that when illuminated with a focused beam from He-Ne lasers for times around 10 s, the films with thickness of exhibit a thickness expansion up to which is approximately 10 times as great as that expected from the conventional photo-expansion phenomenon. The photon energy is located in the Urbach tail This effect was called giant photo-expansion. The expansion enhancement was explained by the photorelaxation of strains generated by photo-expansion. The phenomenon was found in too.

216

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

In addition, it was found that when is illuminated with focused Ar laser light of lOmW, the sample surface takes a concave shape in contrast to the expansion induced by illumination of the He-Ne laser light. The convex structure with a typical dimension of in height and in diameter can be formed with exposure to the Ne laser light. d) Photo-induced softening and hardening and photo-induced deformation. Both softening and hardening of the amorphous films in Ge-As-S system, subjected to ultraviolet irradiation have been observed [65]. Amorphous films with the thickness of were prepared by vacuum thermal evaporation. The films were irradiated by UV light and power density 0.05 W/cm2). Long time irradiation days) was used in order to ensure the saturation of the transformations induced in films. The sample heating did not exceed 40 °C during irradiation in flowing air. In the system the hardness increaseswith the germanium content and exhibits a maximum at which corresponds to the threshold of the topological transition from two-dimensional to three-dimensional amorphous network in these alloys [66]. For strong softening was obtained by UV irradiation, while for a hardening effect was revealed. Both effects tends to saturation for of irradiation. Figure 3.1 shows the results. The explanation of the hardness results is based on the sulphur loss during UV irradiation, and by specific modifications of the noncrystalline network that accompany the chemical changes. Hisakuni and Tanaka [68] demonstrated that light or electron beam exposures deform the chalcogenide glasses (e.g. As2S3). This phenomenon may be referred to as photo-induced softening (or photo-induced glass transition). The glass becomes viscous in the illuminated area. This unique phenomenon is due to the fact that the noncrystalline chalcogenide behaves as a soft semiconductor. The softness is due to the two-fold co-ordinated chalcogen atoms, which are susceptible to exhibit electro-atomic responses. This chalcogenide glass exemplifies a flexible electron-lattice coupling system.

Figure 3.1. The micohardness of amorphous virgin films o after UV irradiation reported values on bulk samples [67].

films,

CHAPTER 3

217

Photo-chemical modifications

a) Photo-decomposition and photo-amplified oxidation. These are processes characterised by the modification of the chemical composition of the material. Electromagnetic radiation approximately equal to band-gap energy has been established as responsible for

the dissociation of amorphous

and

The dissociation is accompanied by an

optical identification observable as a “photografic” effect in thin films of these materials. The densification is reversible by thermal cycling to higher temperatures. After Berkes et

al. [2] the photolysis of arsenic trisulphide follows the reaction: (3.1)

A photo-dissociation in sulphur rich regions and arsenic regions takes place. The amorphous sulphur produced during photo-decomposition transforms into rhombic sulphur. For the photolysis process the photon energy should be larger than the semiconductor band gap, and this is e.g. the case of the argon laser used for irradiation. In the next step is produced the oxidation of free arsenic in the presence of oxygen and

moisture: (3.2)

The electron microscopy studies assessed the formation of the the surface of

crystallites on

film exposed to light.

In the case of the photolysis process exhibits peculiar features [2]. Initially the decomposition follows the scheme: (3.3) where

Firstly, a non-stoichiometric composition is formed. Then the process

proceeds as in the case of arsenic trisulphide. In the case of arsenic trisulphide we are in fact dealing with a photo-catalysed oxidation of arsenic which creates a depletion in arsenic in glass and determines the release of sulphur rings that crystallise. The photo-chemical reaction produces the optical darkening of the glass, in direct connection with the presence of the arsenic phase [69]. By thermal annealing, a decrease of the precipitate arsenic results and, consequently, a partial recovery of the optical transmission coefficient is obtained. Some authors [70] consider that, in the case of compounds, as e.g. the energy received during exposure is enough for the release of the arsenic atoms, which will form finally small

crystallites. The enrichment in sulphur of the material will give rise to the bleaching due to the fact that the optical absorption edge of sulphur is situated at smaller wavelengths than the absorption edge of As-S.

Matsuda and Kikuchi [71] have analysed the photobleaching effect in the crystalline films of and concluded that this effect is conditioned by the formation of

218

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

amorphous regions with an excess of sulphur as a result of the decomposition of the crystallites: (3.4)

for

The photochemical transformations have been observed in amorphous Ge-S [72] and Ge-Se [73,74] films. The occurrence of such transformations is confirmed by the presence of the Ge-O bonds evidenced in infrared spectra. These bonds accompany the irreversible photobleaching of the material. Therefore, in ambient conditions, the photo-oxidation process is triggered during exposure to light. The composition [72] is one of the most sensible. The photo-amplified oxidation is very important in the applications of chalcogenide thin films. Exposure to UV radiation determines the effusion of chalcogen from chalcogenide films or bulk glasses. Popescu et al. [65] have shown that exposure to UV radiation of thin films in the system Ge-As-S induces the decomposition of the films with the release of sulphur. The process seems to reach saturation after days of irradiation at the power density of 0.05 W/cm 2 . Recently, the author of this book has found that bulk glass is strongly transformed when is UV irradiated at temperatures near Selenium is released and AsSe crystallites appear at the surface of the sample. The irradiation in air determines the appearance of large size (arsenolite) crystallites. For the chalcogenide films deposited on silicon wafers the photooxidation can be enhanced due to a chemical reaction of the chalcogenide material with silicon [119]. b) Photo-dissolution and photo-doping. The photo-dissolution of the metals into the chalcogenide alloys, mainly the diffusion of a metal layer deposited on the surface of a chalcogenide film has been observed as early as 1966 [75]. Later, there was discovered a great variety of thin metallic films (Ag, Cu, In, Zn) and metallic alloys that can be dissolved in the amorphous chalcogenides under the light exposure. The dissolution is followed by the diffusion of the metal ions through the chalcogenide layers along the direction of the incident light. The diffusion inside the unexposed regions is very weak. It is worthwhile to mention that the diffusion of the metals in chalcogenides occurs in the absence of the light too, but the diffusion rate is considerable diminished and no preferential direction of diffusion was observed. Among the metals with the largest diffusion rates in chalcogenide glasses is silver. The silver concentration in chalcogenide films can reach 29.1% [76]. The first explanation of the phenomenon of photo-diffusion of silver in was given by Kokado et al. [77]. The incident light excites the silver atoms: (3.5)

and the excited atoms (ions) are dissolved in the amorphous chalcogenide film thus forming an amorphous photo-doped solid solution:

CHAPTER 3

219

(3.6)

A possibility exists for a non-specific excitation of

(see [78]) (3.7)

followed by the reaction with silver of the excited chalcogenide:

(3.8) Kluge [76] considered that the photo-dissolution may be described as a solid state reaction, which requires the intercalation of the silver ions in the amorphous chalcogenide. According to this he supposed two diffusion coefficients: and for the concentration of silver and for with is a certain threshold concentration of the silver in the doped chalcogenide). Wagner and Frumar [79] interpreted the metal dissolution process as a photoenhanced diffusion in a diphasic system with immiscibility gap. This interpretation is based on the existence of two regions of glass formation in the system Ag-As-S [80] with different silver content at.% and at.%). On this basis they succeeded to explain the steepness of the diffusion edge and supported the existence of two diffusion coefficients. The lateral diffusion rate (diffusion rate in the plane of the substrate) is much higher for the chalcogenide films deposited on conducting substrates than for those deposited on insulators [81-83]. The time dependence of the position of the diffusion edge follows a parabolic law, [81-82] or varies linearly with the time, These experimental results have been interpreted in [79] by the diminishing of the spatial charge created during the diffusion of the more mobile silver ions (Ag+). J. P. de Neufville et al. [13] observed that photo-doping does not lead to the formation of crystalline as required if the light should produce a segregated sulphur rich phase. The photo-enhanced oxidation and the silver diffusion, can be described by the following reactions: (3.9) (3.10) These reactions are thermodynamically favourable (they are accelerated at high temperatures) and are strongly inhibited in the absence of light of energy near Therefore, the light can be regarded as a catalyst that diminishes the activation energy, which prevents a thermodynamically possible but kinetically inhibited chemical reaction. The is not in fact a true catalyst because it participates in both reactions. Nevertheless, its photo-enhanced reactivity might indicate enhanced catalytic activity with respect to some chemical reaction in which cannot participate.

220

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Elliott [84] described a detailed model of the photo-induced dissolution of silver, exemplified by Ge-S. The general reaction between Ag and Ch during photo-dissolution can be described as a redox reaction: (3.11)

where the metal atoms oxidise and become positive ions and the chalcogen atoms become negative ions with non-bridged configurations. For the germanium chalcogenides one can write the following reaction: (3.12) where the breaking of a Ge-Ch bond leads to a negatively charged non-bridged centre and a half-filled orbital on the germanium atom (Ge*). Although the reaction (3.11) shows that the concentration limit for the silver dissolution under light exposure is controlled by the amount of chalcogen in the glass, in fact, this conclusion can be invalidated if the steric effect is important. Thus, it would be more favourable from the energetic point of view, to transform the dangling bonds of the incompletely co-ordinated germanium atom in a new configuration, eventually with the participation of some homopolar Ge-Ge bonds: (3.13)

This reconstructive transformation will be facilitated by the decrease of the connectivity of the network that gives rise to layer flexibility due to the appearance of the non-bridged chalcogens. If the microphase separation during photo-dissolution is only bounded to the formation of homopolar Ge-Ge bonds, then the limit structure of the ternary photo-doped material saturated by silver will correspond to the case of two silver atoms for every new Ge-Ge bond. Thus, the structural mechanism of photo-doping predicts that the maximum amount of Ag dissolved in glass is equal to that of germanium and this was experimentally confirmed [85,86]. In the case of the photo-dissolution of silver in arsenic chalcogenides one supposes a mechanism similar to that in the germanium chalcogenides. With the complete segregation of chalcogen, the maximum silver amount in the glassy network corresponds to the total amount of chalcogen necessary for the formation of as observed experimentally [87]. Therefore, the maximum content of photo-dissolved silver (e.g. in corresponds to the chemical formula of the glass: The experimental values found for the limit composition are situated between [88] and [87] and they prove that only partial phase segregation occurs during photo-dissolution. The segregation of arsenic during photo-doping will lead finally to the formation of amorphous domains based on arsenic. Amorphous arsenic has been evidenced in the photo-doped materials by Raman scattering experiment [89].

CHAPTER 3

221

In conclusion, the redox mechanism described by the equation (3.11) seems to lay at the basis of all the silver photo-dissolution processes in amorphous chalcogenides. Nevertheless, the type of the reaction products and, therefore, the maximum amount of silver able to be dissolved will be dependent on the type of chalcogen. In the case of the germanium chalcogenides the tetrahedral co-ordination will reduce the flexibility of the local structure and the phase segregation will be bounded by the formation of homopolar Ge-Ge bonds. In the case of the arsenic chalcogenides the low arsenic co-ordination as compared to germanium will facilitate the mobility of the arsenic atoms and, the dominant factor for controlling the maximum amount of silver able to enter into the glassy matrix will be the chalcogen concentration. A novel photo-effect related to photo-doping has been observed in bilayers and alternately deposited films consisting of glassy chalcogenides by Tanaka et al. [90]. The phenomenon is remarkable, in films. Light exposure induces an increase in the photocurrent (photoconductivity response) in the bilayer and disordered multilayer structures. The mechanism seems to be due to the photo-induced diffusion of selenium atoms at the heterojunctions. It is useful to compare the photo-diffusion at heterojunctions with the photo-doping phenomenon observed, typically, in Ag/As-S structures. A remarkable difference is that the motion of silver is much more dramatic than that of selenium. In fact, it is not difficult to dope silver atoms into a depth of 1 µm but for selenium this depth has been estimated to be less than 100 Å. This difference seems to originate from different removal mechanisms. Although the photo-doping mechanism has not been completely elucidated, a common idea is that ionised silver atoms migrate electrically. However, since selenium has a mixture of chain-like and ring-like molecular structures, photo-induced breakage of the molecules into isolated single atoms may occur with difficulty. The ionisation of the selenium fragments may not occur. Then one speculates that the neutral Se fragments, which could be produced by illumination, thermally diffuse into the neighbouring amorphous regions, with much faster rater than in conventional thermal diffusion. The phenomenon of photo-induced surface deposition of metallic silver is a photochemical reaction in which a large number of Ag particles deposit on the surface of Ag-rich chalcogenide glasses or films following illumination with band gap light. Possible application to optical recording devices has aroused much interest, since this allows direct positive patterning to be achieved with high contrast. The mechanism, especially the force inducing the migration of silver toward the illuminated surface, has been discussed from the viewpoint of the electrical properties of the Ag-rich chalcogenide glasses being mixed ion-electron (hole) conductors. Furthermore, from a thermodynamical consideration, this phenomenon has been explained to be a photochemical reaction toward a thermodynamically stable state with segregation of excess Ag+ ions. After Kawaguchi et al. [91], the origin of the compositional dependence of the photo-induced surface deposition of metallic silver in Ag-As-S glasses can be accounted for by the combined effects of diffusivity and insolubility of Ag+ ions in the Ag-rich phase. The photo-migration of Ag+ ions in Ag-As-S glasses was observed by Tanaka and Itoh [92]. Migration over distances larger than 1 um was induced by light illumination and dramatic changes in atomic compositions were revealed.

222

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

c) Photo-polymerisation. The polymerisation of the glass during exposure to light, accompanied by a red shift of the optical absorption edge, is a typical irreversible phenomenon, which was observed both in evaporated As-Ch films [93] and simple chalcogen films [94]. In the case of the amorphous evaporated films of composition and De Neufville et al. [13] have shown that both annealing and exposure to light of photon energy corresponding to determine structural transformations, that are essentially identical, excepting the appearance of a reversible component for illumination. In both cases the films were irradiated in vacuum in order to protect them against oxidation. The film structure and its relaxation after illumination are undoubtedly related to the molecular nature of the vapour phase from which the film is born. By evaporation are formed mainly (or molecular species. These molecules knock the deposition substrate, migrate, coalesce and form a molecular glass. Other molecular species are also present in the vapour phase in significant concentrations, as e.g. and [95] or selenium homologues. Therefore, the amorphous film must consist initially of distorted packing of molecules. During relaxation of the film (by heating or by illumination) the molecules polymerise i.e. form large molecules by bond breaking and interconnection. Because the photochemical activity of the arsenic sulphide is accompanied by structural modifications, the two effects are closely related. The response of the material to the light consists on one hand in the photo-polymerisation and on the other hand in the creation of trapped non-equilibrium carriers (holes and electrons). In both cases it occurs a perturbation of the equilibrium concentration of the holes, of the trapped electrons and of the dangling bonds (in fact the free uncharged radicals). An excess of concentration of broken bonds is necessary in a given stage of the polymerisation process for the triggering of the reconstructive transformation of the network. The catalytic effects of the semiconductors are also associated with the free charge carriers on the surface, which are influenced by light and impurities [96]. The enhanced chemical reactivity of which accompanies the exposure to light of band-gap energy (e.g. the oxidation in a 10–6 Torr vacuum or the reaction with a thin silver layer) is related to the structural transformations by the intermediary of the localised electron defects that include trapped electrons and broken bonds. Onari et al. [97] have studied the photo-induced modifications in the amorphous evaporated films of by irradiation with a Hg-lamp (38 mW/cm2). The chopped infrared light with was used. The observed polymerisation process was explained by the formation of defect pairs D+ and D– during the absorption of phonons and subsequent rearrangement of the local configurations (Fig. 3.2) by switching the As-As bonds and the D+D– pairs which lead to As-S or As-Se bonds [98]. In the model of configurational co-ordinate [99] the polymerisation is explained as follows. Firstly, electron-hole pairs are created by the excitation of the non-bonding electrons of the chalcogen atoms. After the relaxation of the network by electron-lattice interactions do appear trapped pairs D+D– and the final configuration of the network is obtained by switching the interatomic bonds. During the polymerisation process the rate of variation of the optical absorption coefficient of grows with the temperature and

CHAPTER 3

223

this fact suggests an activated process. The variation of the number of polymerised molecules, m, can be expressed by the equation: (3.14)

where P is the probability of the polymerisation process that can be related to the activation energy and the temperature T by the equation: (3.15)

Figure 3.2. Schematic representation of the polymerisation processes. a. initial configuration b. structural rearrangement of the pairs D+D–. c. switching of the As-S bonds and of the D+D– pairs.

In the hypothesis of a linear dependence between m and where is the modification of the light absorption coefficient, after a very long irradiation time and is the modification of the absorption coefficient after the time t, it is possible to calculate the probability P for every temperature. From an Arrhenius plot In there was obtained for an activation energy of 0.008 eV [97]. Polymerisation causes densification and, in general, after prolonged exposure, the density and the structure of the thin film become virtually identical with those of meltquenched glasses and well-annealed films [100, 101]. Photo-polymerisation of [102] has been reported in crystals of and [71], so that this particular photo-induced effect is not unique to the amorphous state. The photo-polymerisation of the molecular fraction was shown to have the major contribution to photodarkening during the first stages of illumination of by CW and

pulsed laser radiation [103]. Under the action of light the depolymerisation can occur in certain conditions. It is suggested that by simultaneous annealing below and illumination a partial depolymerisation is possible.

224

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

3.1.2. REVERSIBLE MODIFICATIONS

The reversible photo-structural transformations, which will be discussed in this section, are modifications induced by light in thin amorphous films annealed below [13] and in bulk glass [21], which can be eliminated by an appropriate thermal treatment. The reversible modifications are independent of the sample history and, therefore, they are of interest for understanding the fundamental properties of the amorphous solids. The class of the reversible phenomena comprises the small but significant decrease of the optical gap, the photodarkening. The reversible photodarkening is accompanied by

reversible modifications of hardness, softening temperature, density, rate of dissolution in various solvents and elastic, dielectric, photoelectric and acoustic constants [4]. The reversible modification of the hardening during exposure to light has been observed in many chalcogenide compositions. The effect was called photo-hardening. The photo-hardening is typical for stoichiometric Figure 3.3 shows the change in microhardness and in optical properties during several exposures - annealing cycles applied to an film.

Figure 3.3 The dependence of microhardness in bulk (a), of optical absorption edge (b) and of the refractive index (c) of a thin film on the annealing – illumination cycles. 0 – initial state 1,3,5,7 – after annealing 2,4,6 – after illumination (after [104,105])

The light irradiation determines the decrease while the annealing determines the increase of the microhardness. The influence of the substrate seems to be very important [69]. As a function of substrate the photo-hardening effect can change its magnitude and sign. For arsenic sulphide films deposited on quartz, glass or CaF, the annealing of the sample exposed to light leads to a decrease of microhardness and the

CHAPTER 3

225

next light exposure determines a new increase of the hardness. For KCl, NaCl or even crystalline supported films, for free-standing films and for bulk glasses, the annealing after exposure leads to the increase of microhardness and the next exposure determines the decrease of the hardness. It is supposed that the effect of the substrate is related to the magnitude of the expansion coefficient of the material. For the first type of substrate materials the expansion coefficients are lower than that of the film, while for the second type the situation is reversed. After annealing near and cooling at room temperature it seems that in the chalcogenide film are developed mechanical strains and these strains are dependent on the substrate properties. As opposed to films, the light exposure of films leads to the increase of microhardness and density. The maximum effect in As-Se system is observed for the compositions with the maximum disordered structure [70]. The reversible phenomena seems to be caused by the excitation of the As-As and As-Se bonds, and their switching with the consequence of the formation of the charged defects and The optimum condition for bond switching is the presence of one As-As bond for every Se atom in the system [70]. The change of the rate of dissolution of amorphous chalcogenides (in particular in solvents under light exposure and by thermal annealing is practically the same. By light irradiation of the annealed films, i.e. in the reversible cycles, the change of the rate of dissolution in and solutions is no more than two times while in the first cycle starting from the virgin films this change is by a factor of The change of the dissolution rate after photo-structural transformations and by annealing seems to be related to the variation of surface absorption ability of the solvents. The solvent molecules are activated by fragments and by oxidation, which creates As-As bonds and finally leads to the formation of thio-arsenates [106]. There were evidenced reversible optical anisotropy effects induced by the polarised light [107] with energy identical to that used for triggering the photodarkening. In spite of the close relation between the optical edge shift and the modifications of the macroscopic properties of the chalcogens, many papers point out the fact that the mechanism of some photo-structural changes may be different from that which produces the shift of the optical absorption edge.

Photodarkening and photobleaching. a) Photodarkening. The reversible photodarkening (PD) phenomenon was reported by De Neufville et al. [13] as early as 1973. If an amorphous chalcogenide film (e.g. is irradiated by light with the photon energy near (2.4 eV) then one observes a shift of the absorption edge towards lower energies up to a saturation limit. The new state called darkened state due to the higher transparency of the film when regarded in white light can be erased by annealing the glass near the softening temperature. Because the photoconductivity spectrum shifts also towards lower energies (red shift) [108] it was

226

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

assumed that the photodarkening phenomenon is a photo-induced decrease of the optical gap which can be recovered by annealing. The recovery was called photobleaching (PB). The photobleaching is the reverse process of photodarkening. The photodarkening is a common phenomenon in many glassy materials and especially in chalcogenide ones. The chalcogens themselves show such phenomena

(amorphous selenium is a typical example) but they exhibit specific features [109,110]. The hydrogenated As-S and As-Se films show also photodarkening [111]. The amorphous tetrahedral materials, as e.g. a-Si:H do not exhibit photodarkening. As regarded the pnictide materials, the experimental results are controversial: the arsenic does not show darkening [112] but phosphorus exhibits such effect although with different characteristics when compared to the amorphous chalcogens [113].

The oxygen belongs to the same chalcogen group of the Periodic Table. The studies of and glasses have not issued in a definite answer concerning the existence of the photodarkening in these materials [114,115]. The photodarkening can be practically detected only in amorphous films thinner

than because this effect enhances the absorption coefficient for the light of wavelength corresponding to optical gap and, consequently, the effective depth of light penetration in the sample is diminished. Thus, although the intense, long-time illumination can, in principle, lead to the photodarkening of the thick samples, the limited exposure time (practically no more than one week) puts a limit to the sample thickness [116]. The most experiments have been carried out on thin films (or very thin plates) which were previously annealed below for stabilisation. It is remarkable that the bulk glasses are preferred for the investigation of the mechanism of the photo-induced phenomena because the films deposited by evaporation, sputtering, etc.. contain a high amount of defects that are not completely eliminated by annealing [117]. The structural nature of the reversible photodarkening phenomena is demonstrated by the observation that only the disordered materials show reversible photodarkening [21]

and also by the observation that the photodarkening is sensible to hydrostatic pressure [118]. This is because the photon energy necessary to induce structural changes in crystals is much higher than those needed to induce photodarkening. The non-crystalline state exists in multiple structural configurations and is characterised by local minima of the

distortion energy, not very different one from another, so that the photon energy can be enough for triggering a transition to a metastable neighbouring structural configuration.

A fundamental feature of the materials with photodarkening properties is the presence of a significant concentration of at least one element with non-tetrahedral bonds. This gives more freedom for steric arrangements as a consequence of the change of the atomic-scale interactions.

For layers deposited on silicon wafers, an interaction with the substrate was demonstrated during illumination. In this case, too, the reversible structural changes induced by light are accompanied by reversible photodarkening [119].

Finally, we must remark that, recently, Tanaka and Nakayama [120] have shown that the fundamental photoconductive edge in

glass is located at nearly the same

CHAPTER 3

227

photon energy with the absorption edge in the corresponding crystals. This correspondence implies that the mobility edges in the glass are located at the band edge positions in the crystals. Such an electronic similarity must reflect the structural similarity in amorphous and crystalline chalcogenides. Therefore, the crystalline features should be important for the explanation of the photodarkening effects. Kuzukawa et al. [121] have studied, recently, the effect of band-gap illumination and annealing below the glass transition temperature on the thickness and the optical band-gap of As-based and Ge-based obliquely deposited chalcogenide films. It was observed that in the case of arsenic-based glasses, illumination increases the thickness (expansion effect) and the band-gap decreases (darkening effect), while for germanium based glasses, both thickness and band-gap show an opposite behaviour to that of arsenic based glasses. By annealing the samples, before and/or after illumination, the trends of the changes in thickness and band-gap are reversed. These changes have been explained on the basis of ordering of the structure by annealing, and of repulsion and slip motion by illumination, the latter processing being due to the negative charging of layers by electron accumulation in conduction band tails [122]. An important observation is that the changes in thickness and gap are higher for the obliquely deposited amorphous films than for the case of normally deposited films. In the following we shall analyse the photodarkening phenomena observed in chalcogens and in the most important chalcogenide compounds and alloys. i) Chalcogens. The optical absorption edge of the amorphous selenium films deposited at 80 K is

modified by the exposure to the light of a xenon lamp and the effect exhibits saturation. As can be seen in figure 3.4 the optical absorption edge shifts from the position characteristic for as-prepared films (a) to some energies (c) and after thermal annealing at 300 K (near a partial recovery occurs (b). Between (b) and (c) the annealingillumination cycles can be repeated indefinitely because the photodarkening is fully reversible. The shift between (a) and (c) is irreversible. The selenium layers deposited onto the substrates situated at temperatures between room temperature and 320 K show the same reversible photodarkening as for the layers evaporated on substrates maintained at 80K while, the films whose substrates exceed 300 K during deposition do not exhibit irreversible photodarkening. This peculiarity is due to the stabilisation of the film structure. If the absorption edges for the states a, b, c are recorded at room temperature one gets a unique spectrum ,d (Fig. 3.4 I) [63]. Figure 3.4 II shows the optical transmission spectra of the amorphous sulphur films deposited at 80 K. The reversible and irreversible photodarkening appear in amorphous sulphur, too, and the characteristics seem to be similar to those in amorphous selenium.

228

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Figure 3.4, The optical absorption edge of selenium (I) and sulphur (II) films (measured at 80 K). a. initial state b. after annealing c. after illumination d. at room temperature

The reversible photo-induced modification in a-S has been studied by Tanaka [94] at 240 K He found that at this temperature no photo-effects are induced in evaporated films. Figure 3.5 shows the dependency of the reversible photodarkening, in a-S and a-Se upon of the incident photon energy.

Figure 3.5. The dependence of the reversible photodarkening on the photon energy (at 80 K). a. selenium b. sulphur (Ar laser irradiation, 10 mW/cm2, and He-Ne and Hg-lamp provided with interferential filters)

CHAPTER 3

229

In both cases the results are similar and the maximum photodarkening effect is induced by light of photon energy near The sub-band excitation or supra-

band excitation

induce smaller shifts of the optical absorption edges.

ii) Arsenic chalcogenides. The system As-Ch is the paradigm of the chalcogenide glasses and comprises the most important materials with photodarkening effects. The systematic research on the stoichiometric compositions and demonstrated the importance of the As-Ch bond in the control of the photodarkening. When one goes in the series from to then one passes from a disordered covalent network to a network with significant metallic character. The effect of the substitution of chalcogen in can be followed in figure 3.6 [123]. The figure shows the magnitude of the maximum reversible shift of the optical absorption edge, measured at different temperatures, as a function of for various chalcogenides. All the compounds with the same chalcogen are situated on the same curve. is a function of the type of chalcogen and decreases in the series Te with the optical gap Tanaka [123] concluded that the photodarkening is determined by the type of the chalcogen and is independent of the particular structure of the glass. The photodarkening is not independent of structure because for a given compound is closely related to the structure [124]. From Fig. 4.5 it results a strong dependency of the photodarkening upon temperature. The plot shows that by moving

Figure 3.6. The photodarkening in various chalcogenide glasses.

from the room temperature to high temperature, decreases and completely disappears at This is because the thermal annealing rate is at least equal or even larger than the rate of induction of photodarkening processes. At very low temperatures increases abruptly down to 80 K. [125]. Illumination at still lower temperatures does not induce a significant increase of photodarkening. It is remarkable that the photodarkening effect in

volume by

is accompanied by an increase of the sample while in the case of films of composition Ge was observed a

230

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

decrease of the sample volume by 0.2 % [126]. The shift of the absorption edge in (at T = 14 K.) reaches the value -0.2 eV and is accompanied by a change in the refractive index with [13]. Although the general photodarkening properties of the stoichiometric arsenic chalcogenides are similar, the amorphous alloys and show important differences. The structural differences are, without doubt, the key for understanding the particularities of the photoinduced effects. For As-rich compositions increases with x both for sulphur glass and for selenium glass [127,128]. The sulphur-based system exhibits an important feature. In the composition range no homogeneous phase exists due to the easiness of formation of the molecules that leads to a compositional splitting. For the arsenic-rich compositions the amorphous evaporated films are unstable above room temperature and probably a phase separation occurs during annealing. In the literature there was evidenced photodarkening in the compositions with [94] due, probably, to a strong enhancement of the disorder in the As-As bond system around x = 0.45. Because does not exhibit photodarkening [129] it is expected a gradually decrease of when the composition approaches x = 0.50. In the selenium-based alloys the photodarkening effect is encountered in a larger range of compositions, because, as opposite to the sulphur-based system, where the glassforming region extends only up to 43 % As, in the selenium-based system is possible to get glasses with up to As. The plot of the optical gap versus As concentration (figure 3.7) shows a monotonous increase of up to a maximum at x = 0.6; then a rapid decrease follows for higher arsenic concentrations [127,130].

Figure 3.7. The photo-induced reversible shift of the optical absorption edge in the composition x.

as a function of

The experimental data in the system [131] reveal a decrease of sulphur rich compositions. The experimental points in the plot of versus the curve of the sulphur compounds, evidenced by Tanaka [123].

in the follow

ii) Germanium chalcogenides. In the Ge-Ch systems the photodarkening phenomena are weaker than in As-Ch systems. The data on Ge-Se films are somewhat contradictory. Some authors observed a photodarkening effect. Other authors reported a photobleaching effect. A comparative

CHAPTER 3

231

study of the photo-induced effects in and films [132] has shown that the photo-induced processes are specific. If the films are annealed, they show photodarkening irrespective of composition and, after annealing, the initial state is recovered. As opposite to films, in the freshly prepared films were revealed both a photodarkening effect and photobleaching effect The type of effect that is observed depends on the temperature where the light exposure was performed [133] (see figure 3.8). The different behaviour of the two types of films was explained as follows: In the virgin films there is a high number of wrong bonds, Ge-Ge, As-As and Se-Se which are transformed by illumination and/or annealing in Ge-Se and As-Se heterobonds. In the Gebased films this process leads to an increase of the optical gap, i.e. to photobleaching, while in the As-based films, due to heterobond specificity, the photodarkening is produced. Such irreversible processes are, of course, produced only in fresh, nonequilibrated films. The photodarkening effect, in the evaporated films of and is 0.04 eV and respectively 0.05 eV [134] (for long-time illumination by weak green light /5145 Å/ emitted by an Argon laser). No significant effect light was observed in the films of composition and The photodarkening effects in the Ge-Se films are weaker than in the ternary system As-Se-Ge [134]. In is 0.12 eV. In the compositions where sulphur is added is larger. For example, in the alloy is 0.18 eV.

Figure 3.8. The optical transmission in amorphous

1 - fresh samples

and

2 - samples annealed at 300 K

3 - samples illuminated at 77 K 4 - annealed samples

Takahashi and Harada [135] reported an irreversible variation of the optical gap at room temperature of for the germanium dichalcogenides and when the films are illuminated by a Hg-lamp of P = 100 W for 100 min. They have not observed any photo-induced effect in the germanium monochalcogenides GeS and GeSe.

232

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Tanaka et al. [136] and Tichý et al. [137] observed irreversible photo-induced changes (photobleaching) in amorphous GeS films. Since the magnitude of photobleaching depends on the ambient air pressure during the illumination, it was suggested that the photobleaching be due to the photo-oxidation of Ge atoms. Kawaguchi et al. [138] support this explanation of the effect with additional studies on films.

iv) Phosphorus chalcogenides. The system shows a maximum photodarkening effect at the composition corresponding to [139]. The reason behind this effect seems to be breaking of P=Se double bonds due to light irradiation. In fact, a double bond is likely to be broken first and easily. It appears that units break up the tetrahedrally co-ordinated units into threefold co-ordinated units and twofold -Se-Se- units. The well-annealed films do not show any tetrahedrally co-ordinated units in the IR spectra. It appears that the co-ordination number of phosphorus decreases on exposure and that of selenium increases.

Obliquely deposited films in all cases show more pronounced photoeffects and have been interpreted in terms of the difference in the structure of 80° films compared to 0° films (normal incidence) films. The maximum photoeffect in is due to the maximum number of four-fold coordinated units at this particular composition. For other compositions there occurs breaking into three-fold and two-fold co-ordinated units on illumination. A reversible photodarkening effect was observed in the system Ge-Se-P by exposure to light of wavelength 5115 Å from an Argon laser [99]. For the compositions and the change of the optical absorption edge induced by light is and 0.076 eV respectively. These values are larger than those measured on the germanium chalcogenide glasses but lower than those measured in arsenic glasses.

b) Photobleaching. In most chalcogenide films was observed a bleaching phenomenon under the influence of light: photobleaching. The photobleaching is the reverse effect of photodarkening. In order to get photobleaching it is important to illuminate the sample at temperatures a little but larger than those, which favour the photodarkening [140]. The effect is illustrated in figure 3.9 [133]. The line AB corresponds to the photodarkening at room temperature. The lines AC and BD describe the modifications in the transmission of the not illuminated and photodarkened films with temperature, respectively. The light irradiation of a photodarkened film at the temperatures T’ and T” produces an increase of the transmission (lines G’T’ and G”F”). It is remarkable that the illumination of the virgin films at the same temperatures T’ and T” gives rise to a photodarkening effect (lines H’F’ and H”F”) with the same resultant transmission. If a film photodarkened at room temperature is heated under the light irradiation its transmission coefficient changes following the line BEF’F”C. Thus,

CHAPTER 3

233

the final value of the transmission is determined only by the temperature corresponding to the last irradiation but not on the irradiation sequence or by the heating processes.

Figure 3.9. The modification of the transmission in an film during irradiation by a He-Ne laser light at various temperatures.

The above-described phenomena are typical for a great number of glasses in the systems As-S and As-Se. In some cases, as e.g. in the films, the photobleaching is characterised by a thermal threshold and this means that the bleaching can be observed only above a given temperature called the optical bleaching threshold, that is lower than the thermal bleaching temperature, (Fig. 3.9). It is worthwhile to mention a higher sensibility by a factor of of the photobleaching process compared to the photodarkening one. Many experiments [133] have shown that the value of the final transmission depends on the intensity of the incident light (Fig. 3.10). The curves 1 and 2 in the figure correspond to the photodarkening of an film during light irradiation at 75 °C, using a He-Ne laser for the intensities and If a sample, previously illuminated by is irradiated with then the photodarkening is produced (curve 3) up to of the value of the transmission. In the case of films the illumination at room temperature by red light (bandband energy) can bleach the material if a previous illumination by blue light (in-band energy) was performed. As a conclusion we can affirm that after successive photodarkening and photobleaching of an amorphous chalcogenide film the value of the transmission and also of the refractive index, are defined by the temperature at which is carried out the last irradiation, by the wavelength used and by the intensity of the light.

234

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Figure 3.10. The modification of the optical transmission in a film of intensities.

for various illumination

Photo-induced anisotropy. The amorphous materials are essentially isotropic undoubtedly while the crystals show anisotropy i.e. most of the properties depend on the crystal orientation. The photoinduced anisotropy was originally discovered by Weigert [141] in 1920. Then, the phenomenon was observed in organic polymers including liquid crystals, in oxides and in phase separated systems. Besides the temporary anisotropy induced by electrical and magnetic fields or by mechanical forces which disappears when the inducing factor is switched off, the chalcogenide amorphous films [142-144] and bulk glasses [145] show quasi-stable optical and electronic anisotropy when illuminated by polarised light. The anisotropy induced by the electric field of the light wave is maintained after switching off the illumination. The main photoinduced anisotropy phenomena, called also vectorial phenomena are dichroism and birefringence. The other are: difference in the intensity of the photoluminescence and difference in the fine structure of the X ray absorption edge for polarisations of the control light beam parallel and perpendicular to the direction defined by the polarisation of the beam used in photodarkening. The optical anisotropy induced by exposure to polarised light are observed not only in annealed glasses but also, and essentially undiminished, in glasses photodarkened by light exposure. The photo-induced anisotropies can be reversibly reproduced after annealing, and the axes defining the optical anisotropy can be rotated by turning the polarisation direction of the exposing light. These anisotropic phenomena differ from the isotropic photostructural changes in their induction and annealing kinetics as well as in their dependence on temperature and on photon energy.

CHAPTER 3

235

The dichroism induced by light is experimentally easily accessible and is more directly related to the scalar photodarkening (i.e. isotropic photodarkening) than

birefringence. The plot of the variation of the absorption coefficient in the Urbach tail as a function of the photon energy is given in figure 3.11 for the case of films. The absorption

Figure 3.11. The dichroism. The variation with the incident photon energy, of the optical absorption coefficient of an film photodarkened in polarised light.

coefficient in the direction parallel

and perpendicular

to the polarisation plane

defined by the polarised beam which gives rise to photodarkening, are different. The absorption along the perpendicular direction is higher than along the parallel direction while both are lower than in the case of irradiation by unpolarised light [146].

By studying the difference between and during the photodarkening under polarised light, there was observed an oscillation dependent on the light intensity and on temperature, between the states with positive dichroism and those with negative dichroism [147-149]. The most interesting feature is the cycle negative-positivenegative dichroism produced when the annealed sample is photodarkened sequentially to 80 K and to 300 K [149]. The dichroism phenomenon was also observed in the chalcogenide films that do not photodarken at room temperature [147]. The photo-induced dichroism can be re-oriented by changing the polarisation of the light. The existing dichroism is destroyed within a time span much shorter than that

236

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

necessary for its creation. A similar dichroism is created afterwards in an orthogonal direction, i.e. reversal of dichroism is possible. Such re-orientation can be performed several hundred times without any sign of a decrease in the effect. This light induced reorientation of erasure is not associated with any change in the scalar photodarkening. The largest value of the photo-induced dichroism was observed in and [150], i.e. in materials, which do not exhibit reversible photodarkening. A further proof that photo darkening and photo-induced anisotropy are different phenomena is the fact that thermal destruction of the photodarkening and photo-dichroism are also characterised by different behaviour [150]. The photo-induced anisotropy can be not only reoriented by switching the polarisation of the beam which produces the photodarkening but can be eliminated by irradiation with unpolarised light too, without influencing the photodarkening [151]. The comparison between the magnitude of the effect as a function of the excitation energy of the two effects (dichroism and photodarkening) shows that the dichroic effect reaches a maximum value in the region of the Urbach tail where the threshold of the efficiency of the photodarkening is reached. Thus, the dichroism will be largest when the electrons are excited with highest probability from the states situated at the top of the valence band originated from the orbitals of the lone pair. The photons of higher energy excite the electrons from the lower states and give rise to PD without anisotropic characteristics, associated to the excitation of the lone pair orbitals. Based on this argument Lee [149] suggested that PD is a more general effect which implies global changes in the atomic network at the level of the medium range order while the anisotropy is induced as a special case where the polarised light of appropriate energy excites the lone pair electrons leading to the reorientation of the local anisotropic structures. Thus, the two effects take place simultaneously in certain conditions but some of their properties are different (Figure 3.12).

Figure 3.12. The photo-induced darkening and anisotropy in as a function of the energy of the photon which induces the effects. Above the photo darkening is nearly independent of energy while the photo-anisotropy decreases with the increase of the energy.

CHAPTER 3

237

The polarisation image can exhibit a very high contrast, which is limited only by the homogeneity of the polarizer elements [114]. The comparison between polarisers must be done in the transparency range where dichroism is absent and the photo-induced birefringence is high. In figure 3.13 are shown the photo-induced birefringence spectrum in AsSe and films, where is the refractive index for polarised light along

Figure 3.13 The spectra of the photo-induced birefringence in amorphous chalcogenide films.

along the optical photoinduced axis and is the refractive index for the case of the polarisation perpendicular to this axis [143]. The photo-induced anisotropy is not a property specific to the amorphous chalcogenide films. Similar phenomena have been observed in bulk glasses too [153] and in photosensitive emulsions dispersed in gelatine [154]. The anisotropy has been revealed also in As-(S,Se), in [144] and in other amorphous materials. The maximum value of the anisotropy can reach ~1/10 of the intensity of the PD effect and of the modification of the refractive index related to this effect [155,156]. Lyubin et al. [157] have revealed the possibility of re-orientation with the linearly polarised light not only of defects and scattering centres but also of interatomic covalent bonds in various chalcogenide glasses. Recently Tikhomirov and Elliott [158] remarked that ordered chirality is a property of crystalline analogues of chalcogenide and oxide glasses (e.g. spirals in and or right and left-hand modifications in in contrast to the disordered chirality in glasses. The glasses can be nevertheless ordered by irradiation with polarised light and this explains the photoinduced anisotropy.

238

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Metastable photoinduced anisotropy generated by linearly polarised sub-gap light was shown to appear also in the undoped and Pr doped Ge-S-I, Ge-Ga-S and As-Ga-S glasses with varying composition and Pr content [159]. Compositional trends of the photoinduced anisotropy are related to the light-stimulated reorientation of intrinsic anisotropic centres and their environments, which vary with composition. These centres can be modelled with pairs of over- and under-co-ordinated sulphur atoms: the valence alternation pairs (VAPs). Kolobov et al. [160] have investigated the photoinduced anisotropy in a model of chalcogenide glass using reflectance difference spectroscopy. They found that the anisotropy can be induced in the energy range much exceeding the energy of the photons of the exciting light and that not only defects but also main covalent bonds of the glass are reoriented by linearly polarised light. The sign of the photoinduced anisotropy, especially at higher energies, strongly depends on the photon energy of the exciting light. This feature was explained by the photoinduced change in the bond topology involving a conversion between bonding and nonbonding electrons. Preirradiation of the glass by unpolarised light increases substantially the magnitude and creation rate of the photoinduced anisotropy indicating that both native and photoinduced defects play a role.

The processes occurring in bulk glasses and thin films are essentially identical and the observed difference in reflectance difference spectroscopy is caused only by an interference phenomenon. Tanaka et al. [161] have recently discovered that linearly polarised light can produce an anisotropic surface corrugation in amorphous chalcogenide films of Ag-As-S. The corrugation resembles a mouth whisker consisting of narrow fringes, which are parallel to the electric field of light, and streaks, which radiate from the illuminated spot to directions nearly perpendicular to the electric field. Optical birefringence of about 0.01 appears with this pattern. Fritzsche [162] has shown that optical isotropic materials such as chalcogenide glasses can become optically anisotropic because they consist and contain entities, which are anisotropic. The original macroscopic anisotropy originates from the random orientations of the microscopic anisotropic entities. A recombination event, which leads to a structural change of a microscopic anisotropic entity, will change the orientation or nature of this anisotropy. This constantly happens everywhere in the material during illumination without, however, necessarily producing a macroscopic anisotropy. For this to happen it is necessary that the recombining electron-hole pair be excited in the same microscopic anisotropic entity, which undergoes the structural change. This means essentially that macroscopic anisotropies result from geminate recombination of electronhole pairs, which do not diffuse out of the microscopic entity in which they were created by absorbed photons. The lack of electron-hole pair diffusion and the geminate nature distinguish the recombination events leading to anisotropies from all the other events, which yield isotropic (or scalar) photo-induced changes. In order to explain the optical induced anisotropy in bulk glasses it was supposed that the microscopic mechanism comprises two parts: the optical irreversible scalar component due to creation of randomly formed dipole moments, and the reversible vectorial component caused by the re-orientation of intrinsic dipole moments (structural units) according to the electrical vector of the inducing light [163].

CHAPTER 3

239

Emelianova et al. [164] have developed a model for explaining some basic characteristic features of photo-induced anisotropy in glassy semiconductors. The model assumes the occurrence of correlated pairs of localised states for electrons and holes and relates photo-induced anisotropy to generation of geminate electron-hole pairs trapped by these localised states. Other reversible photoinduced effects. Recently, Lyubin and Tikhomirov [165,166] have shown that the photo-anisotropy in chalcogenide glasses can be induced not only by light of energy above the optical gap but also by light of energy below the gap. There were discovered new phenomena: the anisotropy of the transmittance without PD, the photo-induced girotropy, the photoinduced scattering of light accompanied by depolarisation. These phenomena were revealed in glasses by irradiation with polarised light of sub-gap energy. a) The anisotropy of the transmittance without PD. If one measures the variation of the relative transmittance of the linearly polarised light during irradiation one observes that the transmittance decreases with more than one order of magnitude in the case of a prolonged irradiation (Fig. 3.14). On the other hand, if one follows the time evolution of the transmittance of the anisotropy defined by then the kinetics of the process looks differently.

Figure 3. 14. The kinetics of the evolution of the optical transmittance (1) and of the

transmittance anisotropy (2) for a control beam, induced in a sample of of 2.5 cm thickness by a control light beam of power density 5

240

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

The curve of evolution is not monotonous (Fig. 3.14, curve 2). Firstly the anisotropy of the transmittance increases then maximum follows and finally decreases and even changes the sign. The maximum value of this anisotropy is 0.4 and this exceeds the value observed in non-crystalline chalcogenide films [147]. By heating the sample, the transmission anisotropy disappears around 120 °C, therefore at a lower temperature than the of the (185 °C). The disappearance of the anisotropy by heating at temperatures substantially lower than has been observed in amorphous films too [147]. The oscillatory character of the kinetical curves of the transmittance anisotropy with the alternative change of sign is typical and, because there is no reason to consider apriori that the photoinduced dichroism can have alternated signs, then, in order to explain the non-monotonous character of the variation of the transmission anisotropy, Lyubin and Tikhomirov [166] suggested that in fact we are dealing with the rotation of the polarisation plane of the polarised light during sample penetration. Hájto and Ewen [167] have shown that the glass rotates the polarisation plane of a polarised ray before light irradiation (therefore exhibits natural optical activity) and, after irradiation, the rotation angle changes (therefore exists photo-induced optical activity). Lyubin and Tikhomirov [166] have shown that simultaneously with the rotation of the polarisation plane, an ellipticity of the transmitted light does appear. The ellipticity is enough low before the sample irradiation but during irradiation changes the sign and increases up to a saturation value situated well above the initial values. Moreover, there was shown that the transmitted light is completely polarised before irradiation and the depolarised component appears during irradiation and its weight increases up to a saturation value. A significant variation of the optical activity, of the ellipticity and of the degree of the depolarisation of the transmitted light, as a function of the time of action of the incident beam, were observed when the incidence position of the laser beam on the sample was changed.

b. Photoinduced girotropy. Starting from the results obtained from the investigation of the optical activity and of the photoinduced ellipticity, Lyubin and Tikhomirov [166] have

suggested that the appearance of the photo-induced girotropy in the chalcogenide glasses, i.e. the photoinduced circular birefringence, which leads to optical activity and photoinduced circular dichroism that leads to ellipticity, therefore the optical properties of the investigated chalcogenide glasses are essentially determined by the spatial dispersion. There was observed and studied the photoinduced circular dichroism [166]. Figure 3.15 (curve 1) shows the kinetics of the transmittance girotropy induced by the linearly polarised light. The sample exhibits initially a very small girotropy of the transmittance that differs in magnitude and sign for various regions of the sample. The transmittance girotropy changes its sign and increases up to large values. Figure 3.15 (curve 2) shows the kinetics of the transmittance girotropy induced by right-hand

CHAPTER 3

241

circularly polarised light. The circular dichroism reaches in this case values three times greater than in the previous case.

Figure 3.15. The kinetics of the transmittance girotropy induced in an

sample of

thickness 2.5 µm by the linearly polarised light (1) and circular polarised light (2).

c) The photoinduced scattering of the light. There was observed [166] that the light of energy below the gap exhibits a strong photoinduced scattering. This effect is revealed by the change of the shape of the cross-section of the transmitted laser beam. Before the irradiation the image of the transmitted laser beam on a screen is circular, as for the initial beam. During irradiation a diffuse halo is formed around the original spot. The image is gradually eroded and finally is stabilised as a nebula covered by spots (speculae).

The above-described photo-induced effects, evidenced in the glass, have been revealed also in other chalcogenide compositions. There was found that for the values of the transmittance anisotropy and girotropy exceed those found in and are reached in a shorter time. It is remarkable that the sign of these effects is negative and their kinetics has a monotonous character. The amplification of these effects in the iodine glass can be ascribed to the following causes: i) The introduction of the single-valent iodine atoms leads to the formation of polymeric chain structures with enhanced micro-anisotropy as compared to the stoichiometric This feature was stated as early as 1973 on the basis of the conductivity anisotropy in As-Se-I glasses [168]. ii). The addition of iodine atoms determines the diminishing of the In this case the radiation used in experiments will be situated in the energy range of the Urbach tail of the absorption spectrum where the effects can be very different. The photo-induced scattering of the light is only partially thermo-reversible, as opposite to other photo-induced effects, which are completely reversible.

242

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

3.1.3. MODELS FOR THE PHOTO-INDUCED PROCESSES The PD and PB processes are essentially photostructural processes. They are based on the modifications of the spatial co-ordinates of the atoms in the amorphous network. Many models, more or less elaborated, have been proposed up to day with the purpose to

explain as exact as possible the huge amount of experimental data on chalcogenide glasses. The models may be classified in two main categories: a. models based on the hypothesis of breaking and reforming of bonds between atoms. b. models that suppose modifications of the atomic co-ordination spheres with the exception of the first co-ordination sphere while maintaining unchanged all the bonds between the atoms. In the following section we shall discuss shortly the main models advanced for the explanation of the PD mechanism. Street’s model (1977) Street [98] explained the PD in elemental chalcogenide materials by the electrical charged defects generated by breaking and reforming the covalent bonds. The Street’s model consists in three different configurations (Fig. 3.16). Starting from the fundamental state, an electron-hole pair is created via photon absorption and, consequently, a bonding state

Figure 3.16. The diagram of the configurational co-ordinate for the recombination processes in

the Street’s model [98]. It exists two ways of recombination for the excitons generated by optical excitation of a electron-hole pair: i) the direct recombination to the fundamental state (way I) and ii) the creation of a metastable pair as a self-trapped exciton (way II) which necessitates the thermal excitation for the returning to the fundamental state.

CHAPTER 3

243

or an exciton does appear. The energy of the free exciton is decreased by self-trapping and this leads to a metastable state which represents exactly the photodarkened state. The exciton anisotropy is produced due to the electron-photon coupling which dissipates the energy of the exciton before the radiative recombination did. There was suggested that some self-trapped excitons should be in a metastable state that would define the PD state. The thermal excitation would be necessary for returning to the fundamental state because the radiative recombination is not possible. In the opinion of Street, the shift of the optical absorption edge is determined by the enhanced absorption due to the excitation of the self-trapped excitons. In order to relate the excitonic model to the structural features of the PD it is important to note that the atomic configuration of a self-trapped exciton is exactly the same as that of a defect pair and , If the concept of pair is extended to the chalcogenide alloys (e.g. the model will predict the formation of several configurations of defects of the type “ valence alternation pairs” (VAP) because the photoinduced breaking of the As-Ch bonds can lead to the formation of homopolar bonds. Elliott [169] took into consideration four of the most probable configurations and concluded that only one of them, consisting from a tetra-coordinated arsenic atom and a single-coordinated chalcogen atom can explain the observed volume expansion. The reconfiguration of the chemical bond for the purpose to create an pair leads to the formation of an As-As bond per every pair. If the formation of the pairs would be the only mechanism responsible for PD, it would be expected to occur a

strong increase of the concentration of these bonds. The true increase is in fact smaller and this leads to the conclusion that the excitonic process can be a factor, which contributes to PD but cannot be the determining factor. Therefore, the Street’s model explains the PD by the transition from a fundamental state to an excited state, without any description of the local structure of these states. Tanaka’s model (1980) The model was initially proposed for the explanation of the photoinduced phenomena in non-crystalline arsenic sulphide [107]. Tanaka supposed that the PD effect is due to some special positions or localised defects whose density was estimated on the basis of optical and structural studies as at.% from the chalcogen atoms in the material [94,170]. In every local configuration the illumination induces transformations from a stable configuration to a quasi-stable one by the intermediary of an electronic excited state. The annealing induces a structural relaxation directly from the metastable to the stable state. Thus, one can build an energetic diagram (Fig. 3.17) where X, Y and Z is the stable state, the metastable state and the excited state, respectively. The fundamental state must be an asymmetric double potential well in order to explain the thermal bleaching. As concerned the excited state, the simplest hypothesis is a unique potential well. Tanaka assessed the minimum of the state Z exactly above the metastable state Y because only in this way it is possible to explain the dependency of on the temperature at which is induced the phenomena and the dependency on the energy of the quantum light. The barrier height and the asymmetry energy for the adiabatic

244

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

potential can be estimated quantitatively from the thermal data in the hypothesis where is the density of the positions in Y configuration. Tanaka found for

Figure 3.17. The configurational diagram for the local centre, which contributes to PD. There are shown the dominant photo-induced processes (X - Z’) and the thermal processes (Y - X).

the following values: and for selenium The variation of the energies reflects the distribution of the heights of the energy barriers, which varies from site to site due to the atomic disorder.

In order to justify his energetic diagram Tanaka [94,123] developed a mechanism of bond twisting and studied the structural modifications that satisfy the requirements of the model with simple and double potential wells (Fig. 3.18). The annealed state corresponds to the state X of the configurational model. The atoms have double coordination and form molecular chains or rings. The intramolecular bond is covalent and the intermolecular interaction is Van der Waals. The state Z’ is the structure immediately after the excitation of the electron on the atom A. The interaction between the atom A and B is dramatically changed. In X this interaction is of Van der Waals type and in Z’ becomes ionic and large attraction forces appear. Then, the atom A has the possibility to twist in the conjugated position A’ and this corresponds to the configuration change from Z’ to Z. The movement is realised in s. After s the electron recombines and the structure is frozen in the configuration Y. This twisting movement is equivalent to a distortion of the network with the dihedral angle specific to the helicoidal chain. By processes, which are similar to that described above, the

CHAPTER 3

245

Figure 3.18. The mechanism of bond twisting in the Tanaka’s PD model [94].

amorphous network becomes more disordered. A’ goes nearer to B, leaving the old position. The structure Y is metastable because the structural relaxation from the state A’ to the fundamental state A requires the distortion of the intramolecular chemical bonds, especially the angle between the covalent bonds which costs much energy. Nevertheless this distortion becomes possible when the sample is heated at In other words, the molecular “flow” induced at relaxes the structural distortion. As a consequence, the initial structure X is rebuilt. In conclusion the model of molecular twist postulates that the inter- and intra- molecular distortions are those which produce the asymmetry energy and the height barrier The configurational model of Tanaka cannot be applied to the stoichiometric germanium chalcogenides because Ge is tetra-coordinated and the bond twisting is not possible due to the absence of the conjugated positions. Tanaka tried to save his model by introducing the hypothesis of the existence of the chalcogen trimers. In this case the trimer chalcogen atom configuration can be twisted and, therefore, some conjugated positions appear. Thus, the photodarkening can be triggered.

246

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Grigorovici’s model (1981)

This model explains the reversible PD on the basis of a photo-polymerisation process [170]. The model was proposed for the explanation of PD in thin films of and is based on the idea of phonon and photon assisted transitions between a fundamental molecular state and an excited metastable state, by means of an intermediary excited state. The energy barrier which separate the fundamental state and the excited state is more higher when the polymeric structure is more rigid (in the case of alloy this structure becomes more rigid when more As is added). The process at the atomic scale is schematically presented in figure 3.19. Two neighbouring molecules use the energy of two absorbed photons for breaking the intramolecular bond in every molecule. The four atoms will reorient and a dimer will be formed. The first absorbed photon breaks an intramolecular bond thus forming an intermediary state. In order to produce the dimerization it is necessary a secondary photon to act a time interval shorter than the lifetime (temperature dependent) of the intermediary state. At room temperature the transition from the intermediary state to the fundamental state has negligible probability. The bi-photon dimerization process (elementary PD) does not need important changes of the atomic positions. This is demonstrated by

Figure 3.19. The elementary mechanism of the photo-dimerization in the PD model developed by Grigorovici [171].

the very faint modifications of the diffraction patterns during photodarkening. The biphotonic processes do not lead to major modifications of the MRO because lower ones and vice-versa substitute the higher interatomic distances. The polymer can be dissociated by light. The breaking of one of the newly created bonds will release enough energy to be used for the breaking of the second bond, thus rebuilding the initial molecular configuration. The depolarisation by light, which signifies the transition from the excited state to the fundamental state assisted by photons, and which has as effect the PB, is therefore a single-photon process. The Grigorovici’s model explains satisfactorily the PD, PB and the thermal bleaching. The model succeeds to explain qualitatively the microhardness and solubility changes that accompany the PD and the photoinduced anisotropy phenomena. Although the model offers a mechanism for PD it does not explain in details the structural modifications which go beyond the simple increase of the atomic disorder, at the level of the change in MRO as resulted for example from the shift of the FSDP in the X-ray diffraction patterns.

CHAPTER 3

247

Kolobov’s model (1981) Kolobov et al. [172] developed a model of configurational diagram somewhat different from that of Tanaka. In this model the configurational states implied in the PD process are represented by specific isomers. Thus, in vitreous selenium, which is built from a mixture of rings and chains, the interatomic distance in a ring or in a chain is much shorter than that between these molecular entities. Under the influence of light the bonds within the chains can be broken and bonds between chains can occur (Fig. 3.20). In such cases it is possible to build a configurational diagram where the electronic energies of the stereo-

Figure 3.20. I - Isomer structure in the amorphous selenium, possibly implied in the

PD process. II - The scheme of the structural modifications in amorphous selenium. (within the circle: a. fundamental state

b. metastable state)

isomer stable in the fundamental state, in the excited state and those of the metastable isomer in the same states are concurrent. The photostructural rearrangements both in elements and in the compounds do not necessarily imply changes in the character of the chemical bonds. A typical example is [173]. The structure of the selenium arsenide is probable a packing of disordered layers (after Kolobov). When an excess of arsenic is added, the As atom is inserted in the network so that the third As bond is free and can be used for bonds between adjacent layers, which gives rise to a distortion accompanied by the decrease of the interlayer distance. During the light absorption the breaking of the bonds within the layers and the formation of bonds between adjacent layers, which determine a distortion is accompanied by the decrease of the interlayer distance. During the light absorption the breaking of bonds within the layers and the formation of bonds between the layers is possible because, by the intervention of the AsAs bonds that link two layers, these layers approach on to another. Thus, the light can form additional bonds between layers. The induced bonds are stretched and, therefore,

248

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

less energy is necessary to ionise an atom implied in such type of bonds. For this reason, the observed optical gap is shifted towards lower energies and the PD effect appears. The model convincingly explains the increase of the microhardness of the samples during illumination [5]: the supplementary bonds between the layers give rise to a more rigid network. One of the shortcomings of the model is the impossibility to explain the volume expansion during PD. The photobleaching is explained like in the Tanaka’s model. Later, Kolobov [174] considerably improved his model and gave a detailed explanation of the PD phenomena. The annealed film, as a consequence of the preparation procedure is characterised by the accumulation of strains. If the annealed film will be illuminated, the network will be weakened and the structure will be able to relax. Since the strains originates from the bond conversion As-As to As-S, it is likely that As-As bonds will be formed again, the probability being higher in the sites where As-As bonds were initially present. Kolobov believes that exactly the changes in such strained regions are responsible for the photostructural changes since it seems rather unlikely that the matrix of an “ideal” glass will undergo structural changes.

It should be stressed that the formation of As-As bond, although inherent to the photo-structural changes, cannot explain the darkening since any deviation from the stoichiometry in the As-S(Se) systems leads to the increase in the optical gap. The authors believe that As-As bonds act as nuclei for a co-operative process involving chalcogen atoms. Let us imagine that, on annealing, an above-average local density occurs in one of the layers. In a relaxed structure this layer is curved in order to accommodate the extra density. On annealing the overall structure becomes more ordered and this layer is forced by its surrounding to fit into a smaller space. On illumination, the bond weakening allows the layer to relax back to a more curved shape approaching one of the neighbouring layers and thus facilitating the inter-layer As-As bond formation. In a relaxed structure the lone-pair orbitals take the orientations that allow improving the contribution of p-bonding, thus lowering the network free energy. The increased interaction between the relaxed lone-pair electrons gives rise to a broader valence band. The results of the XPS study of the photo-structural changes carried out by Kolobov et al. [175] have shown that the PD is due to the shift of the valence band edge [175] and this fact supports his model. In As-rich films of As-Se compositions the photostructural changes are believed to be caused by mutual reorientation of molecules in the glass, causing a transition between structures analogous to and AsSe crystals having identical unit cells but different d-spacing. A phase transition within the amorphous state observed by DSC (differential scanning calorimetry) [176] seems to support the above idea. Recently, a nanometer scale mechanism for the reversible photostructural change in amorphous chalcogenides has been developed by Kolobov et al. [177]. The process in amorphous selenium is assumed to proceed as follows: In an annealed sample, each selenium atom in a chain (Fig. 3, left) possesses two electrons in the bonding state and two lone-pair (LP) electrons that form the top of the valence band. The electronic configuration is shown below the schematic representation

CHAPTER 3

249

Figure 3.21. Creation of threefold co-ordinated pairs cross-linking selenium chains in the photoexcited state. Electronic charges corresponding to the structures shown are given at the bottom. For the case of a similar picture represents neighbouring layers with As species shown as shaded circles.

in figure 3.21. Following the notations of Kastner et al. [178], such atoms are referred to as where C stands for chalcogen, the subscript indicates the coordination number and the superscript denotes the charge state. Upon photoexcitation, one of the LP electrons is excited into the antibonding state, while the other one is left unpaired in the former LPorbital (Fig. 3.21, middle). Provided the lifetime of the photoexcited photoelectron is long enough, additional bonding may take place between the two neighbouring chains (Fig.3.21, right) through one of the two possibilities. First, a second electron (on a neighbouring chain) may be photoexcited within the lifetime and the structure can be stabilised by forming an interchain bond. Second, just one electron may be enough to make an interchain bond as can clearly be seen from a comparison of the energy of a pair (a photoexcited atom next to an atom in the normal configuration) with the energy of a pair cross-linking the two chains. The comparison clearly shows that the pair is energetically more favourable and such pairs will be formed, provided the lifetime of the photoexcited electron is long enough. Following the photoexcitation, the photoexcited carriers recombine but, since the formation of the third bond removes a corresponding empty state from the former LP orbital, recombination in a threefold co-ordinated geometry is impossible and bond breaking takes place. The process proceeds via different paths. First, the newly formed bond can break returning the system to its original state, second, bonds other than the newly formed one may break, in which case electron spin resonance (ESR) active defects (neutral valence alternation pairs (VAPs) should be formed. The corresponding structures are shown in figure 3.22. In a-Se, the bond breaking process results in a recovery of the initial average co-ordination but, in a binary glass, this results in a decrease in the coordination of the As species, which explains the observed photo-induced decrease in the average co-ordination number for arsenic. The threefold co-ordinated defect in a-Se is not stable and further bond breaking takes place. In a binary glass, the latter process may

250

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Figure 3.22. Initial (top) and metastable (middle and left) structures for a-Se and case the shaded circles represent arsenic.

In the latter

result in the creation of wrong bonds (Fig. 3.22, bottom), since the nearest atoms in the initial state are species with a different chemical nature. In figure 3.22 this is illustrated by making one kind of species shaded. Finally, the question as to why such changes cause photodarkening has to be answered. Kolobov et al. [177] explains the photodarkening as follows. In the annealed state, the LP orbitals of selenium, forming the top of the valence band, are oriented perpendicular to each other in order to minimise the energy. Any change in the atomic

positions, such as the creation of threefold co-ordinated pairs, or VAP-like defects, or a topological change, results not only in displacements of the particular atoms involved but also in displacements of their neighbours from their original positions. This increases the repulsive interaction between the LP electrons and results in a shift of the top of the valence band upwards in energy thus leading to a decrease in the forbidden gap, i.e. photodarkening.

Malinovski’s model (1982)

Malinovski et al. [180-182] have suggested that the heating of a spatial region during the photon absorption process induces the photo-structural changes. The model is based on the

CHAPTER 3

251

hypothesis according to which the non-radiative recombination of the photoexcited carriers is carried out by a multiphotonic transition. The vibration energy is strongly concentrated in a spatially limited zone because by transition are generated high frequency phonon modes predominantly localised. The lifetime of the localised phonons is enough long to allow the heating of the microvolume implied, at an effective temperature above At this temperature the material viscosity is low and the atomic displacements bring the structure towards a more disordered state. Because the heating is well localised, the cooling takes place very rapidly so that non-equilibrium, disordered structure is frozen without any relaxation. A crucial aspect of the model is the volume of the micro-region heated by one photon and the relation with the structure at the atomic scale of the glass. The heated volume is estimated and therefore tens of atoms are implied. The size scale for the local heating is near to that proper to the intermediary order. Nevertheless the temperature dependence of the PD between 100 and 300 K seems to contradict the model. The basical concept of the model agrees with the experimental data which confirm that PD is accompanied by an amplification of the structural disorder: a red-shift of the absorption edge, as in the case of PD, appears also in the quenching experiments of the disordered material [124]. Popescu’s model (1983) Popescu [183,184] succeeded to simulate the structural modifications, which appear during PD starting from structural models where molecular configurations and spatially extended configurations do exist. There were simulated structures of composition where molecules coexist together with the disordered layers. The connection to the disordered layer of a molecule determines major distortions in the second and the third co-ordination spheres, thus minimising the additional concentration of As-As homopolar bonds. Thus, it is possible to explain satisfactorily the variation of the structure at the level of the intermediary order and the only slight variation of the homopolar bond concentration into the material. The model is appealing but its full verification as regarding the peculiar aspects of the photostructural phenomena has yet to be done. The model is essentially structural and, in the same time, it can explain simply and suggestively the anisotropy phenomena without introducing special anisotropic centres supposed by several authors. The unified model of Elliott (1986) Elliott [169] proposed a model for the photoinduced reversible processes in the chalcogenide glasses, which is based on the breaking of the intra- and inter-molecular bonds. For the intra-molecular bond breaking have been proposed two mechanisms: the creation of the self-trapped excitons and bond switching. The first mechanism, it seems, do not play a dominant role in the arsenic chalcogenides. By bond switching it would be possible to reach a homopolar bond concentration of ~7 % for and of ~55 % for while, experimentally, have been determined much lower concentrations

252

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Elliott assessed that the breaking of the weak inter-molecular /Van der Waals type/ bonds is the dominant process. According to the model, the photon absorption creates a hole in the lone pair orbital of a chalcogen atom or promotes an electron in an antibonding state thus, strongly weakening the attractive Van der Waals interaction with a neighbouring chalcogen. Then will dominate the repulsive interaction, which will lead to the shift away of the two chalcogens. The non-radiative recombination of the electronhole pair situated in excited state reforms the initial force equilibrium but without returning the two atoms in the original equilibrium position. Thus, the unique configuration from the annealed state is transformed in several metastable configurations of the fundamental state. The Elliott’s model offers an explanation for the modifications of the intermediary order in the photoinduced processes, a fact experimentally demonstrated by the experiments of Yang et al. [185] which have shown that the most important structural changes during PD are produced in the second co-ordination sphere and that the raising of the homopolar As-As bonds is low. As concerned the concentration of the As-As bonds the Elliott’s model gives significantly lower values than those observed.

The model of Lee, Paessler and Sayers /LPS/ (1989) In the LPS model [138] the authors take as a basis the experimental results according to which the photo-structural modifications imply an explicit local atomic configuration. The structural modelling has been performed starting from a piece of amorphous network (Fig. 3.23). The photon density used for the irradiation of such sample corresponds to a photon for 7 atoms of the cluster. Such cluster is

Figure 3.23. A typical configuration with 11 atoms, centred on a sulphur atom in the

amorphous structure of

used in the LPS model [186].

an unity, which consists from a sulphur atom, situated between two pyramidal units When four arsenic atoms are interconnected so as to bind a 7-atom cluster to the rest of the network, it results an 11-atoms configuration as that from the figure 3.23. The authors believe that the medium range order is extremely important for the description of the photo-induced structural modifications.

CHAPTER 3

253

On the basis of the 11 atom cluster, the PD mechanism is the following: the photon absorption promotes an electron from the lone-pair of the central sulphur atom of the cluster, either to a state of localised exciton, or to a conduction band state. The nature of the chemical bond during excitation is fundamentally different from that before and after excitation. During excitation the absence of the lone-pair electron of the central sulphur atom modifies the character of the chemical bond of the neighbouring chalcogens in the direction of amplification of the character p and weakening of the character When the electrons recombine the nature of the bond returns to its normal character of bond but the eventual distortions, which are produced during a short time interval of excitation, can be frozen. Thus, it is adopted a model based on rapid thermalization processes which follow immediately after the photoexcitation of an electron of the lone-pair which leads to changes of the chemical bond from to p. The resulted structures, either relaxed or frozen, are considered to be representative for the PD state. Using the data for the kinetics of the PD and the X-ray absorption spectrometry measurements, Lee, Paessler and Sayers have described with much more scientific rigour the structural details of the PD process. The exemplified material is The process of structural transformation starts with a rotation of the pyramids around the common sulphur atom. Then gradually develops an intermediary state strongly anisotropic in some illumination conditions. From this model it follows that rather an anisotropic centre defined by bridged sulphur with helicoidal neighbourhood very probably gives the local structure associated to PD. The helicoidal structures gets an increased distortion by exposing to light of energy near Fritzsche’s model (1993) Fritzsche [187] developed recently a model, which accounts for the multiple features of the photoinduced transformations allowing for the qualitatively superior explanation of the phenomenology of the chalcogenides under the influence of light. He observes that all the experimental data converge to demonstrate that the modifications in chalcogenides are not limited to isolated active sites or defects but, practically, are affected all the atoms and all the configurations. The general structural change seems to be the consequence of a high number of local bond modifications, which are produced after the creation by light of an electron-hole pair and during the recombination process.

The structure of the chalcogenide is an open structure and, therefore, it allows to the weakly co-ordinated atoms and to the co-ordination defects to pass into the neighbouring regions and to form new bond configurations. Such induced modifications of the chemical bonds, determined by recombination and diffusion, which are repeated in the same vicinity, produce a structural major change. The model supposes a common mechanism for the reversible and irreversible photo-induced modifications. According to this mechanism, in chalcogenides take place under the influence of light, changes of the atomic configurations and of the spatial positions via transition self-trapped excitons. The state saturated by illumination (a state more disordered and with weaker intermediate order than the thermally relaxed state) results from a dynamical equilibrium between the configurational modification induced by light and the thermal relaxation. The decrease of the optical gap and the broadening of

254

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

the Urbach edge, which represent the reversible PD, arise mainly from the interactions of the lone-pair electron, stronger in the saturated state than in the thermally relaxed state. The photoinduced modifications affect the concentration, the homopolar bond concentration and the average strength of the bonds and thus contribute to the observed phenomena. The Fritzsche’s model is at variance with other models where the reversible photostructural modifications are related to specific sites, to defects or to wrong bonds in the material and can be considered as a general model whose souppleness allows for the explanation of the photoinduced transformations. Recently, Meherun-Nessa et al. [188] have shown that the fundamental optical absorption in amorphous chalcogenides, should be described by taking into account the fractals that are known to dominate many physical properties in amorphous semiconductors [189]. They introduced the concept of density of electronic states on fractals and pointed out that the presence of disorder can greatly influence the nature of the density of electronic states even for the extended states.

Shimakawa’s model (1998) Prolonged photo-irradiation induces a volume expansion and causes a decrease in the optical band-gap; i.e. produces photodarkening, in well-annealed chalcogenides. Shimakawa et al. [190] proposed a model for both photodarkening and volume expansion processes in amorphous The amorphous chalcogenides contain disordered layers as a constituent element of their structure. Recently, Bradaczek and Popescu [191] have shown that a paracrystalline model for the medium range order is realistic for these materials. The layers are packed on a short-range scale similar to the case of corresponding crystals (Fig. 3.24 [190]).

Figure 3.24. Schematic illustration of the layers in The arrows E and S indicate expansion and slip motion, respectively. Note that the S motion is a relative movement of each sheet.

The model can be described as follows. During the illumination the layers which absorb photons become negatively charged, giving rise to a repulsive Coulomb interaction between layers which produces a

CHAPTER 3

255

weakening of the van der Waals forces, and hence the interlayer distance increases (volume expansion). This process is indicated by the arrows E in figure 3.24. The experimentally observed widening of the valence angle subtended at sulphur atoms within a layer and hence subsequent increase in the distance between two arsenic atoms bridged by a chalcogen atom on illumination [185] can be explained by the repulsive force involved in the process E; the reaction of repulsive forces between layers acts as a compressive force for each layer. Practically, may take the form of clustered layers, which are shown in figure 3.25 and a slip motion along the layers should also take place with the occurrence of the E process between neighbouring clusters. The slip motion is shown by the arrow S. As the energy required for a slip motion along layers is expected to be greater than that for expansion normal to layers, the rate of S may be lower than the process E. Both processes E and S occur owing to the same repulsion force between the layers, but only the process S is expected to be directly related to photodarkening.

Figure 3.25. Layered clusters in amorphous The expansion and slip motions are indicated by arrows E and S, respectively.

It is known that the photodarkening effect disappears when a certain amount of a group I metal is introduced into [192]. This may be explained as follows: atoms of the metal (group I) (e.g. copper) may act as bridging atoms between the layers and hence reduce the flexibility of the layer network. Such bridging will then reduce the ability of both volume expansion as well as the slip motion, leading to photodarkening. 3.1.4. MODELS FOR THE PHOTOINDUCED ANISOTROPY

Although Tanaka [193] affirmed that today the mechanism of the photoinduced anisotropy is speculative, it is quite remarkable that already several models have been elaborated in order to explain the anisotropic effects induced by light in amorphous compositions and these models succeed to describe satisfactorily the observed phenomena. Grigorovici et al. [194] proposed a model for the vectorial (anisotropic) reversible transformations based on biphotonic atomic processes where the VAP dipolar state is the

256

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

same with that in the case of the scalar (isotropic) transformations, i.e. of the PD. The vectorial and scalar transformations differ only by the second stage of the induced modifications. In the scalar processes the polymerization has not a directional character (Fig. 3.26, a). The two types of processes are necessarily limited to the materials where are possible the light induced transitions towards a metastable excited state with a different degree of structural connectivity. In the vectorial transformations, the second

stage leads to a preferential orientation of the covalent bonds (Fig. 3.26, b, c) due to the directional characterof the interaction between the VAP dipole and the electric field of the

Figure 3.26. Model for the reversible photoinduced structural modifications in the As-Se amorphous system. a. reversible scalar effects due to the dimerization in As-Se b. reversible vectorial effects in c. reversible vectorial effects in AsSe

incident light wave. Terefore, there are not compositional restrictions for the appearance of such anisotropic transformations. The model developed by Popescu [183,184] for the anisotropic transformations, using the structural simulation by computer, offers a simple and suggestive explanation, in agreement with the experimental observations. Popescu started form the observations of Treacy et al. [195] according to which the structure of As-S and As-Se systems, built essentially from disordered layers of the type and is accompanied by molecules, as a result of the compositional fluctuations in the non-equilibrium state. There was built a model for an initial structure of disordered as a piece of As-Se layer (62 atoms As and 40 atoms Se) (Fig. 3.27). The PD process has been simulated by the integration of the molecule in the layer, as a consequence of the polymerization phenomenon. During polymerization, as a consequence of the biphotonic process, two As-As bonds are broken and the quasi-planar

configuration of the open molecule is included in the network of the disordered neighbour layers. The new total configuration (layer + inserted molecule) preserves the number of

CHAPTER 3

257

homopolar As-As bonds. In the general case new As-As bonds can appear, a fact in perfect agreement with the experiment which indicates a small increase of the density of As-As bonds during PD.

Figure 3.27. The modelling of the structural effects in I - initial state (transparent) P1 - photodarkened state P2 - the new polymerization direction (for polarized light).

If for PD is used the polarized light, then the polymerization takes a directional character. In the computer simulated case (starting from a plausible physico-structural model) the direction of bonding of the opened molecule to the layer can be drastically changed (even by a 90° turn) without much expense of distortion energy (Fig. 4.22, P1,P2). The model developed by Lee, Paessler and Sayers [186] has some common points with the Popescu’s model [183]. In the case of the optical absorption corresponding to the optical gap, Eg, i.e. at photon energies situated above the PD edge but below the photoconduction threshold, the photocarriers instead to diffuse and to shift away one from another, form localized excited pairs and determines a high rate of PD [4,196]. Such excitation is an excitation from the band formed by the lone pair complex) towards the top of the valence band. The PD model considers the mechanism in the terms of the structural modifications which follow the excitation. In the case of the anisotropic structural modifications the relation between the photon polarization and the orientations of the lone-pairs is controlled by the optical excitation rate. After excitation with polarized radiation the combined effects of several changes of local orientation lead to a general effect which is manifested as a global anisotropy. By studying a cluster built from 11 atoms of the network, relaxed by computer, Lee et al. [186] suggest the way to reach a dynamic equlibrium state by anisotropic structural modifications. By excitation, that is by elimination of an electron from the lone pair, does appear an intermediate oscillation of at every sulfur atom and a twisting of the two pyramids of (Fig. 3.28). These twistings lead to the local

258

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Figure 3.28. The bond of two pyramids in the network. I As(1)-S(1,2,3) and II As(2) - S(1,4,5)

anisotropic PD and, globally, to a distortion of the amorphous network. The experimental data of X-ray absorption show that PD determines the opening of the bond angle while the second distance As-S remains constant and this corresponds to the indicated rotation in figure 3.28 of the pyramid I around the bond axis As(l)-S(l). In fact the structure of a chalcogenide glass, particularly the arsenic glass, is defined by the bonds between the pyramids. The modification of the configurations of MRO implies, for example, variations of bond angles or rotations of two pyramids with a common sulphur atom. The relative rotations of the pyramids can be described in the terms of the dihedral angle rotations. The sulphur atoms, with double co-ordination, give high flexibility to the network so that the local modifications can lead to changes in the

MRO with a low energy cost and without bond breaking. In the modelling of the anisotropic effects the attention is focussed on the chalcogen atom. Because any chemical bond in the material is in relation with the second order neighbours by means of the dihedral angle, then the position of the chalcogen atom can be defined by two dihedral angles, one for every As-Ch bond. In the crystal case (as e.g.

orpiment, it exists two distinct surroundings of sulphur: the first is defined in the helix -As-S-As-S-As-S- situated along the axis c and the second is specific to the position of connection of two neighbouring helices. The sulphur atoms from the first category have

been called “sulphur atom in helix” while those from the second category have been called “bridged sulphur atoms”. In the case of the cluster with 11 atoms figure 3.29 shows the configurations B and H (bridge and helix) for the case of the crystal and of the

amorphous solid. The amorphous configurations are represented after a structural relaxation and, therefore, it represents the thermally annealed state.

The two configurations of sulphur behave differently. Thus, in the configuration B (symmetrical) the potential energy is lower than in the configuration H (asymmetrical). When an electron of the lone-pair is expounded from the chalcogen of the configuration H, it results a twisting of the two pyramids on the both parts of the bridging atom. On the other hand, for the model of bridged sulphur atom such twisting it is not observed but a symmetrical oscillation of the bridging atom does appear.

CHAPTER 3

259

Figure 3.29. The bridge and helix neighbouring of the chalcogen S in the crystal (1) and in amorphous solid (2). In the amorphous solid the modelling shows that the two configurations dominate although they are characterised by a random structure.

The structural photo-induced anisotropic modification can be related to the anisotropy experimentally observed. Taking into account the fact that in the glasses take place rotations of pyramids and deformations of the local positions during the PD process, as the experiment shows the structural data can be corroborated with those obtained by computer simulation. The conclusion is that the anisotropic centre is given with maximum of probability by the sulphur atom situated in helix and not that bridged. Fritzsche [197] proposed a model, which explains qualitatively a large variety of experiments regarding the anisotropy induced by light in chalcogenides and predicts new phenomena as e.g. the possibility that the non-polarised light induce optical anisotropy in chalcogenide materials. Fritzsche starts from the observation that, at the molecular scale, the structure of the chalcogenide glass is strongly anisotropic. It exists a minimum isotropic volume, which increases together with the size of the anisotropic structural units. The recombination process of a photo-excited electron-hole pair is able to change the local configuration of the bonds and therefore, the optical properties. When the electron and the hole diffuse at larger distances than the scale of the anisotropic microvolumes before recombination, then the local structural changes will be not produced in the initial absorption microvolume but in the microvolumes situated at higher distances which have different orientations as to the original one. Such diffusive recombinations will not lead to a general anisotropy of the material because the anisotropy of the different microvolumes are not correlated. The cumulative effect of the distribution events will, nevertheless, reduce the absorption for the polarisation direction of the incident light and will enhance equally the absorption along the other two perpendicular directions thus creating an anisotropy. A consequence of the redistribution is the possibility to produce the optical anisotropy in the case of illumination by non-polarised light, also. The nonpolarised light, which propagates in a given direction, will be absorbed only in the microvolumes, which react to the polarisation in the other two fundamental directions in

260

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

space. The recombination, the local structural changes and the redistribution must lead to a decrease of the absorption for the polarisation directions x and z and to an increase of the absorption along the direction of incidence of the non-polarised beam, y. The photo-induced anisotropy data have been interpreted in the frame of a model by Tikhomirov and Elliott [158]. The model was proposed to explain the data on the changes in the kinetics of the appearance and of the relaxation in dark of the metastable photo-induced anisotropy as a consequence of the repeated on/off switching of the inducing light. The model is based on structural changes of intimate valence alternation pairs (IVAPs) and of their local environments. These pairs, as well as non-intimate valence alternation pairs, were postulated earlier as [198] or [199] pairs to account for the coexistence of diamagnetism and pinning of the Fermi level in the gap of the chalcogenide glasses. The structural changes proposed are due to photo-induced charge transfer between negatively and positively charged atoms of IVAP’s resulting in changes of their micro-anisotropy (and that of their environment) and chirality in accordance with the polarisation of the inducing light. The model is able to explain the photo-induced gyrotropy (circular dichroism and birefringence) when chalcogenide glasses are illuminated with circularly polarised light [166,200]. Due to the lack of translation symmetry, the structural origin of metastable macroscopic photo-induced anisotropy in glasses is different from the anisotropy of crystals and is more similar to the structural anisotropy of liquid crystals. In order to understand the microscopic origin of the photo-induced anisotropy in chalcogenide glasses, it is reasonable to look for strongly anisotropic structural elements, sensible to the Polarisation State of the light and specific to these materials. In this respect a model of intrinsic (or native) valence alternation pairs in glassy chalcogenides was suggested [198,199]. Kastner et al. [199] denoted these pairs pairs, where C stands for a chalcogen atom, the subscript is the co-ordination number and the superscript is the charge of the chalcogen. Obviously, VAP’s are anisotropic structural elements due to their electric dipole moment. In order to develop the Kastner’s model Tikhomirov and Elliott [158] recall that the only structural elements of the chalcogenide glasses known for certain are the pyramids in As(Sb)-Se(S) systems, or the tetrahedra in glasses of tetrahedra in glasses of Ge-Se(S) systems, where the subscripts stand for the co-ordination number of the central cation. This fact has been well established by Raman, IR-absorption, X-ray absorption and scattering studies. In addition it was observed by Lyubin et al. [166,200] that when exposed to circularly polarised light, photo-induced gyrotropy (circular dichroism and birefringence /optical activity/) appears in As-Se(S) glassy systems. This means that the active centres responsible for this effect must consist of at least four atoms to be able to yield one degree of chirality. Let’s consider the most simple case of a pyramid (Fig. 3.30) where P stands for a pnictide atom (e.g. As, Sb) and C stands for a chalcogen atom, and the particular example of a pyramid containing an IVAP, meaning that positive and negative charges are located at two corners of the pyramid. We place the IVAP’s on C sites since the vectorial effects of photo-induced anisotropy and photo-induced gyrotropy depend only very weakly on the concentration of pnictogen atoms [147,200] i.e., chalcogen atoms must play the main role in the vectorial effects.

CHAPTER 3

261

Before irradiation, IVAP’s are located randomly with respect to each other, since glass is generally overall isotropic. Linearly polarised light excites most effectively those centres where the lone-pair (more strictly speaking the plane defined by two type lonepair orbitals of a atom, normal to the bond) is parallel to the electric vector of the inducing light, but not those centres where the lone-pairs are orthogonal to the electric vector. As a result, a macroscopic anisotropic structure is formed in the glass after prolonged irradiation by linearly polarised light with the presumed orientation of IVAP lone-pairs being orthogonal and, as seen from figure 3.30, with the presumed orientation of the electric dipoles being parallel to the electric vector of the light.

Figure 3.30. Schematic illustration of the centre responsible for the photo-induced anisotropy and gyrotropy in pnictogen-chalcogen systems (e.g. As-Se) before (a) and after (b) optical excitation. Open and solid circles are pnictogens and chalcogens, respectively, solid lines are bonds and dashed

lines delineate the

pyramids. Centres (a) and (b) are mirror images of one another in the plane shown by dotted line.

262

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Obviously, when changing the polarisation of the linearly polarised inducing light by 90°, a repeated exchange of charges inside IVAP’s and reorientation of the optical axis to an orthogonal direction can take place, thus permitting the high degree of optical reversibility of photo-induced anisotropy observed experimentally [147,166]. A very interesting aspect of the suggested particular structural change shown in figure 3.30 is that it is accompanied by a change of chirality of the pyramid and its environment. The chirality itself is due to the pronounced asymmetry of this pyramid caused by different coronations of and sites. The structural change suggested is also able (uniquely it seems) to account for the observed photo-induced gyrotropy (circular dichroism and optical activity) when a pnictogen chalcogenide glass is illuminated by circularly polarised light [166], since pyramids of opposite chirality have different magnetic-optical dipole moments for optical transitions. In particular, prolonged irradiation by right hand circularly polarised light will produce an excess of left hand pyramids by means of charge transfer inside the IVAP shown in Fig. 3.30 due to the higher probability of excitation of right-hand pyramids by right-hand polarised light {right and left-hand pyramids are defined to be those which have magnetic dipole moments for optical transitions parallel and anti-parallel, respectively, to the light wave

vector [201]}. Interestingly, the effect of the photo-induced gyrotropy (optical activity) in amorphous semiconductors was predicted theoretically by DiVincenzo [201] as a nonlinear optical effect. It was noted [201] that the time scale for decay of photo-induced optical activity should be related to the time required for carriers to diffuse from the place of excitation (called domains of right- or left-hand chirality), this time being dependent on the size of these domains. After many-repeated on/off switching of the linearly polarised inducing light, one may reach a more or less stable macroscopically anisotropic glass with domains aligned with respect to one another. This seems to be in contradiction with the general accepted idea of isotropy of disordered solids, but nevertheless may be reasonably if one recalls that stable anisotropy (different from the anisotropy of usual crystals due to the lack of translation symmetry) can be a property even of liquids (e.g. liquid crystals) due to the self-consistent interaction of anisotropic molecules carrying electric-dipole and -multipole moments. In the model developed by Tikhomirov et al. [202] the anisotropy is the result of bond flipping at intimate valence alternation pair (IVAP) sites. The concentration of native defects present in amorphous chalcogenides is sufficient to explain the experimentally observed values of the anisotropy. Kolobov et al. [203] have shown that the sign of the anisotropy depends on the energy of the probing photons. It is argued that anisotropical formation of dynamic interchain bonds is followed by the redistribution of the spatial orientation of bonding and non-bonding electrons [204] Other explanations have been also advanced for the photo-induced anisotropy. The explanation involving quasi-molecular defect [205] may provide a better explanation for the photo-induced gyrotropy. Tanaka et al. [206] have shown that remnants of crystalline layers taking various orientations can account for the observed photo-induced anisotropy. Lyubin and Klebanov [207] demonstrated that excitation of bulk chalcogenide glass by the non-polarised, linearly-polarised or circularly-polarised sub-band-gap light is

CHAPTER 3

263

accompanied by the appearance of isotropic, anisotropic or gyrotropic light scattering due to the generation of new scattering centres or by the transformation of scattering centres already existing in the glass. The different states of scattering centres reflect the possibility that certain structural fragments in the glass exist in several quasi-stable states. This seems to be the key of all effects induced in chalcogenides by the sub-band-gap light, including phenomena as photo-induced optical activity and ellipticity [208]. Arkhipov et al. [209] has formulated a new model that allows to explain some basic characteristics of photo-induced anisotropy in glassy semiconductors. The phenomenon is considered as purely electronic. The model assumes the occurrence of correlated pairs of localised states for electrons and holes and relates the photo-induced anisotropy to the generation of geminate electron-hole pairs trapped by these localised states. Photogeneration, trapping and recombination of geminate pairs control kinetics of photoinduced anisotropy. The model is not intended to propose some specific sort of defects or some particular configuration of structural units that could provide such correlated traps. However, the negative-U centre is obviously one of possible candidates for this role. Fritzsche [210] gave an overview regarding the explanation of the photo-induced anisotropies. Optical isotropic materials such as chalcogenide glasses can become optically anisotropic because they consist of and contain entities, which are optically anisotropic. A recombination event, which leads to a structural change of a microscopic entity, will change the orientation or nature of this anisotropy. This constantly happens everywhere in the material during illumination without, however, producing necessarily a macroscopic anisotropy. For this to happen, it is necessary that the recombining electronhole pair be excited in the same microscopic anisotropic entity, which undergoes the structural change. This means essentially that macroscopic anisotropies result from geminate recombinations of electron-hole pairs, which do not diffuse out of the microscopic entity in which they were created by absorbed photons. The lack of electronhole pair diffusion and the geminate nature distinguishes the recombination events leading to anisotropy from all the other events, which yield isotropic (or scalar) photo-induced changes. This important difference is the cause for the fact that the dependencies on temperature, light intensity, photon energy among other parameters are different for the anisotropic (vectorial) and isotropic (scalar) photo-induced effects. One kind of anisotropic entity is represented by the intimate valence alternation pair defects. After a photo-structural bonding change, its dipole moment will be changed. Other anisotropic entities are also possible. Optical excitations from the valence band to the conduction band are transitions from lone-pair electron states to antibonding states. These transitions are polarisation dependent because of the low covalent co-ordination of chalcogens. Hence, any tiny microvolume of the material is optically anisotropic and is altered with a photo-structural change. Therefore, all optical transitions, i.e. interband, Urbach tail, and defect transitions, are polarisation dependent. Each elemental step of photo-structural change alters the local anisotropy. However, a macroscopic anisotropy can result only from non-radiative recombinations of electron-hole pairs that have not diffused away. Which of the various optical excitations fulfils this requirement can be determined by measuring the quantum efficiency for a given anisotropic effect as a function of photon energy.

264

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

3.1.5. EXOTIC PHOTO-INDUCED MODIFICATIONS

Photo-induced fluidity Photo-induced fluidity or photo-induced glass transition has been discovered by Hisakuni and Tanaka [211]. Sample of (optically processed film: thickness of 50 size 0.2 was annealed at the glass-transition temperature (~450 K) in order to stabilise the glass structure. The free flake was obtained by peeling off from a substrate. Next, one end of the sample was pasted to a glass slide and the other end was bent with a stick. The bending was confirmed to be elastic. Then, the curved flake was illuminated locally by a laser-focused light beam (the spot diameter: from a He-Ne laser (633 nm, 10 mW) for two hours, a permanent deformation occurred. The irradiated part became fluid. The phenomenon becomes more conspicuous if illumination is provided at lower temperatures. This indicates that the photo-structural fluidity occurs through athermal processes. If the temperature rise induced by light were responsible, the phenomenon would become more prominent at higher temperatures (in the case of the estimated temperature rise is less than 0.1 K). The photon energy during illumination is 2.0 eV and is substantially smaller than the Tauc optical band-gap of ~2.4 eV in [212]. In this sense, 2.0 eV photons can be regarded as sub-band gap light, or more precisely Urbach tail light, since at this photon energy exhibits the so-called Urbach tail. It was suggested that the Urbach tail light is responsible for the photo-induced fluidity. Koseki and Odajima [213] have demonstrated that photo-induced stress relaxation is observed in a-Se when subjected to band-gap illumination. A photo-conductive measurement of glass using the constant photo-current method showed that the photoresponse induced by 2.0 eV light is smaller by than that by 2.4 eV light. That is, the number of photoexcited carriers by the Urbach tail light is considerably smaller. However, the ratio merely reflects the absorption coefficients. If the numbers of generated carriers normalised to an absorbed photon are evaluated, no appreciable difference exists between 2.0 and 2.4 eV photons. This fact implies that, since holes are responsible for photo-currents only free electrons are needed for the photo-induced change, irrespective as to whether electrons being trapped (for Urbach ail excitation) or free (for band gap excitation). Microscopically, some electroatomic processes follow the photoexcitation of holes, and then slipping of molecular clusters will occur, which appears as the photo-induced fluidity. Giant photo-expansion Tanaka [213] observed that if glass is illuminated by tightly focused He-Ne laser light under unstressed condition, macroscopic surface expansion occurs. The thickness expansion for 50 thick annealed sample illuminated for 1000 s by 10 mW laser light is ~4%. This is greater by an order of magnitude than that observed (~0.4 %) in the conventional photo-expansion phenomenon in [21]. The giant photo-expansion has been observed also in and when excited by 1.6 eV and 2.7 eV light, respectively. The phenomenon is specific to chalcogenide glasses subjected to Urbach tail illumination.

CHAPTER 3

265

The expansion becomes greater if the sample is illuminated at lower temperatures. This means that the process is an athermal one. Figure 3.31 shows that the maximum expansion observed amounts to When L is determined by the self-focused depth of light Accordingly

Figure 3.31. Giant photo-expansion in a 04 mm thick As2S3 as a function of temperature at which the sample is illuminated. The exposure time is indicated. The light source is a 25 mW He-Ne laser, focused by a ×5 objective lens. Illumination is performed at room temperature.

As illustrated in figure 3.32 we can assume that the giant photo-expansion is induced through a combination of the conventional photo-expansion and the photo-induced fluidity. That is the irradiated volume with a thickness of L and a diameter of 2r is able to expand with a ratio (~0.4 % at room temperature) of the conventional photo-expansion. However, the lateral photo-expansion is practically impossible due to the existence of not

Figure 3.32. A model for giant photo-expansion (cross-sectional view).

266

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

illuminated region and accordingly the strain components will flow to the vertical direction through the photo-induced fluidity. The apparent expansion becomes That is the conventional photo-expansion is seemingly amplified by L/r [65]. This model is consistent with the observations that the giant photo-expansion is prominent only when L >> r. As to the explanation of the photo-expansion effect this is related to an increase in structural randomness of amorphous networks. It is known that a glass is less dense than the corresponding crystal by ~10 %. It is also known that the density of can be modified by ~1 % under some treatments such as annealing. Accordingly, illumination may also be capable to increase the specific volume by which is comparable to the magnitude observed in the giant photo-expansion. Kuzukawa et al. [214] studied the effect of illumination in obliquely deposited thin films of and It was observed that, on annealing, the thickness decreases and the band-gap increases while illumination is found to increase the thickness and decrease the band-gap. Annealing the samples before or after illumination always shows an effect, which is opposite to that of illumination. The authors observed giant changes in both thickness and band-gap with illumination. It was also observed that post-illumination annealing causes the changes to revert to nearly the initially conditions. These giant changes have been explained on the basis of the presence of voids and an easy motion of layers in the films, resulting in an easier expansion and slip motion of the layers. Anisotropic opto-mechanical effect. Krecmer et al. [215] discovered a reversible anisotropic volume change induced by polarised light in a thin film of Contraction occurs along the direction of the electric field vector and dilatation perpendicular to that direction. Light from a He-Ne laser was used whose energy (~2 eV) falls into the Urbach absorption region. This experiment shows that the anisotropies produced extend to other material properties besides the optical tensor. The magnitude of the effect suggests that the anisotropic microvolumes of the whole material are involved and not only IVAP species. These new results imply that the elastic properties, sound propagation and probably many other material parameters of chalcogenide glasses become anisotropic with light exposure.

3.2. Modifications Induced by other Electromagnetic Radiations in Chalcogenide Glasses

The photo-stimulated processes in the elemental chalcogens and in their non-crystalline alloys are of very great diversity and are specific to the amorphous state of the material. Several researches have been carried out for the purpose to see if other electromagnetic radiations, as e.g. UV, X and gamma radiations influence the properties of the chalcogenide glasses.

CHAPTER 3

267

3.2.1. MODIFICATIONS INDUCED BY ULTRAVIOLET RADIATIONS

Hulderman et al. [216,217] have studied the modification of the mechanical strength of fibres exposed for 30 days to the combined action of the UV light, humidity and vacuum. The maximum radiation fluency was They observed a strong degradation of the material. The irradiation was chosen so that the increase of the sample temperature is negligible (under 5 °C). The humidity was 95 %. The effect of the degradation of the mechanical properties of the UV irradiated fibres is stronger in the presence of the moisture. After irradiation there were identified at the fibre surface some defected crystals. How can be explained these phenomena? By UV irradiation and hydrolysis is formed the ortho-arseniate acid at the exposed surfaces: (3.16) (3.17) (3.18)

For the samples irradiated in vacuum there was evidenced some oxidation due to residual oxygen. There was shown that is necessary a vacuum better than Torr in order to eliminate the photo-oxidation of the thin films [13]. The deep internal degradation of the material is probably determined by photo-structural transformations, which do not imply the oxidation. It is well known that the surface tension (compressive) induced during the process of fibber drawing (here is implied the drawing speed and the quenching processes) contributes to the raising of the mechanical strength of the fibber. It is supposed that the ultraviolet radiation induces in fibbers photo-structural transformations of the type present during photo annealing when breaking and reforming of bonds take place. The UV radiation determines also the modification of the rheological properties as e.g. the viscosity [218] and the material flow [2], the visible effect being the smooth topography of the amorphous freshly prepared films. From the studies carried out by different groups it results that the water vapours which come in contact with the chalcogenide surface attack the material in the presence of the UV radiation, which induce bond breaking and facilitate the chemical reactions with the molecules from the environment. Barasch et al. [219] have studied films of thickness 0.1 ÷ 0.3 µm, of composition As2S3 and AsSe prepared by thermal evaporation in vacuum on silicon and glass substrates. The irradiation was carried out by using an excimer laser ELI-72 (russ.) working on the wavelengths 308, 248 and 193 nm at pulse length of 20 ÷ 30 ns and repeating frequency 10 ÷ 150 Hz. The dependence of the relative darkening dT = [T(H)-T(0)]/T(0), where T(0) is the transmittance before and T(H) is the transmittance after irradiation, on the exposure dose, measured at nm is shown in figure 3.33.

268

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

A large darkening (more than three orders of magnitude higher than observed during exposure to continuous wave laser radiation on the wavelength ), starting from several should be explained by local heating of the films by irradiation with strong pulses, including the thermal mechanism of the photo-induced reactions, which proves to be more effective in comparison with the photo-transport reactions which take place also for low intensities of the continuous irradiation. Nevertheless, the researches have evidenced the fact that the spectral composition of the radiation influences essentially the efficiency of the reaction, which points out on the non-thermal character of the phenomenon. For example, in order to reach the same value of the photodarkening in AsSe film, measured on nm, by irradiation with laser pulses with the same peak power and the same pulse length but at 308 nm, there was necessary 10 times higher exposure than in the case of the UV irradiation by an ArF laser (193 nm).

Figure 3.33. The photodarkening versus exposure relation for the case of AsSe film, (thickness: ; f=10 Hz, L0= 193 nm); E H ’: 1 ÷ 2 mJ/cm2 and 2 ÷ 4 mJ/cm2.

The T(H) relation is linear up to a value of then it fully saturates to the level of T~ 0.2 and, further, after reaching the value abruptly decreases with the change of the mechanism towards photobleaching. The experiments have been performed at fixed energies in the pulse, and the dose has been chosen on the account of the pulse number being considered as By increasing the pulse energy, the photodarkening takes place more intensely and reaches the saturation at somewhat lower exposures. Thus, at pulse lengths in the nanoseconds range, the law of reciprocity is broken, i.e. photodarkening depends not only on dose but also on the pulse energy. The bleaching tendency of the chalcogenide films at higher doses (more than 500 pulses) is, evidently, related to the fact that the corresponding part of the curves have been obtained in the regime of multipulse irradiation. In this regime, firstly, the repeating time of the pulses is comparable with the relaxation time of the photo-excitation (for the compound As-Se is of s) and the quasi-stationary regime of the photodarkening does not succeed to stabilise. Secondly, when the dose is increased there appear photo-ablation processes (the elimination of a part of the material under the action

CHAPTER 3

269

of the radiation). With the increase of N and v these processes develop as a result of the local heating and the thickness is lowered in the process of exposure to radiation. Thus, for a full UV ablation of the AsSe film at and v = 50 Hz there was necessary 1250 pulses or an exposure time of 25 s. The effects induced in amorphous chalcogenides by ultraviolet radiation have been studied by Hayashi et al. [220]. A quasi-monochromatic undulator-radiation from an electron storage ring was used as a vacuum ultra-violet light source. On irradiation, the optical absorption edge shifts to lower energies. This change is similar to the case of photodarkening with band-gap radiation. The initial state is recovered by annealing near the glass transition temperature. The rate of change in the optical absorption edge for the UV irradiation is two orders of magnitude higher than that for the band-gap irradiation. While in the case of band-gap irradiation the photodarkening is caused by changes in atomic co-ordinations and positions resulting from exciting lone pair electrons, in the case of UV irradiation an Auger process may participate. When UV radiation is used inner core electrons can be excited (As 3d), i.e. generation of inner core holes is produced. The inner core holes can be immediately filled by outer electrons with Auger processes which could induce more holes in upper states (bonding and lone-pairs states), since one Auger process creates two holes (vacancy cascade process). In this situation, bond breaking or ionisation of atoms is easy to occur, leading to a change in local structural order in amorphous network. Thus, the higher rate of photodarkening during UV irradiation is understandable.

Popescu [221] has studied the amorphous chalcogenide films in the system The modifications induced by UV light in the films have been studied by X-ray diffraction, optical absorption and hardness measurements. The ultraviolet radiation determines a relaxation of the general structure with the formation of more stable structural units in fresh films and modifications implying some structural disordering in annealed films. During irradiation was observed a softening of the films with x < 19 and a hardening of films with The light saturated states arisen both from fresh and annealed state of the films converge towards a unique state, which has been called photo-amorphous state.

3.2.2. MODIFICATIONS INDUCED BY X-RAYS The irradiation of the amorphous chalcogenides by X-rays produces a high photoconduction state. Only a few studies were carried out and up to day no changes in the other properties of the chalcogenides have been observed (excepting photoconduction), as a consequence of X-ray irradiation [222]. Only the synchrotron radiation (a very intense radiation) produces the change of the optical properties and of the rate of dissolution of the material.

270

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

3.2.3. MODIFICATIONS INDUCED BY GAMMA RAYS In the selenium glass the irradiation by gamma rays of high-energy leads to modifications of the mechanical parameters [223]. The shear modulus, the microhardness and also the internal friction and the linear dimensions increase. Domoriad [223] has studied the effect of gamma radiation on several very different amorphous chalcogenide compositions: As-Se-Ge, and AsSe-I. The irradiation has been performed at dose of 200 ÷ 750 R/s, by using a source. The temperatures of the samples have varied between 15 and 35 °C. The maximum integral dose was The linear dimensions of the gamma irradiated samples shrink and this means that a densification effect occurs. Figure 3.34 illustrates this effect as a function of the irradiation dose for selenium and for The densification leads to the increase of the

Figure 3.34 The modifications of the linear dimensions (I) and of the shear modulus (II) a.

in chalcogenide glasses subjected to gamma irradiation. Composition Irradiation dose b. Se c. molten quartz 1. 0.25; 2. 0.5; 3.

1.0; 4. 2.0

chemical stability as related to the dissolution in solvents, a fact experimentally demonstrated [224]. It is interesting to point out that all the glasses from the systems and show a maximum after an initial increase of the linear dimensions. A similar maximum appears also in the glassy selenium. The dose at which appears the expansion maximum (the minimum of density) depends on the glass composition. There was shown that the variation of the shear modulus is significantly reduced with the increase of the irradiation dose. When, in the systems and the sulphur is gradually substituted by selenium and selenium by sulphur,

CHAPTER 3

271

respectively, then decreases. Thus, for the glass, the maximum change of is of 5.4 %, while for is only 0.52 % (Fig. 3.35).

Figure 3.35. The modification of the shear modulus and of the microhardness in amorphous chalcogenides during the irradiation - heating cycles.

The effect of the gamma irradiation on the parameters L and G in the system As-Se-Ge is smaller than in the case of As2Se3 glasses. For and for is only 0.6 %. In the system As-Se-Ge increases with the dose up to saturation. The saturation dose depends on the germanium content. Thus, it is confirmed the stabilising effect of the germanium atoms, as shown in the paper [225], The structural modifications induced by gamma radiation in the chalcogenide systems studied up to day are mostly reversible (see figure 3.35). The return to the initial state can be obtained by thermal annealing. It is interesting to remark that the effect of the gamma radiation decreases from cycle to cycle (when cycles irradiation-annealing are carried out). Thus, at is 5.4 % in the first cycle and only 4.9 % in the second one. The gamma radiation induces significant microhardness modifications, which are mostly reversible. There was demonstrated [226] that the relative modification of the microhardness, in thin amorphous films does not differ from that in bulk glasses. The annealing at 150 °C for 1520 hours reduces by ~ 40 % the induced modifications. The irradiation-annealing cycles produce a gradual decrease of the maximum in the relative variation of the microhardness during irradiation (Fig. 3.35). The structural phenomena, which accompany the gamma irradiation, have been explained by the concurrency of two opposite processes: a destroying process and a polymerisation process, with the consequence of the gradual re-structuring of the amorphous network. In every irradiation stage the polymer structure may take one of several possible metastable states. The metastability is the cause of the returning to the initial parameters. The effect was also observed in irradiated selenium [227]. In the system As-Se-I the iodine interacts with both components of the glass but the bonding energy with these components is different: E(Se-I) < E(As-I). It is supposed that initially the gamma radiation breaks the Se-I bonds. Then, takes place the diffusion of the

272

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

free iodine atoms at the sample surface. As a consequence, the microhardness of the sample and increase by the re-structuring of the amorphous network towards a network with improved bonding. Amorphous films of with the thickness of 100 nm, prepared by thermal evaporation, have been irradiated by gamma rays with doses up to 22 Mrad emitted by a source (dose rate ) [228]. The effect of of -irradiation with different doses is shown in the table below:

One can observe a slight diminishing of the band gap for pure and low doping of arsenic sulphide films and a slight increase of the band gap for higher indium doping. No explanation of these effects has been advanced up to now. Shpotyuk et al. [229] have shown that the gamma-irradiation of vitreous with absorbed doses exceeding Gy is accompanied by strong changes of their physical properties (microhardness, fundamental edge absorption, longitudinal acoustic velocity, acoustic loss coefficient, acoustooptical figure of merit). These effects are reversible as the following thermo-annealing near the glass transition point restores the initial state. The irreversible component of the observed changes achieves 20 % in the fifth cycle of irradiation and annealing. The reversible component of the gamma-induced structural transformations in is accounted for by intrinsic defect formation, in particular, the formation of anomalously co-ordinated centres, and the irreversible one by impurity radiolysis. 3.3 Modifications Induced by Irradiation with Particle Beams

3.3.1. MODIFICATIONS INDUCED BY ELECTRON BEAMS

There were observed structural modifications and modifications of the physical properties in the and amorphous films subjected to electron irradiation, by Macarevici et al. [230] and by Averyanov et al. [231]. As a consequence of the irradiation by electrons of energy 20 ÷ 100 keV of films with thickness 3 ÷ 10 µm, there was observed the photodarkening phenomenon. During irradiation takes place the increase of the refractive index but is observed a high dispersion of the refraction medium in the irradiated and non-irradiated zones.

CHAPTER 3

273

The penetration depth of the electrons and, therefore, the film depth where the interaction with the amorphous material occurs depend on the acceleration potential for the electron in the following form: (3.19)

where d is the penetration depth of the electron beam, U is the acceleration potential and A is a constant.

The darkening effect under the action of the accelerated electrons is reversible. It can be eliminated by heating the glass near the softening temperature or by using electron beams of power larger than that used in darkening, able to heat the already irradiated zones up to the temperatures where the effect disappears. The electrono-stimulated effects have been observed also in other compositions from the system As-Se [222]. The parameters of the electrono-induced modifications, as in the case of the photo-stimulated processes depend essentially on the composition of the films. They strongly change (increase) with the increase of the arsenic content [222]. The electron-induced modifications as also the photoinduced ones are specific to the amorphous state and do not appear in the crystalline analogues. A detailed study of the effects of the electron beam on the amorphous As2S3 has been carried out by Andreichin [232]. He has shown that the irradiation by electrons of both thin films and bulk glasses produce a statically electrization of the surface. The breaking of this electrization under the action of light (by stimulating the photoconduction between the electrized external surface and the internal earthen layers) allows for the formation of a potential relief usable in xerography. The trapped electrons will be localised and do not move anymore. An intense electron beam produces similar effects to the case of irradiation with intense light. A low electron dose can induce the melting of the irradiated surface and a part of the material will be evaporated. Then, a non-uniform solidification will take part and a dark spot will be formed. A more intense electron beam creates a hole whose walls show melting and decomposition effects. In the walls have been detected crystallites of arsenic and arsenic oxide. Because the energy of the electron beam is not only of thermal nature, complementary effects due to the electric charge do appear. Bulk, polished samples of composition As-S and As-Se with the thickness have been irradiated by electron beams of energy E = 2 MeV at doses of [233]. As a consequence of the irradiation are changed the magnitude and the spectral distribution of the optical transmission. These changes depend on the sample composition and on the irradiation dose. The shift of the spectral characteristic of the transmission and the ratio between the maximum value of the optical transmission before irradiation and the value after irradiation (T/T') for the electron dose are given in the table below [233]:

274

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

For all the compositions (with the exception of As2Se3, which practically is not sensitive to the electron irradiation,) remains practically unchanged for irradiation doses in the range and the optical transmission diminishes with the dose. The modification of the transmission saturates when the dose is reached. In the case of the irradiated by doses situated in the range there was found a linear dependence on dose (D) of the transmission variation The electron-stimulated modifications of the optical characteristics of the amorphous chalcogenides are partially or totally reversible (as a function of composition) for temperatures near and behave as reversible for multiple irradiation-heating cycles. The modifications are supposed to be related to the atomic scale structure. As argument serves the study of the structure dependent parameters as e.g. microhardness and the dissolution rate of the material. In the case of as a result of the irradiation with the electron dose of the microhardness increases by 30 % and the dissolution rate in a solution of 3 % M KOH increases by 1.8 times. Kalmnici et al. [234] have analysed the electron-stimulated processes in Sb-Ge-Se (evaporated films). At electron energies of more than 0.6 MeV and for doses higher than rad (corresponding to a fluency of the defect formation is efficient. After irradiation the microscope analysis does not evidence morphological modifications for resolutions down to nm. Nevertheless, the structural measurements performed by X-ray diffraction and the calculation of the radial distribution function indicate some reversible modifications in the second, third and even in the higher co-ordination spheres, a fact related to the change in the intermediate range order in the material. In the compositions during irradiation takes place a certain diminishing of the permittivity ε, a fact explained by the structure densification and the diminishing of the dipolar moment of the heteropolar bonds. The modifications are partially reversible. In the composition the changes exhibit a more pronounced character. In the films of thickness ~80 nm there were observed by electron microscopy an additional small number of crystallites with the structure approaching that of the cubic selenium As a result of the electron-stimulation, there was observed, besides the processes specific to the photo-stimulation, a more deep re-structuring of the network whose

CHAPTER 3

275

mechanism is related to the appearance of the of the domains with excitation of the thermal nature. [235]. The raising of the stresses around the domains leads to the diminishing of the mechanical resistance of the polymeric skeleton and to the increase of the internal potential fields. This explanation is supported by the correlation of the electron effect with by the increase of the sample density during irradiation and, also, by the reduction of the inclination angle of the absorption edge measured on bulk samples. In the region of the excitation domains takes place a more efficient clustering of the amorphous matrix which makes more difficult the reconstruction of the metastable phases and facilitates the increase of the crystalline phase for electron irradiation and, also, raises the crystallisation rate during annealing [234]. Electron beam induced deformation has been observed, recently in chalcogenides [236]. A film of which has been exposed to a line-scanned electron beam of 20 keV expands in the irradiated area while the periphery, is grooved. Closer inspection has shown that the volumes of the expanded and the grooved regions are comparable, which fact implies that the electron beam exposure induces a flow of _ _ toward the irradiated area from the periphery. Only scanned or pulsed electron beams can produce the deformation. If a focused electron beam spot continuously irradiates a point, deformation hardly appears. Deformation produced under fixed pressure is the greatest in and and has been not detected in No deformation was observed in

crystalline (orpiment). These observations have suggested that the surface deformation be induced through an athermal mechanism. If deformation should be a thermal process, then it would be greater in than in since the thermal effect becomes important at which lower in than in Tanaka [236] suggested the electrostatic force and electro-induced fluidity to be responsible for the surface deformation. Thus, for pulsed spot irradiation, a hemispherical region with a diameter of is envisaged. In this area the incident electrons are scattered (Fig. 3.36, a). In the region a number of electrons and holes are exited by the incident electrons. Since is nearly insulating, the incident electrons make the region negatively charged. These electrons are trapped at localised states giving rise to Coulomb repulsion among the sites, which forces to expand the hemispherical region. Moreover, the electron beam exposure can also enhance the fluidity through generating a number of free carriers. As illustrated in figure 3.36 b the free surface expands with accompanying peripheral depression, which is necessary to maintain the material density. Meanwhile electrons and holes will undergo dielectric relaxation and/or recombination since holes are mobile in the material of interest. If Ag-As-S glass is exposed to electron beams, only expansions do appear. This

phenomenon was observed in Ag-As-S(Se) glasses with the Ag content of 15 ÷ 30 at.%. The phenomenon has not been detected in Cu-As-Se glasses. The geometrical change hardly appears in crystalline Ag-As(Ge)-S samples. The relief pattern can be reduced with thermal annealing. A flood exposure of electron beams gives blurred deformation.

276

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Figure 3.16. Model of the electron beam induced deformation. Solid and open circles are the electrons and holes Arrows show the material flow direction.

In the model shown in figure 3.37 the volume expansion induced by electron beams is assumed to be caused by the accumulation of ions into an irradiated region. The glasses Ag-As(Ge)-S(Se) with the Ag content higher than 10 at.% are the ion-conducting amorphous semiconductor, in which the electric conductivity is governed by ions and holes being more mobile than the electrons [237,238]. When such glasses are exposed to electron beams, ions can migrate toward the irradiated region from surroundings and, in consequence, the Ag content at the irradiated region dramatically increases which has been actually demonstrated through compositional studies. Therefore, the Ag-accumulated region will be expanded.

Figure 3.37. Model for the electron-beam induced chemical modification (cross-sectional view). • electrons; o holes;

CHAPTER 3

277

It might be assumed that volume contraction would occur at the region where ions are depleted, while such contractions have not been detected. It is plausible that the glass network consisting of covalent As-S bonds is fairly rigid, and accordingly, Ag+ ions can solely migrate in the As-S network without leaving any traces. Only -accumulated region will expand. It is remarkable that chemical modifications can also be induced by light, while no geometrical changes have been detected [237]. It is suggested that the lack of geometrical changes is due to a small changes in the Ag content upon light illumination.

3.3.2. MODIFICATIONS INDUCED BY ION BEAMS The ion beam bombardment is used currently for the transformation of the crystalline materials into the non-crystalline forms. Orton and Riviere [239] have shown that, as a consequence of the irradiation of the amorphous selenium by argon ions, take place significant modifications of the properties of the amorphous material. The bulk a-Se samples have been irradiated by Argon ions produced at 0.9 kV and There was found that the bombardment produces reversible changes of the structure of the XPS spectra in the region of the valence band of the vitreous selenium. Thus, the maxima corresponding to the 4p-type bonding and antibonding bands merges into one broad peak at an energy approaching to that of the maximum given by the non-bonding band. This effect has been explained by the fact that the dihedral angle between the adjacent bonds can take any value (as shown by the theoretical calculations for the electron state density). Consequently, there was concluded that the bombardment by ions introduces a random dihedral angle which relaxes in a time interval no more than 6 days, at room temperature and stabilises at a fixed angle but exhibiting a random direction, this feature being characteristic to the glass structure, as shown by the diffraction experiments and by the calculation of the electron state density. 3.4. Modifications Induced by Electric and Magnetic Fields

From start we shall exclude from this paragraph the category of modifications related to the ovonic switching implying the use of very high fields, because the problem is discussed in chapter 4. The problem of the influence of the electrical field on the structure and properties of the chalcogenide glasses has been approached by Borisova et al. [240], The influence of the electrical field on the crystallisation process of the amorphous films of composition has been studied in the temperature range The films annealed under do not crystallise in the absence of the electrical field. If an electric field of intensity 200 V/cm is applied the films crystallise at above 80 °C Under the action of the same field, at temperatures above the crystallisation rate of the film increased by a factor of four as compared to the case of the crystallisation in the absence of the field. At 100 °C and 80 V/cm the complete

278

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

crystallisation of the film takes place in 45 seconds. In the absence of the field the film crystallises at 240 °C in 45 seconds.

The amorphous films crystallise under a field of 200 V/cm in 8 minutes 1 at while under a field of 50 V/cm the film is completely crystallised at 150°C in 60 minutes. From the thermal dependencies of the time necessary for complete crystallisation of the amorphous films, it results that the crystallisation process, which depends on the structural defects, is characterised by small activation energy of 20 kcal/mol. When, a 200 V/cm field is switched on, the crystallisation rate increases more than 30 times. The activation energy of the film crystallisation in the presence of this field, at temperatures of 140-190 °C is of 37 kcal/mol the same as that of the thermal crystallisation of the bulk samples. Therefore, the crystallisation process of the glasses in the presence of electrical fields is limited by the As-Se bond switching as in the case of the vitreous flowing. Thus, the electric field eliminates the kinetic limitations of the crystallisation process at temperatures below Borisova et al. [166] consider that the stimulating action of the electric field on the crystallisation of the amorphous films is a consequence of the orientation modifications induced by the electric field in the molecular units and Dembovski et al. [241,242] have discovered a significant effect of the weak electrical fields V/cm) on the viscosity of vitreous selenium at temperatures above This fact has been explained by the existence in glasses and, particularly, in selenium, of some special defects able to interact with the weak electric and magnetic fields. Lyubin [150] has shown that even under high electrical fields able to produce significant injection currents, there was not possible to put in evidence the influence of the fields on the photo-stimulated processes, particularly the rate of changing of the optical parameters during illumination. Some authors [243,244] have studied the influence of the electric field on the sensibility and quality of the holographic and optical recordings. Morikawa et al. [243] have shown that the application of an electric field to an amorphous semiconductor increases the sensibility of the material to the optical recording i.e. the limit energy of the laser necessary to control the optical effects is diminished. The experiments have been carried out on amorphous films of thickness The information inscription energy when the laser beam is acting onto the material subjected to an electrical field is lowered by a factor of 10J as compared to the case when the electrical field is absent. There was also observed that the combined action of the electrical field and of the light determines a strong increase of the domain transformed by illumination. Thus, Okuda [ 170] has shown that the diameter of the point inscribed by laser irradiation increases by a factor of three and this decreases the resolution of the optical recording. An other chalcogenide system studied from the point of view of the influence of the electrical field is P-Se [240] as bulk samples. The glasses do not show any modification of the structure, electrical conduction or other physico-chemical properties modification for long time annealing (more than 100 hours) both in the absence and in the

CHAPTER 3

279

presence of fields of 300 V/cm. Nevertheless, in the PSe glass one observes significant changes under electrical field, even at room temperature. The electrical conductivity increases by three orders of magnitude and the activation energy is appreciably diminished. There was observed on the surface of the sample subjected to the action of the electrical field, a red colour deposit, made from fine crystallites. The same crystalline compounds do appear also in the case of the system P-Ge-Se [240]. The influence of the magnetic field on the amorphous chalcogenides has been studied by Chechetkina et al. [245] and by Dembovskii et al. [246]. There were studied the mechanical properties (ultrasound propagation velocity) and the optical properties (absorption edge) of the glasses Se, and formed in a magnetic field in the glass formation interval (where is the melting temperature) and in alloys at The measurements have been carried out at room temperature. The applied field was 240 Oe and was oriented parallel to the sample. It was demonstrated that a magnetic field applied at temperatures above influences the propagation speed of the ultrasounds and the microhardness and, therefore, directly the structure of the material. The largest modifications in magnetic field have been observed for samples quenched under field for 10 ÷ 15 minutes from the regime temperature down to The induced changes gradually disappear (the material relaxes) during heating in the absence of the field at around Thus, within a month, all the selenium samples completely relax while the modifications from and are partially frozen. The action of the magnetic field on the mechanical properties of the chalcogenide consists in the hardening of the structure of the alloy or of the metastable liquid, which is manifested in the tendency to diminish the relaxation rate of the selenium microhardness and in the tendency of increasing the microhardness and US propagation velocity for and in comparison with their thermal analogues. The hardening is manifested also by the raise of the resistance to mechanical stresses which appear by cooling: the samples prepared from the melt in the absence of the magnetic field are more easily destroyed by polishing than those prepared in the absence of the magnetic field. There was studied the influence of a constant magnetic field and of a variable one directly in the process of viscous flow and crystallisation of selenium at temperatures above A very strong effect has been observed in variable magnetic field [247]. It was observed that by changing the sample temperature and the direction of the magnetic induction relative to the direction of the viscous flow, it is possible to get a modification of the magnitude of the viscosity coefficient of about 6 times. The presence of such effect at a given frequency, only in the narrow temperature interval around 48 °C clearly indicates a resonance character. In constant magnetic field (0 ÷ 1.6 kOe) the absolute effect is smaller and represents around 10 ÷ 30 % from the maximum effect and, in given conditions, an essential role is played by the distribution of the temperatures in the sample. The analysis of the kinetics of the crystallisation of selenium in magnetic field has shown that the magnetic field gives rise to a freezing effect in the process of formation of the crystalline nuclei and not an effect of freezing of the crystallite growth process.

280

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

3.5. Mechanically Induced Modifications

3.5.1. MODIFICATIONS INDUCED BY PRESSURE

The primary effect of the pressure on the solids is the decrease of the inter-atomic distances with consequences on the physico-chemical properties of the material. In the chalcogenide glasses the changes produced by pressure are dramatical because the glasses are mostly molecular compounds formed by covalent clusters linked by weak forces which originate from the Van der Waals type interaction. By applying a pressure the weaker bonds are preferentially compressed while the covalent bonds remain. Thus, the pressure can separate the effects of different categories of bonds in solids. In the chalcogenide glasses there were observed a semiconductor-metal phase transition for pressures above 100 kbar [248]. Weinstein[249] has carried out luminescence studies at high pressures, low temperatures in crystalline and on the amorphous alloy The result has been explained starting from the existence of clusters with layered structure.

Several researchers [249-253] have studied the effect of the pressure on the absorption edge of the chalcogenides. They have observed a significant red shift of the absorption edge (photodarkening effect). For the bulk glass at 293 K it was found a shift towards lower photon energies with with the pressure (5 GPa). After a light exposure of 5 hours (at 293 K) the optical induced densification by pressure observed in the absorption spectrum, is partially recovered with about 0.05 eV. After thermal annealing at 185 °C for one hour the sample completely recovers. Figure 3.38 shows the shift of the absorption edge by illumination for glassy subjected to pressure. The shift increases significantly at low pressures and exhibits weak pressure dependence at high pressures.

Figure 3.38. The shift of the optical absorption edge by illumination

of glassy

subjected to pressure.

CHAPTER 3

281

Therefore, the application of the pressure produces the optical densification simultaneous with the densification of the solid and this optical densification is not

completely eliminated when the pressure is switched off. It exists, therefore, a similitude between the modifications induced by light and those induced by pressure. It is well known that the increase of pressure determines the raising of and this mechanism has been discussed in the terms of the free volume model [251]. The effects induced by pressure in the glasses Se, and have been recently studied by Tanaka [254]. Tanaka observed that the parameter decreases in these materials in th e order: that is, with the mean coordination number, Z. This effect was ascribed to the weak intermolecular interactions [255]. It is known that the mean intermolecular distance from is the highest in this group of materials. It results that the interaction between the molecular clusters from is low at normal pressure and ,therefore, An other possible explanation should be related to the rigidity of the lone pair electrons, which outgo in the intermolecular space. In the tetra-coordinated germanium atoms fix the direction of the lone pair orbitals of the sulphur atoms. Thus, the superposition of the wave functions of the electrons cannot be appreciably amplified when a moderate pressure is applied. Because the valence bandwidth in chalcogenides is mostly determined by this superposition it is quite normal to look for a weak dependence with the pressure of Based on the effect on and on the first sharp diffraction peak (FSDP) intensity (decrease with increasing pressure) Tanaka [255] concluded that during pressure a transition takes place from the layer structure to the continuous random network structure, thus increasing the glass dimensionality. The transformation is accompanied by a very

probable growth of the density of dangling bonds. The densification of the glasses and polymers, in general, may be controlled, as well known, by thermal annealing or by plastic deformation. The densified glasses of and have been prepared by sample annealing at temperatures below and high pressure [256,257] (around 20 ÷ 40 °C under ). After about 20 hours the temperature was decreased to the ambient one and the samples were depressurised. There were obtained density variations in the range (1.5 %) for and in the range for The electrical conductivity of the As2Se3 sample has been changed from As a function of pressure the densification varies. Thus, in for 2000 kbar the density increases by 0.7 % and for 4000 kbar the increase is by 1 %. The densification process at constant temperature and pressure is a process of relaxation of the internal stresses created by elastic deformation. During densification are produced many atom blocks shifting accompanied by switching, distortion and breaking of a number of bonds in the glass network. In this process the connectivity of the network and the network compactness is changed. During elastic deformation the network connectivity remains the same. It is supposed that the difference in the behaviour of the glass during elastic and plastic deformation is related to the appearance of additional distorted and broken bonds. The decrease of during densification indicates a diminishing of the network rigidity, which fact confirms the above interpretations. The breaking and distortion of the chemical bonds are reflected in the material properties by the increase of the number of charged

282

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

electronic states in the gap. In conclusion, we can state that a significant difference exists

between the properties of the densified glass and those of the glass under pressure. 3.5.2. ANISOTROPY INDUCED BY MECHANICAL DEFORMATIONS

While the crystals show anisotropic properties, the amorphous materials are, essentially, isotropic materials. The production of anisotropic amorphous materials is of high fundamental interest. The induced anisotropy accompany the oriented atomic structures. There exists several procedures to get anisotropy in the amorphous material: cooling form the melt under thermal gradient, extrudation, oblique deposition, annealing in electrical and magnetic fields, mechanical deformation. The anisotropy induced by uniaxial compression up to 100 kbar has been studied by Tanaka [258] in Se, and It was shown that the elastically deformed amorphous show photo-elastic birefringence and dichroism as well as a monotonous red shift of the optical absorption with the pressure. A great part of these modifications can be understood starting from the effect of volume change. In axially compressed samples unique properties do appear. Thus, the red shifts of the absorption edges are greater than those induced by hydrostatic compression. By the elimination of the uniaxial compression the frozen birefringence is accompanied by isotropic optical modifications which can be relaxed by illumination and thermal annealing. In order to explain these effects Tanaka [258] suggested that the chalcogenide glasses can be considered as structures of reduced dimensionality consisting in covalent and Van der Waals bonds and, as a consequence, two possible types of plastic deformations can be considered. The first type is given by the gliding of the molecules by eliberating the Van der Waals bonding and the second one is given by the breaking of the covalent bonds. These movements need different energies: 1 eV for intermolecular gliding and 10 eV for intra-molecular breaking [258]. The gliding lines for strong uniaxial compression are parallel to the compressed surface and thus one can suppose that the molecules with layer

structure are able to pack orderly. 3.6. Effects Induced by Ultrasounds The injection of ultrasound waves with the frequency of Hz in bulk samples leads to the modification of the optical properties of the material [259]. There was also

observed the decrease of the slope of the exponential region of the absorption edge, which accompanies the decrease of the propagation velocity of the ultrasounds. Heating the samples in the vicinity of the softening temperature can eliminate both effects. It is, therefore, very probable that these effects be conditioned by the structural transformations of the glass network. These structural transformations have the same nature with the photostructural ones and seem to be related to the presence in the glass of two metastable configurational state [260],

CHAPTER 3

283

3.7. Thermally Induced Modifications

When heat is acting on the amorphous chalcogenide alloys several types of modifications can occur: a. thermo-crystallization b. thermo-structural effects c. thermo-dissolution effects d. decomposition a. The thermo-crystallization, as a difference of photo-crystallisation is a wellknown process, which can take place in every amorphous substance by heating in some conditions. Without insisting on the thermo-crystallisation, which is situated outside the amorphous field, we shall point out several features. The kinetical studies are related to the concept of activation energy. The value of this energy for the case of the glass crystallisation phenomena is associated to the nucleation and growth processes, which dominate the de-vitrification of the majority of the glassy solids. In a process it is possible to identify several activation energies for various nucleation and growth steps but, in general, they combine into representative activation energy for the whole crystallisation process. Thus, for example, in glasses of

composition [261] there are in fact two amorphous phases and the crystallisation process leads to two crystalline phases of different composition. One of them is and the other is The activation energy of the glass transition calculated for every component is and In general, the glass properties vary continuously with the temperature. In a given temperature range, the so-called transformation domain, this variation is more rapid. With the diminishing of the temperature take place changes in the molecular disposal and of the bigger structural units, which lead to the change of the liquid structure and, implicitly, to the change of its properties. The curves, which illustrate the dependency propertytemperature, are of the type shown in figure 3.39. As a function of the rate of variation of the property during transition from the solid to the liquid state, on the curve can be identified three zones. The first zone is situated between and the second zone is between and and the third zone is situated above is the temperature of the liquid, above which crystallisation does not occurs, i.e. is the melting temperature of the crystalline substance (the viscosity reached for this threshold is poise). Above the domain A represents the melt only. In the first zone one observes a very slow variation of the properties with the temperature, described by a straight line. Because in this zone the viscosity is enough low, the mobility of the molecules is enough large to allow for reaching the structural equilibrium, during temperature decrease. That is why above the glass structure is equilibrated, the relaxation time being short. The relaxation time is defined as the time necessary for the glass to reach the equilibrium structure at a given temperature. The transformation temperature is In the temperature range situated between and (marked B) the glass can be considered as an undercooled liquid. The transition from the liquid state to the state of undercooled liquids is practically difficult to observe, while the transition from the liquid state to the crystallised state supposes so large transformations that the observation can be made even by eyes. Below the glassy state is installed.

284

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

Figure 3.39. The curve property - temperature for a glass. A = melt; B= undercooled melt; C= the equilibrium curve for: B, D = glass, E = crystal = the liquidus temperature = the transformation temperature

and

are the low and high limits of the transformation domain)

In the second zone, corresponding to the transformation domain of the glass (between and the variation of the properties is strong during temperature lowering. To every temperature in this domain corresponds an equilibrium structure where the

magnitude of the physical properties of the glass does not modify anymore in time, if the temperature does not change. This equilibrium structure is attained only if the relaxation rate is enough small in order to reach the corresponding relaxation time. Due to the

decrease of the temperature and of the rapid increase of the viscosity, the relaxation time of the structure increases very much and thus the realisation of the equilibrium remains behind the temperature. This means that during the cooling of a glass, for a given

temperature it is produced a structural equilibrium, which does not correspond to the given temperature but to a higher temperature. The transformation domain of the glass is situated between (the vitrification temperature, i.e. the temperature corresponding to the start of the transformations, and For one considers a viscosity of poise. For higher viscosity the glass is solid and fragile. For

is usually ascribed the value of

shows properties specific to fluids. In the range

poise. Below this value the glass

the glass is in the plastic state. This

is the domain of transformation (or hardening) of the glass. In this interval is established, usually in the middle of the domain, which is called the transformation point

CHAPTER 3

285

for the glass, the glass temperature or the transformation temperature of the glass (the name comes from the german word: “Transformationgebiet” = transformation domain). The temperatures for various chalcogenide glasses are given in the table 3.5.

b. The thermo-structural effects appear in the amorphous phase and are due to the re-ordering of the structural units during heating. These units have been frozen during preparation. The effects of stabilisation and elimination of the internal stresses by annealing below are important for improving the mechanical properties of the chalcogenides.

286

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

c. The thermo-dissolution effects have been studied by Kolobov et al. [198]. They have shown that, in the case of besides the photo-stimulated dissolution effects, the effects of thermo-stimulated dissolution of this glass at higher temperatures (T > ~80 °C) is significant. The introduction of zinc in chalcogenides leads to the modification of all the properties of the amorphous material. The conductivity increases strongly and a negative photoconduction is observed in some samples of There was observed the thermo-dissolution of zinc in As-S-Se, Ge-Se and Ge-S also. d. The decomposition effects are related to the transformation of the chemical units in the amorphous material. Examples are the systems As-S(Se)-I and Se-I. During heating, the materials from these systems eliminate part of the iodine element. By heating, sulphur also can be partially eliminated. Finally, we would like to remark that the bulk glasses behave differently, if compared to thin amorphous films, from the point of view of thermal transformations. Thus, e.g., the crystallization temperature of the evaporated films of is around and these values are significantly lower than the crystallization temperature

of the melt quenched glass ( 450 °C) [273]. REFERENCES

[1] S. A. Keneman, Appl. Phys. Lett., 19, 205 (1971). [2] [3] [4] [5]

J. S. Berkes, S. W. Ing, W. J Hillegas, J Appl, Phys., 42, 4908 (1971). A. D Pearson, B. G. Bagley. Mat. Res. Bull., 6, 1041 (1971). K. Tanaka, J. Non-Cryst. Solids, 35&36, 1023 (1980). B. T. Kolomiets, S. S. Lantratova, V. M. Lyubin, V P. Puh, M. A. Taghirdjanov, Fiz. Tverd. Tela (russ.), 18, 1189 (1976). S. R. Elliott, J. Non-Cryst. Solids, 81, 71 (1986). B. T. Kolomiets, V. M. Lyubin, V. P. Shilo, Fiz. Him. Stekla (russ.), 4, 351, (1978). A. E. Owen, A. P. Firth, P. J. S. Ewen, Phil. Mag., B52, 347 (1985). S. C. Agarwal, H. Fritzsche, Phys. Rev., B10, 4351, (1974). K. Shimakawa, K. Hattori, S.R. Elliott, Phys. Rev., B36, 7741 (1987). K. Shimakawa, S. R. Elliott, Phys. Rev., B38, 12479 (1988). Ke. Tanaka, Y. Ohtsuka, J Appl. Phys., 49, 6132 (1978).

[6] [7] [8] [9] [10] [11] [12] [13] J. P. de Neufville, S. C. Moss. S R. Ovshinsky, J. Non-Cryst. Solids, 13, 191 (1973/1974). [14] S. A. Keneman, J. Bordogna, J. N. Zemel, J. Appl. Phys., 49, 4663 (1978). [15] M. Balkanski in Physics of Disordered Materials, Eds. D. Adler, H. Fritzsche and S.R. Ovshinsky, Inst. Amorph. Studies, Plenum Press, 1987, p. 405. [16] G. Pfeiffer, M. A. Paessler, S.C Agarwal. J. Non-Cryst. Solids, 130, 111 (1991). [17] S. A. Keneman, Thin Solid Films, 21(2), 281 (1974) [18] V. M. Lyubin in Non-Silver Photographic Processes, Ed. A. Kartushanski, Himia, Leningrad, 1984, p. 193. [19] S. Rajagopalan, K. S Harshavardhan, L. K. Malhotra, K. L. Chopra, J. Non-Cryst. Solids, 50, 29 (1982). [20] I. I. Turianitsa, A. A. Kikineshi, D. D Semak. Ukr. Fiz. Jurn., 24, 534 (1979). [21] H. Hamanaka, K. Tanaka, S Iizima, Solid State Comm., 23, 63 (1977). [22] T. Igo, Y. Noguchi, H. Nagai, Appl. Phys. Lett., 25, 193 (1974). [23] I. Shimizu, H. Fritzsche, J. Appl. Phys., 47, 2969 (1976) [24] M. Janai, P. S. Rudman, Proc. 5-th Intern. Conf. on Liq. and Amorph. Semicond., Eds. J. Stuke and W. Brenig, Taylor and Francis, London, 1974.

CHAPTER 3

287

[25] R. G. Brandes, F. P. Laming, A. D. Pearson, Appl. Optics, 9, 1712 (1970). [26] A. V. Belykh, O. M. Efimov, L. B. Glebov, Yu. A Matveev, A. M. Mekryukov, M. D. Mikhailov, K. Richardson, J. Non-Cryst. Solids, 213&214, 330 (1997).

[27] V. M. Lyubin, V. K. Tikhomirov, JETP Lett. (russ.), 55, 26 (1992). [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44]

V. M. Lyubin, V. K. Tikhomirov, J. Non-Cryst. Solids, 164&166, 1211 (1993). V. K. Tikhomirov, S. R. Elliott, J. Phys.: Cond. Matt., 7, 1737 (1995) A.V. Kolobov, Ka. Tanaka, J. Optoel. Adv. Mat., 1(4), 3 (1999). M. Janai, P. S. Rudman, Proc. 5-th Intern. Conf. on Liq. and Amorph. Semicond., Eds. J. Stuke and W. Brenig, Taylor and Francis, London, 1974, p. 425. M. Janai, J. Phys. (Paris), 42, 1105 (1981); Phys. Rev. Lett., 47, 726 (1982). J. Feinleib, USA Patent, No. 3636526. J. Dresner, G.B. Stringfellow, J. Phys. Chem. Solids, 29, 303 (1968). P. Chaudhari et al., J. Non-Cryst. Solids, 8&10, 900 (1972). J. Feinleib, J. P. de NeurVille, S. C. Moss, S. R. Ovshinsky, Appl. Phys. Lett., 18(6), 254 (1971). D. Adler, J. Feinleib in The Physics of Optoelectronic Materials, W. A Albers jr., Ed. Plenum Press, N.Y. 1971, p. 233. I. A. Paribok-Alexandrovitch, Fiz. Tverd. Tela, 11(7), 2019 (1969). K. Weiser, R. J. Gambino, J. A. Reinold, Appl. Phys. Lett., 22(1), 48 (1973). R. J. von Gutfeld, Appl. Phys. Lett., 22(5), 257 (1973). E. Haro, Z. S. Xu, J.-F. Morhange, M. Balkanski, G. P. Espinosa, J. C. Phillips, Phys. Rev., B32, 969(15) (1985) D. Jecu, Ph. D. Thesis, University of Bucharest, 1986. D. Jecu, S. Zamfira, M. Popescu, Proc. Intern. Conf. “Non-Cryst. Semicond. ‘89”, Uzhgorod Ukraine, 1989, Vol. II, p. 195. V K. Tikhomirov, P. Hertogen, C. Glorieux, G. J. Adriaenssens, phys. stat. sol. (a), 162, R1 (1997).

[45] V. K. Tikhomirov, G. J. Adriaenssens, Phys. Rev , B55, R660 (1997).

[46] O. Matsuda, H. Takeuchi, Y Wang, K. Inoue, K. Murase, Proc. 7-th Intern.

Conf. on the Structure of Non-Cryst. Materials. 15-19 Sept. 1997, Sardegna, Italy (under press). J. Gaszó, J. Hajtó, G. Zénai, Preprint, Central Res. Inst. Budapest, KFKI-76-4, 1976. A.V. Kolobov, V. A. Bershtein, S. R. Elliott, J Non-Cryst. Solids, 50, 116 (1992). A. V. Kolobov, J. Non-Cryst. Solids, 189, 297 (1995). M. Frumar, A.P. Firth, A.E. Owen, J. Non-Cryst. Solids, 192&193, 447 (1995). A. Roy, A.V. Kolobov, Ka. Tanaka, J. Appl. Phys., 83(9), 4951 (1998). Bhanwar Singh, S. Rajagopalan, P.K Bhat, D K. Pandhya, K L. Chopra, J. Non-Cryst. Solids, 35&36, 1053 (1980). [53] H. Hamanaka, K. Tanaka, A. Matsuda, S. lijima. Solid State Comm., 19, 499 (1976). [54] K. Kimuta, H. Nakata, K. Murayama, T Ninomya, Solid State Comm ., 40, 551 (1981). [55] M. D. Mikhailov, E. A. Karpova, Z. Cimpl, F. K.osek, phys. stat. sol.(a), 117, 467 (1990). [56] H. Hamanaka, S. Minomura, K. Tsuji, J. Non-Cryst. Solids, 137&138, 977 (1991). [57] W. Zhou, D. E. Sayers, M. A. Paesler, Phys. Rev. B., 47, 686 (1993). [58] K. Tanaka, Rev. Solid State Sci., 4, 641 (1990). [59] S. R. Elliott, J. Non-Cryst. Solids, 81, 71 (1986). [60] A. Matsuda, N. Yoshimoto, J. Appl. Phys., 46, 2334 (1975). [61] H. Koseki, A. Odajima, Jpn. J. Appl. Phys., 21, 424 (1982) [62] K. Tanaka, Phys. Rev. B, 30, 4549 (1984). [63] Ya. A. Teteris, I. P Manika, Fiz. Tverd. Tela (Sov. Phys. Solid State), 32, 1441 (1990). [64] H. Hisakuni, K. Tanaka, Appl. Phys. Lett., 65(23), 2925 (1994). [65] M. Popescu, F. Sava, A. Lorinczi, E. Skordeva, P. –J. Koch, H. Bradaczek, J Non-Cryst. Solids. 227-230, 719 (1998). [66] E. Skordeva, J. Non-Cryst. Solids, 192&193, 665 (1995). [67] Z. U. Borisova, Chalcogenide Semiconducting Glasses, Leningrad Univ. Press, 1983, p. 185, 189. [68] H. Hisakuni, K. Tanaka, Science, 270, 974 (1995). [69] M. Terao et al., Jap. J. Appl. Phys., 41, 61 (1972) [70] K. Tanaka, M. Kikuchi, Solid State Comm., 13(6), 669 (1973). [71] A. Matsuda, M. Kikuchi, Solid State Comm., 13(3), 285 (1973).

[47] [48] [49] [50] [51] [52]

288 [72] [73] [74] [75]

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES L. Tichý, H. Tycha, K. Bandler, J. Non-Cryst. Solids, 97&98, 1227 (1987). Ke. Tanaka, Y. Kasanuki, A. Odajima, Thin Solid Films`, 117, 251 (1984). L Tichý, A. Triska, H. Tichá, M. Frumar, Phil. Mag., B54(3), 219 (1986). M. T. Kostishin, E. U. Mihailovskaia, P. F Romanenko, Fa. tverd. Tela, (russ.), 8, 451 (1966).

[76] G. Kluge, phys. stat. sol.(a), 101, 105 (1987). [77] M. Kokado, I. Shimizu, E. Inoue, J. Non-Cryst. Solids, 20, 131 (1976).

[78] D. Goldsmidt, P. S. Rudman, J. Non-Cryst. Solids, 22, 229 (1976). [79] T. Wágner, M. Frumar, J. Non-Cryst. Solids, 116, 269 (1990). [80] Y. Kawamoto, M. Agata, S. Tsuchihashi, J. Ceram. Assoc. Jap., 82, 502 (1974). [81] M. Frumar, A. P. Firth, A. E. Owen, Proc. Intern. Conf. “Amorph. Semicond. ‘84”, Ed. Farchi-Vateva and A. Buroff, Sofia, 1984, Vol. I, p. 216. [82] T. Wagner, M. Frumar, L. Benes, J. Non-Cryst. Solids, 90, 517, 1990. [83] A. P. Firth, P. J . S. Ewen, A E. Owen, C. M. Huntley, in Adv. in Resist. Technol. and Process., Intern. Soc. Opt. Eng., 539, 160 (1986). [84] S. R. Elliott, J. Non-Cryst. Solids, 130, 85 (1991). [85] J. H. S. Rennie, S. R. Elliott, C. Jeynes, Appl. Phys. Lett., 48, 1430 (1986). [86] J. M. Oldale, S. R. Elliott, S.R. Elliott, J. Non-Cryst. Solids, 211, 187 (1997). [92] K. Tanaka, M. Itoh, Optoelectronics, 9, 299 (1994). [93] J. P. De Neufville, in Optical Properties of Solids - New Developments, Ed. B.O. Seraphin, North-Holland, Amsterdam, 1975, p. 437

[94] K. Tanaka, Jpn. J Appl. Phys., 25, 779 (1986). [95] A. I. Buzdugan, I. I. Vataman, V T. Dolghier, K. M. Indricean, A. A. Popescu, Proc. Intern. Conf "Non-Cryst. Semicond. '89", Uzhgorod, Ukrayne, 1989, Vol. I, p. 175. [96] G -M. Schwabb in Semiconductor Surface Physics, University of Pennsylvania Press, Philadelphia, 1957, p. 283 [97] S. Onari, K. Asai, T Arai, J. Non-Cryst. Solids, 76, 243 (1985) [98] R. A. Street, Solid State Comm., 24, 363 (1977). [99] N. F. Mott, A. M. Stoneham, J. Phys. C: Solid State Phys, 10, 3391 (1977). [100] M. S. Chang, J. K. T. Chen, Appl. Phys. Lett., 33, 892 (1978). [101] M. D. Kolwicz, M. S. Chang, J. Electrochem. Soc., 127, 135 (1980). [102] E. J. Porter, G. M. Sheldrick. J. Chem. Soc. Dalton, 1972, p. 1347. [103] M. Bertolotti, F. Michelotti, V. Chumash, P. Cherbari, M. Popescu, S. Zamfira, J.. Non-Cryst. Solids, 192&193, 657 (1995).

[104] Ia. A. Teteris, I. P. Manika, Proc. Intern. Conf. “Non-Cryst. Semicond. ‘89”, Uzhgorod, Ukrayne, 1989, Vol. II, p. 213. [105] I. Manika, J. Teteris, phys. stat. sol. (a), 80, K121 (1983). [106] S. A. Zenkin, K.V. Khirianov, A.B. Lobanov, M. D. Mihailov, I. Iu. Iusupov, O. A. Iakovuk, Proc. Intern. Conf. “Non-Cryst. Semicond. ‘89”, Uzhgorod, Ukraine, 1989, Vol. II, p. 186. [107] K. Tanaka, Solid State Comm.. 34, 201 (1980). [108] M. Babacheva, S.D. Baranovski, V.M. Lyubin, M.A. Taghirdjanov, V. A. Feodorov, Fiz. Tverd. Tela (russ.), 26, 1331 (1984). [109] V. N. Bogomolov, V.V. Poborcii, S.V. Holodkevici, S.I. Shagin, Pis’ma v JETF (russ.), 38, 532 (1983). [110] Y. Katayama, M. Yao, Y. Ajiro, M. Inui, H. Endo, J. Phys. Soc. Jap., 58, 1811 (1989). [111] H. Fritzsche, V. Said, H. Ugur, P. J. Gaczi, J. Physique (Fr.), 42, C4-699 (1981). [112] E. Mytilineou, P. C. Taylor, E. A. Davis, Solid State Comm., 35, 497(1980). [113] K. Kawashima, H. Hosono, Y. Abe, J. Non-Cryst. Solids, 95&96, 741 (1987). [114] W. M. Pontushka, P. C. Taylor, Solid Slate Comm., 38, 573 (1981). [115] R. N. Schwartz, G. R. Blair, J. Appl. Phys.. 65, 710 (1989) [116] K. Tanaka Y. Ohtsuka, J Optique (Paris), 8, 121 (1977) [117] K. Tanaka, Phys. Rev., B36, 9746 (1987). [118] K. Tanaka, J. Non-Cryst. Solids, 90, 363 (1987). [119] R. Prieto-Alcón, E Márquez, J M. Gonzales-Leal, J. Optoel. Adv. Mat., 2(2), 139 (2000). [120] Ke. Tanaka, S. Nakayama, J. Optoel. Adv Mat., 2(1), 5 (2000). [121] Y. Kuzukawa, A. Ganjoo, K. Shimakawa, J. Non-Cryst. Solids, 227-230, 715 (1998). [122] Y. Kuzukawa, A. Ganjoo, K. Shimakawa, Phil. Mag., B 79(2), 249 (1999).

CHAPTER 3 [123] [124] [125] [126] [127] [128] [129] [130] [131] [132]

K. Tanaka, J. Non-Cryst. Solids, 59&60, 925 (1983). K. Kimura, H. Nakata, K. Murayama, T. Ninomyia, Solid State Comm., 40, 551 (1981). J. Hantala, W. D. Ohlsen, P.C. Taylor, Phys. Rev., B38, 11048 (1988). Ke. Tanaka, AIP Conf. Proc., USA, No. 31, 148 (1976). M. Vlcek, M. Frumar, J. Non-Cryst. Solids, 97&98, 1223 (1987). Ke. Tanaka, S. Kyobya, A. Odajima, Thin Solid Films. 111, 195 (1984).

W. Zhou, J. M. Lee, D. E. Sayers, M. A. Paessler, J. Non-Cryst. Solids, 114, 43 (1989). J. Teteris, phys. stat. sol. (a), 83, K47 (1984). V. K. Malinovsky, A. P. Sokolov, V. G. Zhdanov, Solid State Comm ., 51, 647 (1984). A. V. Kolobov, B. T. Kolomiets, V. M. Lyubin, N. Sebastian, M. A. Taghirdjanov, J. Hajto, Fiz. Tverd. Tela, (russ.), 24, 1062 (1982).

[133] V. M. Lyubin in Physics of Disordered Materials, Eds. D. Adler, H. Fritzsche and S.R Ovshinsky, Inst. Amorph Stud., Plenum Press, 1985, p. 673 [134] K. Oe, Y. Toyoshima, H. Nagai, J. Non-Cryst. Solids, 20, 405 (1976). [135] I. Takahashi, Y. Harada, J. Non-Cryst. Solids, 35&36, 1041 (1980). [136] Ke. Tanaka, Y. Kasanuki, A. Odajima, Thin Solid Films, 117, 251 (1984). [137] L. Tichý, A. Tríska, M. Frumar, Phil. Mag., 54, 219 (1986). [138] T. Kawaguchi, S. Maruno, K. Masui, J. Non-Cryst. Solids, 97&98, 1219 (1987). [139] A. Kumar, L. K. Malhotra, K. L. Chopra, J. Non-Cryst. Solids, 107, 212(1989). [140] V.L. Averyanov, A. V. Kolobov, B. T. Kolomiets, V. M. Lyubin, phys. stat. sol.(a), 57, 81 (1980). [141] F. Weigert, Ann. Phys. IV, 63, 681 (1920). [142] V. G. Jdanov, B.T. Kolomiets, V. M Lyubin, V. K. Malinovski, phys. stat. sol.(a), 52, 621 (1979). [143] V G. Jdanov, V. K. Malinovski. Proc. Intern. Conf. “Amorph. Semicond. ‘80 ”, Vol. “Physical Phenomena in Non-Crystalline Semiconductors”, 1980, p.226. [144] J. Hajtó, I. Jánossy, G. Forgács, Preprint KFKI-1982-29, Budapest. [145] V K. Tikhomirov, S R. Elliott, J. Phys.: Cond. Matt., 7, 1337 (1995).

[146] K. Murayama in Physics of Disordered Materials, Eds. M.A. Kastner, Q. A. Thomas, S R. Ovshinsky, Plenum Press, 1987, p. 185 [147] V. M. Lyubin, V. M. Tikhomirov, J. Non-Cryst Solids, 114, 133 (1989). [148] J. M. Lee, G. Pfeiffer, M. A. Paessler, D E. Sayers, A. Fontaine, J. Non-Cryst. Solids, 114, 52(1989). [149] J. M. Lee, Ph. D. Thesis, North-Carolina State University, 1990. [150] V. M. Lyubin, V. K. Tikhomirov, Fiz. Tverd. Tela (Soviet Phys.: Solid State), 32, 1069 (1990) [151] J. Hajto, I. Janossy, G. Forgacs, J. Physique, C15, 6293 (1982) [152] V. G. Jdanov, V.K. Malinovski, Avtometria, 4, 116 (1976). [153] M. G. Serbulenko, Iu. N. Tiscenko, V.G Nenashev, 7th Conference of Coherent and Non-linear Optics, Tashkent 1974. Abstract Booklet, p. 127. [154] V. G. Jdanov, V. K. Malinovski, L. P. Nikolova, T. A Todorov, Optics Comm., 30(3), 329 (1979). [155] K. Kimura, K. Murayama, T. Ninomiya, J. Non-Cryst. Solids, 77&78, 1203 (1985). [156] A. J. Lowe, S. R. Elliott, G. N. Greaves, Phil. Mag., B54, 483 (1986). [157] V. Lyubin, T. Yasuda, A. V. Kolobov, K. Tanaka, M. Klebanov. L. Boehm, Physica B, Cond. Matt., 245, 201 (1998). [158] V. K. Tikhomirov, S. R. Elliott, Phys. Rev., B51 (8), 5538 (1995). [159] V. Tikhomirov, P. Hertogen, G. Adriaenssens, V. Krasteva, G. Sigel, J. Kirchhof, J. Kobelke, M. Scheffler, Proc. Intern. Conf. on Amorphous and Microcrystalline Semiconductors, ICAMS 17, Budapest, August, 1997; J. Non-Cryst. Solids, 227, 694 (1998). [160] A. V. Kolobov, V. Lyubin, T. Yasuda, M. Klebanov, K. Tanaka, Phys Rev. B; Cond. Matt., 55(14), 8788 (1997).

[161] [162] [163] [164]

Ke. Tanaka, T. Gotoh, H. Nakayama, Appl. Phys. Lett., 75(15), 2256 (1999). H. Fritzsche, Phys. Rev. B52, 15854 (1995). V. K Tikhomirov, S. R Elliott, J. Phys.: Cond. Matt., 7, 1737 (1995). E. V. Emelianova, P. Hertogen, V. I Arkhipov, G. J. Adriaenssens, Homage Book Andrei Andriesh, INOE&INFM Publ. House, Bucharest 1999, p. 153; Rom Rep Phys., 51(3-4), 265 (1999).

289

290

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES

[165] [166] [167] [168]

V. M. Lyubin, V .K. Tikhomirov, Pis'ma v JETF, 52, 722 (1990). V. M. Lyubin, V. K. Tikhomirov, J. Non-Cryst. Solids, 135, 37 (1991). J. Hajto, P. J. S. Ewen, phys. stat. sol.(a), 54, 385 (1979). B. T. Kolomiets, V. M. Lyubin, V. P. Shilo, Pis’ma v JETF(russ.), 17, 577 (1973). S. R.Elliott, J. Non-Cryst Solids, 81, 71 (1986) R. Grigorovici, A. Vancu, J. Physique (Fr.), Coll. C4, suppl. au No. 10, T. 42, p.C4-391 (1981). R. T. Phillips, J. Non-Cryst. Solids, 70, 359 (1985). A. V. Kolobov, B. T. Kolomiets, O. V. Konstantinov, V. M. Lyubin, J. Non-Cryst. Solids, 45, 335 (1981). V L. Averyanov, A. V. Kolobov, B T. Kolomiets, V. M. Lyubin, J. Non-Cryst. Solids, 45, 343 (1981). A. V. Kolobov, J. Non-Cryst. Solids, 164-166, 1159 (1993). A. V. Kolobov, Yu. P. Kostikov, S. S. Lantratova, V. M. Lyubin, Fiz. Tverd. Tela (russ.), 33, 444 (1991), V A. Bershtein, L. M. Egorova, S. R. Elliott, A V Kolobov, Proc XVI Intern. Congress on Glass, Madrid, 1992, p 285. A.V. Kolobov, H. Oyanagi, A Roy, Ka Tanaka, J. Non-Cryst Solids, 232-234, 80 (1998). M. Kastner, D. Adler, H. Fritzsche, Phys. Rev. Lett., 37, 1504 (1976). A. V. Kolobov, M. Kondo, H. Oyanagi, R. Durni, A. Matsuda, K. Tanaka, Phys. Rev B 56, R485 (1997). V. K. Malinovski, V. K. Jdanov.J. Non-Cryst. Solids, 51, 31 (1982). V K. Malinovski, V N. Novikov, J. Non-Cryst. Solids, 85, 93 (1986). V. G. Malinovski, J. Non-Cryst. Solids, 90, 37 (1987). M. Popescu, J. Non-Cryst Solids, 90, 533 (1987) M. Popescu in Physics of Disordered Materials. Eds. D. Adler, H. Fritzsche, S. R. Ovshinski, I.A.S , Plenum Press 1985,p 123. C. Y. Yang, M. A. Paessler, D. E. Sayers, Phys. Rev., B, 36, 9160 (19870. J M. Lee, M. A. Paessler, D E. Sayers, J. Non-Cryst. Solids, 123, 295 (1990). H. Fritzsche, Phil. Mag. B, (UK), 68(4), 561 (1993). Meherun-Nessa, K. Shimakawa, A, Ganjoo, J. Singh, J Optoel. Adv. Mat., 2(2), 133 (2000). R. Zallen, The Physics of Amorphous Solids, John Wiley & Sons, New York 1983, p. 135. K. Shimakawa, N. Yoshida, A. Ganjoo. Y. Kuzukawa, Phil. Mag. Lett., 77(3), 153 (1998). H. Bradaczek, M. Popescu, J Optoel Adv Mat., 2(2), 153 (2000). J Z. Liu, P. C Taylor, Phys Rev Lett. 59, 1938 (1987). Ke, Tanaka, Rev. of Solid State Sci., 4(2&3), 641 (1990). R. Grigorovici, A. Vancu, L Ghita, J Non-Cryst. Solids, 58&60, 909 (1983). D. J Treacy, U. Strom, P. B. Klein, P. C Taylor, T P Martin, J. Non-Cryst. Solids, 35&36, 1035 (1980) K. Tanaka, A. Odajima, Solid State Comm., 51, 123 (1983). H. Fritzsche, J. Non-Cryst. Solids, 164&166, 1169 (1993). R. A. Street, N. F. Mott, Phys. Rev Lett., 35, 1293 (1975) M. Kastner, D. Adler, H. Fritzsche, Phys. Rev. Lett., 37, 1504 (1976). V M. Lyubin, V K. Tikhomirov. M M. Chervinski, Semicond Sci. Technol., 6, 807 (1991). D. P DiVincenzo, Phys. Rev., B37, 1245 (1988). V. K. Tikhomirov, G. J Adriaenssens, S. R. Elliott, Phys. Rev. B, 55, 23 (1997). A. V. Kolobov, V. Lyubin, T. Yasuda, K. Tanaka, Phys. Rev. B, 55, 23 (1997). A. V. Kolobov.V. M. Lyubin, T Yasuda, M. Klebanov, K. Tanaka, Phys. Rev. B, 55, 8788 (1997). S. A. Dembovsky, E. A. Chechetkina, J. Non-Cryst. Solids. 85, 346 (1986). Ke. Tanaka, K. Ishida, N. Yoshida, Phys. Rev. B, 54, 9190 (1996). V. M. Lyubin, M. Klebanov, Rom. Rep. Phys., 51 (3-4), 225 (1999). V. M. Lyubin, V K. Tikhomirov, J Non-Cryst. Solids, 135, 37 (1991). V, I. Arkhipov, E. V Emelianova, P Hertogen, G J, Adriaenssens, Phil Mag. Lett., 79, 463 (1999). H. Fritzsche in Physics and Technology of Semiconductors, World Scientific, 1998.

[169] [170] [171] [172] [173] [174]

[175] [176] [177] [178] [179] [180] [181] [182]

[183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196]

[197] [198] [199] [200] [201] [202] [203] [204]

[205] [206] [207] [208] [209] [210]

CHAPTER 3

291

[211] H. Hisakuni, Ke. Tanaka, Science, 270, 974 (1995) [212] S. R. Elliott in Chalcogenide Glasses, ed. J. Zarzycki, Mat. Science and Technology, Vol. 9, VCH, Weinheim, p. 375-454, 1991 [213] H. Koseki, A. Odajima, Jpn. J. Appl. Phys., 21, 424 (1982). [214] Y. Kuzakawa, A. Ganjoo, K. Shimakawa, Phil. Mag. B, 79(2), 249 (1999). [215] P. Krecmer, A. M. Moulin, R. J Stephenson, T Rayment, M E. Well, S. R. Elliott, Science, 277, 1799 (1997). [216] F. Hulderman, J. S. Sanghera, J D. Mackenzie, J. Non-Cryst. Solids, 127, 312 (1991). [217] J S. Sanghera, J. D. Mackenzie, F. Hulderman, Mater. Lett., 8, 409 (1989). [218] D. K. Tagantsev, S. V Nemilov, J. Stekl. Nauk (russ.), 11, 352(1986). [219] E. G. Barasch, A. Iu. Kabin, V. M. Lyubin, R. P. Seysyan, Jurn. Techn. Fiz. (russ.), 62(3), 106 (1992). [220] K. Hayashi, D. Kato, K. Shimakawa, J. Non-Cryst. Solids. 198-200, 696 (1996). [221] M. Popescu, Proc. 10th Intern. School on Condensed Matter Physics, Varna, September 1998 (Eds. J. M. Marshall, N Kirov, A Vavrek, J M. Maud) World Scientific, 1999, p. 101. [222] V. M. Lyubin, Proc. Intern Conf. “Amorphous Semicond. ‘78”, Pardubice, Czechoslovakia, 1978, Vol. I, p20.

[223] I A. Domonad, J. Non-Cryst. Solids, 130, 243 (1991). [224] I. A. Domoriad, V. M. Lyubin, B. T. Kolomiets, V. P Shilo, Fiz. Him.Stekla (russ.), 11, 595 (1985). [225] Z. U. Borisova. Himiia Stekloobraznih Poluprovodnikov(russ.), Leningrad Univ., SSSR, 1972 [226] Radiationnie Stimulirovannie Protchesi v Tverdih Telah. (russ.), Ed. O. R. Niiazova, Ed. FANM SSSR - Uzbek, Taskent, 1969, p. 57. [227] I. A. Domoriad, M. G. Spirina, N I. Timochina, J. Non-Cryst. Solids, 90, 525 (1987). [228] S. A. Fayek, M. Elocker, S. S Fouad, M. H. El-Fouly, G. A. Amin, J. Phys. D: Appl. Phys., 28, 2150 (1995)

[229] O. I Shpotyuk, A. P. Kovalskii, M.M. Vakiv, O. Ya. Mrooz, phys. stat. sol. (a), 144, 277 (1994). [230] V. S Makarevici et al., Izv. Akad. Nauk, SSSR, Ser. Fiz. Mat. (russ.), 5, 44 (1978). [231] V. L. Averyanov, B T. Kolomiets, V. M Lyubin, S I. Nesterov, V. P Shilo, J. Techn. Fiz. (russ.), 49, 865 (1979). [232] R. Andreichin, Proc. Intern. Conf “Amorph. Semic ‘78”, Pardubice, Cehoslovakia, 1978, Vol. II, p.627. [233] S. S. Sarsembinov, R M. Guralnik, S. la. Maximova, T. A. Fedorenko, Proc. Intern. Conf. “Amorph. Semic. ‘82”, Bucharest, Vol. 2, 1982, p.211. [234] S. I. Kalmnici, L.G. Kesler, N. D Savcenko, V.P. Ivanitski, A. M. Fradkin, Proc. Intern. Conf. “Amorph. Semic. ‘84”, Gabrovo, Bulgaria, 1984, Vol. 1, p.54 [235] L. S. Smirnova, Fiz Tverd. Tela (russ.), 2, 1669 (1960). [236] Ke. Tanaka in Physics and Applications of Non-Crystalline Semiconductors in Optoelectronics, Eds. A. Andriesh, M. Bertolotti. Kluwer Academic Publishers, Dordrecht-Boston-London, 1997, p.3l. [237] N. Yoshida, K. Tanaka, J Appl Phys., 78, 1745 (1995). [238] K. Tanaka, M. Itoh, N. Ioshida, M. Ohto, J. Appl. Phys., 78, 3895 (1995). [239] B. R. Orton, J. C. Riviere, J. Non-Cryst. Solids, 37, 401 (1980). [240] Z. U. Borisova, M. D Mihailov, T. S. Rikova, Proc. Intern. Conf. “Amorph Semic. ‘76”, Balatonfuered, 1976, p. 349. [241] S. A. Dembovskii, E. A. Chechetkina, S. A. Koziuhin, Solid State Comm., 44, 1561 (1982). [242] S. A. Dembovskii, S. A. Koziuhin, E. A. Checheitkina, Mat. Res. Bull., 17, 801 (1982). [243] I. Morikawa, T. Nakajima, K. Sakurai, Appl. Phys. Lett, 23(7), 405 (1973). [244] M. Okuda et al. , Appl Optics, 13(4), 79 (1974). [245] E.A. Chechetkina, S.A. Dembovskii, S.A. Koziuhin, V.A. Sidorov, Proc. Intern. Conf. "Amorph. Semic. '84", Gabrovo, Bulgaria, Vol. 1, 1984, p 88 [246] S. A. Dembovskii, S. A. Chechetkina, S. A. Koziuhin, Proc Intern. Conf. "Non-Cryst. Semicond. '89", Uzhgorod, Ukraine, Vol. 1, 1989, p. 142. [247] S. A. Dembovskii, E. A. Chechetkina, S. A. Koziuhin, Pis'ma v JETF (russ.), 41, 88 (1985). [248] S. Minomura in Amorphous Semiconductors, Ed. Y. Hamakawa, Ohm, Tokyo, 1982, p.245.

292 [249] [250] [251] [252] [253] [254] [255] [256] [257] [258] [259] [260] [261] [262] [263] [264] [265] [266] [267] [268] [269]

[270] [271] [272] [273]

MODIFICATIONS INDUCED IN CHALCOGENIDES GLASSES B. A. Weinstein, Phil. Mag., B50, 709, 1984. J. M. Besson, J. Cernogora, M. L. Slade, B. A. Weinstein, R. Zallen, Physica, 105, 319 (1981). K. Tanaka, Phys Rev., B30, 4549 (1984) K. Murase, T. Fukunaga, AIP Conf. Proc., No. 120, 1984, p. 449. H. Tsutsu, K. Tamura, H. Endo, Solid State Comm, 52(10), 877 (1984) Ke. Tanaka, Proc. Intern. Conf. “Amorph. Semicond. ‘86”, Balatonszéplak, Hungary, J. Non-Cryst. Solids, 90, 363 (1987). K. Tanaka, J. Non-Cryst. Solids, 77&78, 1207 (1985). B T. Kolomiets, E. M. Raspopova, S. S Lantratova, Proc. Intern. Conf. “Amorph. Semicona. ‘74”, Reinhardsbrunn, Germany, Vol. 2, 1974, p.284. E. M. Raspopova, E. B. Ivkin, J Non-Cryst. Solids, 22, 437 (1976). Ke. Tanaka, J. Non-Cryst. Solids, 119,243, 254 (1990). S. Okano. SolidState Comm., 28, 369 (1978). Stekloobraznii Sulfid Mishiaka i evo Splavi, (russ.), Ed B. T. Kolomiets, Stiintza, Khishinau 1981, p. 186. N. Afify, J. Non-Cryst. Solids, 126, 130 (1990). Ke Tanaka, Jap. J. Appl. Phys., 25(6), 779 (1986). A Eisenberg, J. Phys. Chem. Solids, 67, 1333 (1963) S. A. Dembovskii, N. P. Lujnaia, J Neorg. Him.(russ.), 9(3), 660 (1964). Ke. Tanaka, Phys. Rev., B30(8), 4549 (1984) Ke. Tanaka, Solid State Comm., 54(10), 867 (1985). B. A. Joiner, J. C. Thompson. J. Non-Cryst. Solids, 21, 215 (1976). M. B. Myers, E. J. Felty, Mat. Res. Bull., 2, 715 (1967). R. Andreichin, P. Simidtcheva, M. Nikiforova, Dokl. Bulg. Akad. Nauk, 18(12), 1079 (1965). G. C. Das, N.S. Platakis, M. B. Bever. J. Non-Cryst. Solids, 15, 30 (1974) R. Ota, T. Yamate, N. Suga, M Kunugi. J. Non-Cryst. Solids, 29, 67 (1978). M. S. Gutenev, Izv. Akad. Nauk SSSR - Neorg. Mat., 22(3), 477 (1986). A. V. Kolobov, B.T. Kolomiets, V M Lyubin, M A. Taghirdjanov. Solid State Comm., 54(5), 379(1985).

CHAPTER 4

293

APPLICATIONS

4. 1. Ovonic Devices 4. 1. 1. PHASE-CHANGE SWITCHES Among the most important applications of the electrical conduction in chalcogenide glasses are the switches with and without memory. The switching phenomenon is an old observation. The current controlled negative resistance, a related feature, was observed in boron as early as 1913. Switching has been found in polycrystalline materials such as nickel oxide and antimony sulphide, in single crystals of GaAs, tin sulphide and in magnetic oxides. Amorphous switching was discovered and applied by Ovshinsky in 1959 [1]. Starting from 1962 Pearson [2] remarked the possibility to get a switching effect in glasses from the system As-Te. Kolomiets and Lebedev [3] found a switching effect in glassy TlAs(SeTe)2 in 1963. Southworth [4] approached this problem in 1964 and Eaton [5] published the results of a detailed study on switching in As-Te materials. The interest in the switching phenomena in chalcogenide materials rose considerably from 1968, as a consequence of publication by Ovshinsky [6] of details on the switch later called “ovonic switch”. The ovonic switching discovered by Ovshinsky was evidenced in a thin film of a non-crystalline alloy based on Te, As, Ge and Si prepared by vacuum evaporation. The switching effect consists of a sharp transition from high resistivity to low resistivity of the material when an enough large electric field is applied. Two kinds of switches were described: threshold switches and memory (bistable) switches. A switch is called threshold switch if the return to the initial conduction state is triggered by the decrease of the electrical field or by switching off the field. If the return to the initial state of the device is realised by a short current pulse then we are dealing with a memory switch. The I(U) characteristics of a threshold switch and of a memory switch are shown in figure 4. 1 [7,8]. In the case of the threshold switch (Fig. 4.1 a) the high conductivity state is characterised by an electrical resistance of at low electric fields. Above the current rises exponentially with increasing voltage. The threshold field is of the order of and is largely independent of thickness. This makes the threshold voltage proportional to thickness and it can be made to vary from 2V to 300 V. On the other hand the holding voltage is independent of thickness and can be varied from 0. 5 V to 1. 5 V depending on the material. Switching time is of the order of 100 ns. Between the application of the threshold field and the occurrence of switching there is a

294

APPLICATIONS

Figure 4.1. Current-voltage characteristics of a threshold switch (a) and of a memory switch (b) in thin films of chalcogenide glasses.

delay that decreases exponentially with the voltage above threshold. This delay can be reduced from to ns. The switching parameters depend on the composition of the material. One important property of the switching materials is the absence of any significant phase transition up to The memory switching (Fig. 4.1 b) was observed in chalcogenide glasses containing small percentages of As and Sb. The characteristic curve resembles that of the threshold material except that no holding voltage is necessary and the switching is bistable. The switching comprises two stages. The first stage is characterised by a delay of for triggering the transition to an unstable state. The second stage consists of the effective switching in the “on”-state (with or without memory) in The first amorphous chalcogenide memory with a 256-bit electronically alterable read-mostly memory (RMM) was built at Energy Conversion Devices on substrate from INTEL Corp. (Fig. 4.2) [9]. In the 1970’s ECD company commercialised non-volatile semiconductor memory devices based on chalcogenide phase-change materials [10,11], These 1024 bit read mostly memory devices (RMM) used a tellurium-rich alloy, which phase segregated upon crystallisation to provide an electrically conducting current filament. The crystallites dissolved back into the amorphous matrix when data are erased. Memory cell accesses devices and drivers and decoders for the RMM were provided by conventional crystalline silicon devices. The nature of the switching in chalcogenide glasses was the subject of controversy and of numerous investigations [12,13]. The materials for memory switches exhibit usually high conductivity (e. g. in [14,15] the room temperature conductivity is about and set in a matter of a few milliseconds.

CHAPTER 4

295

Figure 4.2. The read-mostly memory cell built from a thin film deposited on a single crystal silicon substrate in series with a silicon diode (a). The metal stripes -the x-address lines- connects the doughnut-shaped ovonic switches. Running perpendicular to the stripes the y-addresses buried in the silicon connect the figure-eightshaped silicon diodes (b). The isolated diodes and ovonic switches are connected in series across an x- and y- address line. The RMM consists of 256 such combination (c) (after [7]).

296

APPLICATIONS The materials for threshold switches exhibit usually low conductivity (e. g. in

[16] the room temperature conductivity is and requires much larger times than a millisecond to set. The major difference between memory and threshold switches is, therefore, the time required to set when excited by pulsed or continuous current. The first experimental observations have shown that the switching effect in thick films around room temperature is characterised by large switching times and, therefore, it was considered that switching mechanism is of thermal naure. For thin films and low temperature, the switching time is low and this feature convinced the physicists that the electronic processes are dominant. Haman and Badluev [17] have explained the switching in complex crystalline chalcogenides of composition and in glassy by a crystallisation effect. The transition in the conducting state was explained by the appearance in material of a continuous crystalline “channel” whose cross section increases with the power dissipated within the material. At the voltage corresponding to the transition in the on-state the temperature of the boundary of the crystalline channel reaches the value and, according to the theory of thermal breakdown, the following relation holds: (4. 1)

where B is a constant, is the Boltzmann constant, is the environmental temperature and is the activation energy of the electrical conduction. Borschevskii et al. [18] reached the same conclusions after studying the memory switching mechanism in the glasses and They have shown that the dynamics of the formation of the current channel exhibiting high conductivity depends on the glass structure and on the parameters of the current pulses, which go through the material. During the switching process the temperature in the current channel raises and as a result a phase transition or occurs. Petersen and Adler [19] have shown that the radius of the conducting channel grows when the current increases in the process of switching. The calculations based on the electro-thermal theory of switching have shown that the diameter of the channel obtained experimentally exceeds that estimated theoretically. Moreover, the calculated increase in temperature is not realistic. The experiment gives much lower temperatures (the difference is of Since such a temperature increasing could not even begin to account for the increase in conductivity greater than eight orders of magnitude observed in the on-state, this result unambiguously proves, after the authors’ opinion, the electronic nature of the conductivity state in the switching materials. The following picture for the structural basis of the switching phenomenon emerged. In the high conductivity state the current flows through filaments, some microns in diameter, formed in the glassy matrix. The filaments are built from dendrite crystals, sometimes grown radially around the amorphous core. The ratio of the resistivities in the states “on” and “off” is much lower than the ratio of the resistivities in polycrystalline and in amorphous state of the same material. In the memory switch the transition to the non-conducting state (off-state) corresponds to the melting of the filament and the return to the homogeneous amorphous state.

CHAPTER 4

297

Some authors believe that a mixed electronic and thermal mechanism is more appropriate for the explanation of the experimental data. The thermal, as well as the electronic conditions have to be satisfied for switching, both being necessary but neither

sufficient on its own. Some arguments in favour of this mechanism were set forth by Kolomiets et al. [20]. They pointed out that the electronic mechanism of switching in chalcogenides should be both by tunnelling breakdown and ionisation by collisions because both mechanisms are related to the existence of high internal fields (up to V/m) and “tails” of localised states. The experimental values for the temperature of the conducting filament, measured with various methods in films of some micrometers in thickness, at currents of A fluctuate in the range K. The temperature is maximum in the centre of the filamentary channel [21]. In order to explain the experimentally observed dependencies (the relation between the maximum filament temperature and the intensity of the electrical current; the relation between the direct voltage in the on-state and the filament temperature) the electrono-thermal theory takes into consideration the phase transition that occurs when high temperature and intense electric fields are acting simultaneously. The presence of the phase transition allows taking the current density in the filament independent of the filament radius. Thus, the total current, I, will be the product of the current density, i, by the cross-section area of the filament, S. The current density is obtained from the relation [22]: (4. 2) For typical values: 300 K, W/m, one gets a value which fits rather well the experimental data. Nowadays, the electronic nature of the threshold switching is nearly unanimously accepted. Experiments with ultrashort pulses [23] gave a strong confirmation. The electronic nature of the threshold switching is well established under microsecond switching regime [24]. The switching effect in amorphous semiconductors is used for producing electronic devices with great advantages over the analogous devices produced from crystalline semiconductors due to simple and cheap preparation, resistance against nuclear radiation, cheap raw materials with moderate purity [19]. Kostylev et al. [25] have studied the properties of the thin films based on compositions and, as a result, they recommended them for threshold switches with high filamentary current, because the electronic part is large and the thermal processes does not play an important role. They succeeded to get stable and reproducible switching diodes. In the next sentences we give some technical data on the devices based on amorphous chalcogenides. The threshold voltage of the commercial switches from is situated in the range 2 300 V. The switching time is s. The initial resistance of the switch is the nominal current is A and the switching frequency is Hz. A memory cell exhibits in the non-conducting state a resistance of In the conducting state the resistance is and the switching time is several nanoseconds. The control voltage of the memory cell varies in

298

APPLICATIONS

the range 0. 1 250 V. The bistable switches produced at ECD with Ge-As-Te alloys can be maintained in every state for very long time without power consumption. The temperature range for operation is -180 °C 180 °C. The parameters of the electrical switches based on chalcogenide glasses have been improved by using new amorphous materials [26]. The main requirements for the parameters of a glassy material to be used in the switching devices have been established: the resistivity must not exceed the activation energy for conduction must be lower than 0. 55 eV, the thermal conductivity of the material must be higher than 2mW/cm, the softening temperature must be higher than 410 K, the crystallisation temperature must be situated in the range 490 520 K and the Young modulus must exceed Moreover, at the transition the residual crystallisation must be absent. Himinets [26] has shown that the above presented requirements are satisfied most thoroughly by the glassy compositions based on doped at a concentration of 1 12 at. % by chemical elements from the group Ga, In, Ge and Si and at a maximum concentration of 4 8 at. % by metallic elements from the group Cu, Ag and Ni. The most important physico-chemical parameters of several glasses recommended to

be used for the threshold switching devices are given in table 4. 1.

New memory devices have been produced [27] taking the advantage of the materials advances achieved in the development of current rewritable optical memory disks.

Through the use of the new congruently-crystallising ternary chalcogenide alloy materials and new device designs, the new phase-change memory devices provide six orders of magnitude faster programming time (less than 10 nanoseconds) than the original RMM

devices. The new devices have demonstrated stable long-term data storage and an endurance in excess of operations with low voltage programming capability. The basic device structure of the phase-change semiconductor memory cell, shown in figure 4. 3, is very simple. Essentially, it is comprised of upper and lower metallic contacts separated by a dielectric spacer containing a region of chalcogenide alloy material whose lateral dimensions are defined by a circular via etched in the spacer. The chalcogenide alloy layer is typically 50 nm thick and the diameter of the via that defines the active pore of the device ranges from 100 nm to 1000 nm.

CHAPTER 4

299

Figure 4.3. New chalcogenide memory cell structure (after [11]).

Memory devices are fabricated by RF magnetron sputtering of the dielectric, chalcogenide semiconductor and metal contact layers using conventional photolithography and ion etching processes. These devices can be incorporated in a wide range of integrated circuits ranging from simple memory to microprocessors and microcontrollers. 4.1.2. HYBRID AMORPHOUS/CRYSTALLINE TRANSISTORS Petersen and Adler [28] have developed electronic devices based on chalcogenide glasses with switching properties. A typical device (transistor) consists of an amorphous film of composition (emitter) deposited on crystalline Si or Ge diode (Fig. 4. 4).

Figure 4. 4. Sketch of chalcogenide-glass/p-Si/n-Si hybrid transistor structure (after [28]).

300

APPLICATIONS

Commercial silicon and germanium wafers (n/n and p/p ) were used as substrates. Aluminium contacts were realised by evaporation on the base and collector areas and the glass-emitter was deposited by rf sputtering over the entire wafer after a brief sputter etching of the base surface. The molybdenum upper contact was then sputtered, followed

by aluminium evaporation. As long as the amorphous emitter in the hybrid transistor is in the off- state, it has a very high resistance and the maximum emitter current level is at a base-emitter voltage of 20 V (just below the threshold). Since the off-state carrier concentration in the chalcogenide glass is very low compared to the concentration in the base extremely small common-emitter gains are expected. A typical plot of small-signal gain as a function of the collector current, for a glass/p-Si/n-Si device is shown in figure 4. 5.

Figure 4.5. Small-signal current gains as a function of collector current for the low-gain states of chalcogenide-glass/p-Si/n-Si and chalcogenide-glass/n-Si/p-Si transistors (after [28]).

Off-state gains are smaller in the chalcogenide glass/p-Si/n-Si configuration than in

the chalcogenide-glass/n-Si/p-Si devices, since hole conduction predominates in the off-state of the glass and it is much easier to inject holes than electrons into the glass in the off state. This reverse hole injection reduces the already low emitter efficiency even farther. Consequently, the emission currents are higher and the loss currents are lower in glass/n-Si/p-Si devices than in glass/p-Si/n-Si devices. Once the voltage between the base and emitter exceeds the threshold voltage of the amorphous material, the emitter switches into the on-state. When glass/p-Si/n-Si devices were operated with the glass in the off-state, the current gains are of the order of 0. 05. Upon switching the glass, the gain increases more

CHAPTER 4

301

than two orders of magnitude. Small-signal current gains as high as 15 have been observed in non-optimized configurations. The chalcogenide glasses are covalent semiconductors that exhibit thermally activated electrical conductivities analogous to intrinsic crystalline semiconductors. The electronic and optical properties of these glasses are almost always dominated by defects that are influenced by strong electron-lattice interactions. In tetrahedral semiconductors the Fermi level can be moved to near the band edges with the introduction of large number of “dopants”. In chalcogenide glasses the defects keep the Fermi level pinned near the midgap independent of impurity species. Although the chalcogenide glasses are employed in switching and memory devices they have been excluded from most device applications because of their inability to be doped. As shown by Hautala [29] the most likely way the chalcogenide glasses can be doped is by increasing the average coordination of the chalcogen atoms. If the average coordination of the chalcogen atoms can be increased to near tetrahedral values then the more “rigid” material may exhibit positive electron - electron correlation energy and doping might be possible. One structural model which predicts the transformation from a low coordinated system to a higher coordinated system is the formal valence shell model developed by Taylor et al. [30]. Electron spin resonance (ESR), light induced ESR, photodarkening, photoluminescence and conductivity measurements [29] on bulk Cu-As-S, Cu-As-Se and Ag-As-S glasses indicate that a transformation takes place from glasses with low chalcogen co-ordination number (Z 2) where doping is not possible to glasses with higher chalcogen co-ordination number where doping might be possible. Thin films were produced with the aim to introduce large metal concentration in the amorphous phase. Sputtered amorphous films of Cu-As-Se were produced with higher Cu concentrations than were possible for bulk glasses. These films included those situated at compositions where the Se atoms should be tetrahedrally co-ordinated. Conductivity and absorption data indicate that by incorporating oxygen and perhaps hydrogen, some deep gap states are removed or passivated and that a potential doping mechanism involving oxygen may exist. 4.1.3. THERMO-SWITCHES

The thermo-switches are devices, which operate as thermistors on the basis of the exponential dependence of the electrical conductivity on temperature. Ternary ovonic glasses with tellurium have been tried in preparing thin film switches: and The highest thermostability was obtained for thermo-switches based as This material allows for switching in the conducting state without the change of the electrical current parameters up to 10 mA for cycles of operation [31].

302

APPLICATIONS

4. 2. Xerography and X-ray Radiography

The copying of documents on normal paper by an electrophotographic dry process is well known and largely used. The procedure of reproduction of graphical documents (reprography) is usually called xerography (from the Greek xerox = dry and graphos = writing). The xerography is essentially based on the use of photoconducting chalcogenide materials, particularly amorphous selenium (a-Se). The electrophotographic process consists in the selective photo-discharge of an insulating photoconducting film electrostatically charged in dark and the formation on it of an electrostatic image of the document to be copied. The image is revealed with thermo-plastic powder (toner) whose particles are electrized and attracted by the unexposed zones of the charged film. By transferring the toner particles on paper and after fixation of the image by thermal treatment at the xerographic copy is ready. The photo-receivers are usually thin films of amorphous selenium and alloys based on selenium, deposited in vacuum on aluminium cylinders or plates.

In the absence of light the selenium films behave similarly to the insulator materials because they do not possess free charge carriers. The lack of carriers in bulk selenium facilitates the electrostatic charging of the film surface by corona-air discharge and the long-term stability of this charge, in dark. During illumination are generated electron-hole

pairs at the surface of the photoconducting film that split off under the action of the electric field created by the previously deposited electrostatic charge. If a light beam is projected on the film as an optical image, then an electrostatic image will be formed by selective discharge of the film. As mentioned above, the classical photoconducting material used in reprography is amorphous selenium. In this material the electrical charges can be trapped and a residual potential will appear. The traps have been ascribed to the specific defects in the material: the valence alternation pairs (VAP). The defects are pairs of atoms: a selenium atom with triple valence that is positively charged and a selenium atom with single valence that is negatively charged We mention that the normal state of the divalent selenium atom is in the notation of Kastner [32]. If an atom in the state retains one electron then it becomes neutral in non-equilibrium excited state If the atom in the state retains a hole in transit through the selenium film an other neutral nonequilibrium state is formed, The defects and are intrinsic defects in thermal equilibrium with the lattice. The defects and are non-equilibrium defects that relax and return to the initial defect states and by releasing the retained charge carriers. By the above-described mechanism the charge carriers are trapped or released and thus the residual potential is controlled. A reversible fatigue of the material, characterised by the increase of the dark current and of the residual potential, was observed during many operating cycles in the automatic xerox-machines. After switching-off the apparatus and long time storage of the photoreceiver in darkness, the residual potential and the dark current will be recovered.

CHAPTER 4

303

The increase of the residual potential and of the dark current during the operation of the Xerox machine was explained by the formation of non-equilibrium defects in the selenium film. These defects arise, as discussed above, from the interaction of the charge carriers with the non-crystalline network that induces the breaking and reforming of some distorted bonds between selenium atoms situated in chain or ring configurations. Thus, during long time operation of the photoconducting film a new metastable state of the material is reached. If the perturbing process is interrupted, then the transition to the initial state occurs by thermal activation. The density of local defects and in the photoconducting film depends on the material history, on the maturation time and, especially, on the nature and concentration of the impurities. If selenium is doped by arsenic the arsenic atom will link three selenium chain ends thus forming a bridge very similar to that encountered in the case of defect state. For low concentration of arsenic impurity the number of free chain ends diminishes and, as a consequence, the residual potential and the dark current in the selenium films will decrease. For high concentration of arsenic impurity the selenium will fix a large amount of arsenic atoms with unsatisfied valence requirements that behave as traps and will catch easily the positive charge in transit. If the impurity concentration exceeds 5 at. % As the film will be inappropriate to be used in electrophotography because all the transit charge will be retained and therefore no more discharge will be possible. In the case of doping by bromide it must be observed that Br atom has one electron more than selenium atom and will play the role of chain terminator defect). Since the state is an excited state, then it can easily lose an electron. As a consequence, the dark current in the selenium film must increase with the Br concentration. For concentrations higher than 40 ppm the photoconducting films will stop to operate due to the rapid decrease of the positive charges that are important for dark sensitivization [33]. In the case of doping by tellurium it must be observed that tellurium atoms introduced within the selenium chains will form Te-Se bonds that are weaker than Se-Se bonds. Thus the distorted Se-Te bonds will be easily broken and thus will appear local trapping states. Due to the large amplification of the local defects the doping by tellurium is not advantageous. The amorphous selenium has been used from early times in xerography and its limitations were evidenced: i) relatively low sensitivity ii) instability of the material that leads to the formation of defects in films during preparation and to the change in their properties during long time storage. The operating parameters of the selenium films were improved by doping with various elements. Thus, in the selenium films doped by 5 10 at. % Te there was obtained an integral sensitivity with more than one order of magnitude higher than in pure amorphous selenium. In the spectral range 300 500 nm the spectral sensitivity reaches Unfortunately, these films exhibit low thermal stability and rapid decrease of the dark voltage. The crystallisation of the film can occur even at room temperature. The doping by phosphorus increases the microhardness, the softening temperature and the resistivity. For phosphorus concentrations up to 5 at. % the thermal stability of

304

APPLICATIONS

the alloy increases, the rate of decrease of the dark voltage is monotonously reduced, the integral photosensitivity is significantly lowered and the residual potential is raised [34]. The Se: Te: As films exhibit fully reversible light exposure and thermal history effects. These effects are metastable decaying away during isothermal dark annealing at a rate, which decreases dramatically with where is the annealing temperature [35]. In the last years some ternary chalcogenide materials have been tried. In the system Se-Te-P the main feature is related to the fact that, independently of the phosphorous concentration (up to at. % P) and for high amount of tellurium in the alloy, the spectral distribution range of the photosensitivity fits to that of the binary Se-Te alloy and covers all the visible spectrum with high absolute values (Fig. 4. 6).

Figure 4.6. The spectral distribution of the photosensitivity in the positively charged electrophotographic films based on selenium.

The decrease of the tellurium concentration down to less than 5 at. % or lower leads to a sharp decrease in photosensitivity. The range of the spectral distribution approaches that of selenium films. The increase of the plastic properties of the vitreous selenium was obtained by doping with sulphur and iodine. In these cases is lowered and the expansion coefficient is raised. The ternary compositions Se-S-I with up to 5 at. % sulphur and up to 2 at. % iodine (I) show high stability against crystallisation. Nevertheless, the reprography based on these materials is characterised by low photosensitivity. Montrimas and Lukatskaia [34] used a combination of two layers based on Se-Te-P and Se-S-I amorphous materials. They succeeded to get a new improved xerographic film

CHAPTER 4

305

with high elasticity, high photosensitivity, and stability against the external factors and, especially, lacking the residual potential even after long time cycling (Fig. 4. 7).

Figure4.7. The kinetics of the photo-discharge of the positively charged electrophotographic films 5 based on - bottom layer top layer 6 - film No. 5 after 100 cycles of operation.

Other types of films were based on arsenic chalcogenides: and The films show the maximum photosensitivity, the highest limit of the charging potential and a relatively low rate of the dark discharge. As a result of pre-illumination and under the action of high electric fields, within the films are formed electron-hole pairs and their excitation leads to the increase of the dark discharge potential by a factor of 10 15. The photosensitivity of the films with the composition approaching exceeds 10 15 times that of the most qualitative selenium films and increases with the substrate temperature up to 140 150 °C (Fig. 4. 8). It is possible to get considerable increase of film photosensitivity and broad spectral range by doping with Sb, Bi, Te, and Tl. The integral photosensitivity of the film increases with the concentration of the doping element and the limit potential decreases while the speed of its decay in darkness increases. Therefore, according to the electrical requirements, the doping concentration must not exceed 10 at. % or 5 at. % or 2 at. % (because at these concentrations the decay time of the positive potential decreases down to s). The films with these compositions exhibit higher sensitivity (as compared to in the whole visible spectrum and the long wavelength boundary is extended towards the near IR (Fig. 4. 8).

306

APPLICATIONS

The most panchromatic and the most sensitive films for monochromatic light belong to the system The excellent spectral properties of these new films give the possibility to develop materials for high quality colour xerography.

Figure 4.8. The spectral distribution of the photosensitivity of positively charged electrophotographic films.

It is worthwhile of note that Chernov [36] used As-S-Br glasses with maximum transparency in the range of photoactive absorption in order to get high bipolar sensitivity layers on transparent substrates with good properties of adherence and barrier layer. Buzdugan et al. [37] have shown that glasses in the system with high content exhibit high integral photosensitivity, large range of spectral sensitivity and high thermal stability. Recently, conceptually new materials for electrophotography have been designed and prepared. In 1985 Rianeli and Kaplinskaia [38] have demonstrated that it is possible to get by thermal evaporation of Se and Te, special structures later called varigap structures (from variable gap). These structures are characterised by the change of the width of the optical gap along the film thickness with 0. 6 0. 8 eV for thickness of These structures contain an additional organic layer.

CHAPTER 4

307

Later, Rianeli et al. [39,40] have shown experimentally that the existence of the Eg gradient leads to the appearance in such structures of quasi-electric internal fields of 105 V/cm. The research related to the influence of the internal field on the monopolar photoinjection of the charge carrier in the organic layer has shown that the internal field has influence on the hole injection only. The efficiency of the photo-injection of the electrons practically does not depend on the orientation of the internal field. The photosensitivity of the multilayer xerographic photo-receivers is also strongly modified when the internal quasi-electric fields are present. The injection field increases the xerographic sensitivity in weak fields. If the internal field is directed from the metal towards the organic semiconductor, then, even in the absence of the external electric field, in the structure metal - varigap semiconductor - organic semiconductor there is observed an effective photo-injection of holes into the organic semiconductor. On the other hand, a strong decrease of photo-injection is observed in relatively large external fields, if the structural field is oriented from organic semiconductor to metal. The creation of varigap and, implicitly, the structural fields in multilayer films, allow solving simultaneously three problems. The first one is the broadening of the spectral range of the photosensitivity as compared to the homogeneous structures. The second one is the diminution of the injection from the metal into semiconductor on the account of the blocking contact on the side with large gap of the varigap structure. The third one is the increase of the photosensitivity in weak fields on the account of the increase of the quantum yield of photo-generation and photo-injection when the recombination is diminished in the range of generation. The chalcogenide materials are materials suitable to replace conventional semiconductor crystals in X-ray detection. Recently, there were devised integrated radiographic systems using the capacitance readout of X-ray induced charge images on amorphous selenium layers [41]. In these X-ray sensitive systems the image sequences are taken at short time intervals with one and the same image receptor. The electronic recording of radiographic images extended the range of sensitivity of a-Se layers. In view of the thermal stability of the layers, a-Se: As alloys are used in practice, with an As concentration of typically 0. 5 at. % [42]. The X-ray sensitivity of hydrogenated a-Se: As photoreceptors for electro-radiography, which depends on the thickness of the amorphous layer and on the applied electric field, was discussed by Aiyah et al. [43]. The most important problem to be solved is the enhanced dark decay and the reduced X-ray sensitivity (fatigue) upon irradiation. These are important for the minimisation of the observed memory effects in X-ray detector applications based on amorphous chalcogenide films. Researches on operating a-Se photoconductors at high electric fields have revealed the feasibility of X-ray imaging devices of new performance. There are two reasons for operating at high fields. First, the electron-hole pair creation energy decreases with electric field, which leads to an improved X-ray-to-charge conversion. Second, a-Se at sufficiently high fields (in excess of 80 exhibits avalanche multiplication. Although the avalanche effect in a-Se has made possible to have high-gain avalanche rushing photoconductors (HARP), its physics in amorphous semiconductor remains poorly understood.

308

APPLICATIONS

Recent developments in flat-panel display technologies have enabled flat-panel detector systems to capture the whole X-ray image and to provide readout in digital form. A flat panel X-ray image detector based on selenium has been envisaged [44, 45]. The flat panel detector has potential for use in fluoroscopy, that is, real-time interactive X-ray imaging. In fluoroscopy, a video image on a monitor enables the radiologist to see a moving X-ray picture of the inside of the human body. This capability facilitates the diagnosis of function and guides such therapies as the use of baloon catheter to reopen blocked coronary arteries. With these new developments, teleradiology and telemedicine will be possible. A “filmless” radiology department has been implemented in the Montreal General Hospital in Canada [46]. 4. 3. Holography

The holography is a two step process for getting three-dimensional images of the objects. In the first step the object is illuminated by coherent light and the interference pattern of the reflected (or transmitted) light is recorded on a photosensitive material. To this purpose the light beam that falls on the object is split: part of the beam is directed onto the photosensitive material and the other part is firstly reflected by the object and thereafter is directed onto the same photosensitive plate. The superposition of the two beams determines the appearance of the interference pattern of the hologram. Thus, the hologram is inscribed in the photosensitive material by an interference image. In the second step of the process, i. e. the reconstitution of the image, the inscribed hologram is illuminated by a reference beam that will be scattered as in the case of the real object giving rise to a spatial image of the object: the “optical copy”. The recording of the hologram can be performed by amplitude or by phase. In the case of holograms inscribed by amplitude, the basic mechanism for hologram inscription is the change of the absorption coefficient of the material. For the hologram inscribed by phase, the recording is carried out by the local change of the refractive index. The general principle of the hologram recording is shown in figure 4. 9. The diffraction efficiency is a parameter that characterises the luminosity of the holographic image and is usually defined as the ratio of the square of the amplitude of the diffracted wave, to the square of the amplitude of the reference wave, that penetrates the medium with the hologram: (4.3)

A high efficiency is obtained in the case of phase inscribed holograms. Most research in chalcogenide based holography was performed on amorphous films of arsenic sulphide and arsenic selenide. Real-time holographic information storage is one the most important of all applications of amorphous films [43]. Diffraction efficiency exceeding 80 %, high spatial resolution of 9700 lines/mm reached recently [47], the simplicity of the direct

CHAPTER 4

309

recording process for practically permanent holograms which need neither fixing nor development, reversibility and the availability of large samples are specific features. Holographic recording in films is based on photo-induced structural changes, a unique phenomenon, which is found only in non-crystalline chalcogenides.

Figure 4.9. The general scheme of the holographic recording.

The properties of afilms differ depending on whether they are annealed. For instance, the highest diffraction efficiency can be obtained in as-evaporated non-annealed films whereas annealed films are fully reversible [48]. The changes are due to the photoinduced breaking of the As-As bonds by band-gap light followed by the formation of phonon-assisted As-S bonds. Because holograms are sensitive to the temperature and light, at room temperature they should restored in the dark. In the were inscribed holographic gratings with the maximum relief component of the order of 10-20% depending on the spatial frequency, at a relief depth of [49]. Many theoretical and practical results have been obtained on sulphur rich As-S compositions [50]. Particularly, the anisotropy effects induced by polarised light [51, 52] and the amplification of the efficiency of the hologram inscription on As-S glasses is mostly determined by the composition and the structure of these materials. On the other hand, the dependence of the diffraction efficiency of elementary holograms on the temperature of the plates gives the possibility to record the holograms at high tempe-

310

APPLICATIONS

ratures close to the softening point of the material where the sensitivity and the diffraction efficiency are the highest. The procedure enables to perform non-destroying reading down to the room temperature. Recordings of holograms were carried out in germanium-telluride, seleniumtelluride and also chalcogen-metal systems. The choice of the compositions based on Se-Te and Se-As is justified by the particular character of these systems with a range of possible transformations from the chain-like and ring-like structures, characteristic to amorphous selenium, towards a two-dimensional network that occurs at a not too high amount of Te(As) introduced in Se (10 15 at. %) [53, 54]. As opposed to Se-Te and Se-Te systems, in the S-As system there was exploited the effect of separation of the structural elements in the presence of a halogen and the preferred compositions were those with iodine The measurements were carried out on films of thickness prepared by vacuum evaporation on usual glass substrates. The recording of the phase-amplitude grating of the elementary hologram was performed in two-beam configuration by the use of a He-Ne laser radiation for As-Se films and Argon laser radiation for As-S-I films at average power densities of for illumination in the plane of the sample. The diffraction efficiency has been determined by reading the hologram with pulses or with continuous light of the same laser used for writing and for As-Se films and with He-Ne laser radiation for As-S-I films. An efficiency of 0. 1 0. 6 % in a-Se and 4 8 % in As-Se and As-S-I has been found. Heating the active medium above erases the holographic recording. One of the most important feature related to the hologram writing in some materials is a limited increase of the refractive index followed by a slow relaxation that occurs during the reading of the hologram by pulses in attenuated light. The effect appears both in darkness or under weak illumination after the holographic exposure and it has been observed in amorphous selenium, in Se-As films with up to 15 % As in films and in films. In Se-As films there were observed various relaxation rates in darkness. The effect seems to be absent in and in AsSe films at room or at higher temperatures [55]. The effect described above is probably related to the formation in the layered structure of these materials of new configuration elements with the relaxation time proper to the intermediary states. The dominant role is probably played by the orientation characteristics in the illuminated zone. An analogous effect appears during uniform illumination. Nevertheless, in this case the rate of the process is higher. This fact was explained by the significant influence of the electrical fields that appear in the amorphous network with anisotropic structural elements (layers, chains). Dembovski [56] discussed this problem in connection with the behaviour of a-Se. If this is the case, then the transition probability from the intermediary configuration to the metastable oriented configurations is increased and the mobility of the chain-like configurations after reorientation decreases. This behaviour can be stimulated by introducing small amounts of As and Te in a-Se, that create structural ramifications and facilitate the formation of defects where the electron-hole pair

CHAPTER 4

311

excitation will be located. In the case of the system As-S it is necessary either to introduce a halogen element or to shift the compositional ratio towards higher sulphur concentrations, in order to give more mobility to the chain fragments. Therefore it is important to get oriented nuclei (in this case chain configuration nuclei) during the recording of the hologram and to ensure their self-activated growth under the action of the orienting fields,

on the account of the intermediate states, enough stable, formed as a result of the excitation of the material by the polarised light. This conclusion is supported by the variation of the diffraction efficiency with the composition (Table 4. 2).

In order to reach high diffraction efficiency it is recommended to use the phaserecording method. Nevertheless, this type of recording limits drastically the range of amorphous chalcogenide compositions to be used in holography. Table 4. 2 shows the main chalcogenide alloys used in the holographic recording together with their basic characteristics related to this application. It is remarkable that the most holographic writings have been performed in arsenic sulphides and selenides. In these materials the phase recording has been successfully applied. In other chalcogenides the holography was realised by amplitude recording. The phase recording is related to the restructuring of the network at the atomic scale under the action of the light. This reorganisation is effective only for the first alloys in Table 4. 2. The doping of As-S alloys by germanium strongly diminishes the maximum of the diffraction efficiency. The holographic recording has been successfully performed on metal - glassy chalcogenide structure by exploiting the effect of metal diffusion from surface to bulk. The recording is obviously irreversible but the diffraction efficiency and the photosensitivity reaches very high values. Thus, % in the case of structures [57]. As regarding the resolution of the holograms inscribed in glassy chalcogenide films, Keneman [58] reported for films 2860 lines/mm, Ohmachi and Igo [59] reported in

As-S films 2500 lines/mm and Sakuma et al. [60] reported a value of only 1500 lines/mm in the system chalcogenide-metal. All these values speak in favour of the possibility to write high quality holograms in amorphous chalcogenide films. Mandrosov et al. [61]

312

APPLICATIONS

have estimated the maximum density of lines to be written in thick chalcogenide films: lines/mm.

In the study of the processes of holographic recording of information and reversibility of the registration media Gurevich et al. [70] have shown that the diffraction efficiency decreases after the first recording-erase cycle and remains constant for the other cycles. This effect is explained by the change of the chalcogenide structure, stimulated by laser irradiation, and is strongly dependent on the glass composition. The kinetics of the diffraction efficiency on fresh As-Se-Sn films and the influence of glass composition and film storage conditions have been investigated by Buzdugan et al. [71]. Freshly prepared films are characterised by a maximum magnitude of diffraction efficiency (up to 18 %), which is achieved at a rather short recording time (t = 1. 0 to 15.0 s). For recording media kept for a long time in the darkness or in the light, the type of kinetics and their parameters change, in particular the registration time, necessary to achieve the (less) maximum magnitude or diffraction efficiency increases and the sensitivity of the media decreases.

CHAPTER 4

313

Contrary to this behaviour, specific to As-Se films, in tin containing fresh films takes place an increase of the diffraction efficiency and the maximum of the diffraction efficiency is reached after a longer time. For long time keeping in the dark an significant decrease of diffraction efficiency occurs and, at the same time, the sensitivity of the media increases as a result of the change of recording kinetics. In freshly prepared films the sensitivity decreases. Experimental data show that the optimum thickness for the recording film is situated in the range and the Sn concentrations which prevent the decrease of the diffraction efficiency is 1 3 at. %. The process of thermal erasure of information both in the case of fresh and annealed films shows that at a temperature of 60 °C the diffraction efficiency quickly decreases by 70 %, and at longer annealing the information is completely erased. These features are explained by the formation of tetrahedral units in the amorphous films. As a result microhardness and melting temperature increase. These factors considerably influence the real time recording and degradation processes in As-Se-Sn films. If these films are stored in dark for a long time, then the initial non-equilibrium structure relax and new structurochemical units may appear, thus determining the change of the diffraction efficiency. The authors conclude that the Sn-doped As-Se films are more promising media for holography than the alloys without tin. It is necessary to point out that the holographic recording in amorphous chalcogenide films possesses as unique property, the self-enhancement of holograms. An increase of diffraction efficiency after holographic recording over time without any special treatment. The changes of diffraction efficiency in amorphous were studied as a function of ageing time, initial diffraction efficiency, recording light intensity, temperatures and type of substrate material [72]. The dark self-enhancement process of diffraction efficiency of holographic gratings can be accelerated either by heating at temperatures up to 100 °C or additional illumination after holographic recording [73]. For the holographic gratings with the initial diffraction efficiency of the enhancement factor up to has been obtained [74]. The self-enhancement phenomenon has been explained by the presence of internal mechanical stresses in amorphous films, arisen during film preparation by thermal deposition in vacuum. Despite the conclusions on the impossibility of stable optical recording in a-Se at room temperature [75], stable holographic gratings were recorded on a-Se layers at 294 K [76]. Kikineshi et al. [77] discovered that the curve of diffraction efficiency versus the recording time in a-Se films has two distinct parts. The first component of the optical recording is rather small and unstable at room temperature because of the vicinity of the selenium softening temperature. Further recording at longer time allows to reach much higher efficiency of the hologram 1. 3 %), which is very stable at 294 K. The surface relief of the hologram, exposed up to s, can be completely erased by annealing at T 310 K, and the recording may be repeated several times until the crystallisation effects will not restrict the relaxation and destroy the surface quality. Photo-crystallisation is distinct and rapid in the repeated third-fifth cycle of optical recording - thermal erasing. An important discovery was the reversible, single step surface deformation in a-Se films

314

APPLICATIONS

during hologram recording at temperatures close to Tg. The cause of this effect seems to be the photo-induced local ordering and film contraction. Ovshinsky and Close [78] pointed out the relation between the resolution and the value of the exposure. Various amorphous films based on Se or Te allow to get a resolution of 100 lines/mm for an exposure of and 500 lines/mm for an exposure of They have shown that it is possible to improve the quality of the inscribed image by thermal annealing of the films. A different principle for the hologram recording has been successfully used in the special heterostructures with chalcogenide materials. The application of an electric field (constant or variable field) to a heterostructure metal-chalcogenide-metal having the thickness of the chalcogenide layer greater than leads to the appearance of mechanical deformations as prominences and dips. The size and the density of these deformations depend on the technology of preparation of the heterostructure, on the thickness of the chalcogenide film and on the upper electrode. Moreover, the frequency and the magnitude of the applied field are important parameters. The main cause of the deformation effect seems to be the heating of the structure up the influence of the electromotive forces. If the current density exceeds then a high amount of heat is produced and this is enough for softening the chalcogenide film. The surface deformations appear in those parts of the layer where the temperature reaches because there the viscosity forces are weaker. Two aspects are important. Firstly, the deformation is a function of the field intensity and, therefore, in certain limits, the deformations are reversible. This effect allows for the realisation of the modulation of the light beam reflected by the system [79] because the deformations are produced on those parts of the film with largest electrical conductivity. In the presence of the internal photoelectric effect the conductivity is controlled by the light beam and, thus appears the possibility to control the light by the mechanical deformations. Secondly, in the irradiated zones the current density is higher and, therefore, the film heating is stronger and the mechanical deformations are greater [80]. In the above-discussed case we are dealing with the residual deformations that are maintained for a long time after switching off the field: they are therefore irreversible deformations. In the heterostructures for hologram recording the current density necessary for reaching the softening point of the material is of the order of and can be obtained at high electric fields, high illumination and enough large temperatures. The most appropriate heterostructures for holography are those with crystalline chalcogenide films. In such structures the current can be raised on the account of the injection of the charge carriers across the heterostructure. By application of a field of negative polarity on chalcogenide film (forward bias), the illuminated zones will acquire electro-induced deformations of the surface of the chalcogenide film [81, 82]. In comparison with the monolayer structure, the use of heterostructures based on chalcogenides allows for growing significantly the contrast of the optically recorded image and to make more efficient the photographic process by the appropriate choice of the strength of the applied field and of the irradiation intensity.

CHAPTER 4

315

The effect related to the surface deformations controlled by the electric field and by the light intensity was used for the recording of holographic information. The hologram recording is performed in the two-ray scheme of the Fourier holography without lenses. The read-out of the holograms was carried out by reflection during the holographic writing on the wavelength of the He-Ne laser by measuring the changes in time of the first diffraction maximum. During illumination of the surface of the sample with two coherent light beams from the chalcogenide side and by simultaneous application of an electric field and negative polarity on chalcogenide, the film surface is strongly deformed along the illuminated interference bands, thus forming a spatial (threedimensional) diffraction grid based on relief changes. The holographic parameters depend essentially on the chemical composition of the chalcogenide film and on the recording regime that determines the physical processes in bulk and at the surfaces in the case of more complex configurations. In the recording systems were used heterostructures with various thickness ratios New chalcogenide hetero-structures with p-Inp and p-GaP was investigated and new chalcogenide compositions (e.g. As-S-Ge) were tried [67]. In heterostructure, the sensitivity of the recording structure and the diffraction efficiency of elementary holograms depend on the strength and polarity of the electric field, light intensity, electrode material and density of the recorded information. An increase of the sensitivity and diffraction efficiency was demonstrated for the case when the applied voltage pulses is reversed [83]. 4. 4. Photo – recording 4. 4. 1. PHOTO-THERMAL RECORDING

The modern recording on compact disks is based on the direct irreversible process, i. e. on the formation of discrete zones of very small area printed in the material, whose optical properties are irreversibly modified. These special zones are domains of transparency at variance with the material matrix, or hollows, prominences, dips. Special recording procedures are used by taking advantages of the formation of bubbles with gases or vapours embedded into the material or the local ablation of the material. The recording processes based on the above mechanisms are called photo-thermal recording processes. In many cases the phase transitions induced by light have a thermal character too. The photo-thermal recording of optical information is carried out in the following steps: 1) light absorption 2) material heating 3) melting, evaporation, local ablation of the active layer or crystallisation. The formation of holes in the material is energetically favourable in the melt where the surface tension plays the leading role. For writing and reading are used various lasers. The increase of the efficiency of the photo-induced reactions is obtained in the pulsed regime of irradiation. The average power for the information recording is 5-10 mW and the recording time (pulse length) is

316

APPLICATIONS

10 100 ns. Transmission or reflection of a laser beam through the recording material carries out the reading. In order to use the reflected beam, a special reflecting layer is deposited onto the memory disk (Fig. 4. 10).

Figure 4.10. Multilayer systems for information recording a. absorbing layer (1) without compensation of reflection b. poorly absorptive layer (1) used as transparent layer c. transparent layer (3) for the high absorption layer (2) and for the layer sensitive to light (1) d. transparent layer (3) for the recording system with bubbles; metallic reflector (2), volatile material (4). The thickness of the transparent layer is

The light signal that impinges onto the photoelectric receiver is transformed into an electrical signal and then enters into the computing system or in the electronic system of the device for the audio-visual reproduction. In order to diminish the reflection of the photosensitive film and for increasing the sensitivity during the optical recording, transparent layers are currently used. The cross-sectional view of an optical disk is shown in figure 4. 11 [84].

Figure 4.11. The optical disk with memory. 1- recording medium 2- air layer 3- coatings 4- protection layer/ substrate

CHAPTER 4

317

The schematic view of the way the optical discrete recording is carried out is shown in the figure 4.12.

Figure 4.12. The principle of the discrete optical recording. a. general view: 1-semiconductor laser 2-focusing system 3-optical disk with memory 4-photodetectors b. structure of the point by point recording: 1- initial spot due to focused laser radiation; 2 - inscribed point (small area)

The amorphous semiconductors and, particularly, glassy chalcogenides are used, beside metallic films and organic compounds, as photosensitive materials for photothermal recording. In order to reduce the energetic losses during the light reflection, of great importance are the interference nodes for reflection compensation. The compensation of the reflection and the improvement of the parameters of the photosensitive films are made with the multilayer configurations. The multilayers comprise: a) the substrate (glass, polymethylmetacrylate/PMMA/) b) photosensitive layer c) transparent layer d) reflector (only in the case of the recording with reflected signal) e) volatile layer (for bubble inscription) f) additional layers for increasing the mechanical strength of the whole sandwich and for the decrease of the diffusion rate (barrier layers) g) protection layer. Sometimes the same layer accomplishes several functions. The layer 3 from figure 4.10, that is a transparent layer, plays simultaneously the role of protection layer. In figure 4.10 the layer 4 (volatile layer) raises the mechanical strength of the absorptive layer 1. The substrate is the mechanical carrier of the multilayer system and its thickness is higher than mm. The thickness of the absorption layer (the photosensitive layer or the layer of evaporable material) is chosen as a function of the magnitude of the absorption coefficient. For strongly absorptive materials the penetration depth for the

318

APPLICATIONS

light is lower than the wavelength of the radiation and this allows for using photosensitive films as thick as The thickness of the transparent layer is determined by the interference conditions in thin films (the minimum thickness at nm and n = 1.5 is The reflection layer ensures an effective reflection at a thickness greater than 50 nm. The coating layers have usually some tens of micrometers in thickness. Every layer has a specific structure, according to the purpose: - the protection layer protects the photosensitive layer against the atmospheric factors in general and against dust in particular ; the optical system is protected against the vapours of material produced in the multilayer device. This layer is from siliconic resins, etc. and has the thickness of - the transparent layer increases the photosensitivity of the device with an order of magnitude. It is situated below the photosensitive layer and its composition is chosen according to the active material. Thus, for tellurium photosensitive layer, is preferred to PMMA. - the intermediary layers are used for increasing the mechanical strength of the multilayer system. To ensure a good mechanical quality, the neighbouring layers must be isomorphous. Usually, in multilayer systems are used and other materials. - the barrier layers are used for the insulation of the elements in the system. They reduce the diffusion phenomena between different layers and give stability to the whole system. Among the most interesting and most investigated photo-thermal materials are the amorphous chalcogenides: The material recently studied, is a good material for photo-thermal reversible recordings. The recording by photo-thermal procedure in Te films of thickness deposited on PMMA substrate, has been studied by Chen et al. [85]. The virgin Te film was crystalline. Under the action of the pulses of an Argon laser (average power: 8 mW, illumination from the substrate side) a phase transition from crystalline tellurium to amorphous tellurium takes place. The contrast coefficient (K) is determined by the relation:

where and are the intensities of the transmitted light in the recording point and of the transmitted light in the non-illuminated part of the material, respectively. The process is

reversible but, from the practical point of view does not present any interest due to the low contrast coefficient The formation of an open zone starts only at powers of P > 4 mW. For powers of the laser radiation well above 8 mW the phase transition is blocked and, as a consequence of the melting of Te, holes appear whose diameters increase with the power of the incident light. The contrast coefficient for powers higher than the threshold value (8 mW) quickly saturates and for 14 mW reaches the value 0.65.

CHAPTER 4

319

Usually the photo-thermal reversible recording at room temperature is not possible in the amorphous films of Ge-Te (8 mol.% Ge) deposited on PMMA substrate. Nevertheless, when the threshold power is exceeded, then open zones do appear. For GeTe the threshold is sharper than for Te and the saturation (for K = 0.68) appears suddenly in the neighbourhood of the threshold power. In the films of composition Te-Bi (6 at%) the recording processes are similar to those in the tellurium films. The recording threshold and the contrast coefficient are the same. The substrate strongly influences the threshold power during the photo-thermal recording. Low thermal conductivity of the substrate leads to low energy losses by heat dissipation. The threshold power strongly decreases if the system is situated on and/or selenium substrate. Among the elemental photosensitive films, the tellurium films exhibit the most favourable parameters for information recording. The thermodynamic characteristics of the chalcogens, important for the photothermal recording are given in table 4.4.

Dual -layer recording method combined with the land and groove recording method and the mark-edge recording method was recently used for optical disks of the new generation (rewritable versatile disks DVD-RAM). To this purpose a new composition based on Te-O-Pd was optimised. The best material was found [86]. The asdeposited material is amorphous. Upon crystallisation during recording, the reflectivity increases and transmittance decreases. The maximum reflectivity difference between the amorphous and crystalline state occurs for a thickness of 55 nm. The recording bit length is 0.41 using the land and groove recording method. Disks with 5.2 GB recording capacity on one side have been produced. The recording by crystallisation is an important way for photo – recording. Stoichiometric GeTe amorphous films can be crystallised with laser pulses of 30 ns duration or less and this effect can be used for the recording media in the direct overwrite mode. The addition of antimony extends the performance of GeTe to a wider

320

APPLICATIONS

compositional range, improving manufacturability [87]. The control of crystallisation speed is of crucial importance for the development of high speed, high density optical recording materials. was used as recording layer. Sb doping of only 5 % raises the crystallisation time from several seconds to 165 °C for the stoichiometric material to more than 5 minutes or the doped material [88]. It is confirmed that recording sensitivity,

erasability and overwrite cycles are improved by doping with nitrogen the Ge-Sb-Te recording layer [89]. The action of nitrogen is connected with the suppression of the micro-material flow. Ohta [90] explained this effect by the formation of some nitrogen compounds in the recording layer. 4.4.2. MATERIALS FOR RECORDING WITH BUBBLES.

Feinleib et al. [91] took into consideration the following mechanism for recording with bubbles. In the chalcogenide films the softening temperatures are situated above the ambient temperature with some tens of degrees and the temperatures for vapour formation

are with some hundreds of degrees above That is why, at high illumination intensity the materials melt rapidly and evaporate. The vapour pressure increases gradually during the light absorption and vapour bubbles start to form. If the illumination stops, then the quenching of the material around the vapour bubbles occurs. Due to the vapour condensation the internal space of the bubble becomes a free space. The effect depends on the magnitude of the refractive index of the material that must be situated around the value 3.2 for a good efficiency of the absorbed light. As an example, the optical recording was performed in the amorphous alloy [92]. To get bubbles of diameter it is necessary an energy of 800 nJ. This energy was provided by a powerful Kripton laser working in pulses of 200 mW with the duration of 4 The erasure of the inscribed information was performed by a pulse of 280 nJ A weaker pulse than that used for erasure effected the reading. It is mentioned that the wavelength of the laser radiation must be tuned to the range of spectral sensitivity of the material because the light must be not absorbed outside the borders of the chalcogenide film. The possibility to inscribe the information by bubbles with a good control was demonstrated for the case of the eutectic by Jecu et al. [93]. S-Se thick films have been irradiated by ruby laser pulses with the energy per pulse of ~1J. The image of the laser spot (dark zone) has a diameter of A pulse energy of 1.4 J has been necessary for bubble recording. Figure 4.13 shows the pulse image as recorded in the material. The high energy and high power density allowed for the production of compositional modifications as e.g. sulphur diffusion and partial evaporation. This is proved by the change in transparency of the material due to the shift of the absorption edge. The dark spots around the main laser spot (for films of thickness less than 10 are stable even during heating up to Then, around the secondary spots do appear transparent bubbles (Fig. 4.13 b). Thus, by fine control of the pulse power and by appropriate thermal annealing, round bubbles of uniform size can be obtained. The bubbles make the film locally

CHAPTER 4

Figure 4.13. Bubble formation in

321

thick films,

a. immediately after laser pulse impact b. after thermal annealing at

transparent and can be, therefore, used as information support. The range of the bubble diameter is The authors concluded that the optical memory with bubbles is possible to get as stable, non-volatile memory. The information storage has been estimated as 15000 While the crystallised material of the film can be easily amorphized by heating above the bubbles are thermo-stable [94]. 4.4.3. THERMO-PLASTIC RECORDING

The thermo-plastic materials are used for the transformation of the charged parts of an electrical image in a visible relief by deformation of the surface of the sample. The effect is used in reprography holography and in memory media. The thermo-plastic process is enough complicated and includes the surface charging, illumination, heating of the substrate up to and cooling. For the thermo-plastic recording experiments have been used chalcogenide glasses as and heterojunctions and alloys on their basis [95]. The substrate was a polyethylene terephtalate ribbons covered by Cr. By thermal evaporation in vacuum have been deposited thin films of vitreous arsenic chalcogenides on metallic films and onto these films has been deposited from solution (with toluol) the thermo-plastic film. As opposite to the classical electrophotographic process, the formation of the charged relief and its transformation in a visible image occur during the charging of the surface of the recording film. The parameters for optical recording in various materials are given in table 4.5. In order to improve the photographic characteristics of the thermo-plastic film it is necessary to maximise the photo-structural transformations in the material. Thus the ratio signal/noise will be raised and the half tones will be better reproduced.

322

APPLICATIONS

According to the change of the optical properties, the highest diffraction efficiency in the thermo-plastic films is observed for the wavelength of 625 nm. The efficiency of the image formation for this wavelength is lower than for shorter wavelengths. This indicates that the modification of the electrical properties of the chalcogenide glasses during illumination by light of shorter wavelengths occurs more strongly. The thermo-plastic substrates based on heterojunctions exhibit the highest photosensitivity The application of the thin film heterostructures in the systems for optical recording of information offers several advantages: high photosensitivity, large spectral sensitivity range and high dark resistance in comparison with the single layer materials [96]. Heterostructures from antimony sulphide and arsenic sulphide has been produced as photosensitive elements for the electrophotographic and photo-thermal carriers with high thermal stability [97]. These qualities were obtained on the account of the splitting in the functions among the layers of the structure, where the layer is blocking and passivating while the layer ensures the sensitivization of the structure by carrier injection. Varying the thickness of the layer can vary the characteristics of the heterostructure. They are determined not only by the sum of the bulk component of the properties, but, to a great extent depend on the injection and blocking of the charge carriers at the heterostructure boundaries which fact leads to the additional increase of the photosensitivity and of the dark resistance [98].

CHAPTER 4

323

4.4.4. MATERIALS FOR PHOTO- AND ELECTRON-BEAM RECORDING The reversible photosensitive materials

For the reversible recording of the optical information it is recommended the use the photo-induced transformations as e.g.: The photo-crystallisation mechanism has been investigated by Chaudhari et al. [99]. The existence of metastable states in the amorphous material gives the possibility to make a repeated recording after a previous erasure of the inscribed information. The transition from an amorphous state (type I) to another amorphous state (type II) and the reverse transformation are both possible but the last transformation has lower probability. The erasable optical recording is under development in various laboratories worldwide [100] and at ECD, Troy, and USA under the leadership of S.R. Ovshinsky. For erasable recording a reversible phase transition in a thin layer of an alloy of InSb and Te is used. The optical constants of the material depend on the structure of the film. It is possible to switch the layer selectively between the crystalline and amorphous state by using different laser powers. The original state has a fine crystalline structure. During writing, localised areas of the material are melted with the aid of a powerful laser pulse of a short duration The material then cools off so rapidly that the recrystallisation does not take place, leaving an amorphous area on the disk with a different reflectivity. For erasure, a laser pulse with a lower intensity (in order to ensure that the temperature of the layer remains just under the melting point) and slightly longer duration is used. This treatment causes the material to return to its original, thermodynamically more stable crystalline state. Wang et al. [101] have shown that the reflectance of the Te-In-Sb films of composition increases by about after the specific and complicated phase transition. The films exhibit a large absorption coefficient at nm and may constitute a suitable medium for new optical disks. The ability to modify the structure of thin films of chalcogenide alloys using light allows the construction of several types of imaging systems, most notably phase change optical memories. The benefits of phase change optical memories include performance, and, as a result of the inherent simplicity of the read and write processes, low cost. The key to achieving high speed is the design of an alloy where the crystallisation process involves diffusionless crystal growth in a system that does not phase segregate. These materials can be crystallised with laser pulses of 30 ns duration or less, and such high crystallisation speed allows the use of the recording media in the direct-overwrite mode. Ge-Sb-Te alloys show excellent characteristics for such application, including fast transition times and excellent stability [102]. Stoichiometric GeTe also shows good performance, but the addition of Sb extends the performance to a wider compositional range, improving manufacturability. The technique, which uses the photo-crystallisation effect, is called ovography (after the name of the inventor: Stanford R. Ovshinsky). In the ovographic record the thickness of a recorded point is and this size gives a density of (for 3 mm distance in the row and 6 distance between the rows). The rates of for the

324

APPLICATIONS

recording of information are possible when the phase transition from polycrystalline state to glass is used. [103]. For photographs, a stable selenium glass film of thickness 100 nm on transparent substrate is used. After illumination with the image is seen in reflection as positive (contrast 20:1) and in transmission as negative. One light quantum transforms about 100 selenium atoms in the crystallite. The resolution is lines/mm. By application of tellurium based films it is possible to improve the image contrast by a careful thermal processing. The scientists and technologists from Matsushita Electric Industrial Co. have discovered an interesting material with the composition having very good properties of reversibility (they succeeded to operate one million of cycles recording-erasurerecording in the same film) stable properties and high photo-sensitivity [104]. The optical disks with films were obtained by evaporation of the active material from two sources (Te and and deposition on PMMA and thus good heterogeneous films Tewere formed. The optimal compositions for optical recording are those with x = 1.1 and 1.2. The films are amorphous. The annealing of the virgin (as-prepared) films does not modify very much the structure although the stability of the film parameters is considerable improved. The recording in the films is based on the change of the transmission through the film. The transition temperature in varies from 350 K for x = 0.8 to 400 K for x = 1.2. For practical purposes are used films of thickness 140 nm deposited on PMMA. The writing is carried out with nm at a pulse power of 8 mW and the inscribed point has the size of 0.8 × 0.8 For the composition TeO1.1 the ratio signal/noise is 59 dB and the energy necessary for writing is 1.6 nJ. There were investigated complex compositions with reversible properties as e.g. Amorphous films of nm in thickness were prepared and the recording was performed on the basis of the transition. For writing are necessary high power and short pulses: mW and ns, the energy being nJ. For erasure must be used pulses of power mW and pulse duration of at the energy of 50 nJ. For reading are used semiconductor lasers with the power of 0.1 mW. The information is stored at the room temperature during some months and the number of possible cycles is of Table 4.6 shows some usual materials for photo-thermal recording. Reversible phase-change optical recording materials such as GeSbTe [106] and InSbTe [107] ternary systems have been developed and good practical performance has been reported. Recently, the Ag-In-Te-Sb quaternary alloy has been studied and Iwasaki et al. [108] obtained complete erasability with the composition at linear velocity of around 7.0 m/s and Handa et al. [109] obtained good performance with the composition at the CD linear velocity. Matsushita et al. [110] studied recently similar ternary compositions for use as computer memories and erasable compact discs and found the highest number of cycles for which exhibits the largest reflectivity difference between the as-deposited and the annealed state at nm. The recording regime for some photo-thermal materials is given in table 4.6.

CHAPTER 4

325

Nippon T&T Public Corporation developed recording media based on and glassy films [113]. The recording is based on the photo-induced structural transformations in these glasses. Although phase-change technology has long been recognized for the simple record and read processes it uses, the commercial success of erasable phase-change optical memories has been predicated on their capability to be used in a direct overwrite mode. Direct overwrite is the replacement of pre-existing recorded data with new data in a single exposure to a laser beam. Compared to currently available magneto-optical storage disks, which require one revolution to erase existing data, followed by a second revolution to record new data, information can be recorded up to twice as fast using direct, overwrite phase-change media. In order to qualify for use as a direct overwrite media, phase-change optical memory materials had to be developed, which could be crystallised using the same duration of laser exposure, which was used to make them amorphous. This requires the crystallisation process to be very rapid. Gonzales Hernandez et al. [114] have studied the transformation kinetics induced by diode laser pulses in Ge-Sb-Te films having compositions on the GeTe – pseudo-binary line and concluded that they exhibit adequate crystallisation speed to be used as media in optical memories. The highest crystallisation speed is achieved in materials, where the atomic configurations in the amorphous and crystalline phases are similar. Films with compositions Ge: Sb: Te / 50.5: 3.5: 46/ 50.5: 1.5: 48/ 32: 12: 51/ 24/21/55/ 14: 29: 57/ exhibit crystallisation when exposed to infrared diode laser exposures for less than 50 ns and are stable at ordinary temperatures. The irreversible photosensitive materials. The non-erasable recording is realised by irreversible changes in the recording medium. The media for irreversible recording using glassy chalcogenides can be situated in two categories: 1. Materials whose recording properties are based on chemical and physical changes . These changes can be: a. removal of the material (ablation)

326

APPLICATIONS

b. photo- and thermo-stimulated diffusion of the metal (Ag,Cu) in the chalcogenide glass. 2. Materials for recording based on the thermo-stimulated reaction between the elements. In the case of ablative recording the laser pulse forms holes in the active layer of a tellurium alloy. Because the mirror layer is locally removed, the reflection decreases at such points. The laser recording based on chalcogenide medium, developed by Konstantinov and Starbov [115] provides stability of the optical disks both on storage and on multiple reading. The laser recording medium is designed on the principle of the thermochemical reaction between a chalcogenide glass and a metal or compound during illumination with short laser pulses (Fig. 4.14).

Figure 4.14. Chalcogenide media for recording with thermo-chemical reaction. a. removal of the product of the reaction and formation of small pits b. thermo-chemical reaction of the elements included in the recording medium

CHAPTER 4

327

The laser recording medium offered by the authors is especially suitable for the spectral region above 700 nm. Only the metal absorbs the laser pulse but this effect leads to the melting of both substances, to their mixing and interaction. By quenching the melt one obtains a metal-chalcogenide compound. The combination metal-chalcogen must be chosen so that the final compound is transparent for the laser beam (above 700 nm). An example is the combination that by illumination with laser pulses gives rise to As-S-Bi glass. If the metal is dispersed into the chalcogenide matrix (Fig. 4.15), then the thermal losses will decrease strongly since chalcogenides are poor heat conductors.

Figure 4.15. The laser recording medium with metal dispersed in the chalcogenide matrix a. the configuration with metal layer (Bi) b. the configuration with metal dispersed in the chalcogenide matrix.

The experiments have shown that with the combinations or it becomes possible to reach energies for recording below 0.1 It is expected that with refractory but resistant to oxygen and moisture materials (e.g. Ge) is possible to achieve energies of recording in the normal required interval of without encountering problems associated with the stability of optical disks to moisture and oxygen. Materials for reversible electron-beam writing Direct electron-beam writing have been reported in amorphous Ag-Ge-S films: [116]. Oldale and Elliott [117] reported the extreme sensitivity to electron-beam radiation of thin Ag-Ge-S films with low Ag content at.% Ag), e.g. They demonstrated the feasibility of reversible electron-beam writing on a submicron scale. The writing process entails the e-beam-induced depletion of Ag in the irradiated area. The Ag displaced from irradiated region piles up in the form of a halo of Ag-enriched amorphous material around the illuminated areas. Such an accumulation is reversible: e-beam irradiation of the halo region causes Ag depletion in a normal way. The phenomenon can be understood on the basis of thermal (beam heating) and electronic effects: the thermal gradients resulting from the electron beam incident on the film cause

the lateral displacement of Ag due to the different ionic mobilities in the temperature gradient. The process is rapid because of the electron-induced enhancement of the electronic conductivity (akin to photoconductivity) which is the rate-limiting quantity.

328

APPLICATIONS

4.5. Photolithography

The need for imaging at the submicron scale has imposed new research in the field of the photolithography, mainly in two directions: a. the increase of the resolution of the exposure systems by turning away from the visible region of the spectrum to far UV, X-ray, electron and ion microlithography b. the development of new photoresist systems. The problem with the photoresist is the conflicting requirements: high resolution and low defect level. High resolution imposes the use of thin resist layers while low defect level (and also etch resistance) require thicker films On the other hand thick resist films require the use of high image quality, good contrast and, therefore, the optics must be adequately tuned. By reducing the thickness of the photoresistive part of the resist film down to 0.05 the requirement for image contrast could be lowered by a factor of Thus, the cut-off spatial frequency increases, the defocusing tolerance increases too and it is possible to get a higher coherence factor, i.e. shorter exposure times. With existing lenses of numerical aperture NA = 0.3 the resolution is limited in that case by the wavelength only. Lower image contrast allows for using lenses with low numerical aperture, therefore greater field size, the placement of multiple chips in a single field and larger chip sizes. The defocusing tolerance that is important for copying on relief surfaces is increased accordingly. After the discovery of the effect of photo-stimulated modification of the solubility of the chalcogenide [118] various applications in technology emerged. Thus, the way was opened for producing new offset plates and a new photoresist for high-resolution microelectronics circuitry. Due to the sensitivity of the chalcogenides to electronic and X-ray beams it was possible to get special resists of very high resolution. Excepting microelectronics, the photoresists can be used successfully for making precise diffraction gratings as well as amplitude and phase zone plates in optics. The increase of the solubility rate in ammonia with by illumination was used in the photo-masking technique [119]. A resolution of lines/mm was reached. The effect can be enhanced when the chalcogenide films thickness) are covered with thin metal films (e.g. Ag or Cu: thickness). The activation of the chalcogenide glass by irradiation in the range of optical hole band through the semitransparent metal film is related to the photo-induced diffusion of metal in chalcogenide films. Special patent for photomask was deposed by Canon [120]. The chalcogenide glasses show mostly a strong durability against acid solutions, whereas these glasses are dissolved easily in alkaline solution. Yoshikawa et al. [121] have developed an inorganic photo-resist by utilising the sputtered Ge-Se films which can be used both for positive and negative-type photo-resist by applying the two photo-chemical effects: i) the dissolution rate into alkali solutions gets higher by illumination of the film and ii) the chalcogenide glasses become almost insoluble into alkaline solutions by the photo-doping with silver. The amorphous films exhibit the necessary properties for the preparation of imaging layers sensitive in the wavelength range nm. The low crystallisation

CHAPTER 4

329

aptitude allows getting thin homogeneous and stable films. In the system As-S the photoinduced modifications are very strong. The optical absorption edge shifts towards red part of the optical spectrum with the refractive index increases by the dissolution rate increases by 40% and the photo-induced vaporisation is [122]. The luminous image obtained in film is stabilised by selective dissolution in alkaline solutions of the exposed and not exposed areas. As a result, a positive or negative image appears. Some studies [123] pointed out that the increased arsenic content in layers causes an enhancement of the stability against alkaline solutions. In applications are used both pure films and combinations with metals as e.g. silver. The arsenic sulphide is very resistant against acids, especially fluoro-hydric acid and this property allows for preparing chromium-based masks. The advantages of the inorganic resist over the organic one are: good uniformity on large areas and small thickness. Moreover, the technology is able to ensure the purity conditions in the fabrication of resist films in the same vacuum cycle. By applying a novel selective developer [124] that does not dissolve the unexposed parts of the film but dissolves completely the exposed ones, the interest in photoresist considerably increased. Extremely thin layers (0.08 - 0.09 could be used, thus achieving simultaneously high sensitivity similar to that of the resist AZ1350 [124] and high resolution. Buroff [125] succeeded to reproduce the details of 0.35 and achieved the theoretical limit of the lens resolution. The electron micrographs reveal no traces of standing wave effect in spite of the high reflectivity of the chromium layer. Simultaneous printing of large and small features were demonstrated. The system developed by Buroff provides a number of technological advantages. The rate of condensation, the substrate temperature the composition and the residual pressure in the vacuum chamber Torr) do not affect the resist properties. An important factor is the evaporation temperature that can be easily controlled. The 0 development temperature and the pH of the developer can vary in the limits C and respectively. Printing plates of the type “offset” with positive sensitivity have been prepared in the Laboratory of Photo-processes of the Institute of Physics of the Bulgarian Academy of Sciences in Sofia, in 1980. A positive image has been obtained by selective dissolution of an film evaporated on an anodised aluminium surface. The not dissolved areas are hydrophobic and therefore are sensitive to ink, while the free aluminium surface is hydrophilic and does not attract the ink. The printing is carried out with a perfect quality and the life of the plate is printing cycles. Since is not resistant to dry etching, other systems were invented. For this purpose the interaction metal-vitreous chalcogenide, firstly proposed by Kostyshin [126] was used. A smaller thickness of the photoactive zone is achieved on the metal/semiconductor boundary (0.02 [127]. The high resolution and contrast is facilitated by the concentration of the reaction products in the photoactive zone and, therefore, the diffusion of the metal only in the illuminated areas, with the result of and an edge-sharpening effect. The reaction product is very resistant to plasma etching, displays

330

APPLICATIONS

an excellent protective ability and avoids the standing wave effect due to high absorption efficiency in the UV region. Negative offset plates have been produced on the same principles, by using the combination In this case, after irradiation, the unreacted components are selectively dissolved. The product of the photochemical interaction remains on the substrate and thus the printing element is created. For such printing system a more complicated treatment is needed and still the life of the plates, expressed in number of copies, is relatively small. The selective dissolution of the chalcogenides gives the possibility to use these materials for getting phase objects in relief (holograms, phase filters, optical integrated units). Figure 4.16 shows a micrograph of relief grating obtained on a chalcogenide film by a holographic method, with subsequent etching [128].

Figure 4.16. Micrograph of the relief structure printed on an film by selective etching. a. diffraction grating of period 0.46 b. the fingerprint (in projection) of a RX mask (nickel grating, wire width: 7 )

A special multilayer design for a photo-resist with excellent resolution and more simple processing as compared to the conventional microlithographic resist was recently developed by Kozicki et al. [129]. The modern trend towards higher integration level in the monolithic circuits is strongly related to the realisation of enhanced packing density of the electronic components on very low areas. The reduction of the size of the integrated circuit plates is essentially dependent on the advances in the lithographic technique. The gradual reduction in size of the circuitry is in close relation with the capacity of the optical systems combined with the contrast possibilities of the photoresist. Today, the optical conventional resists (organic) are inferior to the organic ones as regarding the contrast. An additional limitation of the resolution stems from the fact that the organic resists uses wet substances in the developing stage and these are unable to dissolve the material in narrow spaces. That is why it is suitable a photo-resist system with very high contrast and with the possibility to be dissolved in dry conditions.

A multilayer-planarized system of the type As/S/Ag, called PASS has been developed on non-photo-active substrate. The active film was As-S with a very thin layer

of silver deposited on it by evaporation or sputtering. When exposed to light (or other

CHAPTER 4

331

radiation of energy close to the optical gap of the chalcogenide glass: 2.5 eV) silver diffuses in arsenic sulphide and forms a ternary compound with properties very different from those of the basic material. The complex layer of arsenic sulphide and silver has the thickness of nm while the silver upper layer is only 40 nm thick, enough for permitting the light to pass through. As-S is stable in the plasma and the ternary compound is highly insoluble. When the image is developed, after the elimination of the unreacted silver by a wet procedure, then the As-S regions are eliminated by plasma etching or by reactive etching with in order to expose the underneath planarized layer. The remained (unetched) ternary layer zones are used for selective protection of the planarized layer. Thus it is possible to develop in a dry procedure the resist. This procedure gives rise to a negative image of the applied mask. A positive image is produced if a different sequence during developing is followed: in the first stage is used sulphur plasma that dissolves the ternary compound but not the arsenic sulphide. In the second stage is performed the etching in oxygen plasma with the purpose to etch the underneath planarized layer. The optimum material for this lithographic scheme is that forms a homogeneous system of ternary glass without phase separation when is combined with the appropriate amount of silver. Because the active material has no large macromolecule components, the theoretical resolution can be reduced down to several nanometers. Other chalcogenides, as for example the germanium selenides exhibit a larger lateral diffusion and, therefore, the metallic elements will diffuse (thermally) in the absence of the light. It results therefore, that the system As-S/Ag seems to be an ideal system for lithography, can be deposited easily by evaporation and its stoichiometry can be controlled without difficulty. In order to reach the optimum concentration of Ag in As-S, the thickness of the As-S layer must be situated after photo doping at a value corresponding to times the thickness of the Ag layer. These results in the formation of a ternary alloy situated in the middle of the glass formation domain as shown in the ternary phase diagram (see Ch. 1) The PASS scheme is very sensitive to the electron beam and, therefore, it is possible to inscribe details below 100 nm for line dose The Ag layer above the chalcogenide one plays an important role in the diminishing of the electrical charging of the layer during electron beam exposure. The PASS method succeeded to eliminate most part of the disadvantages of the conventional photo-resist, including also the problems of the focusing depth. Many multilayer configurations have been tried. In the photo-passive polymer bilayer scheme, a Ge-Se layer is evaporated on a polymer planarizing layer or a photoresist-planarizing layer is used. In both cases a silver layer is deposited onto Ge-Se from one of the following solutions of silver ions and complexes: Ag(alanine), [130]. After image-wise exposure unreacted silver and Ge-Se are removed. The remaining reaction product protects the sublayer during dry etching. With the photoactive system etching of the resist could be carried out (with a developer) after the uniform illumination

332

APPLICATIONS

of the sample (the reaction product serves as a mask). The system has the additional advantage of the anisotropic etching of that possesses a columnar structure and dissolves vertically without undercutting [131]. Thus, the three-component system: planarizing polymer layer - thin photoresist layer, combines high sensitivity and resolution without undercutting and standing wave effects. serve as an insulating layer between the photoresist and the planarizing sublayer as well as a reliable protection upon plasma etching. The basic disadvantage of these systems is the deposition of silver from solution prior to exposure. This is necessary since the thin metal layer is not stable when in contact with the atmosphere and diffuses in the volume of the semiconductor in some tens of hours. Furthermore, the deposition process itself is quite irreproducible and imposes measurements of each batch of samples. The reverse system where the silver layer is deposited between the substrate and the photosensitive layer is more appropriate and more reproducible. In that case, too, the reaction products accomplish their protective function. Another possible solution to the problem is the deposition of the metal in the form of

some chemical compound AgCl,

AgI. The use of AgCl leads to an increase in the

reactivity but the resolution drops due to the formation of clusters with sizes up to 150

nm. During electron beam exposure AgCl decomposes but Ag does not migrate in so that an intermediate UV exposure is necessary (sensitivity of 85 is achieved). The combination provides a possibility for considerable increase of the sensitivity; it rises by nearly two orders of magnitude. After exposure the bromide layer is developed. Upon subsequent illumination the developed silver reacts with The system could be appropriate for use with expensive exposure devices as e.g. X-ray or ebeam units. The addition of a third component from the third group of the periodic table to the Ge-Se system provides new opportunities for the photoresist technique. Films of composition where is Ga, In or Tl have been studied. Films with 5, 10, 15 at.% metals and 5 and 10 at.% Ge were prepared [132]. The film starts to transmit above 500 nm. In the kinetic dissolution curves a negative effect of photo-stimulated dissolution was observed. The non-irradiated films exhibit a relative good resistance against etching, the ratio between the times for complete dissolution of the irradiated and non-irradiated areas being around 5.

Arsova et al. [133] have shown that that thin films of Ge-As-S are suitable to be used as photoresist in combination with a wet developer based on NaOH with an emulgator type surface active substance (gelatine). The films exhibit high sensitivity to band-gap illumination, good adhesion to Si, GaAs, etc... and high resolution. We have achieved by contact printing with Cr-test mask lines with minimum size µ m. the resolution can be improved by means of electron beam lithography and dry etching.

CHAPTER 4

333

Electron beam [134] and ion beam exposure [135] can be used for submicron lithography. It was reported that the lithographic sensitivity could be increased if the chalcogenides are evaporated at an oblique angle [134]. The unusual contraction effect induced by UV radiation in Ge-As-S thick films [136] suggest the application of these materials for photo-plastic recordings of images (printing plates) without the need of selective etching procedures. The sensitivity for visible and X-ray radiation makes the chalcogenide materials very promising for high-resolution photo-resist, even in the lithography. For X-ray imaging are used plates of amorphous selenium on aluminium substrate. Selenium is doped by at.% As and by small amounts of chlorine. In the electroradiography the electrostatic image results by the exposure to X-rays of the object and the image is read by using scanning techniques. Finally, the image is converted in a digital form for storage and computing [42,137]. Ultrahigh resolution photolithography was demonstrated by Saito et al. [138] using Ag-Se/Ge-Se resist and X-ray synchrotron radiation. The modern electron beam lithography enables the realisation of increasingly sophisticated diffractive optical elements, which are based on the diffraction of light by microstructured surfaces. Ideals resist exhibits high resolution and sensitivity, linear dependence of profile depth on electron dose, and good profile-shape repeatability. The resist should be suitable for the preparation of nickel shim masters, which are used in the low-cost replication processes of diffractive optical elements. As evaporated as well as thermally annealed As-S-Se films at glass transition temperature were studied as resist for electron beam lithography. The films of various thickness were tested and various electron energies and doses were used for film exposures [139]. The profile depth obtained for 2.0 thick films was 1.25 It means that, the maximum profile depth is about two read-outs wavelengths for He-Ne laser (632.8 nm) and elements with high diffraction efficiency can be fabricated [74]. Jefimov et al. [139] have shown that the as-evaporated films behave as a negative resist for an organic alkaline developer, but after thermal annealing below the melting temperature they become positive resists. 4.6 TV Pick-up Tubes (Vidicon)

In the years ‘50 there was shown that amorphous selenium is a photoconductive material appropriate for electrophotographic receptors and for TV pick-up tubes. The materials based on Se were the first applications at a large scale of the amorphous chalcogenides. The TV pick-up tube with target from photoconducting material was called Vidicon. The working principle of the tube consists in the accumulation of electrical charge and the formation of an electrical relief on a plate. In front of the Vidicon tube is the entry glass window plate with semitransparent layer from as a signal plate. The metalled substrate is covered by a photoconductive layer from an amorphous semiconductor where the optical image is projected. Across the loading resistance RH is collected the potential difference between the semitransparent electrode and the cathode of the electronic gun (Fig. 4.17) [140].

334

APPLICATIONS

Figure 4.17. The scheme of the Vidicon pick-up tube.

1. Entry window 2. Glass plate 3. photosensitive layer 4. Electronic gun 5. Electron beam 6. Magnetic system for focussing and scanning 7. Objective lens

As opposite to the electrophotographic receivers, in the Vidicon picture tube the photoconducting film is changed by a beam of mobile electrons that scan the screen. Under the action of the light, the local photoconduction currents lead to local photodischarges. The local recharging current that appears during scanning of the screen is a function of local illumination and represents the electrical signal for every position on the photoconducting screen. The photoconducting materials for Vidicon tubes are, usually, chalcogenide materials. Last years, the amorphous hydrogenated silicon has been also used. The selenium photodiodes are heterojunctions between the polycrystalline selenium and the polycrystalline CdSe. Several photo-junctions have been prepared with properties similar to those of the Se photo-diodes, based on a-Se and various n-type materials as e.g. CdS, CdSe, and An example of electrical characteristic curve for the heterojunction a-(Se-As-Te)/CdSe(polycryst.) is shown in figure 4.18 [141]. In order to have good diode characteristics at reverse bias it is necessary to block the carrier injection from electrodes in the high resistivity material a-Se. The CdSe layer can be regarded as blocking layer for holes. Sometimes is used as blocking layer for electrons because the minority carriers have small lifetimes in this amorphous chalcogenide. An ideal blocking diode, reversal biased represents the basic element in the Vidicon pick-up tube. A photoconductor of high resistivity is put between two blocking layers. These layers block the injection of the excess carriers from electrodes while the photogenerated carriers can be easily generated from the photoconducting material. It is suitable that the blocking layer situated on the side where the light impinges have a gap larger than that of the bulk material because the quantum efficiency for low wavelengths increases due to the “window effect”.

CHAPTER 4

335

Figure 4.18. Dark-current-voltage characteristics of an amorphous chalcogenide heterostructure photodiode.

While the Vidicon tube possesses blocking layers with recombination, where the electrons injected by the scanning beam are captured and recombine with the photogenerated holes, the conventional electrophotographic plate from a-Se has an imperfect blocking structure because it does not possess a blocking layer on the charging surface. As we have already shown the a-Se layer is a photoconductor very sensitive and therefore appropriate for TV tubes with image sensitivity. Nevertheless, for high wavelength range in the optical spectrum the selenium sensitivity is low. Moreover,

336

APPLICATIONS

selenium is enough unstable and changes its properties by crystallisation. The addition of tellurium to selenium decreases the optical forbidden gap and enhances the sensitivity for wavelengths situated in the red range of the optical spectrum. The addition of arsenic enhances the stability against crystallisation of selenium and produces a barrier that prevents the thermal diffusion of tellurium. Unfortunately, both doping elements enhance the trap concentration in a-Se and disturb the formation of a photo-junction with stable operation. In order to solve this problem Maruyama and Hirai [141] proposed a new photodiode structure with graded chalcogenide composition and on its basis have been developed and produced the Vidicon tubes under the commercial name Saticon. They are now widely used in home colour camera, as well as in broadcast and industrial cameras [142]. This photodiode blocking structure lays at the basis of the preparation of the photoconducting layer for laser printers [143]. The structure consists of a heterojunction between the transparent n-type

electrode (ITO) and a chalcogenide p-type material as e.g. As-Se-Te (Fig.4.19).

Figure 4. 19. The layer configuration in the Saticon target (after [141]).

CHAPTER 4

337

A porous layer is added to the structure in order to protect the surface and for preventing the secondary electron emission. The layer exhibits a high density of traps that block the injection of the carriers originated from the electron beam. The transparent electrode is directly biased relative to the scanning beam. Since most electron-hole pairs are generated in the region rich in tellurium, the electrons go from the chalcogenide layer to the transparent electrode and the holes pass across the layer dominated by Se and recombine with the electrons from the layer, thus determining the discharge of the surface. In the Saticon structure the chalcogenide layer is the main component and is designed in two variants: a) structure with graded forbidden gap and b) structure with included field effect. In the variant a) the continuous flow of the carrier current through the region with smaller gap is ensured by a variation of the gap in the film determined by the linear increase followed by a linear decrease of the tellurium concentration in the material (Fig. 4.20 a).

Figure 4.20. The compositional distribution of the Saticon chalcogenide film (after [141]). a) graded-gap structure and b) blocking electrodes with built-in-field effect structure

338

APPLICATIONS

In the variant b (see Fig. 4.20 b) a region rich in As follows the region rich in tellurium. This zone is rich in deep electronic traps. The arsenic content as well as the associated deep trap density diminishes from the region rich in tellurium towards the region of almost pure selenium. Due to the accumulation of spatial charge on these traps it appears a strong electric field that extracts the photo-carriers from the zone rich in tellurium. The As-rich region forms also a barrier against the thermal diffusion of Te in a-Se, thus improving the thermal stability of the whole structure.

The plate with chalcogenide semiconductor can be easily integrated with colour organic filters and, therefore, shows great perspectives for the construction of cheap TV

cameras. In the Vidicon tubes are used various glassy chalcogenide films: CdSe, CdS, Se, etc. Very promising are the pseudo-binary systems [144], and others that allow for a significant increase of the range of the spectral sensitivity. The Vidicon tubes can be produced with sensitivity in various spectral ranges: X-ray, UV, VIS and near IR. Jedlicka et al. [145] have shown the possibility to use in Vidicon targets of high sensitivity based on heterostructures of the type In this system the interface between layers is enriched with oxygen. A thin layer of SnO2 on the side of CdSe layer makes contact. The chalcogenide layer was deposited on the baked CdSe layer by vacuum evaporation and the electron beam scans the surface of the chalcogenide film..

These targets of television camera tubes shows a high average value of the integral sensitivity (2750 a very low dark current nA), a low lag (decay-lag current of 15 % after 60 ms) and good resolution that fulfils the requirements of current television transmission. The accumulating layer of the target can be in principle composed of any chalcogenide glass type where From the viewpoint of production technology and reproducibility of the parameters it is advantageous to use pure selenide or pure sulphide, but it seems that sulphides give higher values of lag. The necessary value of the field intensity varies from V/cm, as x changes from zero to one. 4.7. Radiation Sensors and Measuring Devices

4.7.1. DETECTORS OF RADIATION AND RADIATION AMPLIFIERS

A solid emits electrons when is subjected to electron bombardment. The phenomenon was observed as early as 1902 by Austin and Starke [146] and was utilised for producing systems for electron amplification. On this basis have been produced detectors and ion, X-ray and photon amplifiers. From the group of materials with secondary electron emission properties are mainly used the glasses due to simple fabrication procedure, the lack of chemical reactivity and broad range of characteristic properties.

CHAPTER 4

339

When a primary electron beam impinges the surface of a solid, then a certain fraction of the beam is elastically reflected and other fraction enters into the solid. Among the electrons that enters into the material, some of them loss the energy by inelastic collisions with lattice electrons and return to the surface, that they leave as a result of the so-called Rutherford scattering. These electrons are the part of the inelastically scattered

primary electrons. The most part of the primary electrons that penetrate the solid surface losses their energy through the excitation of the lattice electrons to high energy levels. The excited electrons (secondary electrons) move towards the surface and some fraction of them leaves the solid. The physical parameter called secondary yield, is defined as the ratio between the density of the emitted electrons and the incident electrons. In this definition are included both the primary elastically reflected electrons, those primary electrons inelastically reflected and the secondary electrons themselves. The secondary emission is strictly related to the energy of the primary electrons. For low energy increases with the energy and reaches a maximum values and then slowly decreases. The solid materials are characterised by very different surface structure and properties and these have decisive influence on the secondary yield. In some conditions the number of emitted electrons can be higher than that of the primary electrons that knock the surface. In the table 4.7 are given the highest value and the primary energy corresponding to this maximum for several glasses [147].

The insulating materials that form easily glasses exhibit in the glassy state the highest values of the secondary yield. There was shown that the mean free path of an electron into the material is a function of the deviations from the lattice periodicity. A lower mean free path (characteristic to the glasses) leads to lower The experimental results have shown that the processing (by polishing) of the surface has profound effects on the secondary yield. The maximum secondary yield can be intensified considerably by atomic depositions on surface. These effects are related to

340

APPLICATIONS

the phenomenon of decrease of the electron affinity to the surface. They are exploited in commercial electron amplifiers. In glasses, the maximum secondary yield is independent of temperature due to the disordered structure of the atomic network. In crystals Gm decreases when the temperature increases. This is due to the gradual diminishing of the electron mean free path. The most widely spread devices that use the secondary electron emission are the systems of electron multiplication [148], These devices are ideal detectors and amplifiers for electrons, ions, soft X-rays and photons. They can produce multiplications with the factor 103 ÷ 106 in a series of metallic electrodes, called dinodes, coated by a high Gm layer. The dinodes are disposed in a line with increasing potentials so that the number of electrons that collide a dinode be amplified every time the electrons knock a dinode surface in this series. In the modern electron multipliers are used glasses [149]. In the field-plate type multiplier both the field plates and the dinodes are built from glass. An electrostatic or magnetic field guides the electrons down to the channel. Thus there was obtained a multiplication factor of 1.3 × 107. In the variant of multiplier with parallel plates anode the incident electrons are multiplied by successive collisions on both plates. So, multiplications factors of 107 ÷ 108 were obtained, The most sophisticated electron multiplier is the system with tubular dinode. Figure 4.21 shows the sketch of such system [150] that gives a multiplication factor of 108.

Figure 4.21. The scheme of the electronic multiplier with tubular channel. In all these multiplication systems the plates or the glass tubes are coated with glassy materials of high Gm. It was observed that it is advantageous to use the internal surface of the glassy tube. This is because the deposition of good homogeneous adherent layers is a difficult task, There was possible to use also the internal surface of a bulk glass processed as a tube. The glass fibre technology, including that of the chalcogenide glasses, represents an important advanced step in the field of multipliers. Although the values of Gm in glassy chalcogenides are lower than for other glasses, it is possible to get amplifications of 105 ÷ 106 with the condition that the ratio tube length/tube diameter be high.

CHAPTER 4

341

If a photocatode is mounted to the tube entry and a phosphorescent screen to the exit, then one obtains a very efficient light amplifier device. Such amplifiers were fabricated as plates provided with a large number of channels with the diameter of called microchannel plates. The light amplifiers have been initially used for military purposes. In the last time civilian applications have been devised as e.g. the special eye-glasses for the people suffering of the disease named “retinitis pigmentosa”. 4.7.2. RADIOMETRIC DEVICES

The photoelectric properties of the chalcogenides (photoconductivity, photopolarization, and photoelectret state) have been utilised in the development of several methods for measuring the radiation. Among the vitreous materials with applicability in radiometry are pure or doped by metals (Sn, Ge, Cu, Ag...), arsenic sulpho-selenides complex compositions etc. The applications in the memory radiometry are based on the remarkable photopolarization properties of the chalcogenides, i.e. accumulation of electric charges under in darkness and its destroying (by flowing an electrical current in the circuit) in the absence of the external electric field when the sample is illuminated. For ohmic injecting electrodes the dominant effect is that of stationary photoconduction. The photo-polarisation and photo-electret processes become dominant when blocking electrodes are used (these are produced by pressing two metallic lamellae against the sample, sometimes using an intermediary layer from mica or polyethylene). Kolomiets and Lyubin performed the first observations on the photo-polarisation and photo-electret effects in vitreous chalcogenide semiconductors in 1962 [151]. Later Andreichin et al. [152-154] has investigated the polarisation and electret properties in many chalcogenide compositions. One of the compositions widely studied is In this vitreous composition the dark polarisation at high field is obtained both with blocking electrodes and with injecting electrodes. This fact proves that we are dealing not only with a surface component of the polarisation but also with a bulk component. In the case of injecting electrodes the residual currents are larger but the polarisation is lower. The electromotive voltage of the polarisation defined as

reaches only from the applied voltage (here is the residual current and U is the voltage). The relaxation time is of min. and the initial state is fully restored after minutes. The currents, especially the residual one, show a strong dependence on the temperature.

342

APPLICATIONS

In the case of the blocking electrodes the dark polarisation effect is much stronger. The decrease in time of the current is rapid and the residual current is low. The polarisation can reach in the case of the blocking electrodes around % from the value corresponding to the applied voltage and when additional intermediary blocking

layers are introduced this value is exceeded. The distribution of the potential between electrodes is still unknown but there are indications that most part of the electrical charge is accumulated to the surface (at the interface with the electrodes). The time decay of the current follows an exponential law. The relaxation times and the restoring times are times larger than in the case of the ohmic electrodes. The dark depolarisation starts immediately after the switch-off of the electric field and completely disappears after hour. The photo-polarization is the electrical polarisation produced by illuminating the sample. If one takes into account the processes in the sample and, on this basis, one analyses the depolarisation phenomena, then it will be found the following three components: a) Dark polarisation that is independent of illumination. This component starts when

the field is swittched on and disappear immediately after switching-off of the field and lasts for a short time. This polarisation is related to the shallow traps. b) Light polarisation for high field. This component is similar to the dark polarisation but exhibits a higher initial current than that in dark and, also, a larger electrical charge The depolarisation starts immediately after the field switching off and in darkness disappears after a longer time interval. These features are due probably to a higher concentration of occupied energy levels, of similar depth or somewhat higher as compared to the dark polarisation. They are responsible for the disappearance of the polarisation by the thermally stimulated recombination of the

carriers. c) Permanent polarisation of the photo-electret state in darkness. The depolarisation starts only by illuminating the sample in the absence of the electric

field. The photoelectret state is characterised by the maintenance of a part of the photopolarization state and this part has been called photoelectret polarisation. This polarisation is measured by the ratio where is the charge lost during depolarisation and is the total polarisation charge, after long time storage (for the case of the crystalline and vitreous and also for the single crystal and poly-crystalline sulphur this time interval is After hours necessary for the disappearance of the light polarisation and dark polarisation remains almost constant at the level of in the case of vitreous and this value exceeds significantly those of other single crystal or polycrystal compositions. Weaker polarisation phenomena were evidenced in and Similar properties do appear for doped by metals as Sn, Cu, Ag but also Ge as well as in and in alloys of the type The polarisation phenomena and the photo-electret phenomena have been explained by Andreichin [155]. In the arsenic chalcogenides there are not dipoles to be oriented by the applied electrical field and there are not possible ion movements or other processes.

CHAPTER 4

343

The vitreous materials are built from polymeric type chains with many uncompensated free states at the chain ends. In the light of the more recent researches, we are dealing with atomic configurations of disordered layers with dangling bonds at the borders of the amorphitic domains [156]. The conduction is of p-type and therefore the holes will move towards the uncompensated free sites and one electron per atom from the neighbour atoms will occupy the free sites of the atomic chains. Thus the electrical conduction will proceed mainly along the chains and at a low rate between chains. Thus the charge carriers will be situated at the borders from the electrode sides and at the positions of chain bending where, as a function of the strength of the chemical bond, will participate to the dark polarisation, to the light polarisation and to the photoelectret state. In spite of the relatively high photoconductivity the vitreous chalcogenides are not used as photo-resistors for the measuring of the radiation because: 1) they have high resistivities and therefore very sensitive devices are needed for the measurements (electrometers) 2) they have very pronounced polarisation properties. After the applications of the electric field (in the presence or in the absence of the light) the dark current and the photocurrent decrease, this fact leading to the creation in the bulk or in the vicinity of the contacts of an electromotive voltage for reverse polarisation. These polarisation and photo-polarisation processes allow for the implementation of the applications in the

memory radiometry and pulse radiometry. The applications in memory radiometry are based on the photoelectret properties of the chalcogenide glasses. The electrical charge accumulated under the simultaneous

action of the radiation and electric field and the electrical charge released during the depolarisation show the following features: 1) For constant irradiation fluency the charge Q varies linearly with the applied field up to a given value (typically: 3kV/cm) thereafter reaching saturation. 2). When the external electric field is constant, the charge Q is proportional to the irradiation fluency up to the maximum radiance of the incandescent lamp utilised in the

experiment (25

in the visible spectrum)

3). In the case when both the electric field and irradiation fluency are maintained at a constant value, the charge Q varies initially linearly with the irradiation time and then rapidly saturates. The above-described behaviour can be utilised in the radiometric practice. After an enough long time starting from the moment of the switching off the radiation fluency, it is possible to estimate the time interval used for the sample irradiation and/or the amount of radiant energy absorbed into the material. The most important applications of the polarisation phenomena have been developed in the pulse radiometry [157-159] for lasers, pulsed lamps, luminescent diodes, etc. One of the main trends of the modern photo-electronics consists of the production and use of very rapid photoelectronic radiation receivers. The ideal situation in the pulse radiometry corresponds to the case of the relaxation time of the detector much shorter than the pulse duration. Nevertheless, in some cases an opposite situation is preferred and this case can be exploited in a measuring method somewhere analogous to the ballistic method for measuring the electric charge.

344

APPLICATIONS

In the radiometric devices there are two measuring possibilities. One can measure either the initial current in a rapid measuring device or the maximum value, with an inertial device. A more accurate method consists in the measuring of the current-time curve with an oscilloscope with rapid memory. In all these cases the observed effects are produced in a longer time after the pulse action. In a usual dynamic electrometer method is reached at s after the action of the radiation pulse (although the pulse duration is smaller than 0.1 s). The decay to zero of the current needs

s. Therefore, the observable effects are slow and they extend

much above the pulse duration. A method for measuring the radiation pulse is the following: During the dark polarisation of the sensor cell the radiation pulse is recorded as a current pulse that rapidly increases but slowly decreases. The initial current, the maximum current, or the charge Q released in the circuit are parameters proportional to the intensity and the duration of the pulse, i.e. with the irradiation dose (Fig. 4.22 I-a).

Figure 4.22. The measurement principle for the radiation pulses (I) a. with preliminary polarisation in dark b. by using the photoelectret state E - the applied electric field I - the current through the device -the charge accumulated during polarisation - the charge released during depolarisation L - light D - dark LP - light pulse. The receiver scheme (II) with the two pairs of electrodes connected in parallel. Vertical hatching: ohmic electrodes. Horizontal hatching: blocking electrodes.

An other measuring method is shown in figure 4.22 I-b. After complete polarisation in dark and in light, L, and after depolarisation in dark,

D (when only the photo-electret charge will subsist) the light pulse, LP, is recorded by the photo-electret charge and the measured parameters are the same.

CHAPTER 4

345

The receiver scheme with two pairs of electrodes (ohmic - o and blocking - b) connected in parallel is given in Fig. 4.22 II. The experiments with light pulses of and electric field of have shown that these methods are equally sensitive but that the first one is more easy to apply: a much shorter time is needed for the cell to return in the initial state and then to start a new measurement. Because the operation is carried out outside the short-time irradiation limit, the proportionality between the pulse dose and and Q is excellent. Other method of lower sensitivity can be used in measurements. After the total polarisation by light, when the saturation current is reached, the cell is put in dark. The saturation dark current, will be lower than the saturation light current Now if a radiation pulse is impinged upon the cell then the appeared current will exhibit a maximum towards the value (Fig. 4.23). In this case will be measured the same parameters as in the first two methods.

Figure 4.23. The measuring of the radiation pulses on the basis of photo-conduction after complete photo- (L) and dark (D) polarisation. E - applied electric field LP - light pulse I - current through the device

4.7.3. SOLID STATE INTEGRATORS The change in time of the local electrical resistance of the transmission coefficient, of the reflection coefficient and of the physical parameters of the thin film structure AlChalcogen-Conductor during the application of an electric field allows to use solid state integrator of the quantity of electricity in systems for the control and measurement of the working time of various equipments or installations, in automation and computing technique. The integrator allows getting an output parameter proportional to the integral of the action in time of an external parameter with very low power consumption. The reading of the information is carried out without erasure. The solid-state integrator based on chalcogenide materials consists of a conducting and transparent basis, e.g. a thin film of chalcogenide glass with variable thickness

346

APPLICATIONS

and a counter-electrode from aluminium. A source is introduced in the circuit (Fig. 4.24 a) [67].

Figure 4.24. The solid state integrator. a. General scheme b. The variation in time of the electrical resistance 1 - aluminium counter-electrode 2 - chalcogenide glass 3 4 - source

The mode of operation is the following: The power source is connected to the integrator so as the measuring electrode has

positive potential. In this case the electrical field stimulates the chemical transformation of the aluminium counter-electrode in the layer of the transparent dielectric. The transformation rate is determined by the value of the current flowing through the structure, that depends on the properties of the active layer and on the electrical field. The non-uniform thickness of the active layer ensures a non-uniform distribution of the intensity of the electric field along the measuring counter-electrode and the nonuniform transformation of the last one. If the active layer has the shape of a wedge, the amount of electricity is determined visually after the magnitude of the shift of the colour boundary observed on the metal measuring electrode. The estimation of the quantity of electricity that goes through the integrator can be realised by measuring the electrical resistivity, the optical transmission, and the integrator reflection as a whole or as separated elements. The necessary current for the integrator is small and depends on the chalcogenide composition, on its thickness, on the area of the measuring counter-electrode, on the value of the applied voltage and has the order of magnitude of A. Figure 5.24 b shows the characteristic line of the change in time of the electrical resistivity of the integrator when a constant voltage is applied: During long-time field application the electric resistivity of the integrator rises with about three orders of magnitude.

CHAPTER 4

347

4.8 Acousto - optical Devices

The acousto-optic effect provides a method of optical signal manipulation without using high voltages. Its use in the far-IR has been previously limited by the low efficiency of the commercially available materials, but the benefits from acoustically controlled frequency modulators and beam splitters make viable research into new, more efficient acousto-optic materials for mid and far - IR. The effect is set up when an ultrasonic wave is passed through a material, causing variations in refractive index to give an effective acoustic diffractive grating within the glass. A laser beam travelling in a plane perpendicular to the direction of travel of this acoustic wave will be deviated from its original path by an angle depending on the frequency of the acoustic grating. The amount of light diffracted depends both upon the efficiency of the glass and the amount of acoustic power launched into the test sample. The efficiency of an acousto-optic material is characterised by its figure of merit, M:

where n is the refractive index, v is the sound velocity , p is the photoelestic constant and is the density. The greatest effect on M is given by n. A good material must exhibit: a). high optical transmission in the required

wavelength range b). stability under normal application conditions and c). low acoustic attenuation. The material must be fabricated to a high degree of uniformity and enough large size for commercial devices Acousto-optical devices can rapidly modulate the amplitude and direction of a laser beam. An rf field is applied to a block of suitable material and an acoustic grating is thus

generated within the block. The grating can be used to produce two optical beams and the measurements of the particle flow are then possible at the point of recombination of the beams. Far IR chalcogenide glasses, when suitable formulated to optimise the acoustooptical figure of merit, are attractive for operation at The heating effect produced by incident rf power means that application of the glass in acousto-optic devices may be limited by thermal expansion mismatch between components at elevated temperatures and by glass stability. The system As-Se-Te was mainly studied due to its higher stability and diminished phase separation aptitude. Sb, Ge and Pb have been used as dopants [160]. For shorter optical wavelength operation there were investigated glasses in the system Ga-La-S [161]. Glasses in the chalcogenide system based on has been found to exhibit relatively high acousto-optic figures of merit, low acousto-optic losses and high acoustic velocity. These properties combined with the transparency both in visible and infrared high refractive index, non-toxicity, high softening temperatures enough high damage threshold, ease of fabrication, optical quality, chemical stability and isotropic properties make it attractive for acoustooptic device applications. The interest in this glass system was awakened in the last years due to its low phonon energy, hence acting as a promising host for rare earth elements for

348

APPLICATIONS

the development of an efficient optical fibre amplifier operating in the near infrared region. The glass can therefore be suitable for building efficient in-fibre acousto-optic devices. The As-Se-Te-(Ge-Pb) system transmits over a wide range covering three laser wavelengths at The transmission levels are near the theoretical maximum. The reported figure of merit relative to silica, seems to be very large although the results are characterised by large error bars. The amorphous is transparent in near infrared and exhibits a high figure of merit: Addition of germanium to As-Se system increases the connectivity of the network, hence raising both TG and acoustic velocity. Refractive index is lowered and M values at wavelength fall from for glasses [162]. The base glass As-Se-Te with 5 mole % Ge, 10 mole % Pb and n = 3.205, developed by Laine and Seddon [160] shows an outstanding figure ofmerit: at The base glass As-Ge-Se (n 10.6 = 2.878) with additions of 20 mole % Te and 5 mole % Pb gave : at the wavelength 4.9. Chemical Sensors and Other Devices 4.9.1. CHEMICAL REMOTE SENSORS The wide IR windows and the high resolution of Fourier transform infrared spectrometry together, permit remote sensing of several gases or liquids using their absorption due to

the fundamental absorption modes, for instance by means of fibre optic-evanescent field spectroscopy [163]. The unclad GeSeTe fibre was used to this purpose. It was possible to measure the existence of small amounts of organic and acid species in aqueous solutions quantitatively. Maeda et al. [164] introduced a new approach for the qualitative and quantitative analysis of gas species such as CO in the coal combustion furnace. The emission spectrum of the furnace was delivered through the fibre cable to the FTIR spectrometer. The in-situ quantitative measurement of CO gas as low as 100 PPM was carried out at The unstable gas CO was also successfully detected at during the reduction of by graphite. Chalcogenide glass fibres (sulphides) were used instead of KRS-5 fibres due to higher flexibility and thermal stability of these fibres with length of more than 10 m required for the applications in the pilot plant combustion furnace. 4.9.2. ION-SENSITIVE ELECTRODES.

As in the case of oxide glass electrodes for pH measurement, the chalcogenide glasses based on As-Se or As-Se-Te with up to 25 mol % Cu or Pb content have been

CHAPTER 4

applied for the detection and measurements of the [165,166].

349

and

ions, respectively

4.9.3. ELECTRICAL RESISTANCE ELEMENTS.

Several chalcogenide glasses have been used as thermistors in the regulation devices for the nuclear reactors due to high neutron fluency [167]. Kirsanov et al. [168] have studied several compositions for thermo-switches based on Te in planar thin film devices: The best thermo-stability has been observed for the switches based on This material allows for switching in the conducting state at without change of the current parameter up to 10 mA for cycles. The presence of an exponential dependence of conductivity versus temperature allows using several chalcogenide glasses for making good quality thermistors. 4.9.4. OPTICAL FIBRE SENSORS.

The progress in the study of the optical fibres offers new opportunities for the elaboration of new family of devices for the measurements of the physical parameters. New sensors and transducers of high performance have been produced on the basis of some properties of the fibres, which make possible the modulation of the transmission parameters by the physical parameter to be measured. Big companies are interested in the research and production of sensors based on optical fibres in general and chalcogenide fibres in particular. At the end of this millennium the total volume of production for sensors based on optical fibres is estimated to 4.88 billions of dollars (Proceedings SPIE, Fiber Optic Sensors, Vol. 586, p. 2-13, 1985). A fraction of 84 % is used in military technique and 14 % in industry and medicine. The most important companies involved in this field are: AEG Telefunken, Allied Bendix, Boeing, Hitachi Ltd, IBM, Martin Marieta. 70 % of optical fibres production is by North America, 15 % is by Europe and 12 % is by Japan. There are two categories of sensors: sensors, which use the fibre as sensing element and sensors which use the fibre as transmitting medium from the object to the device. The parameter to be measured is transformed in some optical signal following the

interaction of the optical guide with the source. The high sensibility of the transducers based on optical fibres is due to high optical distance wherein the interaction with the source is realised. The working principle of the optical fibre sensors is based on the recording of several parameters of the light transmission in fibres: -light propagation time (interferometric sensors /phase or frequency modulation/) -intensity of the transmitted light (amplitude modulation sensors) The sensors have the advantages of the non-sensitivity to electromagnetic fields. Moreover, they do not induce electromagnetic noise in other electronic systems. The transmission of the information can be realised at large distances and the sensors have low dimensions and low prices.

350

APPLICATIONS

4.10. Solid State Batteries The properties of vitreous ionic conductors are of fundamental and practical interest.

Glasses are suitable for a wide range of electrical and electrochemical applications. In addition there is a renewed interest in understanding the mechanisms of the ionic conductivity. Chalcohalide glasses are of prime interest. ion - containing chalco-

halide glasses are particularly interesting because they possess high conductivity that can reach at room temperature. In the ternary system the best conductivity obtained is for molar composition with an activation energy around 0.15 eV. There was proposed the application of chalcohalide glass in solid-state batteries. The thio-iodide glasses show high conductivity, are very stable versus moisture and can be shaped easily as bulk material or thin film. Therefore they are good candidates for electrochemical applications. The behaviour of a silver thioiodate - silver phosphate glass was studied in a test solid state battery. A galvanic chain phenotiazine was developed. The performances were comparable to that obtained with (a solid electrolyte) with the highest ion conductivity at room temperature leading to batteries capable of considerable discharge. Various silver thioiodate glasses based oh have been discovered. They possess a very high mobility and solid-state batteries have been developed using these materials as electrolyte [169]. The amorphous has been investigated with the purpose to use it as a highenergy cathode material in ambient temperature alkali metal cells [170]. Amorphous prepared by chemical or thermal decomposition of ammonium thiomolybdate, has been found to react readily with n-butyl lithium and sodium and potassium naphtalides to give compositions The cathodes used in lithium electrical batteries (cells) react with up to 3.8 Li per under constant current conditions above 1.40 V. The mean discharge voltage of is close to 2.0 volts which, together with high coulomb capacity results in a cell with high theoretical energy density of 1,040 Wh/g. This energy density is significantly higher than the values reported for crystalline transition [171]. cells show good reversibility for several cycles.

4.11. Infrared Optical Media.

Over the past decades chalcogenide glasses have been studied in order to assess their suitability as passive bulk optical component materials for IR applications. The applications are closely related to the transmission of the IR radiation in ambient atmosphere. Fortunately, the main atmosphere components, nitrogen and oxygen are homopolar molecules and possess neither a permanent nor an induced dipole moment,

CHAPTER 4

351

and hence do not exhibit infrared active molecular vibrations, which would result in the absorption of IR radiation. In the initial region of their spectrum from (near infrared), another from (middle infrared), and a third from to (far infrared). From the black body emittance curves it results that in order to detect relatively hot objects window is most suitable and to detect objects at room temperature the window is most suitable. There is a major interest in thermal surveillance systems at present and these practical uses of IR are concerned with wavelengths up to about

In order to focus this IR radiation onto detecting systems, windows, lenses and telescopes are required made from materials exhibiting adequate transmission in this wavelength region. For low power surveillance or thermal imaging applications where chalcogenide glasses will be useful, materials are required in sizes of up to 150 mm diameter and up to 20 mm thickness. Materials with infrared transparency in the wavelength range are selenide and telluride glasses. The arsenic sulphide and also the chalcogenide glasses with silicon are transparent up to wavelength. Ge-As-Se glasses (big pieces of kg) have been produced in Russia under trademark IKS-29 [172]. Hilton [173] developed technical glasses of composition (trademark Ti20) and (trademark Ti 1173) under the form of discs of diameter 70 mm and platelets of area The applications of these glasses are: energy management, thermography, temperature monitoring, electronic circuit detection, laser technique and IR-spectroscopy as well as high-resolution optics in night image technique. AS an example of the latter, the blackbody radiation emitted by room temperature objects such as the human body is situated in the range where selenide – telluride – based glasses are applicable for thermal imaging [174]. As opposite to alkali-halogenides and alkaline-earth fluorides, the great majority of chalcogenide glasses are very stable in normal atmosphere. Windows for recipients used in infrared spectroscopy study of acid solutions are preferably made of chalcogenide glasses. Other optical applications are based on the high refractive index of the chalcogenide glasses. They exploit the increase of the transmission of coatings with decreased reflection. The chalcogenide glasses are used as antireflection coatings for IR filters. Depending on the composition, one can change continuously the refractive index “n” of these glasses from the values situated around 2 for non-stoichiometric arsenic sulphide with large excess of sulphur up to for systems with tellurium [175,176], The mechanical strength and the thermal stability of the chalcogenide glasses are significantly lower than those of the oxide glasses. The thermal expansion coefficient, the temperature coefficient of the refractive index and the elastic-optical constants are higher than those of the oxide glasses. Because of the comparatively small mechanical strength and chemical softness of these materials, care has to be taken to avoid damage. and [ 179] have been proposed as new materials with improved mechanical properties and good infrared transmittance in the wavelength range appropriate for window glasses.

352

APPLICATIONS

Finally, we must mention the use of photo-induced anisotropy effect in

films

to realise non-linear image processing like image addition and subtraction [180]. The same effect was applied [181] to expose Damman gratings for multispot array generation.

4.12. Components for Integrated Optics

The development of the integrated optics is basically related to the creation of the light conductive lines for transmitting optical information. This was possible due to the appearance of laser sources of light and light conductors with low losses. The optical micro-technologies, including new techniques and materials for micro-switches, microlenses, etc. are mainly directed toward the infrared optics where the chalcogenide materials started to play a leading role. 4.12.1. OPTICAL FIBRES.

Optical fibres are important as conducting elements in the optical integrated circuits or devices. Optical materials useful for IR fibres generally need to possess the following optical properties: low material dispersion, Rayleigh scattering, high energy band gap and long wavelength multiphonon edge. There is a wide need for infrared fibres operating in the middle infrared regions with potential applications in transmission of YAG: Er, CO and laser energy as well as in the detection of thermal radiation emitted by warm bodies and thermal imaging. Through the 1980s, attention was focused on the fabrication of ultra-low loss IR fibres for telecommunication signal transmission to compete with silica optical fibres [182,183]. The first report on the electrophysical and optical properties of the optical fibres with the composition was published by Abashkin et al. [184]. Very promising results were achieved with ternary glasses as e.g. Ge-As-Se [185] and Ge-Se-Te [186]. Glasses of composition Ge-Se as well as Ge-Sb-Se ternary chalcogenides are transparent in the range and have good thermal, mechanical and chemical properties. Lezal et al. [187] and Lucas et al. [188] have described a new class of chalcogenide halide glasses for infrared fibres. The main advantages of this class are the decrease of the intrinsic absorption around wavelength and the improving of the mechanical properties. Fibres of glassy arsenic chalcogenides with the diameter of embedded in protective coating and without coating were produced by pulling from the glass heated up to the softening temperature or from the melt as in the case of and Considering the low mechanical strength of the fibre it results that a buffer material must protect the fibre. The most usual method is cladding with Teflon (perfluorinated ethylene propylene). Nishii et al. [163,189] proposed a preparation method for a fibre with glass cladding using a silica crucible. By this method was possible to draw a telluride glass which is less stable against crystallisation but transparent above

CHAPTER 4

353

selenide glasses draw into unclad fibre confirmed that a loss less than 0.1 B/m could be attained. The measurement of the spectral transmission of the fibres has shown that for the wavelengths larger than the light damping coefficient is less than 0.5 dB/cm. The theoretical minimum losses of mid-IR chalcogenide fibres are probably limited to >10 dB/km. Extrinsic losses of far IR chalcogenides are too high [190]. Andriesh et al. [191] studied the optical losses in fibres in the wavelength range and the influence of the humidity and of the neutron radiation on the optical losses of the fibres. Fibres of diameter and more than 10 m in length were obtained from melt. The investigations in the range have shown the essential dependence of the optical loss spectra on technological defects as microinclusions, cracks, and non-uniform diameter size that all increases the light damping. By neutron irradiation the optical losses in the entire investigated wavelength range increase and, therefore, the defect concentration is increased on account (probably) of the stoichiometry changes. Development and application of chalcogenide optical fibres with high performance features has been impeded by two factors: insufficient degree of purity of the glasses and their tendency towards crystallisation. The data on the effect of impurities on optical losses in optical fibres are very scarce. Churbanov [192] has shown that there are a large number of impurities that greatly affect the optical properties of chalcogenide glasses. The spectral range of transparency of glasses comprises the absorption bands of oxygen, hydrogen, and carbon. The minimum optical losses in the chalcogenide glasses optical fibres fabricated up to day are 23 dB/km at at at [193,194] much higher than the theoretical values. The excessive optical losses in the range are due to hydrogen impurities bonded with oxygen and elemental macro-components, hetero-phase inclusions of submicrometer size and defects in the glass structure. The experimental fact that the optical loss minimum in glass fibres is observed at may be explained by the predominant effect of hydrogen impurities on the optical absorbency. The optical losses in the range are predominantly related to impurities of hydrogen bonded to the glass network and of oxygen bonded to carbon. Impurity absorption in the range is due to carbon and oxides of elemental macro-components. The production of chalcogenide optical fibres with lower losses requires an increase in the degree of purity of the glasses and the development of a drawing technique for optical fibres that precludes the crystallisation of the glass. Due to their prospective optical properties e.g. high transparency in the mid-IR region up to and relatively low refractive index by comparison with chalcogenide glasses the chalcohalide glasses have applications as IR transmitting materials [195]. Their better stability towards crystallisation and chemical durability towards water and moisture show that chalcohalide glasses appear to be good candidates for optical fibre preparation. During the past decade the most extensive research on chalcohalide glasses has focused on Ge-S-X systems Br or I) [196] and on the TeX glasses [197].

354

APPLICATIONS

Lezal et al. [198] focused their attention on the preparation and optical studies of fibres based on As(Ge)-S(Se)-I(Br) glasses. The compositions and have been prepared by synthesis in glass ampoules. Due to a low gflass transition temperature, a new method has been developed for fibre drawing. This method is based on sucking and pumping of the glass melt in capillaries. Teflon capillaries and glasses have been used. Fibres of in diameter and 1 m in length were obtained. The optical losses of the chalcohalides fibres were from 0.9 to 12 dB/m in the diameter range of The glasses have good infrared transmission and are stable. The fibre core is formed from soft glasses at room temperature and therefore the fibres cannot be damaged by a mechanical deformation. Praseodym - doped La-Ga-S - based glasses are at present investigated for all-optical fibre amplification [199]. The LGS glasses are amongst the lowest phonon energy glasses that are suitable for applications as optical fibre amplifier (low phonon energy host discourages non-radiative relaxation of the excited state of Pr). Medeiros Neto et al. [200] claim that an optical fibre amplifier based on LGS

can easily achieve in excess of 20 dB gain across the useful wavelength range

Advantages of using this glass system are: a) a high solubility of the rare earth dopant and b) a low phonon energy host which discourages non-radiative relaxation of the excited state of praseodym. A single lifetime of about for the level in this system was recently observed. Laser power transmission has applications for industrial welding operations and also for microsurgery. surgical lasers operating at have relied on bulky articulated arms to deliver the beam to a micro-manipulator and onto the size for surgery. These may be replaced by more compact fibre optic systems based on chalcogenide fibres. Nishii et al. [163] showed that at a chalcogenide core /clad fibre of GeSe-Te/Ge-As-Se-Te (of size and 100 cm. long) was able to receive 21.8 W c.w. and the transmitted power density was Churbanov [201] has reported the Ge-As-Se fibre of diameter and 1.5 m length transmitted Er/YAG laser radiation at using pulses, giving an output power density of

4.12.2. PLANAR WAVEGUIDES

Excepting the optical fibres the thin films waveguides are fundamental media for the integrated optics. On their basis is studied the construction of the devices for introducing, processing and extraction of the information and its transmission at various distances. The thin film planar waveguides consist of dielectric layers whose refractive index is higher than that of the substrate and of the other layer. This ensures the use of the total reflection phenomenon for directing light without losses. In the same time we must have low losses Losses as low as 0.4 dB/cm were obtained for the case of amorphous composition [202]. High optical transparency of both film and substrate and the absence of light scattering must be ensured for a good waveguide.

CHAPTER 4

355

The most interesting wavelength range for laser communication lines is VIS, near IR and the radiation of of the laser used for processing and transmission of signals of the laser locators. For laser transmission lines the chalcogenide glasses are very promising because these materials show high transparency in infrared, high values of the refractive index (n and can be easily obtained as wires or thin films. Moreover, it is possible to achieve matching with other waveguides like Al-Ga-As or in hybrid structure. In the applications as thin films waveguides the chalcogenide glasses proved to be different from the other materials because they allow for recording the optical information and, in particular, the phase holograms for unlimited resolution and high diffraction efficiency (>80 %). Thin film optical waveguides have been prepared using amorphous layers. The layers have been evaporated in vacuum on glass substrates crystalline quartz or lithium niobate The measurements show that the minimum losses for light in the thin film waveguide is for and 2.462 for respectively. These values show that although the losses are not very low the can be used for thin planar optical waveguides. The change of the optical parameters of the chalcogenide films under the action of light (in particular the change of n) allows to produce special waveguides shaped as narrow parallel bands. They are prepared with a contact mask illuminated by a He-Ne laser The surface scattering has a considerable contribution in a monolayer waveguide due to surface irregularity. In multilayered As-S structure (e.g. the influence of surface roughness is diminished (Fig. 4.24). High performance planar waveguides with attenuation less than 0.5 dB/cm were obtained, by realising a structure of variable refracting index using the same chemical composition and changing only the evaporation temperature [203].

Figure 4.24. Multilayered planar waveguides.

Andriesh et al. [204] produced waveguide gratings on films and performed an integrated optical spectral demultiplexing. Two basically principles were used in fabrication (Fig. 4.25): in plane Bragg’s type deflection (a) and out of the waveguide deflection (b). The effects were achieved by selecting the angle and period of the grating, written by the interference of two coherent rays of laser. For two wavelengths, 630 nm and 1150 nm of a He-Ne laser an efficiency more than 70 % was obtained.

356

APPLICATIONS

Figure 4.25. Waveguides gratings tor spectral demultiplexing devices (1, 2, 3 – coupling prisms).

The modulation of the infrared radiation in waveguides based on has been recently studied by Popescu et al. [205]. The principle of operation for such light-light modulator is shown in figure 4.26. The modulating beam of the argon laser with the power of 40 mW is directed towards the area

between the prism perpendicular to the planar waveguide plane. The modulation coefficient is linear for pump beam intensity higher than The modulation curves (Fig. 4.26) shows that the characteristics of the modulator behaves much slower with the frequency in the case of films but are sufficiently fast (1 ms) in the case of waveguides. The origin of the additional absorption can be explained in the framework of two-steps transition model. The first step is the excitation of excess carriers into the conducting band by highly absorbed pump beam and their partial capture on localised states. The second step is the re-excitation of trapped carriers and restoration of optical absorption. The materials with narrow forbidden gap (e.g. exhibit, also, low activation energy and, therefore, better dynamic characteristics.

Figure 4.26. Light modulation of waveguide beam. a. Principle of the experiment b. Frequency – amplitude characteristics for 1–

and 2–

4.12.3. KINOFORM OPTICAL ELEMENTS

The change of the refractive index in chalcogenide films under the action of the radiation was used for producing optical elements with thin film phase and elements of integrated optics (kinoform elements).

CHAPTER 4

357

In order to record kinoform optical elements the films of with the thickness were used. Such a thickness is optimum for a phase shift up to in the irradiated regions when the recording is performed by an Argon laser and the reading by a He-Ne laser [206]. In order to produce saw-like change of the refractive index in the chalcogenide films the procedure of successive scanning over the film with a light spot in triangular shape has been used. The Argon laser beam was directed to a mask with triangular cut in it. By successive recording such triangular profiles of width a kinoform prism was obtained. Kinoform cylindrical lenses have been recorded in chalcogenides. A more complicated mask consisting of twelve different triangles was used. An array of kinoform cylindrical lenses has been recorded. With this technique it was possible to reproduce kinoform optical elements in quantity, to make complicated spatial phase filters, aspherical lenses and other optical elements which are produced by mechanical treatment of glass with difficulty. By means of the contact exposure it has been made also phase gratings, Fresnel zone plates, random phase masks, a phase filter performing Hilbert transform. The chalcogenide materials allow getting important optical processors. It was possible to produce waveguide lenses, geodesic lenses and Lundberg lenses. The geodesic lens is a non-planar part of the waveguide with the shape of a cupola. Lundberg lens is a three-dimensional lens where the refractive index increases smoothly

from the index value of the waveguide border to a maximum value in the centre of the lens. The creation of these lenses is a difficult task. The advantage of the chalcogenide glasses is the possibility to use the laser for the local modification of the optical

parameters. The Lundberg lenses were obtained by deposition through a diaphragm of variable radius. The correction of the lens needed in this case is carried out by exposing the border part of the lens to a laser radiation. By making use of the photo-induced phenomena in chalcogenide glasses it should be possible to produce diffractive optical elements for use at IR wavelengths. these elements have a potentially large range of uses in the waveband (e.g. simple gratings, mirrors, lenses and beam combiners) and may have advantages over conventional IR refractive elements as regards weight, cost and ease of fabrication. Many chalcogenides exhibit a metal-photo-dissolution effect known as photo doping in which illumination causes metal atoms to dissolve into the glass. Both surface relief and phase gratings have been fabricated using the photo-dissolution of Ag in As-S films and diffraction efficiencies of up to 35 % at 632.8 nm and have been measured [207]. The glassy chalcogenide films are promising materials for passive elements of integrated optics. Light waveguides, prisms, lenses, etc. can be produced with and within thin chalcogenide films using a laser or an electron beam. The holography can be used in the fabrication of optical elements such as grids, Fresnel lenses, etc., with high transmission in IR range, very low level of stray light and diffraction efficiency as high as 85 % [208]. The possibility to make microlenses was demonstrated [209]. The chalcogenide glasses can be shaped under light illumination

358

APPLICATIONS

[210]. The fluidity of the glass is increased by the illumination and, under external tension, the shape of the material can be changed. The mechanism of this increased fluidity is athermal. Microlens arrays can be found in an increasing number of optoelectronic applications, such as optical communication and computing, CCD cameras, and faxes, imaging systems and IR technology. They were fabricated by a variety of techniques, including distributed index planar techniques [211], resin thermal reflow [212] and laser beam ablation [213]. In the last years there was proposed a process, which consists in the direct one-step formation of a three-dimensional microlens array using the dependence of the etching rate on the illumination intensity typical of chalcogenide photoresist [214-216]. In spite of many advantages of the method, the microlens arrays exhibited some drawbacks. The maximum sag was limited to when gray scale or continuous tone lithography with classical UV exposure sources was used. In order to overcome these drawbacks Eisenberg et al. [217] used the thermal reflow method that avoids exposure problems and lead to improvement of the shape of the lenses. Using a binary mask containing holes or slits, island of 3D binary shape can be formed that can be then transformed to 3D plane-convex microlenses. This is done by heating the material close to the melting point, causing reflow and formation of the desired 3D shape. chalcogenide films proved to be a suitable material for fabrication of the microlens arrays. The maximum sag limit caused by photodarkening effect is considerably increased.

4.12.4. CHALCOGENIDE MATERIALS IN OPTOELECTRONICS

The main feature of an optical integrated circuit is the planarity of all functional elements that gives compactness and small volume to the devices that must be produced on a single substrate in a single technological cycle. The materials able to be used in optical integrated circuits must fill the following requirements: a. low optical losses in the spectral range of interest. b. broad range of compositions in order to offer the possibility to select materials with convenient refractive index c. the aptitude to be deposited on various substrates without additional optical losses. d. the aptitude to change the refractive index and the optical losses by external factors. e. particular properties allowing to combine the circuit with the radiation sources and photo-receivers f. high fiability and stability of the elements Many materials have been already tried, as e.g. epitaxial layers of ZnO, ZnS, ZnSe, Cds, CdSe, GaP, GaAs, GaAlAs, InAs, films of melted quartz, epoxy resin, etc. The chalcogenide glasses satisfy the major part of the above requirements. The glasses exhibit high transparency in IR. Thus, the As-S planar waveguide shows optical

CHAPTER 4

359

losses of < 1 dB/cm for The refractive indices of the chalcogenides are high. The chalcogenide glasses are lacking inhomogeneity that are characteristic to polycrystalline films and are responsible for the spatial optical fluctuations. The acoustooptical quality factor is high and the materials are therefore efficient for acousto-optical modulation and for scanning of the laser radiation in a broad spectral range. With the chalcogenide glasses it is possible to get thin planar waveguides with

refractive index anisotropy (see Chapter 4). For example, the measurement of the TE and TM modes on the optical guides gave an anisotropy of the refractive index, The anisotropy can be controlled by additional illumination that induces photo-structural transformations in the material (see Chapter 4). The chalcogenide glasses are strongly influenced by various external factors (optical radiation, electron, neutron and X-ray beams). These particular properties allow for creating special band waveguides, grating structures and other passive or active elements. On the basis of the dynamical changes of the intensity at the output of the thin optical waveguide, induced by the action of the radiation in the vicinity of the fundamental absorption edge of the material, an optical switching element has been developed. With yellow radiation n increases. With IR radiation n decreases. Therefore, with the transmission in the optical wave guide can be controlled.

The mechanism of the dynamical change can be explained qualitatively in terms of the kinetics of photo-excited carriers accompanied by lattice distortion or defect creation. In this devices, the propagation of a light beam in optical waveguides constructed with amorphous As-S films is controlled with blue (and red) light illumination that is able to

modify the refractive index of the films. The devices operate as single-throw or doublethrow photo-optical switches, having switching times ranging between 20 ms and 2 s [218]. The switching time decreases considerably if the yellow radiation intensity increases and the IR radiation has low intensity. The switch-off time depends only on the intensity of the red light and decreases if the intensity increases. Nevertheless, the

decrease of the intensity up to 1 W leads to the heating of the film. That is why the minimum switching time to be reached in this switching element is believed to be no less than 1 ms. In order to improve the switching time it is necessary to operate in the pulsed regime of irradiation. Today the interaction of powerful short pulses of radiation with the optical planar waveguides based on chalcogenide glasses is one of the main preoccupations in the field of technological investigations.

Among the elements for the integrated optics of interest for future developments are: the mono and multilayer planar waveguides, the channel-type planar guide, piezo and acoustic transducers and spatial modulator for the acousto-optical waves, active optical elements based on non-linear effects appeared during the interaction of the light in planar guide. A quasi-equilibrium laser technique for obtaining thin films and nanostructures of multicomponent chalcogenides has been, recently, developed by Kacher et al. [219]. A new class of spatially configured materials has been realised: the amorphous chalcogenide-based superlattices. Trodahl et al. [220] prepared Ge/Se multilayers, superlartices were prepared and investigated by Vateva and Georgieva [221 ] (layer thickness from 2 to 50 nm and the maximum total number of layers: 100) and

360

APPLICATIONS

superlattices were prepared by Imamura et al. [222]. CdSe/Se multilayers (layers of 5 nm thick and a total number of 14 pairs) have been reported by lonov et al. [223]. Detailed research has been done on the electrophotographic applications of Se/Se-Te and Se/CdSe multilayers [224-225]. These multilayers exhibit several properties, which make them appropriate for including in photoreceptors: a. the dark conductivity in a direction perpendicular to the layer planes is lower than the conductivity of the constituent materials, b. the stability against thermal and light treatment is good and increases with decreasing sublayer thickness, and c. the spectral response of the selenium-based photoreceptors can be shifted towards the red part of spectrum in order to be used in the same diode laser printer, without decreasing their stability. Most modern photocopying machines employ 0alloy films with tellurium content up to 5 at.%. Te increases the spectral sensitivity in the red spectral region but reduces the charge acceptance and increases the dark decay rate. If one uses a photo-receptors, and only one transport layer, and include a photo-generation layer of combined with multilayers on top of the structure, a fast photoinduced discharge under illumination not only with nm but also with nm and even at nm can be observed. The sensitivity at 750 nm is higher than that of photo-receptor and two times higher than that of a-Si:H photo-receptors for nm. The stability of selenium in nanolayer structures is considerably increased. No structural change was observed after two years from preparation. Giant photo-induced extension was recently observed in multilayered structure [226]. This effect is closely related to optical recording. Internal stresses and stimulated deformations were considered for the explanation of this effect. The expansion of the spectral sensitivity toward the near-infrared region and fast photo-discharge suggest that amorphous-Se monolayer/superlattice structure would be prospective for electrophotographic and laser printer photo-receivers [227]. The superlattices are also of high interest for getting other optoelectronic devices [228] and an intense research in this field already started.

4.13. Applications of Heavy Metal – Chalcogen Compounds and Alloys Molybdenum disulphide is a diamagnetic semiconductor with a modulated structure consisting of Mo-S layers with strong in-plane bonding and a very weak Van der Waals bonding between the layers. The electronic and magnetic properties of can

be improved significantly by insertion of alkali ions into the van der Waals gap of this layered compound. In these cases the electronic conductivity of may be changed from a semiconductor to a superconductor by intercalation [229]. The compound has been considered as a cathode material for high energy density batteries (see 4.10), owing to its electrochemical properties related to the ability to intercalate/deintercalate lithium ions. The specific energy density of a cell is very low, of about 100 Wh/kg

CHAPTER 4

361

for the utilisation as a primary cell. Haering et al. [230] have discovered that a lithiummolybdenum disulphide compound exhibits several distinct stages of operation when used as a cathode in a battery with a lithium anode. In this case, the cell is reversible and the specific energy density is doubled. Julien et al. [231] have observed a high storage-charge capacity and high mobility for lithium ions in the disordered phase of The relative good behaviour of a disordered cathode is very promising and the use of a polymerbased electrolyte instead of borate glass should produce a more convenient system for future development. Tungsten diselenide is a semiconductor, which can act as an efficient electrode in photo-electrochemical solar cells [232]. An efficiency as high as has been obtained with single crystals [233]. Tungsten diselenide thin films were synthesised by Essaidi et al. [234]. When tungsten is deposited by r.f. sputtering, the layers are amorphous, which explains the better homogeneity of the films and the better adherence to a glass substrate [235]. The method of synthesis is promising for efficient thin films for solar cells. Mercury cadmium telluride is one of the most important semiconductor materials used for IR detector applications. By changing the mole fraction of cadmium in the pseudo-binary alloy, the band gap of Hg-Cd-Te can be varied over a wide range and materials corresponding to both the atmospheric windows in the wavelength range can be obtained. For devices it is necessary to passivate the Hg-Cd-Te surface in order to minimise the leakage currents and to provide

both thermal and mechanical stability. Cadmium sulphide films grown anodically have been successfully used to passivate both n- and p-type surfaces of the basical alloy. Gandotra and Gupta [236] obtained a good passivation by disordered anodic sulphide films grown from non-aqueous solution of and ethylene glycol. REFERENCES

[1] S. R. Ovshinsky, Electronics, Aug. 14, p.76 (1959). [2] A. D. Pearson, W. R. Northover, J. F. Dewald, W. F. Peck, Advances in Glass Technology, Eds. F. R. Matson and G. E. Rindone, part I, p. 367, part II, p. 163, Plenum Press, New York 1962 and 1963. [3] [4] [5] [6] [7] [8] [9] [10] [11]

B.Y. Kolomiets, E. A. Lebedev, Radiotehnika i Elektronika (russ.) 7, 2087 (1963); 8, 1941 (1963). M. P. Southworth, Central. Engin. 11, 69 (1964). D. L. Eaton, J. Amer. Ceram. Soc. 47, 553 (1964). S. R. Ovshinsky, Phys. Rev. Lett., 21, 1450 (1968). D. Adler, Electronics, Sept. 28, p.61 (1970). S. R. Ovshinski, H. Fritzsche, IEEE Trans. on Electronic Devices, ED-20, 91 (1973). R. G. Neale, D. L. Nelson, G. E. Moore, Electronics, Sept. 28, 56 (1970). R. Shanks, C. Davis, IEEE Intl. Solid State Circuits Conf., 1978. S. R. Ovshinsky in Homage Book – Andrei Andriesh, Ed. M. Popescu, INOE&INFM Publ. House,

Bucharest, Romania, 1999, p. 15.

[12] H. K. Henisch, J. -C. Manifacier. R. C. Calarotti, P. E. Schmidtin Physics of Disordered Materials,,

Eds. D. Adler, H. Fritzsche and S. R. Ovshinsky Plenum Press 1985, p.79. [13] M. P. Shaw in Physics of Disordered Materials, Eds. D. Adler, H. Fritzsche andS. R. Ovshinsky,

Plenum Press 1985, p. 793. [14] W. D. Buckley , S. H. Holmberg, Solid State Electronics, 15, 1261 (1975). [15] J. Kotz, M. P. Shaw in Proc. 16-th Intern. Conf. on Phys. of Semicond., Montpellier, France, 1982.

362

APPLICATIONS

[16] K. E. Petersen, D. Adler, J. Appl. Phys., 47, 256 (1976). [17] V. I. Haman, A. I. Badluev, in Structure and Properties of Non-Crystalline Semiconductors (russ.), Nauka, Leningrad, 1976, p.495. [18] A. S. Borschevski, S. P. Vihrov, A. S. Glebov, S. I. Maltsenko, in Structure and Properties of Non-Crystalline Semiconductors (russ.) Nauka, Leningrad, 1976, p.505. [19] K. E. Petersen, D. Adler in Structure and properties of Non-Crystalline Semiconductors (russ.), Nauka, Leningrad, 1976, p. 475 [20] B. T. Kolomiets, E. A Lebedev, E. A. Smorgonskaia, Fizika i Tehn. Poluprovod.(russ.), 6, 2073 (1972). [21] V. F. Korzo, Zarubejnaia Elektr. (russ.) 6, 71 (1971). [22] B. T. Kolomiets, E. A. Lebedev, K. D. Tsendin, Fizika i Tehnika Poluprovod., 15, 304 (1981). [23] W. D. Buckley, S. H. Holmberg, Solid State Electronics, 18, 127 (1975). [24] D. Adler, H. K. Henisch, N. F. Mott, Rev. Mod. Phys., 50, 209 (1978). [25] S. A. Kostylev, V. A. Shkut. V. V. Himinets, Proc. Intern. Conf. “Amorph.Semic. ’80”, Vol. Structure. physico-chemical properties and applications of non-crystalline semiconductors, Stiintza, Chishinau 1980, p.277. [26] V. V. Himinets in Kvantovaia Elektronika, Naukova Dumka. Kiev, 1983, part. 3, p.63. [27] S. R. Ovshinsky, Rom. Rep. Phys., 51(3-4), 167 (1999). [28] K. E. Petersen, D. Adler, Structure and Properties of Non-Crystalline Semiconductors, Nauka, Leningrad, 1976, p. 480. [29] J. Hautala, P. C. Taylor, J. Non-Cryst. Solids, 141, 24 (1992) [30] P. C. Taylor, Z. M. Saleh, J. Z. Liu in Transport, Correlation and Structural Defects , World Scientific, Singapore, Vol.3, p 3, 1990. [31] A. M. Loireau-Lozac’h, M. Guittard, J. Flahaut, Mat. Res Bull., 11, 1489 (1976) [32] M. Kastner, J. Non-Cryst. Solids, 35&36, 807 (1980). [33] V. Bogush, Ph. D. Thesis, University of Bucharest, Romania, 1986. [34] Z. A. Montrimas, A. S. Lukatskaia, Proc. Intern. Conf. “Amorph. Semicond. ’84”, Gabrovo, Bulgaria 1984, Vol. 2, p. 212. [35] M. A. Abkowitz, J. Non-Cryst. Solids, 141, 188 (1992). [36] A. P. Chernov, Ph. D. Thesis, IONH, Akad. Nauk SSSR. Moscow, 1970. [37] A. I. Buzdugan, L. I. Zelenina, Iu. D. Ivaschenko, Proc. Intern.Conf. “Amorph. Semic. ’80”, Chishinau, SSSR, 1980, Vol. 1, p. 249. [38] E. F. Rianeli, L. V. Kaplianskaia, Zhurnal Nauchn. i Prikl. Photo. i Kinematogr., 30, 330 (1985). [39] E. F. Rianeli, L. V Kaplianskaia, E. V. Mikubaeva, Zhurnal Nauchn. i Prikl. Photo. i Kinematogr., 33(2), 105 (1988). [40] E. F. Rianeli, E. V. Mikubaeva, Zhurnal Nauchn. i Prikl. Photo. i Kinematogr..(russ.), 33(4), 2461 (1988). [41] W. Hillen, U. Schiebel, T. Zaengel, Proc. SPIE Conf., Vol. 914, 253 (1988). [42] U. Schiebel, T. Buchkremer, G. Frings, P. Quadflieg, J. Non-Cryst. Solids, 115, 216 (1989). [43] A. Ozols, O. Salminen, M. Reinfelde, J. Appl. Phys., 75, 3326 (1994). [44] J. A. Rowlands, S. Kasap, Physics Today, November 1997, p. 24. [45] W. Zhao, J. A. Rowlands, Med. Phys., 22, 1595 (1995). [46] C. J. Henri, R. K. Rubin, R. D. Cox, P. M. Bret, S. Guberman, J. Optoel. Adv. Mat., 1(2), 49 (1999). [47] M. Pvotasova, A. Ozols, Latvian J. Phys. Tech. Sci., 3, 38 (1994). [48] O. I. Shpotyuk, phys. stat. sol. (b), 183, 365 (1994). [49] N. A. Alimbarashvili, A. D. Galpern, G. G. Dekanozishvili, I. A. Eligulashvili, K. L. Mosulishvili, A. A. Paramonov, V. A. Smaev, Proc. Intern. Conf. “Amorph. Semicond. ’89”. 1989, Vol.1, p.216. [50] V. G. Tsukerman in New recording media for holography (russ.), Nauka, Leningrad, 1983, p.45. [51] V. G. Jdanov, V. K. Malinovski, Pis’ma v JETF (russ.), 3(18), 943 (1977). [52] V. E. Karnatovski, V. G. Tsukerman, Kvantovaia Elektron., 5(6),1394 (1978). [53] Z. U. Borisova, Chemistry of Glassy Semiconductors, Izd. LGU, Leningrad, 1983, p.247. [54] G. Parthasarathy, K. I. Rao, E. S. R Gopal, Phil. Mag. B, 50(3), 335 (1984). [55] A. A. Kikineshi, in Properties of Photosensitive Materials and their Applications in Holography, Nauka, Leningrad 1987, p. 82. S. A. Dembovski, E. A. Chechetkina, J. Non-Cryst. Solids, 64, 95 (1984). [57] P. Shavlandjev, J. Stoykova, T. Todorov, Optics Comm., 24, 67 (1978).

CHAPTER 4 [58] [59] [60] [61] [62]

363

S. A. Keneman, Appl. Phys. Lett., 19(6), 205 (1971). Y. Ohmachi, T. Igo, Appl. Phys. Lett., 20(12), 506 (1972).

H. Sakuma et al., J. Jap. Soc. Appl. Phys., 41, Suppl. p. 76 (1972).

V. I. Mandrosov, E. I. Pik, G. A. Sobolev, Optika i Spektr., 34(6), 1196 and 35(1), 131 (1973). K. K. Shvarts, Ia. J. Kristapson, Physics of the Optical Recording in Dielectrics and in Semiconductors (russ.), Zinatne, Riga, 1986. [63] F. P. Laming, A. D. Pearson, US Patent No. 3594167. [64] R. G. Brands, F. P. Laming, A. D. Pearson, Appl. Optics, 9(7), 1712 (1970). [65] S. B. Gurevich, N. N. Iliashenko, B. T. Kolomiets, V. M. Lyubin, M. V. Suharev, Zhurn. Techn. Fiz., 44(1), 232 (1974). [66] V. I. Mandrosov, E. I. Pik, G. A. Sobolev, G. Z. Vinogradova, S. A. Dembovski, A. P. Chernov, Izv. Akad. Nauk SSSR - Ser. Neorg. Mat., 9(8), 1349 (1973). [67] Chapter: Glassy Semiconductors in Photoelectric Systems for Optical Information Recording (russ.), Ed. A. M. Andriesh, Khishinau, Stiintza, 1988. [68] S. C. Agarwal, H. Fritzsche, Bull. Amer. Phys. Soc., 18(3), 453, (1973). [69] L. V. Beliakov, D. N. Goriacev, Iu. I. Ostrovski, L. G. Paritski, Zhurn. Nauchn. i Prikl. Photo. i Kinematogr., 19(1), 54(1974). [70] S. B. Gurevich et al., Struktura i Svoistva Nekristaliceskih Poluprovodnikov, Nauka, Leningrad, 1976. [71] A.I. Buzdugan, M. S. Iovu, A. A. Popescu, P. G. Cherbari, Balkan Phys. Lett. 1(1), 7 (1993). [72] J. Teteris, O. Nordman, J. Opt. Soc. Am. B 14, 2498 (1997). [73] J. Teteris, O. Nordman, Opt. Comm., 138, 279 (1997). [74] J. Teteris, T. Jaaskelainen, J. Turunen, K. Jefimov, Functional Materials, 6 (3), 580 (1999). [75] A. Roy, A. V. Kolobov, H. Oyanagi, Ka. Tanaka, Phil. Mag., B 78, 87 (1998). [76] J. M. Bellisteros, R. Hernandez, J. M. Herreros, C. N. Alfonso, A. C. Petford-Long, R. C. Dole, Appl. Phys., A 65, 463 (1997). [77] A. Kikineshi, V. Palyok, M. Shiplyak, I. A. Szabo, D. L. Beke, J. Optoel. Adv. Mat., 2 (1), 95, 2000. [78] S. R. Ovshinsky, P. H. Close, Journal Non-Cryst. Solids, 8&10, 892 (1972). [79] A. M. Andriesh, S.D. Shutov, D.I. Tsiuleanu, Information Recording on Non-silver Carriers (russ.)

Visshaia Shkola, Kiev, Ukrayna, 7, 55 (1976). [80] A. M. Andriesh, D. Tsiuleanu, V. G. Dogari, Fundamentals of Optical Memories and Media (russ.),

Visshaia Shkola, Kiev, Ukrayne, 10, 51 (1979). [81] V. V. Bivol, B. V. Efremushkin, M. S. Iovu, I. S. Chumakov, Glassy Semiconductors (russ.), Leningrad, 1985, p.324. [82] A. M. Andriesh, V. V. Bivol, B. V. Efremushkin, M. S. Iovu, The formation of the optical image and its processing method (russ.) Stiintza, Chishinau, 1985, Vol. 2, p.57. [83] M. Andriesh, M. S. Iovu, E. G. Kanchevskaia, Balkan Phys. Lett., 4(1), 1 (1996). [84] D. Y. Lou, G. M. Blom, G. C. Kinney, J. Vacuum Sci. Techn., 18(1), 78 (1981). [85] M. Chen, V. Marello, U. G. Gerber, Appl. Phys. Lett., 41(9). 894 (1982). [86] K. Nishiuki, H. Kitaura, N. Yamada, N. Akahira, Jap. J. Appl. Phys., 37, 2163, 1998. [87] S. R. Ovshinsky, J. Non-Cryst. Solids, 141, 200 (1992) [88] C. Trappe, B. Béchevet, S. Facsko, H. Kurz, Jap. J. Appl. Phys., 37, 2114 (1998). [89] R. Kojima, S. Okabayashi, T. Kashihara, K. Horai, T. Matsunaga, E. Ohno, N. Yamada, T. Ohta, Jap. J. Appl. Phys., 37, 2098 (1998). [90] T. Ohta, Optical Data Storage, Dig. Series, 5, 84 (1991). [91] J. Feinleib, J. Non-Cryst. Solids, 8&10, 909 (1972). [92] J. Feinleib, US Patent, No. 3,636,526 (1971). [93] D. Jecu, S. Zamfira, M. Popescu, Proc. Intern. Conf. “Amorph. Semic. ’89”, Uzhgorod, Ukrayne, Vol. II, p. 195, 1989. [94] M. Popescu, I. N. Mihailescu, Proc. 2-nd Intern. Conf. on Photo-excited Processes and Applications, ICPEPA, Jerusalem, Israel, 1995. [95] L. M. Panasiuk, A. A. Forsh, Proc. Intern. Conf “Amorph. Semic. ’89”, Uzhgorod, Ukrayne, 1989, Vol. 1, p. 199. [96] A. M. Andriesh, D. I. Tsiulyanu, E. P. Kolomeiko, Fiz. Tehn. Poluprovod. (russ.), 11, 664 (1977). [97] A. I. Buzdugan, L. I. Zelenina, L. I. Ivaschenko et al., Zh. Nauchnoi i Priklad. Fotogr. i Kinemat, (russ.), 28(5), 384 (1983); 28(6), 440 (1983).

364

APPLICATIONS

[98] A. I. Buzdugan, L. I. Zelenina, Iu. N. Ivaschenko, M. S. Iovu. A. Simashkievich, S. D. Shutov, Izv. Akad. Nauk Moldov. SSSR, Ser. Fiz. Tehn. i Mat. Nauk, 1, 68 (1987) [99] P. Chaudhari, S. R. Herd, D. Ast, M. H. Brodszky, R. J. von Gutfeld, J. Non-Cryst. Solids, 8&10, 900 (1972). [100] D. J. Gravesteijn, Philips Technical Review, 44, 267 (1989). [101] H. Wang, F. Jiang, F. Gan, J. Non-Cryst. Solids, 112, 291 (1989). [102] S. R. Ovshinsky, J. Non-Cryst. Solids, 141, 200 (1992). [103] R. J. von Gutfeld, P. Chaudhari, J. Appl Phys., 43, 4688 (1972). [104] M. Takenaga, N. Yamada, K. Nishiuki, N. Akahira, T. Ohta, S. Nakamura, T. Yamashita, J. Appl Phys., 54(9), 5376 (1983); M. Takenaga, Ceram. Jap., 19 (4), 313 (1984). [105] Fujutsu Lab. Inform. Lasers and Applications, 4(3), 201 (1985). [106] N. Yamada, E. Ohno, K. Nishiuchi, N. Akahira, J. Appl. Phys., 64, 2849 (1991). [107] Y. Maeda, H. Andoh, I. Ikuta, H. Minemura, J. Appl. Phys., 64, 1715 (1988). [108] H. Iwasaki, M. Harigaya, O. Nonoyama, Y. Kageyama, M. Takahashi, K. Yamada, H. Deguchi, Y. Ide, Jap. J. Appl. Phys., 32, 5241 (1993). [109] T. Handa, J. Tominaga, S. Haratani, S. Takayama, Jap. J. Appl. Phys., 32, 5226 (1993). [110] T. Matsushita, A. Suzuki, T. Nishiguchi, K. Shibata, M. Okuda, Jap. J. Appl. Phys., 34, 519 (1995). [111] J. Feinleib, J. de Neufville, S. C. Moss, S. R. Ovshinsky, Appl. Phys. Lett., 18(6), 254 (1971). [112] R. J. von Gutfeld, Appl. Phys. Lett., 22(5), 257 (1973). [113] Nippon T&T Public Corporation. E. C. L. Technical Publ. No. 67. [114] J. Gonzalez-Hernandez, B. S. Chao, D. Strand, S. R. Ovshinsky, D. Pawlik, P. Gasiorowski, Appl. Phys. Comm., 11(4), 557 (1992).

[115] I. Konstantinov, N. Starbov, Proc. Intern. Conf. “Amorph. Semic. ’84”, Gabrovo, Bulgaria, 1984, Vol. 2, [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133]

p.240 T. Kawaguchi, S. Maruno, K. Masui, J. Non-Cryst. Solids, 77&78, 1141 (1985).

J. M. Oldale, S. R. Elliott, Appl. Phys. Lett., 63(13), 1801 (1993). R. W. Hallman, US Patent No. 3762325, 1973. H. Nagai. A Yoshikawa, Y. Toyoshima, A. Ochi. Y. Mizoshima, “Appl. Phys. Lett.”, 28, 145 (1976). H. Canon, BRD Patent No. 2,201,178, 1972. A. Yoshikawa, H. Nagai, Y. Mizushima, “Jap. J. Appl. Phys.”, 16, 67 (1977). A. Burroff, Proc. Intern. Conf. “Amorph. Semic. ’80”, Chishinau, 1980, Vol.1, p.226.

M. Mathias, Proc. Intern. Conf. “Amorph. Semic.’78”, Pardubice, Czechoslovakia 1978, Vol. 1, p. 215. Bulgar Patent Appl. No. 46640/15.02.980. A. Burroff, Proc. Intern. Conf. “’Amorph. Semic.’84”, Gabrovo, Bulgaria 1984, Vol. 2, p. 234.

M. T. Kostyshin, E. V. Mikhailovskaia, P. F. Romanenko, “Fiz. Tverd. Tela” (russ.), 8, 451 (1966). A. Burroff, Ph. D. Thesis, Sofia, 1982, p. 86. A. N. Klimin, V. G. Tsukerman, Proc. Intern. Conf. “Amorph. Semic.’78”, Pardubice, Czechoslovakia, 1978, Vol.2, p.680. N. Kozicki, S. W. Hsia, A. E. Owen, P. J-S. Ewen, “J. Non-Cryst. Solids”, 137&138, 1341 (1991). C. H. Tzinis, C. H. Chen, C. T. Kemmerer, Proc. Symp. Resist. Syst., Eds. A. Doane and A. Heller, The Electrochem. Soc Inc., Vol. 82-9, 1982, p. 157. K. L. Chopra, L. K. Malhotra, K. Solomom, B. Singh, Proc. Symp. Resist. System, Eds. A. Doane and A. Heller, The Electrochem. Soc. Inc., Vol. 82-9, 1982, p. 129. M. Mitkova, Proc. Intern. Conf. “Amorph. Semic. ’89”, Uzhgorod, 1989, Vol. 1, p. 222. D. Arsova, E. Vateva. M. Nikiforova, E. Skordeva, E. Savova, Proc. 6-th Intern. School on Physical Problems in Microelectronics (23-28 May. 1989) Varna, Bulgaria, Ed. J. Kassabov, World Scientific, 1990, p. 267 K. Balasubramanian, A L. Ruoff , J. Vac. Sci. Technol., 19, 1374(1981). R. Klabes, A. Thomas, R. Grotzschel, P. Süptitz, phys. stat sol. (a), 101, K9 (1987).

[134] [135] [136] M. Popescu, F. Sava, A. Lörinczi, E. Skordeva, P. -J. Koch, H. Bradaczek, J. Non-Cryst. Solids, 227, 719(1998). [137] J. W. Boag, Phys. Med. Biol., 18, 3 (1973). [138] K . Saito, Y. Utsugi, A. Yoshikawa, J. Appl. Phys., 63, 565 (1988). [139] K. Jefimov, M. Honkanen, P. Laakkonen, J. Turunen, T. Jaaskelainen, J. Teteris, Functional Materials, 6(3), 569(1999).

CHAPTER 4

365

[140] C. Popescu, T. Stoica in Thin Film Resistive Sensors, IOP Publ. Ltd. 1992, Eds. P. Ciureanu and

S. Middlehoek, p. 37. [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155]

E. Maruyama, T. Hirai, J. Non-Cryst. Solids, 59&60, 1247 (1983). E. Murayama, Jap. J. Appl. Phys., 21, 213 (1982). Y. Taniguchi, H. Yamamoto, S. Hirogome, J. Appl. Phys., 52, 261 (1981). E. V. Gernet et al., Elektronnaia promishlennosti. No. 10, 1979, p. 22.

M.F. Jedlicka, P. Kulhanek, J. Mravinac, D. Lezal, Proc. Intern. Conf. “Amorph. Semic ’78”, Pardubice 1978, Vol. 2, p. 617. L. Austin, H. Starke, Ann. Phys., 9, 271 (1902).

J. D. Mackenzie, J. Non-Cryst. Solids, 26(1-3), 456 (1977) J. Adams, B. W. Manley, Electr. Eng., 37, 180 (1965). T. Imaoka, Mater. Electr. Jap., 51 (1969). B. W. Manley, A. Guest, R. T. Holmshow, Advanc. Electron. Electr. Phys., 28A, 471 (1969). B. T. Kolomiets, V. M. Lyubin, Fiz. Tverd. Tela (russ.), 4, 401 (1962). R. Andreichin, P. Simidtcheva, M. Nikiforova, C.R. Bulgarian Acad. Sci., 18, 1079(1965). R. Andreichin, J. Non-Cryst. Solids, 4, 73 (1970). R. Andreichin, M. Baeva, E. Skordeva, A. Alexandrova, C.R. Bulgarian Acad. Sci., 24, 1465 (1971). R. Andreichin, Proc. Intern. Conf. “Amorph. Semic. ’74”, Rheinhardsbrunn 1974, Akad. der Wiss., DDR, ZIE, Vol. 1, p.83.

[156] M. Popescu, J. Non-Cryst. Solids, 192&193, 140 (1995), [157] R. Andreichin, Proc Intern. Symp. on “Light and Rad. Measur. ’81”, Hájduszóbószló, Hungary 1981, p. 88, 92. [158] R. Andreichin, Proc. Intern. Conf. “Amorph. Semic. ’82”, Bucharest 1982, Vol. E (english) p. 246. [159] R. Andreichin, M. Baeva, L. Yurukova, D. Arsova, Proc. Intern. Conf. “Amorph. Semic. ’76”, Balatonfüred, Hungary, 1976, Ed. I. Kósa-Somogyi, Akadémiai Kiadó, Budapest, p.47. [160] M. Laine, A. B. Seddon, J. Non–Cryst. Solids, 184, 30 (1995). [161] L. Abdulhalim, C. N. Pannell, R. S. Deal, D. W. Hewak, G. Wylangowski, D.N. Payne, J. Non-Cryst. Solids, 164, 1251 (1993). [162] A. B. Seddon, M. J. Laine, Physics and Applications of Non-Crystalline Semiconductors in Optoelectronics, NATO AS1 Series, Vol. 36, Kluwer Acad. Publ., 1997, p. 327 [163] J. Nishii, S. Morimoto, I. Inagawa. R. Iizuka, T. Yamashita, T. Ysamagishi, J. Non-Cryst. Solids, 140, 199 (1992). [164] M. Maeda in Chemical Sensors Technology 3, Ed. N. Yamazoe, Kodansha, Tokyo, 1991, p. 185. [165] A.E. Owen, J. Non-Cryst. Solids, 35&36, 999 (1980). [166] Y. Hamakawa, J. de Physique (Fr.) Coll. C4, Suppl. au No. 10, 42, 1131 (1981). [167] J. T. Edmond, J. Sci. Instr., 21, 373 (1968). [168] A. V. Kirsanov, S. V. Rumiantsev, V.F. Anisimov, A.N. Pokotii, I. Yu. Andreichenko, Proc. Intern. Conf. “Amorph. Semicond. ’89”, Uzhgorod, Ukrayne, 1, 264 (1989). [169] J. Portier, J. Non-Cryst. Solids, 112, 15 (1989). [170] A. J. Jacobson, R. R. Chianelli, S. M. Rich, M. S. Whittingham, Mat. Res. Bull., 14, 1437 (1979). [171] D. W. Murphy, P. A. Christian, J.J . DiSalvo, J. N Carides, J. Electrochem. Soc., 126, 497 (1979). [172] E. A. Kishitskaya, V. B. Nosov, V. F. Kokorina, Fiz. Him. Stekla (russ.), 3, 624 (1977). [173] A. R. Hilton, D. J. Hayes, M. D. Rechtin, J. Non-Cryst. Solids, 17, 319, 339 (1975). [174] R. B. Johnson in Proc. Soc. Photo-Optical Instrumentation Engineers, SPIE, 915, 106 (1988). [175] Z. Cimpl, F. Kosek, phys. stat. sol.(a), 93, K55 (1986). [176] Z. Cimpl, F. Kosek, J. Non-Cryst. Solids, 90, 577 (1987). [177] A. M. Loireau-Lozac’h, M. Guittard, J. Flahaut, Mat. Res. Bull., 11, 1489 (1976). [178] S. Benazeth, M. H. Tullier, A. M. Loireau-Lozac’h, H. Dexpert, P. Lagarde, J. Flahaut, J. Non-Cryst. Solids, 110, 89 (1989). [179] J. A. Savage, K.L. Lewis, Proc. SPIE, 683, 79 (1986). [180] C. H. Kwak, J. T. Kim, S. S. Lee, Appl. Optics, 28(4), 737 (1989). [181] C. H. Kwak, S. Y. Park, H. M. Kim, E. H. Lee, C. M. Kim. Optics Comm., 88, 249 (1992). [182] T. Katsuyama, H. Matsumura, Infrared Optical fibres, Adam Hilger, Bristol, 1989 [183] J. M. Parker, A. B. Seddon in High Performance Glasses, Eds. M. Cable and J. M. Parker, Blackie, London, 1992, chapter 13.

366

APPLICATIONS

[184] V. O. Abashkin, A. M. Andriesh, V. V. Ponomari, A. A. Popescu, Proc. Intern. Conf. “Amorph. Semic. ’80”, Khishinau, SSSR, 1980, Vol. 1, p.255. [185] J. Parant, C. LeSergent, D. Guinot, C. Brehm, 1st Intern. Conf. On Halides and other Non-Oxide Glasses, Cambridge, 1984. [186] T. Katsuama, H. Matsumura, Proc. Intern. Conf. on Optoelectronics and Laser, Moscow, 1986. [187] D. B. Petrovská, G. Kuncová, M. Popišilová, J. Götz, New Materials for Optical Waveguides, SPIE, Vol. 799, 1987, Ed. J. Lucas, p. 47. [188] J. Lucas, G. Fonteneau, A. Bonnagad, Z.X. Hua, New Materials for Optical Waveguides, SPIE Vol. 799, 1987, Ed. J. Lucas, p. 101. [189] J. Nishii, T. Yamashita, T. Yamagishi, Appl. Opt., 28, 5122 (1989). [190] N. J. Pitt, Proc. SPIE, Vol. 799, 25 (1987). [191] A. M. Andriesh, O. V. Bolshakov, I. P. Kuliak, V. V. Ponomari, A. S. Smirnova, Proc. Intern. Conf. “Amorph. Semic. ’84”, Gabrovo, Bulgaria 1984, Vol. 2, p. 244. [192] M. F. Churbanov, J. Non-Cryst. Solids, 184, 25 (1995). [193] G. G. Deviarykh, E. M. Dianov, V. G. Plotnichenko, I. V. Skripachev, M. F. Churbanov, High Purity Substances, 5, 1 (1991). [194] A. V. Vasiliev, G. G. Deviatykh, E. M. Dianov, I. V. Skripachev, G. E. Snopatin, M. F. Churbanov, V. G. Plotnichenko, Kvantov. Elektron. (russ.), 20, 109 (1993). [195] F. Gan, J. Non-Cryst Solids, 140, 184 (1992). [196] J. Heo, H. Nasu, J. Mackenzie, Proc. SPIE, 683, 85 (1986). [197] X. H. Zhang, G. Fonteneau, J. Lucas, Mat. Res. Bull, 23, 59 (1988). [198] D. Lezal, B. Petrovská, R. Vogt, J. Götz, Proc. Intern. Conf. “Amorph. Semic. ’89”, Uzhgorod, 1989, Vol. 1, p. 261. [199] A. B. Seddon, J. Non-Cryst Solids, 184, 44 (1995). [200] J. A. Medeiros Neto, E. R. Taylor, B. N. Samson, J. Wa ng, D. W. Hewak, R. I. Laming, D. N. Payne, E. Tarbox, P. D. Maton, G. M. Roba, B. E. Kinsman, R. Hanney, J. Non-Cryst. Solids, 184, 292 (1995).

[201] M. F. Churbanov, J. Non-Cryst. Solids, 140, 324 (1992). [202] S. Zembutsu, S. Fukunishi, Appl. Optics, 18, 393 (1979). [203] A. Popescu, Homage Book dedicated to Acad. A. Andriesh, Ed. M. Popescu, INOE & INFM Publ. House, Bucharest, 1999, p. 211. [204] A. Andriesh, A. Popescu et al., Deposited report number 02.88.0035155, VINITI, Moscow (1998). [205] A. Popescu, A. Albu, A. Tsaranu, Proceedings of SPIE (USA), Vol. 3405, p. 902 (1998). [206] V. V. Koronkevich, V. G. Remesnik, V. A. Fateev, V. G. Tsukerman, Proc. Intern. Conf. “Amorph.

Semic. ’76”, Balatonfüred, Hungary 1976. Ed. 1. Kósa-Szomogyi, Akadémiai Kiadó, Budapest, 1977, p. 41. [207] C. W. Slinger, A. Zakery, P. J. S. Ewen, A. E. Owen, Appl. Optics, 31, 2490 (1992). [208] I. Z. Indutnyj, A. V. Stronski, S. A. Kostyukevitch, P. F. Romanenko, P. E. Schepeljavi, I. I. Robur, Optical Eng., 34, 1030 (1995). [209] H. Hisakuni, Ke. Tanaka, Optics Lett., 20, 958 (1995). [210] H. Hisakuni, Ke. Tanaka, Science, 270, 974 (1995). [211] M. Oikawa, Appl. Opt., 21, 1052 (1982). [212] Z. D. Popovic, R. A.Sprague, G. A. N. Connel, Appl. Optics, 27, 1281 (1988). [213] M. Terao, K. Shigematsu, M. Ojima, Y. Taniguchi, S. Horigome, S. Yonezawa, J. Appl. Phys., 50, 6881 (1979). [214] N. P. Eisenberg, M. Manevich, M. Klebanov, S. Shtutina, J. Non-Cryst. Solids, 198-200, 766 (1996). [215] V. Lyubin, M. Klebanov, I. Bar, S. Rosenwaks, N. P. Eisenberg, M. Manevich, J. Vacuum Sci. Technol, B 15, 823 (1997). [216] S. Noach, M. Manevich, M. Klebanov, V. Lyubin, N. P. Eisenberg, Proc. Int. Conf. SPIE (USA). Vol. 3778, 151 (1999). [217] N. P. Eisenberg, M. Klebanov, V. Lyubin, M. Manevich, S. Noah, J. Optoel. Adv. Mat., 2 (2), 147 (2000). [218] K. Tanaka, Y. Imai, A. Odajima, J. Appl. Phys., 57(11), 4897 (1985). [219] I. E. Kacher, V. M. Zhikarev, N. I. Dovgoshey, N. I. Popovich, Functional Materials, 6(3), 443 (1999). [220] H. J. Trodahl, M. W. Wright, A. Bittar, Solid State Comm., 59(10), 699 (1986). [221] E. Vateva, I. Georgieva, J. Non-Cryst. Solids, 114, 124 (1989) and 164&166, 865 (1993).

CHAPTER 4 [222] [223] [224] [225] [226] [227] [228] [229] [230] [231] [232] [233] [234] [235] [236]

367

S. Imamura, Y. Kanemitsu, M. Saito, H. Sugimoto,. J. Non-Cryst. Solids, 114, 121 (1989). R. Ionov, D. Nesheva, D. Arsova, J. Non-Cryst. Solids, 137&138, 1151 (1991), D. Nesheva, D. Arsova, E. Vateva, Semic. Sci. Tehnol., 12, 595 (1997). E. Vateva, G. Tschaushev, J. Optoel. Adv. Mat., 1(2), 9 (1999). V. Palyok, M. Malyovanik, J. Optoel. Adv. Mat., 1(3), 77 (1999). E. Vateva, D. Nesheva, Amorphous Chalcogenide Superlattices, ISSP, Preprint, Sofia, Bulgaria 1995. R. Ionov, Europhys. Lett., 19, 317 (1992). R. B. Somoano, V. Hadek, A. Rembaum, J. Chem. Phys., 58, 697 (1973). R. R. Haering, J. A. R. Stiles, K. Brandt, US Patent 4,224,390 (1980). C. Julien, S. I. Saikh, G. A. Nazri, Mat. Sci. Eng., B15, 73 (1992). H. Tributsch, Ber. Bunsengesell., Phys. Chem., 84, 361 (1977). G. Prasad, O. N. Srivastava, J. Phys. D, 21, 1028 (1988). H. Essaidi, J. C. Bernede, J. Pouzet. M. Zoaeter, Mat. Sci. Eng., B26, 67 (1994). J. Pouzet, J. C. Bernede, A. Khelil, H. Essaidi, S. Benhida, Thin Solid Films, 208, 252 (1992). V. K. Gandotra, S. C Gupta, Mat. Sci. Eng., B26, L1 (1994)

368

Subject Index

A ablation (laser technique) acoustic (constant, wave) acoustic transducer acousto-optical devices activation energy

Ag

Ag-Se

amorphisation amorphous selenium arsenolite

As-Se-Sn athermal (process) Au Auger (process)

358 224, 272, 347, 348. 359 347 67, 105, 110, 112, 115-117, 126, 129, 132, 135, 137, 138, 142, 148, 149, 151-153, 157, 161, 165, 166, 172, 173, 175, 176, 178, 179, 182, 184, 185, 189, 192, 195, 196,

199, 200, 219, 223, 278, 279, 283, 296, 298, 350, 356. 69-71, 199, 218-221, 275-277, 298, 327, 328, 331, 332, 342, 357. 332. 211,212,214,215. 10, 12, 13, 61, 128, 132, 211, 215, 226, 227, 247, 248, 277, 302, 303, 307, 310, 333. 60, 218. 60, 70, 217-219, 226, 267, 341. 24-26, 29, 54, 60, 63, 69-71, 81, 84, 112, 119, 132, 152157, 160, 162, 163, 166, 169-171, 183, 210-211, 214220, 222-225, 227, 229, 230, 233, 235-241, 243, 244, 251-253, 256-258, 264-267, 270, 272-275, 279-282, 286, 305, 308, 309, 311-313, 319, 321, 322, 328-330, 332, 338, 341, 342, 348, 351-353, 355-357, 359. 24-26, 54, 60, 63, 69, 70, 81, 109, 112, 119, 132, 152157, 160, 162, 163, 170, 171, 183, 210, 211, 215, 216220, 222-225, 229, 233, 235, 237-241, 243, 244, 251253, 256-258, 260, 264-267, 270, 271, 273-275, 279-282, 285, 286, 305, 308, 309, 311-313, 319, 321, 322, 328, 329, 332, 341, 342, 348, 351-353, 355-357, 359, 360. 28, 29, 48, 112, 157, 160, 162, 164-166, 169-171, 229, 275, 285, 306, 339, 352. 27. 26, 27, 29, 158, 222, 223, 230, 251. 27, 29, 158, 248, 257. 312, 313. 214, 264, 265, 275, 358. 13,33,67,71. 269.

Subject Index

369

B

batteries (solid state) biphotonic (process) birefringence bistability (optical) bond energy bulk glass

66, 350, 360. 255, 256. 183, 234, 235, 237, 238, 240, 260, 282. 210 23,43, 105, 107, 108, 127, 157, 160, 176.

31, 40, 109, 153, 160, 161, 163, 169, 183, 185, 189, 192, 194, 209, 218, 224-226, 234, 238, 271, 273, 280, 286, 301, 340.

C

chalcohalide glasses change carrier

350, 353. 75, 116, 117, 161, 195, 212, 213, 222, 249, 251, 257, 262, 264, 275, 302, 303, 314, 322, 334, 337, 338, 343, 356, 359.

cluster

33, 37, 70, 74, 85, 86, 160, 211, 213, 252, 253, 255, 257, 258, 264, 275, 280, 281, 332. 220, 256, 281, 348.

connectivity

continuous random network crystallisation

214, 281.

6, 7, 17-19, 23, 29, 33, 37, 42, 46-49, 51, 53, 54, 68, 69, 71, 79, 90, 104, 107, 110-112, 115, 125, 128, 129, 134, 136-142, 144, 146, 148, 151, 152, 158, 165-167, 171, 173-175, 178, 181, 184, 185, 188, 190, 193, 194, 196, 201, 209-214, 275, 277-279, 283, 294, 296, 298, 303, 304, 313, 315, 319, 320, 321, 323, 325, 328, 336, 352, 353.

D

dangling bond defects deformation density

density of states (DOS)

117, 121, 215, 220, 222, 281, 343. 81, 83, 90, 106, 110, 116, 118, 123, 209, 218, 222, 225, 226, 237, 238, 242, 243, 249, 250, 253, 254, 262, 263, 267, 272, 274, 278, 301-303, 328, 353, 359. 13, 33, 39, 113, 216, 259, 264, 275, 276, 281, 282, 313-

315, 321, 354, 360. 8, 12, 16, 31, 68, 75, 111, 116, 125-129, 131, 134, 139, 140, 144, 147, 148, 150, 152, 156, 158, 159, 163, 164, 167, 170-172, 181, 184, 190-192, 194, 197,200,209, 223-225, 248, 266, 170, 275, 281, 297, 347. 116, 119

370

depolarisation deposition

deposition rate detectors (photo) detectors (x-ray radiation)

device

239, 240, 246, 342-344. 12, 15, 17, 23, 26, 29, 31, 40, 73, 78, 109, 110, 129, 130, 134, 183, 199, 201, 210, 221, 222, 227, 313, 324, 332,

340, 357. 109, 129,210. 317 307, 308, 338, 340, 343, 361, 110, 221, 293, 294, 297-300, 301, 307, 316, 318, 332, 338, 340, 341, 343-345, 347-349, 352, 354, 356, 358-

361. 120-122. 128, 134, 157, 163, 197, 360. diamagnetic 234-237. 240, 241, 260, 262, 282. dichroism 119, 132. dielectric (loss) dielectric (properties, solid) 116, 119, 132, 148, 151, 153, 185, 191, 198,201,275,

dielectric constant dimerization disorder

doping

298, 299, 346, 354. 120, 127, 132, 143, 148, 154, 156, 161, 172, 191, 198, 224. 246, 256. 105, 113, 118, 122, 135, 215, 230, 254. 61, 65, 70, 93, 116, 195, 218-221, 272, 298, 301, 303-305, 311, 313, 320, 328, 331, 333, 336, 341, 342, 354, 357.

E

effusion electrical gap

218 173, 195,

electrical properties electron beam electron spin resonance electrophotography Elliott's model eutectic

33, 60, 65, 110, 114, 221, 322. 216, 272-276, 323, 327, 331-334, 337-339, 357. 249, 301. 303, 306. 252 18-20, 33, 38, 48, 52, 53, 61, 73, 82, 84, 85, 94, 165, 185 192, 193, 196, 197, 213, 320.

F

Fermi level first co-ordination first sharp diffraction peak fractal Fritzsche's model

116, 119, 121, 175, 260, 301. 19, 242. 281, 246. 254 253, 254.

Subject Index

371

G

gamma radiation gamma resonance (NMR) gamma-induced (transf.) gap states Ge-As-S Ge-As-Se Ge-As-Te GeS

266, 270-272. 40, 201. 272. 301. 109, 181-183, 185, 191, 192. 37, 39, 52, 109, 183-185, 190, 191, 285. 46, 47, 72, 114, 216, 332, 333.

47, 72, 90, 351, 352, 354. 48, 57, 89, 111, 298. 34-36, 41, 52, 53, 181, 183, 232. 34, 36, 37, 46, 53, 82, 86, 92, 109, 181-183, 215, 227, 231, 245, 264, 275, 281, 282, 285. 34, 36, 37, 46, 53, 82, 86, 109, 181-183, 215, 227, 231, 245, 264, 275, 281, 282, 285. 48-50.

Ge-Sb-S Ge-Sb-Se Ge-Sb-Te GeSe

37, 48-50.

50, 51, 320, 323, 325. 34-36, 38, 52, 184, 231, 285.

34, 36, 37, 112, 184, 186, 187, 212, 214, 215, 227, 231, GeTe Ge-Te glass transition glass transition temperature grating

Grigorovici's model

237, 245, 275, 283, 285, 286, 332. 37, 38, 41, 48, 53, 84, 90, 91, 188, 189, 319, 323, 325. 33, 37-39, 89, 188-190, 313, 319.

51, 76, 112, 113, 115, 139-141, 147, 148, 158, 167, 169, 172, 174, 176, 177, 181, 184, 190-193, 199, 201, 216, 227, 264, 269, 272, 283, 285, 333. 76, 112, 113, 115, 139, 147, 148, 167, 174, 176, 177, 181, 191-193, 199, 227, 269, 285, 333. 309, 310, 313, 328, 330, 347, 351, 355-357, 359. 246.

H

hardening hardness

heat capacity heteropolar bond heterostructure high-pressure (form) high-pressure (phase) holography

165, 216, 224, 269, 279, 284. 65, 75, 111, 114, 126, 129, 145, 150, 159, 177, 190, 191, 209,216,224,225,269. 114, 115, 130, 139, 140, 159, 167, 169. 274. 314, 315, 322, 338. 15, 19, 27-29, 107. 41, 127, 152, 158, 280, 281. 308,311-315,321,357,362.

372

homopolar bond hopping

220, 221, 243, 251, 252, 254, 257, 350. 116, 119, 120.

I

index of refraction

information recording infrared radiation integrated optics interlayer distance IR-absorption irreversible (process)

122, 125, 127, 133, 135, 144, 145, 149, 156, 157, 161-163, 169, 170, 183, 186, 187, 191, 194, 210, 224,

230, 233, 237, 272, 305, 308, 310, 323, 329, 347, 348, 351, 353-359. 315,316,319. 350, 351, 359. 352, 354, 356, 357, 359. 24, 247, 255. 260 209, 210, 211, 213, 218, 222,227, 231, 232, 238, 253, 272, 311, 314, 315, 325.

K

kinoform (optical element) Kolobov's model

356, 357. 247.

L

laser irradiation light modulator localised states lone-pair (electrons) long-range order luminescence

160, 228, 278, 312. 356. 116, 118, 121, 239, 263, 275, 297, 356. 106, 248, 253, 254, 257, 258, 261, 263, 269.

26, 108, 115, 208, 209, 213. 234, 301.

M

magnetic properties Malinovski's model mechanical properties medium range order melting point memory switch

149, 150, 152, 360. 250.

107, 140, 148, 267, 279, 285, 351, 352. 236, 252, 254. 6, 8, 19, 31, 134, 158, 188, 199, 319, 323, 358. 293, 294, 296.

Subject Index

metastability (of glass) microcrystalline microhardness

373

103,211,271. 70, 127. 65, 112-114, 126, 131, 139, 140, 144, 147, 148, 158, 159, 164-166, 172, 184, 191, 224, 225, 246, 248, 270-272, 274, 279, 303, 313. 357, 358. 116, 122, 227.

microlens mobility edge molar volume molecular crystal Mott law multilayers

67, 122, 146, 147, 167, 168, 172, 173, 181, 200. 14, 15, 22, 24. 116, 135.

317, 359, 360.

N

nano-(layer, structure) non-bonding electrons

73, 359, 360. 47, 209, 222, 262.

O

optical absorption

117, 118, 127, 137, 138, 144, 155, 173, 186, 210, 217, 222, 224, 225, 227-230, 232, 243, 254, 257, 269, 280, 282, 329, 356.

optical anisotropy optical bistability optical disk optical fibre optical gap

225, 234, 259

210. 316, 317, 319, 323, 324, 326, 327. 348, 349, 352, 353, 354, 366. 118, 127, 132, 135, 143, 161, 169, 181, 183, 186, 189, 195, 224, 226, 229, 230, 231, 239, 248, 253, 257, 306, 331. 298, 321, 325.

optical memory optical transmission optical transparency orpiment (cryst. ovonic device

)

117, 173, 217, 227, 231, 273, 274, 346, 347. 60, 127, 210, 225, 237, 306, 315, 320, 347, 351, 353-355, 358. 25, 28, 60, 153, 214, 275. 293.

P P2S3 P2Se3 P2Te 3 paramagnetic (term) particle beam

21, 71, 285.

23, 54, 69, 71, 88, 112, 147, 149. 23, 150. 121, 122, 152, 157, 188, 190. 126, 209.

374

permittivity phase filter phase-change (switch) phonon photo-ablation photo-amorphisation photobleaching photo-chemical (reaction) photoconduction photo-contraction photo-crystallisation photodarkening photo-decomposition photo-dissolution photo-doping photo-evaporation photo-excitation photo-expansion photo-hardening photo-induced anisotropy photo-induced fluidity photo-induced gyrotropy photo-induced scattering photolithography photo-oxidation photo-polymerisation photo-receptor photoresist photo-sensitivity photo-structural (process) photo-thermal (recording) physico-chemical (propert.) pnictide (element) polarization polymerisation Popescu's model

148, 155, 161, 172, 200, 274. 330, 357. 293, 294, 298, 324, 325. 119, 120, 222, 246, 251, 309, 347, 354. 268 211,214,215. 210, 217, 218, 225, 226, 230-233, 248, 268. 217,221,328. 257, 269, 273, 286, 341. 149, 215. 210-212, 214, 283, 313, 323. 162, 199, 210, 223-236, 246, 250, 254, 255, 268, 269. 214. 218-221,357. 70, 218-221, 328. 210. 249, 253, 264, 268. 264 224 213, 234, 236, 237, 239, 260-263. 264-266, 275, 358. 260, 262. 239, 241. 328, 333. 211, 218, 232, 267. 222, 223, 246. 307, 360.

328-332, 358. 301, 304-307, 311, 318, 322, 324. 224, 225, 248, 250, 252, 263, 264, 267. 315, 317-319, 322, 324, 325. 126, 278, 280, 298. 21, 146, 226, 260. 116, 119, 155, 214, 234-237, 240, 259, 260, 262, 263, 341-345. 210,222,223,246,271. 251, 257.

R

Raman (scattering, effect) reflectivity relaxation (process, time)

220, 260.

132, 238, 319, 323, 324, 329. 120, 122, 130, 155, 210, 215, 222, 243, 245, 251, 253, 258, 260, 264, 268, 269, 275, 279, 281, 283, 284, 310,

Subject Index

reversible (process)

375

313, 341-343, 354. 209, 210, 212, 213, 217, 222, 224, 225-230, 232, 236, 238, 239, 241, 246, 248, 251, 253, 255, 256, 271-274, 302, 304, 309, 313, 314, 318, 319, 323, 324, 327, 361.

S

Sb-S Sb-Se

scalar effect sensor (chemical) sensor (optical fibre) sensor (radiation) Shimakawa's model short-range order silver SiS

4, 29, 30, 31, 66, 109, 171, 172, 174, 236, 306, 322, 335, 337, 338, 342, 352, 29, 30, 37, 46, 55, 173, 283, 305, 306, 334, 352. 29, 31, 171, 173. 29, 31, 173, 174. 256. 348. 93, 349. 338, 344. 254. 107, 115, 188. 69-71, 199, 218, 222, 327, 331, 332, 350. 31, 33, 176. 32, 33, 67, 176, 194.

SiSe

SnS Sn-S SnSe Sn-Se

softening temperature

S-Se Street's model sulphur (amorphous) sulphur (plastic) switch

switching (electrical)

31.

32, 33, 58, 112, 176. 193. 40. 38-40, 52, 193. 39,40, 193. 39,40, 193. 39, 40, 52, 53, 193. 39, 193. 39, 40, 51, 94. 19, 23,42, 63, 68, 76, 80, 89, 93, 103, 107, 109, 113,

138, 146, 148, 153, 158, 166, 169, 172, 181, 188, 191, 194, 211, 224, 225, 273, 282, 298, 303, 313, 320, 347, 352. 17, 18, 29, 108, 126, 137, 145, 213, 320, 333. 242, 243. 9, 128,217,227. 6,9, 127. 89, 90, 110, 172, 173, 179, 201, 211, 213, 222, 223, 225, 234, 236, 251, 260, 262, 277, 278, 293-298, 300-302, 314, 323, 342, 343, 349, 352, 359. 89, 90, 110, 172, 179, 201, 293-302, 314, 342, 343, 349, 352, 359.

376

T

Tanaka's model

243, 248. 90, 296, 299.

(memory glass) 90, 212, 325.

(memory glass) (memory glass) Te-S Te-Se thermal conductivity thermo-plastic thin film

39, 90. 18-20, 29, 141, 310. 18, 131, 138, 140-144, 212, 304, 310, 312. 127, 129, 134, 298, 319. 126, 302, 321, 322. 37, 109, 115, 127, 137, 172, 174, 176, 179, 184, 188,

threshold switch Tl-Te transmittance

195, 200, 211, 214, 217, 218, 223, 226, 238, 246, 266, 273, 293, 297, 301, 302, 318, 321-323, 332, 345, 349, 350,354-356,359,361. 293, 294, 296-298. 42, 195. 236, 239-241, 267, 319, 351.

U

ultrasound (speed, wave) ultraviolet (UV) undercooling Urbach (rule, tail, edge) UV-ablation

148, 279, 282. 199, 216, 218, 266-269, 328, 330, 332, 333, 338, 358. 104, 128, 283, 284. 118, 125, 235, 236, 241, 254, 263, 264, 266. 269.

V

valence alternation pair Van der Waals (force, bond) vectorial effect Vidicon (pick-up tube) viscosity

vitreous

238, 243, 249, 260, 302, 250, 255, 256. 9, 12, 14, 15, 20, 24, 26, 31, 85, 145, 165, 244, 252, 255, 280, 282, 360. 256, 260.

333-336, 338. 8, 104, 134, 135, 155, 160, 167, 170, 167, 278, 279, 283, 284, 314. 11, 13, 21, 29, 47, 51, 56, 60, 61, 63, 64, 66-69, 71-73, 75, 78-80, 84, 85, 87-89, 93, 103, 104, 107, 110, 115, 128-130, 136, 139, 140, 142, 143, 145, 149, 159, 163, 164, 172, 178, 247, 272, 277, 278, 304, 321, 329, 341-343,350.

Subject Index

33, 34, 104, 113, 213, 266, 330, 351, 358.

void W

waveguide (optical) wrong bond

354-359. 121, 231, 250, 254.

X

xerography x-ray absorption (EXAFS) x-ray diffraction x-ray image

86, 203, 273, 302, 303, 306. 18, 253, 258, 260. 18, 19, 28, 70, 138, 158, 195, 201, 246, 269, 274. 307, 308, 333.

377

This page intentionally left blank