\v : [CHALCOGENIDE SEMICONDUCTOR PROPERTIES OF AN AMORPHOUS f f ?5 :';'_{ Gel ASZOTBYG x _ Thesrs for the Degree of Ph D > - g * MECHIGAR S’KTE UNIVERSW ' " ‘ - fKELLv PAUL {(30me V v ” . .1971 : - ' ABSTRACT PROPERTIES OF AN AMORPHOUS CHALCOGENIDE SEMICONDUCTOR Ge10A320Te7o By Kelly Paul Golden An amorphous chalcogenide semiconductor, Te70, was produced GeloAszo and examined for physical, thermal, electrical, and optical properties. Material synthesis techniques are described. Important thermal char- acteristics are reported including glass-transformation, crystallization, and melting temperatures. Electrical properties reported include dc conductivity and ac admittance versus temperature, Seebeck coefficient, and reversible conductivity switching. Behavior of the optical absorb- tion coefficient for high and low energies in the absorption edge is reported. Improved four-point probe dc-conductivity measurement appar- atus is described; apparatus under construction for drift-mobility measurement is also described. Amorphous Ge was synthesized from elemental starting 10A320Te7o constituents in evacuated silica ampuls. Mixed lump and powder raw materials were fused at 800 °C with agitation for about 20 hours; the melt was cooled slowly to 400 °C and then quenched rapidly to 0 °C. The resulting glass was dark, shiny, and very fragile; it fractured conchoid- ally and showed no structure by optical-microscopic or x-ray diffraction inspection. Density of amorphous GelOASZOTe7O was 5.8 g/cm3; Mohs hard- ness was about 2.8. Amorphous material crystallized readily; the cryst- allized state appeared less dense but harder than the amorphous state. Microscopic examination of crystallized material revealed a complex KELLY PAUL GOLDEN network of crystallites; crystalline tellurium was identified from x-ray diffraction patterns. Many thermal prOperties were determined. Specific heat for amor- phous material was 0.289 J/(g °C) at 80 °C; glass transformation occur- red at 120 °C. An unusual endothermic reaction accompanying glass trans- formation was correlated with annealing stabilization. Softening occur- red at about 175 °C; crystallization occurred above 190 °C with an exo- thermic heat of reaction of about 6.1 J/g. Crystallized material had a specific heat of 0.314 J/(g °C) at 80 °C and melted at 340 °C. Heated Ge As 10 20T87o atmosphere. solid and liquid reacted only very slowly with the Low-field dc electrical conductivity appeared to be thermally activated; at room temperatures 5><10's mhos/cm conductivity and 0.3 eV activation energy were typical for bulk material annealed about 10 hours at 100 °C. The activation energy was sensitive to temperature and annealing time and ranged from 0.2 to 0.5 eV. Low-temperature ac conductance was proportional to the 0.85th power of frequency. Seebeck coefficient was approximately +300 uV/°C. Threshold and memory switch- ing were observed at a field intensity of 2.65 kV/cm. Room temperature photoconductivity was very small. Electrical conductivity of crystal- lized material was high and practically independent of temperature at about 70 mhos/cm. The optical-absorption edge was poorly defined, but located at about A - 1.7 pm. At low energies in the absorption edge the absorption coefficient was exponentially dependent on photon energy; a critical energy value of 0.075 eV was determined. At higher energies © 1971 Kelly Paul Golden ALL RIGHTS RESERVED KELLY PAUL GOLDEN in the absorption edge the absorption coefficient was proportional to the square of photon energy; an optical-energy gap of 0.79 eV was found and compared to the measured thermal-activation energy for conductivity of 0.41 eV for an identical sample. Index of refraction was determined interferometrically in the absorption edge and at transmitting wavelen- gths; above A = 4 pm the value was about 3.9. A table of properties for GelOASZOTe7O’ comparisons with work by others, theoretical interpretations of results, and suggestions for additional research are all included. Brief literature reviews of experiment and theory for non-crystalline electrical conductors and for conductivity switching are also supplied. All sections are extensively referenced; a bibliography containing over 500 citations is provided. PROPERTIES OF AN AMORPHOUS CHALCOGENIDE SEMICONDUCTOR GeloAsonem By Kelly Paul Golden A THESIS Submitted to Michigan State University in partial fulfillment.of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Electrical Engineering and Systems Science 1971 Copyright by KELLY PAUL GOLDEN 1971 Dedicated to my wife Patricia Elaine Golden ii ACKNOWLEDGMENTS The author wishes to express appreciation to his major professor, Dr. L. J. Giacoletto, for his guidance in the research effort and for his prompt and careful review of the written thesis. He also wishes to thank the other members of his guidance committee: Dr. K. M. Chen, Dr. D. Fisher, Dr. B. Ho, and Dr. D. J. Montgomery for thesis guidance, and Mr. I. 0. Ebert for early guidance in planning the academic program of study. The author also wishes to express appreciation to Owens-Illinois Inc., particularly to Dr. M. S. Hall, Manager of the Okemos Research Laboratory, for providing facilities and financial support for the research program in the form of a cooperative part-time employment fellowship-grant arrangement. He also gives thanks to all of the Okemos Research Laboratory personnel for their encouragement and generous counsel during the several-years long research program; special thanks go to Mr. R. H. Moore for his ingenious design and expert construction of much of the equipment The author wishes to especially thank his wife, Pat, for her aid in preparation and typing of the thesis, and for her love, patience, and understanding during the entire course of this study. 111 TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O I O O O O O O O O O O O O O 0 v1 LIST OF FIGURES O O O O O 0 O 0 O O O O 0 O O O O 0 0 O I O O v11 AN INTRODUCTION 0 O O O O O O O O O O 0 O O O O O O O O O O O 1 1 1 IntrOdUCtion O O O O C O O O O O O O 0 O O O O O C O O O 1 2 Choice of Material an Goals . . . . . . . . . . . . . . 1.3 Literature Review for the Ge-As-Te System . . . . . . . 1 4 Measurement Objectives . . . . . . . . . . . . . . . . . OWNH MATERIAL SYNTHESIS . . . . . . . . . . . . . . . . . . . . . . 8 1 Raw Materials . . . . . . . . . . . . . . . . . . . . . . 8 2 Handling and Weighing . . . . . . . . . . . . . . . . . . 9 3 Evacuation, Sealing, and Fusing . . . . . . . . . . . . . 10 .4 Quench Technique . . . . . . . . . . . . . . . . . . . . 12 5 Sample Preparation . . . . . . . . . . . . . . . . . . . l3 6 Thin-Film Production . . . . . . . . . . . . . . . . . . 14 GENERAL PHYSICAL! PROPERTIES O O O O O O I O O O O O O O O I O 16 1 Introduction . . . . . . . . . . . . . . . . . . . . . . 16 2 Microscopic Examination . . . . . . . . . . . . . . . . . 17 3 Density and Hardness . . . . . . . . . . . . . . . . . . l9 .4 Thermal Analysis . . . . . . . . . . . . . . . . . . . . 20 5 X-Ray Diffraction . . . . . . . . . . . . . . . . . . . . 26 6 Chemical Analysis . . . . . . . . . . . . . . . . . . . . 31 ELECTRICAL PROPERTIES O O I O O O O O O O O O O C O O O O O O 33 4 1 Introduction . . . . . . . . . . . . . . . . . . . . . . 33 4 2 Dc Electrical Conductivity . . . . . . . . . . . . . . . 34 4 3 Ac Electrical Admittance . . . . . . . . . . . . . . . . 44 4.4 Thermoelectric Properties . . . . . . . . . . . . . . . . 46 4 5 Switching Characteristics . . . . . . . . . . . . . . . . 48 4 6 Drift Mobility . . . . . . . . . . . . . . . . . . . . . 50 4 7 Photoconductive Effects . . . . . . . . . . . . . . . . . 56 iv Page OPTICAL PROPERTIES 0 0 O 0 O o o O o o o o o o s o o o o s o o 58 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 58 5.2 Reflectance and Transmittance . . . . . . . . . . . . . . 59 5.3 Absorption Coefficient . . . . . . . . . . . . . . . . 61 5.4 Absorption of Partially Crystallized Material . . . . . . 65 5.5 Index of Refraction . . . . . . . . . . . . . . . . . . . 65 CONCLUSIONS 0 O O O I O O O O O O O I O O O O O O I O O O O O 69 6.1 Summary of Results . . . . . . . . . . . . . . . . . . . 69 6.2 Suggested Additional Research . . . . . . . . . . . . . . 70 APPENDIX A - NON-CRYSTALLINE ELECTRICAL CONDUCTORS . . . . . . 74 Part 1 —- Historical Introduction . . . . . . . . . . . . . . 74 Part 2 -- Theory of Electrical Conductivity . . . . . . . . . 75 APPENDIX B - MATERIAL PREPARATION TECHNIQUES . . . . . . . . . 78 Part 1 "- Silica Amp'J]. MEChOd o s o o o o o s o s o o o o o o 78 Part 2 -"' ROCking TUbe Furnace s o o o o o o o o o o o o o o o 80 Part 3 -- Flash Evaporation of Films . . . . . . . . . . . . . 83 APPENDIX C - DC CONDUCTIVITY MEASUREMENT . . . . . . . . . . . 87 Part 1 -- General Considerations . . . . . . . . . . . . . . . 87 Part 2 -- Four-Point Probe Head . . . . . . . . . . . . . . . 89 Part 3 -- Description of Electronic Subsystems . . . . . . . . 90 Part 4 -— Sample Temperature Control . . . . . . . . . . . . . 100 Part 5 -- System Operation and Data Analysis . . . . . . . . . 102 APPENDIX D - REVERSIBLE CONDUCTIVITY SWITCHING . . . . . . . . 107 Part 1 -- Historical Introduction . . . . . . . . . . . . . . 107 Part 2 -- The SWitChing Phenomena o o o o o o o o o o o o o s 107 Part 3 -- Theory of Switching . . . . . . . . . . . . . . . . 109 BIBLIOGRAPHY O O O O O O O O O O O O I O O O O 0 O O O O O O O 112 Part 1 -— References for Chapter 1 . . . . . . . . . . . . . . 112 Part 2 -- References for Chapter 2 . . . . . . . . . . . . . . 117 Part 3 -- References for Chapter 3 . . . . . . . . . . . . . . 120 Part 4 -- References for Chapter 4 . . . . . . . . . . . . . . 124 Part 5 -- References for Chapter 5 . . . . . . . . . . . . . . 134 Part 6 -- References for Appendix A . . . . . . . . . . . . . 138 Part 7 -- References for Appendix C . . . . . . . . . . . . . 145 Part 8 -- References for Appendix D . . . . . . . . . . . . . 149 Table 1.1 3.1 4.1 6.1 C-l LIST OF TABLES Computer CONVPCNT . . . . Summary of Measured Thermal Pr0perties Computer Program DATAFIT Summary of Properties of GelOAsone7O . Computer Program DATA . . Output from DATA Program vi Page 27 40 71 105 106 Figure 1.1 3.1a 3.1b 3.2 3.3a 3.3b 3.4 3.5 3.6 3.7 3.8 4.1 4.2 4.3 4.4a 4.4b 4.5 4.6 4.7 4.8 5.1 5.2 LIST OF FIGURES Ge-AS-Te System 0 o o o o o o o o o o o o o o o o Amorphous GelOASZOTe70’ Magnified Surface . . . . Crystalline Ge Magnified Surface . . . 10A820Te70’ TMA Penetration Curve . . . . . . . . . . . . . . Typical DTA Curve . . . . . . . . . . . . . . . . Typical DTA Curve . . . . . . . . . . . . . . . . Glass-Transformation DTA . . . . . . . . . . . . Anomalous GelOAsone7o DTA . . . . . . . . . . . Quenching-Process DTA . . . . . . . . . . . . . . X-Ray Diffraction Pattern, Amorphous . . . . . . X-Ray Diffraction Pattern, Crystalline . . . . . Test of Low-Field Approximation for Conductivity Electrical Conductivity versus Temperature . . Conductivity versus Annealing Time at 99 °C . . . Admittance at Higher Temperatures . . . . . . . . Admittance at -l80 °Celsius . . . . . . . . . . . Hot-Probe Technique . . . . . . . . . . . . . . . Drift-Mobility Measurement Method . . . . . . . . Typical Transit—Time Signals . . . . . . . . . . Drift-Mobility Measurement System . . . . . . . . Average 45° Reflection . . . . . . . . . . . . . Average Normal Transmission . . . . . . . . . . . vii Page 18 18 22 23 23 25 25 28 30 3O 36 38 42 45 45 47 51 51 55 60 60 Figure 5.3 5.4 5.5 5.6 5.7 8—1 8-2 B—3a B-3b C-la C-lb C-Za C-2b c-a C-5 C-7 C-8 C-1O Optical Absorption, Exponential Behavior . . Optical Absorption,Square-Root Dependence . . Optical Absorption Spectrum . . . . . . . . . Absorbance: Amor. vs Cryst. Material . . . . Index of Refraction vs‘fiw . . . . . . . . . . Pieces of Amorphous GelOAsone70 . . . . . . Rocking Tube Furnace . . . . . . . . . . . . Flash Evaporation, Baffled Boat . . . . . . . Quasi-Flash Evaporation, Baffled Boat . . . . Four-Point Probe . . . . . . . . . . . . . . Four-Point Conductivity Measurement . . . . . Trial Constant-Current Supply . . . . . . . . Trial Constant-Current Supply . . . . . . . . A Special 0.1 uA Constant-Current Supply . . DC Conductivity - Measurement-System Diagram DC Differential-Preamplifier Circuit Diagram Thermocouple Calibration . . . . . . . . . . Probe-Reversing System . . . . . . . . . . . Function Sequence Controller . . . . . . . . Sample Holder & Temperature-Control System . Typi C81 Raw-Data curve 0 o o o o o o o o o 0 viii Page 63 63 66 66 68 81 82 84 84 88 88 91 91 93 93 94 97 97 99 99 103 CHAPTER 1 AN INTRODUCTION 1.1 Introduction In the past twenty years there has been much activity in solid— state chemistry, physics, and electronics. Much of the activity has been study of properties and applications of crystalline materials. Theories of behavior for crystalline semiconductors have been highly developed. Study of non-crystalline materials has only recently begun, however, and much remains to be learned. Non-crystalline or amorphous materials are of interest theoretically as an extension of the study of more highly ordered crystalline materials. Also, unusual, unexpected, and largely unexplainable electrical characteristics have been noted for some non-crystalline solids. Classes are the most familiar of non-crystalline materials. A well known type of glass, transparent window glass, is fabricated primarily from silica and other metal oxides by fusing the components at high temperature and allowing the melt to cool rapidly enough to become rigid without crystallizing. Many different glass compositions have been produced in this way and by other methods. A few glasses are significantly electrically conductive. Most amorphous electrical conductors, or semiconducting glasses as they are sometimes called, contain a number of the sixth group of elements in the periodic-classification table as a major constituent. Semiconducting glasses may be classified into two groups: transitiondmetal oxide glasses, and chalcogenide glasses. Transition-metal oxide semiconductors (vanadium oxides are good examples) exhibit only relatively poor electrical 2 conductivity under normal conditions. Never-the-less these materials have important and useful electrical characteristics. Chalcogenide glasses contain a major fraction of one or more of the members of group VI excepting oxygen: sulfur, selenium, and tellurium, combined with other elements from roughly the same region of the periodic table. A tremendous number of combinations have been formed as glasses; only a few have been studied extensively. Chalcogenide glasses have relatively high electrical conductivity and frequently show memory switching effects. The past several years have seen investigation of chalcogenide glasses for infrared~optical(l) and electrical-switching(2) applications. Recently there has been much effort to understand electri- cal effects in these non-crystalline or relatively-disordered materials at a theoretical level. However, difficulty in producing chalcogenide glass- es and the awkward task of describing them in familiar crystalline terms has interfered with progress in the theoretical studies. 1.2 Choice of Material and Goals Amorphous GeloAsone7o, a known semiconducting glass, was selected for comprehensive study of its physical, electrical, and optical properties. The research attempted to show that this material could be produced in a relatively easy and repeatable fashion, and that its characteristics could be measured using mostly conventional techniques. Much effort was devoted to the design and execution of experimental techniques; few theoretical interpretations were made. Also, considerable effort was devoted to the task of keeping abreast of recent developments regarding experiment and theory for chalcogenide semiconductors. Evaluations of data and correlations with available theory and similar experimental information are distributed in the text according to subject. 3 A composition diagram for the ternary germanium-arsenic-tellurium 1 system has been published,( ) and is shown in Figure 1.1. Binary phase diagrams and other chemical characteristics for materials from this system are also available.(3—6) Reported compounds, eutectics, and glasses are indicated on the diagram. The glass-formation region, outlined in Figure 1.1, has been determined theoretically<69> and 1,67 eXperimentally.( ) A computer program, CONVPCNT, used to replot the (7) diagram with weight-fraction scales, is shown in Table 1.1. Many compositions can be synthesized by direct mixture of elements or by mixture of elements and compounds. Similarly, glasses devitrify by separation into several elements or compounds. The computer program was also used to calculate weight and atomic-number fractions for mixtures of two or more elements and/or compounds from the Ge-As-Te system. Several mixtures or possible separations for the composition GeloAsone7o are shown in the composition diagram as dashed lines. 1.3 Literature Review for the Ge-As-Te System Discovery of reversible conductivity-switching effects in semicon- ducting glasses has stimulated study of amorphous chalcogenides? Some of the research reported for Ge-As-Te materials is listed in the references for this and other chapters. Miscellaneous physical properties, including thermal and structural characteristics, have been determined.(1’8-22) Studies of electrical properties, especially switching, have been reported.(19-SO) Finally, optical properties of Ge-As-Te glasses have d.(1’51-61) been publishe Most of this information was very useful, though relatively little was for the specific composition GeloAsone7o. * Brief reviews of theory for non-crystalline electrical conductors and of experiment and theory for conductivity switching can be found in Appendix A and Appendix D respectively. 0‘ \\ I.-- \\ 4-0- . \\\ Ob. \D D U D D/Q «\ O , " \Hz)‘ \\ s a I . \ Ia 6000 GQ/Dvdofl\ 47.0 n. o \anxxd 4.. ch... .e , :AMU..\ m o .\ XXX Q A: x \\ u / \\ . Q «:23 \ u < 5. . Q Q n s 2.0 /. <38 23:3 0 C .82... 3...: 0 05:93.5 0 :33 D 0.25 cutout 259; 29—8228 INV Eggm 0.75450 2 0.52.... 5 I COMMGN FRACT(5)1M(5):C(200IO):N:NN 3] TE8127061A5874o921636E3750593GETE'GEvTEJGETE2'0E*TE*20 36 ASZTE38A5*2o+TE*301 ASQTES'AS*20+TE*50 41 C(IJI)BTEBCC2oI)BASIC(3sI)IGEIC(703)IZo 44 C(4313'GETE1 C(5sl)'GETE2) C(6JI)'A52TE33.CCTaI)'A52TE5 48 C(Io2)8l030(20333I03CCSs4)3IolC(4s2)'IoJC(4o4)3Io 49 C(532)82010(514)8Io3CC6:2)I303C(613382010(702).59 52 ?RINTD" INPUT ' CONSTITUENTS") INPUTJNN 55 PRINT:” WHICH "oNNs" CDNSTITUENTS"! INPUT:(N(K):K'I0NN) 57 PRINT:" IN WHAT PERCENTAGESTJ INPUTQCFRACT(K)0K.I1NN) 60 PRINT:” WHICH CDNVERSION TYPE?”1 INPUTaNH! PRINTJ” ” 63 GO TDCIq2o3)sNMI I CALL MDLWTJ GD T9 IO 64 2 CALL WTMDLIGD TD 1013 CALL UTATIIO STDPBEND 300 SUBRDUTINE MOLNT 301 CDMMON FRACTC5)0M(5)00(209IOIaNsNNI DIHENSIDN "(5) 320 PRINT0"' MOLECULAR E WEIGHT 3” 321 PRINT:" "1 WT=Oo 340 D0 50ILLaIaNNBW(LL)IPRACT(LL)*C(HCLL)01,350 "TIUTOHCLL) 360 00 60 LLlloNNl 60 “(LL)31000*U(LL)/WT 370 PRINT 1000:(PRACT(K)oK3105)o("CKIsKIIsS), 375 1000 FDRMATC" ”aSCFG-Zs" ")0" "05(P6020" "I, 380 RETURN) END 400 SUBRDUTINE WTMOL . 401 CDMMDN PRACT(5):M(5):C(20:10’1N0NN3 DIMENSION "(5) 420 PRINT:” WEIGHT 1 HDLECULAR E” 421 PRINT:" "1 WTBOo 440 DE 50 LL3IJNN3WCLL)IFRACT(LL)/CCH(LL)0I3850 NT'NT‘NCLL) 455 DC 60 LLIloNNI 6O WCLL)81000*N(LL)/NT 465 PRINT 1000s(PRACTCK)0KIIOS)0("(K30KII05) 470 1000 PORMAT(" "05(F6020" ")0" '15CF602:” "I, 480 RETURN! END 500 SUBROUTINE "TAT 501 CGMMGN PRACTCS):MC5)OCC200IOIJNoNNIDIHENSIGN "(5)1NNCSI 512 PRINT)" WEIGHT E TEX ASE GEE 513& (NOTE! ATGMICX)" 516 PRINT:" "1 WTIOO 528 DD 50 LLgloNNt"(LL387RACTCLLIICCNCLLIOI)350 NTOHT¢NCLLI 540 DC 60 LLilsNNi 60 UCLL)‘IOO¢*N(LL)/NT3 DC 70 LL3103 552 WWCLL38WCI)*C(M(I)3LL*I)/(C(H(I)02)*C(H(I)03)OC(H(I’34)) 556 WWCLL)lWUfLL)+W(2)*C(HC2)oLL+I)I 557‘CCCM(2):2)+C(M(2):3)+C(M(2)04)I 560 IFCNNoEQoZI GO TO 70 564 70 WHCLL)INN‘LL)+W(3)*C(H(3)JLL+I)I 565&‘CCMC3):2)+C(M(3)03)+C(H(3)s4)) 572 PRINT 1000:(PRACTCK)0KCIOSIQCNN(K)3KII05) 576 1000 EQRHATC" ”05(76021" ”)0"7"35(E6o20" ”)) 580 RETURN) END Table 1.1 Computer Program COIVPONT 1.4 Measurement Objectives 10‘920Te7o "‘9 desired. Useful determinations which could be conveniently accomplished Complete and general characterization of amorphous Ge using available equipment were done. Apparatus was designed and fabricated for the determination of several additional properties. In a few cases construction of equipment was not completed in time for measurements to be reported here. PrOperty measurements were begun with examination of gross physical characteristics such as density and hardness. Later, thermal analyses<62> were performed, including DTA, TGA, and TMA. Phenomena such as glass transformation, softening, crystallization, melting, quenching, and oxidation were observed by thermal-analysis techniques. X-ray diffrac- tion and chemical-analysis techniques were used to determine composition and structure of glassy and crystallized material. Electrical conductiv- ity and many other electrical properties were investigated. Finally, optical absorption and reflection measurements for bulk and thin-film samples were performed. (68) For simple crystalline semiconductors, electrical current resulting from application of an electric field defines a material property563) conductivity: 0 E J/E - I n1 q1 "1 Contributions to conductivity due to charged current carriers, qi, present with volume density, n1, are proportional to the mobility, "1, of the i'th carrier type. Usually only hole and electron carriers are important, but other mechanisms for current flow have been identified for disordered materials. Understanding of electronic properties of materials requires determination of contributions and energetics for 7 each term and factor involved with conductivity. Much of the information can be obtained by independent measurements of conductivity and mobility versus temperature. Measurement of conductivity versus temperature is fundamentally straightforward but was technically difficult for amorphous GeloAsone70. Special four-point probe apparatus was designed and constructed for measurement of dc electrical conductivity versus temperature. Equipment fabricated by collegues was used for measurements of ac electrical (65) was used admittance and switching characteristics. A simple method for determination of the polarity and approximate magnitude of Seebeck coefficient. Carrier-mobility measurements were complicated and difficult. The Hall effect,(63) often used for determination of carrier polarity and mobility for crystalline materials, was difficult to observe for amorphous GelOAsone7o. Hall mobility for amorphous chalcogenide semiconductors is usually very small and difficult to interpret. Direct determination of drift mobility<63> was considered most appropriate and apparatus was developed which used the electron-beam injection technique<64> for transit-time measurements. Naturally, measurements performed gave results which suggested many additional studies of other characteristics. As time was an important factor, such further study was not practical. The broad collection of measurements made should, however, supply much of the information required for the establishment of simple hypotheses on the nature of electrical conduction and other processes for amorphous Ge As Te 10 20 70 and possibly other amorphous chalcogenide semiconductors. CHAPTER 2 MATERIAL SYNTHESIS 2.1 Raw Materials Acquisition of amorphous GelOAsZOTe70 and other amorphous semicon- ductors from commercial chemical suppliers and from other research organizations was not possible. Most suppliers either could not provide the requested compositions or would undertake their synthesis only upon receipt of an order for substantial quantities at high cost. Energy Conversion Devices, Troy, Michigan, an organization studying chalcogenide semiconductors would not supply the material for proprietary reasons. Information on synthesis technique was requested to aid the unfamiliar task of producing the amorphous materials. At the time of the requests, 1968, this information was highly proprietary, and only information found in open literature was available. Recently at least one organiza- tion* has offered chalcogenide glasses at reasonable cost, purities, and quantities. A first step in the synthesis was selection and acquisition of pure starting materials. Several organizations** offered pure germanium, arsenic, and tellurium; most raw materials were purchased from Alfa Inorganics. Purity and usefulness of materials depended upon the form supplied. Fifty ohm-centimeter resistivity powdered and polycrystalline * Materials Research Corporation, New York, quoted $175 for 100 grams of 99.9992 pure GelOAsone7O, ** Alfa Inorganics Division of Ventron Corp., Beverly, Massachusetts; K 8 K Laboratories Inc., Plainview, New York; Research Inorganic Chemical Corp., Sun Valley, California; Rocky Mountain Research Inc., Denver, Colorado; Apache Chemicals Inc., Rockford, Illinois; Materials Research Corporation, Orangeburg, New York; Atomergic Chemicals Co. Division of Gallard-Schlesinger, Carle Place, L.I., New York; American Smelting and Refining Company, South Plainfield, wa Jersey; and several others. III“ II I [1 {I r 9 germanium was satisfactory. Powder, shot and crystalline lumps of tellurium were available. The shot form was most frequently chosen as a compromise between handling ease and purity. Most powdered or crystal- line arsenic was generally excessively oxidized. Accordingly, pure crystalline lumps packaged in inert gas were purchased and stored in a dessicator under dry nitrogen until used. Arsenic was exposed to the atmosphere for only a short period for weighing. Generally, lumps and shot were used in preference to powders to reduce dusting and to reduce losses due to adherence of powders to weighing tools and ampul walls. Also, lumps had less surface susceptible to oxidation or contamination. Purity and correct stoichiometry of the final compositions was not of great concern; hence purity of starting materials was not too important. A primary consideration was that the synthesis procedure be repeatable and not difficult. This position was justified by the generally accepted thesis that impurities and slight compositional changes in amorphous semiconductors do not significantly affect their (1,33,34) prOperties. However, some workers do not agree that amorphous semiconductors are entirely impurity insensitive, and have been very careful in their preparation technique.(2’3’4’3o) 2.2 Handlinggand weighigg_ A serious consideration in the preparation of materials containing arsenic and tellurium was toxicity. These elements are generally consid- ered highly toxic(5’8) and dangerous in all forms. Of particular concern was possible inhalation of powders and dust. All handling of elements and finished compositions was done initially in polyethylene glove bags (12R Instruments Inc., Cheltenham, Pennsylvania), and later in a glove box (Lab-Con—Co., Kansas City, Missouri). No attempt was made to purge 10 or otherwise control the atmosphere in these systems. This could easily have been done, however, had atmospheric contamination or oxida- tion been a problem. A computer program, CONVPCNT,* was used to determine weight percent- ages of the elements in GeloAsone7o. The weight percentages were calcu- lated as: tellurium-79.85%, arsenic-13.491, and germanium-6.76%. Using these percentages, suitable large lumps of arsenic were chosen and weighed, and remaining ingredients were proportioned appropriately. Weighing was done inside the glove box to an accuracy of about 0.01 g using an analytical balance (Model DWL-S, Torsion Balance Company, Clifton, New Jersey). Initially, constituents were weighed into small dishes and transferred to the quartz ampuls. However, an excessive loss of raw materials occurred in transfer operations, and cumulative weigh- ing of elements directly into the ampuls was finally used. Following this operation the ampuls were tapped lightly to shake down dust on the inside walls. 2.3 Evacuation, Sealing and Fusing, After raw materials were weighed into the quartz ampuls, loose- fitting quartz plugs were dropped in, and the ampuls were transferred carefully from the glove box to a simple vacuum system for evacuation and sealing.** After evacuation to about 10.3 torr and sealing, ampuls containing the raw materials were suspended in a rocking furnace for fusion. Many researchers prefer to backfill ampuls with a low pressure of inert gas. This procedure was not followed as an initial gas fill * The timesharing-computer program, CONVPCNT, was described in Section 1.2. ** Information about the ampuls, the vacuum system, and sealing procedures can be found in Appendix B. 11 would excessively increase pressure in the ampuls at high temperatures. For example, heating to temperatures near 1000 °C could multiply initial pressures by almost forty and possibly burst the ampuls. Near 1000 °C the vapor pressure of tellurium is about 760 torr, and the (6.7.8) vapor pressure of sublimed arsenic is over 20 000 torr. The ampuls were heated carefully in an armored furnace enclosed by a fume hood* to reduce hazards from an explosion.(31) No explosions occurred however, possibly because the volatile arsenic was dissolved or reacted almost as fast as it sublimed. Many researchers(9’10) fuse elements into a homogenous composition by heating to a temperature high enough to melt all constituents. The highest melting point element used, germanium, melts at 937 °C , tellurium melts at 450 °C , and arsenic sublimes at 610 °C and 760 torr.(8) Materials were fused at several temperatures in the range from 500 °C to 1100 °C , and resulting mixtures usually appeared homogenous when sufficiently long reaction periods were provided. Tellurium melted at these temperatures and probably rapidly dissolved the arsenic; the melt- ing point of the arsenic-tellurium alloy was less than 400 °C.(11) The arsenic-tellurium liquid then slowly dissolved the germanium, resulting in homogenous compositions produced at relatively low temperatures. Most GeloAsone70 used was produced by fusing at 800 °C for twenty hours in a rocking tube furnace. A one-half hour warm-up period was provided to allow for solution of the arsenic, preventing build-up of an excessive arsenic vapor pressure. * A detailed description of the rocking furnace used is available in Appendix B. 12 2.4 Quench Technique It is well known that glasses may frequently be created by rapid quenching of the liquid state.(12-15’30) This supercooling process results in a rigid liquid in which crystallization proceeds only very slowly. The nature of the glass formed depends strongly on the quench rate.(16) Many techniques for producing chalcogenide glasses use an air quench.(9’10) To obtain very fast cooling rates, a brine bath chilled to 0 °C was used. Saturated brine had a higher thermal conductivity and tended to boil less violently than plain water.(17’29) The quench- ing solution was enclosed in an armored vessel to reduce hazards from possible explosions. Ampuls were dropped directly from the furnace into the quenching bath. Cooling fused material to 400 °C , a temperature only slightly above the melting point of Ge As 10 20Te7o resulted in faster quenching and lowered fragility of the glass produced. , prior to quenching Boules of material were easily removed from ampuls, probably because differences of thermal expansion do not allow chalcogenide glasses to stick to fused silica. A small hole through the center of the cylindri- cal boules was noted; this probably was produced as the material cooled from the outside and shrunk away from the center.(29) Amorphous materials produced by quenching of melts contained considerable strain, evidenced by extreme brittleness and difficulty in (3) Reduction of strains and modification of (18,19) cutting and polishing. conductivity by annealing has been observed for these materials. Strain and other characteristics of amorphous GeloAsone7o could probably have been modified by changes in quenching and annealing processes. A series of experiments to test this hypothesis was not performed as the primary objective was development of a synthesis 13 technique for a single reproducible material. However, most samples were annealed for about ten hours at 100 °C , slightly below the glass transformation temperature. Annealing at higher temperatures often resulted in crystallization of samples. Even though annealed, the glass was quite fragile and often broke into many pieces at the slightest shock. 2.5 Sample Preparation Amorphous GeloAsone70 semiconductor material produced in the man- ner described was about one centimeter in diameter and about two centi- meters long, or was in fragments thereof. Freshly fractured surfaces of randomly chosen lumps were used as samples on which properties were measured. Techniques for cleaning or modifying surfaces, such as chemical etching and sandblasting, did not significantly affect experi- ments. The only effective etchants found were aqueous solutions of ferric chloride, and concentrated nitric acid. Attempts to cut and polish samples almost always caused fractures, possibly due to the extreme brittleness and strained nature of the glasses. Several thin, flat-sided bulk samples were obtained by grinding down larger lumps of the glassy material. Pieces of amorphous GeloAsone7o were fastened to quarter-inch plate-glass substrates with CrystalBond 509 Adhesive, Aremco Products, Briarcliff Manor, New York, an acetone- soluble, low softening temperature mounting media. During mounting and grinding the temperature of the amorphous semiconductor was kept below 100 °C to avoid additional annealing or possible crystallization. Grinding was done under water, to prevent toxic dusting, on silicon- carbide sandpapers. Both sides of several lumps were flattened in this fashion. Samples thinner than about 250 um broke during grinding or 14 mounting processes; polished surfaces were frequently severely pitted. However, unpolished wafers produced in this fashion were useful for many electrical and Optical experiments. Three additional techniques for producing flat-sided samples were attempted with little success. First, semiconducting glass melted and quenched in flat-sided silica ampuls was always broken during removal from the ampuls. Second, hot-pressing amorphous GeloAsone70 above the softening temperature usually resulted in crystallization of the sample being shaped. Finally, melted material squeezed between chilled silica plates was usually partially crystallized, and was not useful for most purposes. This was probably due to an insufficient quenching rate. Polycrystalline films of GeloAsone7o about 100 pm in thickness obtained in this manner were used for an optical absorption experiment. 2.6 Thin Film Production- The technique described for production of amorphous GeloAsone70 (21) Other quench methods such as has had a long and popular history. splat quenching<22) have been used. Additional techniques for prepara- tion of amorphous compositions include precipitation from solution and condensation from vapor. The latter method has been used with consider- able success in the preparation of amorphous films.(23’28’29) Although it is possible to co-deposit two or more elements simultaneously from (24,25,35) one or more sources onto a cold substrate, an apparently more popular technique is evaporation or sputtering of predmixed and fused materials.<9) Thin films of elemental semiconductors and other materials have been produced by sputtering.(26’28) Several organizations are using 15 this technique for the preparation of compound and alloy films.* Since sputtering apparatus was not available, a modified thermal-evaporation 10Asonem. This method, used by others with success,(9’28) is a pseudo flash-evaporation process<27> was used to produce thin films of Ge technique from an alumina-coated resistance-heated baffled boat source.** Films produced in this fashion ranged in thickness from 0.01 pm to 5 um. Films thicker than about 2 pm were prepared by cascading several flash-evaporated films on a single substrate. All films were deposited on thin soda-lime glass microscope cover slides. Of many instruments (32) the most useful for available for film-thickness measurement, measuring these chalcogenide films were found to be a Varian Interfero- meter and a Zeiss Light-Section microscOpe. It is well known that segregation often occurs if alloys are thermally evaporated. The flash-evaporation technique employed reduced segregation since it allowed easy mixing of vapor species. Frequently, researchers have determined composition of films after deposition by electron-microprobe techniques.(19) Such determinations were not made for flash-evaporated films described here, but properties of films did not differ greatly from bulk source material. Considerable work remains to be done in the field of preparation of chalcogenide alloy films. * Energy Conversion Devices, Troy, Michigan, utilizes sputtering techniques in the production of Ovonic Switches. ** A complete description of the apparatus and process can be found in Appendix B. CHAPTER 3 GENERAL PHYSICAL PROPERTIES 3.1 Introduction Many properties for Ge are reported here. These proper— 10A320Te7o ties are generally non-electrical and non-optical; included are appear- ance, density, hardness, thermal, and structural characteristics. Additional information reported by others is also discussed. Thermal constants for Ge, As, and Te have been determined.(1) Ge melts at 937 °C with a heat of fusion, AH, of 0.468 kJ/g (8.1 kcal/mole). As sublimes to As4 at 610.0 °C and 760 torr with AH - 1.73 kJ/g (31 kcal/mole). Te melts at 449.7 °C with AH - 0.138 kJ/g (4.2 kcal/mole); vaporization to T82 occurs at the melting point with AH - 0.433 kJ/g (13.2 kcal/mole) and vapor pressure 0.18 torr. (2) Mass spectrographic analysis of vapors developed from heated GelOAsone70 showed A34 appearing above 262 °C with a heat of vapor- ization of 1.52 kJ/g (28 kcal/mole). Some arsenic was possibly only dissolved in the glass and appeared upon softening or crystallization. Tez appeared above 386 °C with a heat of vaporization of 1.12 kJ/g (34 kcal/mole). A82T23’ probably present in crystallized GeloAsone7o, was an excellent possible source for Te2 vapors since the melting temperature of this compound is 362 °C. Except for arsenic, dissoci- ation energies from GeloAsone70 were generally greater than those from the covalently bonded elements. Differences between 125Te nuclear magnetic resonances in amorphous do) and crystallized GelSAs4Te81 have been observe and interpreted.(4) The resonance frequencies of 1ste for tellurium dissolved in aqua regia 16 17 for crystalline tellurium, and for crystallized GeloAs4Te81 were all the same. It was therefore concluded that a constituent of crystallized high electrical-conductivity GelSAS4T881 was crystalline tellurium. The resonance frequency for low electrical-conductivity amorphous material was lower than that for crystallized material. This was interpreted as a chemical shift arising from structural differences between the two states. Although the composition studied was not GelOASZOTe7O’ it was assumed that similar characteristics would be observed for this material. 3.2 Microscopic Examination Surfaces of amorphous and crystallized GeloAsone7o were examined with a high quality optical microsc0pe (Ortholux, Leitz-Wetzler Company, Germany). A lump of "silly putty" was used to attach pieces to a microscope slide for observation. Freshly-broken surfaces of amorphous material were dark, shiny, and "glassy" in appearance to the naked eye. Even at very high magnifications, microscope examination showed only stress-fracture lines and smooth surfaces. A freshly-broken piece of amorphous material was crystallized in a dry nitrogen atmosphere by increasing its temperature at a rate of 5 °C per minute until 240 °C was reached. The surface appearance after crystallization seemed metallic and dull as if sandblasted. Microscopic examination showed a complex network of interlocked fibers or crystallites, typically ten micrometers in length, cries-crossing and penetrating a previously smooth and unbroken surface. Photographs of typical untreated surfaces 10 20 is believed that small crystallites of tellurium and other compounds were for amorphous and crystallized Ge As Te70 are shown in Figure 3.1. It formed in the glassy matrix when the amorphous state was crystallized. is? sit: ' “'4 ‘ \ ' ("I p /‘ \5..\~’I'l ' ‘ .1 ' Figure 3.11: Crystalline GeloAsonem, Magnified Surface 19 Additional careful examinations of physical appearance for crystalline and glassy GeloAsone7O would probably yield considerable information. Depending upon quench and annealing treatments, several intermingling glassy and crystalline fractions might possibly be observed and identi- fied. Scanning electron microscopy,(5’31) (6.12) replica electron microscopy and analytical electron-microprobe techniques have been used for detailed examinations of surfaces of chalcogenide glasses from the Ge—As-Te system. 3.3 Density and Hardness A useful and easy to measure property of a material is density. Densities of the elemental constituents of the Ge-As-Te system have (8) been tabulated: germanium-5.32, arsenic-5.72, tellurium-6.24 g/cm3 at 20 °C. A simple homogenous mixture with composition GeloAsone7O should have an average density of 6.08 g/cm3. However, an alloy could have significantly different density. Archimedes' method was used to measure the density of amorphous GeloAsone7o. An analytical scale (Gram-atic Balance B5, Mettler Instrument Co., Hightstown, New Jersey) was used to weigh samples in air and immersed in pure water at 15 °C. A large lump of material was suspended in the water by means of a fine wire hung from the balance. The density of amorphous material was 5.68 g/cm3. Compensation was made for suspension wire weight, and density, but not for surface tension effects on the wire or for the non-unity density of water. (9’10) as about Density of amorphous GeloAsone70 was measured by others 5.6 g/cm3 in a study which showed that the density of chalcogenide glasses was approximately proportional to their molecular weight. Crystallized material was not measured, but was suspected of being 20 somewhat porus. Hardness could have been determined in several fashions, but scratch testing was simple and effective. In this method, the sample was scratched by harder and softer materials to locate hardness between that of two known materials. A usefully-accurate hardness scale was created of minerals with measured hardness, and careful technique. Amorphous GeloAsone70 was found to be relatively soft: about 2.8 on the Mohs scale (about 100 on the Knoop scale), about the same hardness as pure gold. Chalcogenide glasses are generally relatively soft; hardness is approxi- mately proportional to the softening temperature.(9’lo) For comparison, ordinary soda-lime glass was about 5.8 on the Mohs scale. The hardness of GelOAsone7O has been more carefully determined by others(9’11) to be 111 on the Knoop scale. The softness and fragility of these materi- als have severely limited their application to optics and other uses. 3.4 Thermal Analysis (13-16) were used to evaluate thermo- Thermal-analysis techniques dynamic properties of GeloAsone7o. Following suggested procedures, characteristics such as heat capacity, critical temperatures such as melting points, and thermal reactions such as glass transformation, crystallization, and oxidation were observed. Thermal—analysis methods included thermo-gravimetric analysis, TGA, thermo-mechanical analysis, TMA, differential thermal analysis, DTA, and others. A thermal- analyzer system (Model 900, E. I. DuPont de Nemours & Co., Wilmington, Delaware) was used to obtain data reported here. Thermo-gravimetric analysis examined changes of weight versus time and temperature. A typical gravimetric thermogram for amorphous Ge As h . 10 20Te7o was virtually featureless and is therefore not a own 21 Weight changes occurred only in pure oxygen or air atmospheres, began above the glass transformation temperature, and were positive and small. Surface oxidation only was believed to account for this, since weight changes were less than 0.52 for samples heated to 425 °C and held there for long times. TGA has also been used for studies of chemisorption in amorphous germanium.(l7) Thermo-mechanical analysis examined size changes and softening of samples versus temperature. Here, the penetration of a lightly-loaded probe into samples was monitored as a means for observation of soften- ing(19) and melting. A typical penetration thermogram is shown in Figure 3.2. Indicated on the curve are softening, re-hardening upon crystalliza- tion, and melting. Contributions due to thermal expansion were negligible for these measurements. The thermal coefficient of expansion d’(9,1l,l8,19) for chalcogenide glasses has been examine however, and was 18 x 10'6 m/(m X °C ) for amorphous GeloAsone7o.<11) For differential thermal analysis, the temperature of a sample was continuously compared with an inert reference material; differences in temperature due to thermal reactions of the sample were recorded as a function of the temperature of an enclosing furnace used to heat the sample and reference materials. Endothermic and exothermic heat effects (20) in a sample (such as associated with physical or chemical changes glass transformation, or crystallization) were monitored by this technique. A differential-scanning-calorimeter accessory for the thermal-analysis system was used to obtain the typical thermograms shown in Figure 3.3a for heating and Figure 3.3b for cooling. (21,26) Three prominent features of these curves require comment. Glass transformation, at T8, was characterized by an abrupt change in 22 3:50 cozozmcon <2... N.» 953... a on E 238389.. 00. i J 1 Con con CON CON on. . a . . a . . . q as... F1 05:02 - oh Eh eo_.u~_..3ufo1 .20.... Been. Esta _ 23... 9:3»: £53.. 0. .32... E... m l GNP—medbfo 105.333 a... :36 ‘k l L I i I 0. an o, N «r. o. o - - - — (mm) :usmsomdsgo aqond l 3 3l 23 ' I A o .. E's _ k 1' Tu, l 3:) _ Cassy cmromno 1 m N. l. 3 g - G°IoAszd‘Wo ._. L5 . 132 mg Sample < ‘ l' 2 °C/minHearlng Rate W100 :50 200 250 300"" 350' 400 ' Temperature (°C) Figure 3.30 Typical DTA Curve L L 59:0 A“zonro l- l32 mo Sample L 2 °C lmln Cooling Rate Exo - Crystalline ' \& AT (0.25 °C/div) .- Endo :1 0 5 lOO Temperature (°C) Figure 3.3b Typical DTA Curve 24 specific heat, or a baseline shift in the thermogram.(25) Crystalliza- tion began at Tc, and continued over a range of temperatures, liberating heat. Finally, melting of the crystalline phase, indicated by a strong endotherm, occurred at Tm. Examples of thermograms obtained before and after annealing are shown in Figure 3.4. An unusual endotherm in the glass transformation region of the curve occurred only for annealed samples, and was interpreted as a stabilization process.(24) Anomalous thermograms such as in Figure 3.5 were obtained for a few glasses. The complex peak structure was characteristic of samples with several mixed phases,(12) and material showing this property was not used for further testing. Curves in the glass transformation region were similar to others for 1, (24) this materia although the value determined for T was slightly 8 lower than the reported value.(28) Sensitivity of the glass transition (5,18) (28) (24) temperature to quench rate, composition, and annealing are known, however. Carefully calibrated thermograms were also used to (25) and heats of reaction.(26’27’32) determine specific heat, Much thermal investigation of chalcogenide glasses has been done in an attempt to explain threshold and memory electrical conductivity switching as a phase-change effect. Threshold-switching field intensities have been correlated with the glass-transformation temper- (28) Presence of multiple crystallization exotherms and other (5,29,30) atures. characteristics in thermograms are related to switching effects, and attempts to control(18) phase transitions have led to exotic semi- * conducting and switching compositions. * Energy Conversion Devices, Troy, Michigan, produces Ovonic Threshold switches containing films of approximate composition: Te-49Z, As-33Z, A 1" (OJ °C/div) ~Enda Eire—- 25 L h ? Unannealed F- Glassy 7’ Annealed ~ 16 hours T9 1. Gelo ASZOTOm 1;: p 5 °C lmin Heating Rate * ~|OOmg Sample nLnliJLILlellllnln O 20 4O 60 80 IOO 120 I40 160 180 200 AT (l °C/div) e-Endo Temperature (°C) Figure 3.4 GlassTransformation OT A Ezra- Amor haus r stalllzed L Anomalous Ge'oAsoneTO ; 5 °C Anin Heating Rate L l ‘1 l a l 1 l I J L l g I 1 l e 0 50 IOO 150 200 250 300 350 400 450 Temperature (°C) Figure 3.5 Anomalous GsmAsonem DTA 26 Properties of Ge determined by thermal analysis are listed 10A820T87o in Table 3.1. In most cases, broken pieces of as-quenched (possibly annealed) bulk material were used as samples. Some information was taken directly from curves, but calorimetric information was calculated using data taken from the curves and additional calibration information. Considerable time and effort was required to collect this information; thermal analysis was generally a time consuming and complex activity.* An especially-illustrative differential thermal-analysis experiment performed is illustrated in Figure 3.6 as three successive thermograms for the same sample. As shown by the first curve, a piece of un-annealed glassy GelOAsone70 heated slowly exhibited glass-transformation, crystal- lization, and melting phenomena. The melted sample was then rapidly quenched (in the calorimeter) and reheated slowly in a similar fashion. A smaller and shifted crystallization exotherm was exhibited in the second thermogram, indicating that only a fraction of the sample had been returned to the amorphous state. After remelting the sample was allowed to cool slowly, and a third thermogram was run. The absence of a crystallization exotherm in the last thermogram demonstrated that the sample had been completely crystallized. Another experiment,not illustrated,involved rapid quenching from above the crystallization exotherm but below Tm; always produced was a curve similar to the third thermogram. 3.5 X-Ray Diffraction A useful method for study of structure and composition of these (33-35) materials was x-ray diffraction. Experimentally, a sample was * Considerable instruction and contribution from Dr. L. Taylor and J. Tobias of the Owens-Illinois Inc., Okemos Research Laboratory, Okemos, Michigan is gratefully acknowledged. 27 Table 3.1 SUMMARY OF MEASURED THERMAL PROPERTIES Characteristic Temperature Value Conditions* Heat Capacity 80 °C 0.289 J/(g °C) glass annealed 16 hrs Heat Capacity 80 °C 0.314 J/(g °C) crystallized Glass Transformation 120 °C unannealed glass Stabilization 140 °C 2.87 J/g glass annealed 16 hrs Heat Capacity 160 °C 0.382 J/(g °C) glass annealed 16 hrs Heat Capacity 160 °C 0.288 J/(g °C) crystallized Crystallization 190 °C 6.11 J/g heated @ 2 °C/min Fusion 335 °C heated @ 2 °C/min *Determined with 5 °C/min heating rate unless otherwise noted. 28 4.5 3305 9.22.25 QM 059.... Acov 3333.55. 09V 00¢ 0mm 00m 0mm com on. 00. on _ _ d d 1 _ fl _ a q _ _ d _ . 1 3 u D. o f L as 0. v3.80 1 $22.6 32o.mEoo m* 1" l/ 556.. 87.01.! J 33:23.0 32:8 . Nut! ll/ 0 lawn-Iva. _* 5.. \oc. h 1 3 2955 as com 3 M £3. 8:3: 558.. ON . ciao? 28 r (MP/o.) l v 29 illuminated by a collimated source of monochromatic x-rays. Some of the radiation was scattered by electrons of the atoms without a change in wavelength. A diffracted beam.was produced when certain geometrical conditions (the well known Bragg relation) were satisfied. These requirements were satisfied by regular arrangements of atoms in randomly oriented crystallites in crystallized GelOAsZOTe7O. Diffraction patterns of crystals, or powdered crystalline materials, comprising both posi- tions and intensities of diffraction effects, are fundamental properties of samples, and can be used for identification since they are almost as unique as fingerprints. If diffraction conditions are known, informa- tion on structure or the arrangement of atoms in the sample may also be determined. Many techniques existed for creating x—rays, orienting samples, and detecting the diffracted x-rays. An early-model Phillips (Norelco) counter diffractometer was used to perform the measurements cited. A re-plotted diffraction patern for glassy GeloAaone7o is shown in Figure 3.7. Lumps of annealed, as-quenched bulk material were used as samples. Scattered intensity is shown plotted against crystal-plane spacings, d, which could cause diffraction according to the Bragg formula. Absence of sharp peaks in the pattern indicated the sample contained few crystallites of significant size. Broad peaks in the pattern, typical for amorphous materials, could have been used to (44) typically present in (2,9) analyze the small degree of short-range order these materials. A radial-distribution function, specifing the density of atoms or electrons as a function of radial distance from any reference atom or electron, is the maximum structural information that could have been obtained from such diffuse diffraction effects. 30 :00 Gem AsZOTe-m 80 Amorphous 3: '2’ ea- 0 E .0 4o:- 0 E 820- O a) \\\\\“‘-_‘ o—U—fifi—E—fl—m—a‘z I 3 2 LS l.O d (OJnm) Figure 3.7 X-Ray Diffraction Pattern, Amorphous S “no A"20“?0 Partially Crystallized 9.3 a Scattered Intensity 3 I /8 0 l :4 l Lwi _ 1 L- 20 K) 8 6 5 4 3 2 L5 10 d (OJnm) Figure 3.8 X-Ray Diffraction Pattern, Crystalline 31 This technique involving complex calculations, which has seen consider— able recent application in the study of structure<39'41) of amorphous chalcogenides, was not used here. Another pattern for crystallized GeloAsone7o, shown in Figure 3.3. contained many sharp peaks characteristic for crystalline materials. The ASTM x-ray powder-diffraction data file was used to positively identify tellurium as a major crystalline component in this pattern. Other crystalline species were certainly present in significant quanti- ties, but as in many other cases<38> could not be identified, primarily (5.37.38) due to the complexity of the pattern. Many workers have used x-ray diffraction information simply to confirm the amorphous nature of their materials; this was the primary application here. Similar techniques for investigating structure of amorphous (31,42,43) materials which offered advantages over x—ray techniques, were electron diffraction,(45) and neutron diffraction. Advantages of electron diffraction, beyond the utility of the higher-energy radiation source, were that electron-microscopic examination, microprobe analysis, and induced transformations could have been performed essentially simultaneously.(31) 3.6 Chemical Analysis Ordinary wet-chemical quantitative-analysis techniques were capable of determining, with adequate accuracy, the composition of chalcogenide alloys. Wet-chemical analysis was performed* on a sample from an early batch of glassy GeloAsone7o. The results of this analysis were: germaniumr7.832, arsenic-14.41%, and tellurium~77.772 by weight. * Ledoux & Company, Teaneck, New Jersey, examined ten grams from batch #10, using pieces which had been exposed to the atmosphere. 32 The equivalent atomic-number formula was: Gell.44A821.24T267.32’ a slight but definite departure from the composition desired. Possibly the batch from which the sample was selected was not homogenous, or not carefully-enough compounded. Probably later batches were closer to the desired composition as techniques were improved. Oxygen was the only impurity component examined, and was 0.052 by weight including surface- absorbed gas. This significant fraction was possibly due to the use of oxidized raw materials, especially arsenic, early in the research program. CHAPTER 4 ELECTRICAL PROPERTIES 4.1 Introduction Amorphous chalcogenides are interesting for their unusual and potentially useful electrical properties. The science of electrical- property measurement has been organized and studied for crystalline (1-7) materials. Orderly collections of data have been published for (8,9) a few materials, but little definitive lieterature has appeared for amorphous chalcogenide semiconductors. Collection of electrical properties for the relatively new chalcogenide glasses was a goal for this research program. The collection process was routine for some characteristics but very complex for others. An understanding of electrical conductivity required examination of (21) Factors experimentally invest- factors contributing to the effect. igated for GeloAsone7o were electrical conductivity, Seebeck coeffic- ient, conductivity switching, drift mobility,£nd photoconductivity. Dc and ac electrical conductivity versus temperature were measured. Seebeck coefficient was determined approximately. An injected-carrier transit-time drift-mobility measurement technique was adopted for future transport-properties experimentation. Photoconductivity at room temper- ature was briefly investigated. Few switching characteristics were examined in detail since this field was being thoroughly investigated by others. Several additional properties have been determined and reported by others, primarily for chalcogenide glasses other than GeloAsone7o. (lo-12) Pressure effects and piezoresistance, magnetic effects and 33 34 (13,14,95,120) magnetoresistance, and time-dependent effects<41> have been studied. Modification of characteristics by electron bombard- (15) ment and radiation exposure has been observed. Electronic tunneling has been investigated. Recombination and lifetimes of carriers have been reported for a few amorphous materialsglg’lg) 4.2 Dc Electrical Conductivity Many theories have been investigated for electrical conductivity of amorphous chalcogenide semiconductors; at ordinary temperatures a recent mobility—gap model may be accurate.* Examination of electrical conduct- ivity was essential for confirmation of the theories considered for amorphous GeloAsone70. Low-field dc electrical conductivity was measured at temperatures from -lSO °C to +100 °C using four-point probe apparatus of improved design, specially constructed for this application. (22,23) Several standard methods existed for measurement and (21) interpretation of dc electrical conductivity. Special techniques have been developed for certain applications,(24-26) including for use (27-30) (30,31) at very-low temperatures and with very-high resistances. Other unique conductivity measurement methods have been developed for _ as (32 35) The four-point probe technique was unusual applications. adopted here to measure conductivity for unprepared, randomly-shaped samples of amorphous GeloAsone7o semiconductor material. This method was also adapted to minimize the effect of contacts on conductivity measurements. Unfortunatelg four-point probe apparatus for lowbconduct- ivity measurements was not generally available, so special equipment was * A review of the theory for electrical conduction in non-cryst- alline conductors can be found in Appendix A. ** A description of apparatus and technique is in Appendix C. 35 designed and constructed. The lowest temperature at which measurements were made was about 120 °K. Lower-temperature measurements were desired, but were not practical with existing apparatus. The highest temperature was less than 380 °K, avoiding possible change or destruction of samples due to annealing, crystallization, or deformation. Higher—temperature measure- ments, to the softening point, were possible, but were not performed. The conductivity of amorphous GeloAsone7o ranged five orders in magni- tude, increasing monotomically from about 10'8 mhos/cm at ~150 °C to about 10'3 at +100 °C; typical conductivity at 27 °C, room temperature, was about SXIO-S mhos/cm. Azfiot of log conductivity versus reciprocal temperature, shown in Figure 4.1, resulted from tests of the law-field approximation. For semi-infinite samples, probe spacing, 8, test current, I, and meas- ured voltage between probes, V, conductivity was computed: I 2n sV The approximate* electric-field intensity in the sample was: V s Ea: and varied with temperature for the four-point probe method. The high- est field anticipated occurred for a maximumameasurable probe voltage of about 15 V and a normal probe-point spacing of 0.0635 cm; and was about 200 V/cm. This was much lower than values usually associated with non- ohmic effects in amorphous chalcogenide semiconductors.(34’35) Since maximum electric fields occurred only for very-low sample temperatures difficulty with switching and other non-ohmic effects was not expected$36) * This approximation was possibly law; the non-uniform electric field was probably higher at the surface near the probe points. 36 23:26.30 .8 cozofixoaad 30.... I33 no .3» _.¢ 952... 0.3.0.350... 23.034 .32....82 I Axovrxoog 23.. u ”to ea .6 n n I o m .oz .325 69.... 2956 0.8355233: 03... so onood n a . .35.: 32a .anLao... 2.05.8395 323 3:3 nfi. «em—0.u.mmmsd.m-mJ4Im—Fdfiebdfitdimhl1l]:t a 1 A 1 '7 C? use/sow uIJ‘(.o)°'bo-| — mngtonpuoa Ioaguoeua so 001 T F) I 37 Non-ohmic behavior, known for amorphous elemental semi— conductors,(37-39) has been widely reported for amorphous chalcogenides.(40-49) However, electric-field dependency was not apparent for conductivity measurements reported here. The plot shown in Figure 4.1 was measured using a test current of 1.0 us in addition to the most-frequently used value of 0.1 us. As no differences were apparent, test fields used were assumed sufficiently low. Measurements of conductivity at elevated fields were not made. Plots of log conductivity versus reciprocal temperature differed* somewhat from sample to sample. As seen in Figure 4.2, plots converged at higher temperatures and had slightly-different slopes at lower temperatures; greatest curvature was at high temperatures. Variations were attributed to differences in annealing and stabilization for the samples tested. Observed variations might have been due to differences in geometry factors used for computation of conductivity from raw data.** This factor, unity for semi-infinite flat-sided samples, was essentially temperature independent and constant for a given sample; size and shape variations could cause only uniform vertical displacements of logarithmically-plotted conductivity. Absence of significant vertical displacements of data suggested sample size was satisfactorily semi- infinite. Finally, exact magnitude was not as important as the temper- ature dependence of conductivity. * In private conversation with the author, many investigators disclosed having experienced similar difficulties correlating data. ** A description of conductivity-computation techniques can be found in Appendix C. 38 23235.... 322. 3.38350 .3536 Ne 2:2... a: at >0 flowiwzud Aural-Qt ab 3 b E. .m ooaocoaosngooE u on u. .3. :28 n 3.3.3 2.2.. ‘23 u «n.» as: 5:5 u N. .2330 :Zgom ES“. .3353 to. 2:53... a: ..o n u :3 38.0 nu €2.32 2.2a «Eon—I .50... 2.8.8395 26.3554 232.95» 2233 335.8: I Axozuxooo. in? ii... a- A i A '5. l 601 ‘Mgmanpuoo loomoela so no“; 1 L "I’ 4. qu/soqu u! A “(9) 39 Several formula were empirically fitted to observed conductivity versus temperature behavior using computer programs. Conductivity appeared eXponentially proportional to reciprocal absolute temperature over a fairly-wide temperature range, as indicated by the straight-line nature of the log-conductivity versus inverse-temperature curve. This characteristic, typical for amorphous chalcogenides,(37’38’42’48‘55) corresponded to the relation: 0 _ Ooe(-Wa/kT) Conductivity seemed to be thermally activated with an activation energy, We, of about 0.28 eV at 300 °K for bulk amorphous GeloAsone7o. Wa varied from sample to sample and with temperature. Values of 0.2 eV at low temperatures to 0.5 eV at high temperatures were abserved. Electrical conductivity mechanisms for amorphous semiconductors were not well understood. It was suggested that carrier mobility in these materials was exponentially temperature dependent according to a mobility energy-gap model, accounting for observed temperature dependence of conductivity. A computer program, DATAFIT, listed in Table 4.1, was used to fit (56) Values the exponential relation to data in a least-squares sense. for do and Wa were about 1.5 mhos/cm and 0.27 eV respectively. Careful checks determined that observed variations in log- conductivity versus reciprocal-temperature data were not caused by measurement technique. For instance, remeasured samples showed almost identical results; measurements at several locations on the same sample caused only slight variations. Annealing due to thermal cycling during conductivity measurements appeared not to be signicant, possibly Mun—unusu- COLON-O 3O 3! SI 60 7O 80 81 R2 85 90 92 94 96 I00 110 140 ISO 160 200 210 220 22I 230 231 240 241 250 25! 260 261 270 271 280 281 282 290 300 320 340 350 360 370 40 DCUBLE PRECISICN AC5):AH(505)sCCND¢SOIOCCNDLCSOIsTKCSO) DBUBLE PRECISION TKI(SOIsXKaOsPIaXLN2oCCNDLPoAIOA20A3 DGUBLE PRECISIGN A4JA5aTOoEsAA DIMENSION NSCSOIoPHI(SOsSIsYCSO):V(50)¢RH(250) PILENAME INOUUT CONDLPCM)8A(I)*PNI(H0I3+AC2)*PHI(H12) PRINTs" INPILE:UUTPILE”I INPUToINoOUT XK8I0380530°233 0'Io60209D-I9I PI'30I‘I59DO XLN2800693I47DOI XLOGEIOo43429400 LINE'OILZERO'OBERR'OeI READCINOI, LLsSsXIsVQPFoTHIK I FORMATCV) IO READCINsIsEND'IS) LLsTCHVsVV LINEBLINE§I3 NS(LINE)IOI VCLINEI'VDLTCVVsVOPP) TKCLINEI3207SD2O2084SDI*TCMV‘60074D‘I’TCNV*ICHV & ‘I0757D'IiTCMV*TCMV¢TCHV-Io657D-2tTCNV*TCNV*TCHV$TCHV & 06.96I0'3*TCNVtTCMVtTCHV*TCHV*TCHV TKICLINE)‘I¢DO/TK(LINE) IFCIOofiTHIKoLToS) 69 T0 I2 CONDCLINE)I0500*XI/(PI‘S*VCLINE))DBO TC I4 I2 CONDCLINE)3XLN2*XIICPI*THIK*V(LINE)I I4 CONDLCLINE)BDLOGCCUNDCLINE)I PHICLINEoI)3'9*TKI(LINE)/XKI PHI(LINE:2).IODO YCLINE)BCONDL(LINE)I 69 TU I0 I5 NSCI)8LINEI N32 CALL DLSOMMCPHIQYsAsRHpLINEstIsNSoAHoLINEoN) AIBACIII A2BDEXPCAC2)) PRINTo” CUND ' A * EXPC -E*0/CK*T) )" PRINT SaAII PRINT 60A2 5 PaRMATC" E 3 "oD2SeI6s" ELECTRON V9LT$”’ 6 PHRMATC" A I ”oD2SoI60" I/(OHH'CENTIHETER3)”) URITECEUTs2ILZERGoIN 2 PBRHATCII06X0”I/T”o6X0"ERROR”06X0A8) DO 20 L3I393 ERBXLOGEQCCONDLPCL’“CONDLCLII ERCERROERtsR . 2O NRITECUUTsa) LsTKI‘LIsER 3 FORMAT¢II1" "s2CFIO-6)’ D0 30 LSIOsLINEl ERIXLOGEtfcaNDLP(L)‘CCNDLCL’I ERR'ERROEthR 3O HRITECGUTs4) LoTKICL)oER 4 FGRHAT(I212CFI006)) ERR'SQRTCERRIILINE PRINTs” AVE RMS PIT ERR IN LCGCCCND)8”sERR PRINT:",LIST FILE! ”sOUTs” FOR ERROR V50 I/T” 999 STOP} END PUNCTIUN VOLTCVV:VOPP)1 VV'VV'VCPP IPCVVoLToOoOOI2’GO To I001 IPCVVoLToIo2)BC T. IIO VOLT'CVV+¢O3)/I.OO3I RETURN IOO V6LT8VV/lo011 RETURN IIO VOLT=VVIIoOO7I RETURN END Table 4.1 Computer Program DATAFIT 41 because temperatures used were not high enough or sufficiently sustained. Differencesamong samples were therefore real. Boules of GeloAsone70 produced by rapid quenching from the melt in silica ampuls appeared to have been quenched in a non-uniform manner; the outside of the boules were apparently more-rapidly cooled than the center.* Also, samples were annealed for various periods as long as about 16 hours. These treatments could have caused segregation of constituents, anisotropic (57) or non-homogenous strains, or differences in stabilization. Non- uniform characteristics from sample to sample were therefore probable, most-possibly caused by slight variations in annealing. Samples with exactly-known thermal histories were unavailable. Recent measurement of electrical conductivity versus temperature<57> f r h A T o amorp ous Ge10 820 e7O properties to annealing. Curves plotted here were similar to those have indicated sensitivity of electrical reported by others for short annealing periods. Effects of annealing at 99 °C for various periods are shown in Figure 4.3.(57) The slope at low temperatures decreased considerably with annealing periods greater than about ten hours. For no annealing the curve was remarkably straight and had a slope corresponding to an activation energy for conductivity of about 0.5 eV. All curves converged at higher temper- atures, where annealing seemed to have little effect. Exponential dependence of conductivity on temperature has been observed for other amorphous Ge-As-Te compositions<58'59) and for many additional amorphous chalcogenides; thermal activation of conductivity with an activation energy of about 0.5 eV was typical The * Information on preparation of amorphous GeloAsone70 samples can be found in Chapter 2 and Appendix B. Dc Electrical Conductivity - ( mhos/cm ), Log Scale 42 IO'2 N Amorphous GO|O AIonO-po . " Sensitivity to Annealing Time '- at 99 ’Celsius 3 ..¥‘ 0 I— g ‘ °- -\ - ' ' - an: 2‘ ' ° - lChr \ ' l2llr " \. Io“ _ °. __ ' lOllr no .. ”mm b a Results of Iizlma at ct, Ref!” 0.5 I l 1 l 1 I 1 1 1 1 I J as 3.0 3.5 lOOO/T(°K) - Reciprocal Absolute Temperature Figure 4.3 Conductivity versus Annealing Time at 99‘C 43 v;(60-62) activation energy for amorphous GeTe was 0.375 e the value for Asz'l‘e3 was 0.46 eV.(40'63’64) Moat investigators report thermal— activation processes for conductivity were dependent on electric-field intensity, purity, thermal history, and form (bulk or film) of the samples. Attempts were made to explain curvature of log-conductivity versus reciprocal-temperature plots reported here. Modified versions of the computer program, DATAFIT, were used to fit data in a least-squares sense with theoretically—suggested relationa.(73) Apparatus errors such as inaccurate temperature measurement and voltagedmeasurement nonlinearity were eliminated as sources for curvature. Power and polynomial functions of temperature used as pre-exponential factors were not useful for fitting data.(51’54) Also examined were fits to data involving temperature dependence of activation energy. Satisfactory fits were obtained but were not understandable physically and are not listed here. Electric-field dependence of activation energy was examined also, but suggested relations<42’45’47) did not fit data very well. Perhaps functions including terms or factors for several conduction (49) such as simultaneous drift and hopping could have been processes fitted to observed behavior; these were not investigated for lack of time and sufficient data. Conductivity versus temperature measurements were made an amorphous GelOAsone7o crystallized by slow heating to above the crystal- lization temperature.* The four—point probe technique was used with a test current of 0.1 ampere. Conductivity, substantially independent of * Thermal characteristics of GeloAsone70 were reported in Section 3.4. 44 temperature over the range from -150 °C to +100 °C, was about 70 mhos/cm. Measured similarly, the conductivity of 99.9992 pure single-crystal tellurium was 2.1 mhos/cm, slightly lower than for crystallized GelOAsone7o.* A small fraction of metallic arsenic (3XIO” mhos/cm conductivity) could account for the higher conductivity of crystallized GeloAsone70. Relatively few measurements were made on the crystallized state of GelOABZOTe70’ 4.3 Ac Electrical Admittance Electrical admittance versus temperature and frequency were measured for amorphous GelOAsZOTe7O. Ac measurements provided informa- tion on type and transport made for current carriers. Ac measurement techniques were also useful for eliminating certain contact problemsslll) Hopping, a mechanism for carrier transport characterized by conductivity which increases with frequency, was suggested as important for amorphous semiconductors at high frequencies and low temperatures. Information reported for A3233S114) A828e3,(114’116) GeTe,(6o'61) (37) in amorphous form suggests hopping conduction may be and for Ge important for many amorphous chalcogenides. Information reported here**was obtained from a single piece of amorphous GeloAsone7o ground to a thickness of about 0.04 cm; electrode area was not determined. Unusual and generally—unexplainable behavior was noted. As shown in Figure 4.4a, capacitance decreased with frequency increase at higher temperatures; conductance behavior was also complex. As shown in Figure 4.4b, at ~180 °C fi v—w . W‘fiVfiv‘rfi—r T * X—Ray diffraction patterns for crystallized Ge 0A82 Te indicating the presence of crystalline tellurium were discussed in geczgon 3.5. ** H. Boem, Owens—Illinois Inc., Toledo, Ohio, performed the measure— ments. 45 Amorphous GeIOAszd'e-m A 2 : 0‘. °C) '4 5 '06 f'" ‘2‘) .— ' '2 ii 9 . 8 o .9 = 3 3 2 ~ 3 U 0} l- 4 CI 3 T «30,0 ;J E 3'05 - c \ o -' '0' 2 C o '- o o o o I- QQQ '1 all g. a “ 8 :- 1 IO4 4 1 I 1 1 15 i j 4& :03 lo 10‘ to Frequency (Hz) Figure 4.40 Admittance at Higher Temperatures Amorphous Ge|oAsone7o Conductance (nanomhos) Capacitan ce (picofarads) lo2 _. -1 l0| . 't c(-l80 °c) l J . Io° '0 1 1 1 l L Io'r toT «:5 :06 Frequency (Hz) Figure 4.4b Admittance at -I80 °Celsius 46 capacitance was independent of frequency and conductance was dependent on frequency in the following manner: a - so + a1w0'85 Similar behavior has been reported for amorphous SiZSAsste50(113) where hopping was considered important. Negative capacitance for high dc-electric-field intensities has been reported for some amorphous (117) Other anomalous ac characteristics have been (118) chalcogenides. reported for chalcogenide glasses. Additional measurements of admittance may provide useful information for amorphous GeloAsone7o. 4.4 Thermoelectric Properties Information was obtained for polarity and approximate magnitude of Seebeck Coefficient (thermoelectric power) of amorphous GeloAsone7o. Instruments to perform this measurement were offered commercially* but (65) frequently used for determin- were not used. The hot-probe method, ation of carrier polarity, was used to make measurements. This technique is illustrated in Figure 4.5; the temperature difference between a heavy copper heat-sink and heated copper probe caused the generation of thermoelectric voltage between sample contacts. High sample resistance posed no problem as voltage was measured using a 10 GB input-resistance voltmeter (Hewlett-Packard Model 3450A Digital Multimeter). Calibration was established by measurement of materials with known Seebeck coefficient;(66) results indicated accuracy sufficient only for approximate quantative analysis. Experimentation over a range of average sample temperatures showed the Seebeck coefficient was positive and large. Seebeck coefficient, Q, * For instance the Thermoelectric Probe, Electra-Impulse Inc., Red Bank, New Jersey. 47 Hot-Junction Temperature Eal i ng: "Scalamp" SC PSS- 0626-6 Type T Thermocouple l/B'Rigid Copper Probe Heat Adjustment Heating Coil Transformer Step- Transtormer Thermoelectric Voltage A. Digital Multimeter '.Hat Probe" Glass-Tape Insulation Copper Connecting Wires Cold-Junction Temperature Copper Heat -Sinlr Figure 4.5 Hot-Probe Technique 48 at temperature T was computed: Vt Q" AT where AT is the temperature difference, To is the average temperature and Vt is the thermoelectric voltage measured from the hot junction to the cold junction.(65) For TO - 50 °C and AT - 100 °C the Seebeck coefficient was about +300 uV/°C. Strangelysso) the Seebeck coefficient appeared to increase slightly with temperature. Although Seebeck coefficient is useful and easily-measured approximately,(5) little detailed information has been reported for chalcogenides,(67'69’89) except for materials commonly used for thermo— electric applications. Exact determination of thermoelectric coefficient versus temperature is difficult and not often performed. Similar information can sometimes be gathered by other methods. From the theory of semiconductors,(74) polarity of the dominant current carrier is indicated by the polarity of the Seebeck coefficient. Seebeck coefficient for glassy GeloAsone7o, and for practically-all amorphous semiconductors,(50’72) was positive. The dominant carrier was therefore positive as for p-type or hole conduction. The very- high Seebeck coefficient of crystalline and glassy chalcogenide semiconductors such as BiZTe have led to their use in heat-pumping and power-conversion applications.(7o’71’87) 4.5 SwitchingiCharacteristics Considerable excitement has been created by discovery and application of reversible threshold and memory conductivity switching 49 in amorphous semiconductors.* Switching has been observed for amorphous elemental semiconductors,(75'76) for many compound-oxide and chalcogenide glasses, indeed for almost-all thin insulating layers.(77) Simple tests of amorphous GeloAsone70** indicated threshold and memory switching occurred at relatively-low voltages. Switching occurred in and across surfaces of bulk pieces, and through flash- evaporated thin films. Experiments with sharp-pointed electrodes placed about 625 microns (25 mils) apart on either side of a bulk sample showed switching at an electric-field intensity of about 2.65 kV/cm. Memory switching occurred with turn-on currents greater than about 1.0 ma. High, short current pulses returned switched-on devices to an off state. Evidence of filamentary conduction (appearance changes and erosion at electrodes) was apparent. Other organizations have observed switch— ing in Ge-As-Te glasses at an electric field intensity of 2.6 kV/cm.(35) Studies of filament formation accompanying switching<78t79> suggest that memory is accomplished by segregation of crystalline tellurium, or other highly conductive material, from the glassy phase. This process was observed for crystallization of amorphous GeloAsone7o by thermal-analysis, x-ray diffraction, and dc-electrical-conductivity measurements. Few switching measurements were performed as many others were already studying the effect for Ge-As-Te glasses.(36’58’80-86) * A brief review of the phenomena and theory of switching can be found in Appendix D. ** D. Mosher, Owens-Illinois, Inc., Toledo, Ohio. contributed some of the measurements of switching properties. 50 4.6 Drift Mobility Mobility is an important and useful characteristic of semiconductors. Mobility versus temperature information is a key to understanding trans- port and scattering processes.(21) Unfortunately, mobility for high- resistance amorphous conductors is usually low and difficult to measure. Conventional Hall-mobility measurement(5’9o) (26,91,92) apparatus and technique adapted for the purpose has provided anomalous results. Mobility for amorphous chalcogenides, derived from Hall- coefficient and dc-conductivity measurements, is very low (about 0.01 cm2/(VsI)and practically independent of temperature.(49’50’93'96) The sign of the Hall coefficient suggests n-type carriers, contrary to implications from thermoelectric experiments. It has been suggested<93> that amorphous conductors have p-type carriers; a significant (as high as 100 cm2/(Vs) possibly) temperature-activated form of drift mobility(97’51) is thought to account for the observed magnitude and (49-50) A mobility-gap model* temperature dependence of conductivity. has been proposed.(98-loo’122) Measurement of drift mobility in high-resistance amorphous semicon- ductors has been demonstrated using a transit-time technique.(104) This (105-107) or electron(101-103) method uses short-duration light pulses to generate a thin sheet, 6, containing N carriers of charge, q, at one surface of a sample of thickness, d. The technique is illustrated in Figure 4.6. The injected sheet of carriers was swept through the sample by the static electric field, E, established by the voltage source, Va. Transit-time, t,, of the injected sheet of charge was measured by * A brief review of this and other theories for carrier mobility in amorphous chalcogenides can be found in Appendix A. 51 tr8 tt Cur-t 't Time Rc>)t, ec<> Nq 22A (103) where A.was the area of the electrodes. it must be less than 2x107 for: 0.025 mm 0.01 cm? 10 kV/cm so 1.6x10-19 0 nnm>m IIIII wv vii V“‘j w * Relative permitivity of amorphous Ge1 Asone7o,determined by ac- admdttance tests, was possibly as high as 90. 53 An incident 0.1 Fe pulse of 5 kV electrons 10 us in length could generate about 6x10S carriers (assuming 0.5 eV activation energy). A 1% generation efficiency was assumed to account for various losses due to secondary emission and other processes.(103) The space—charge electric field produced by these carriers would be about 10 V/cm. Mability values of about 100 cm2/(V s) are possible<96> and could produce a transit-time of 0.25 us. A peak circuit current of 0.4 Ha would result.(103’108) An R value of 50 0 provided a signal-pulse amplitude of 20 uV. Stray circuit and cabling capacitances added to the sample capacitance and raised C to about 50 pF for the application considered. The resulting RC time-constant of 2.5 ns was less than the transit time. The test electric field, E, was established with a 25 V power supply, resulting in a steady sample current of about 1.0 ma. Blocking electrodes could have been used.(103) Alternatively if R I 20 k0, and added capacitance C - 0.005 "F were used, RC would have been 100 us, much longer than the transit- time. The current pulse could have been integrated and a peak signal of about 20 nV would again have been produced. However, a voltage supply of 50 V would have been necessary to supply the additional steady state voltage drop across the test resistor, R. Other values of N, R, C, and E could have been chosen, observing conditions that excessive fields be avoided, that small relative values of injected charge be used, and that the test signal be detectable. Different sample geometries might also have been chosen, within the constraints listed, provided construction of the necessary sandwich structure was practical. 54 Fabrication of apparatus to measure drift mobility was begun, and included modification of an available demountable CRT system. The CRT system included a high-vacuum pump (Veeco Model VS-400), and electron- beam gun assembly (RCA Model VC2126V4), and electronic circuits for accelerating, focusing, and deflecting the electron beam. A simple 60 pulse-per-second pulse generator was designed and constructed. The pulser controlled the electron gun to provide short jitter-free electron pulses, syncronized to the ac power-line frequency. Modification of the pulsing system may be necessary to provide shorter pulses.(109’110) Use of a transformer in place of the RC circuit for measurement of the transit-time pulse did not provide sufficient sensitivity. An RC network, a wideband amplifier, and an oscilloscope were used for signal detection, amplification, and display. A wide-band amplifier containing two 50 fl impedance-level XlO-gain, and one 1 MR input- resistance Xl-gain, 3 ns-risetime amplifier modules (Keithley Instr- ments Model 105) was acquired for use with both the R - 50 9 and R - 20 k9 cases. A dual-channel 50 fl impedance-level wide-bandwidth higher—sensitivity amplifier system (Hewlett-Packard Model 8447) was considered for use with only the R - 50 9 case. Signal waveforms were displayed with a sampling-oscilloscope system (Tektronix lSl plug-in with a Model 747 Oscilloscope), providing 1.0 cm of deflection for 20 uv signals. The risetime of the amplifier-oscilloscope system was less than about 10 us. A block diagram of the system under construction is shown in Figure 4.8. A suitable temperature—controlled sample-holder and beam—calibration target<110> has not yet been developed. Adaptation of the apparatus for conductivity 55 I lost!“ Ind learn ‘ Pe:oe I can '52:]... Centre! T“ Yoke \\ _ sit-vacuum. svsre’m aecuosoaeli *— ghouls! t ‘7 verse-bond * ‘Z-L . ........... I—- .—'—0" .......][ .........]-- Figure 4.8 peter moemtv Wear mum -Od 56 measurement<32’33) was contemplated. Conditions for measurement of drift mobility in amorphous GeloAsoneyo were established and analyzed. The experiment appeared feasible, but results are not reported here as construction has not been completed. 4.7 Photoconductive Effects Amorphous GelOAsone7o was examined for photoconductive effects. Photoconductivity can be defined as the difference between electrical conductivities measured under illumination and in the dark.(7) Photo- conductive generation of carriers provided a convenient technique for measurement of lifetime and mobility of current carriers(5’20) and therefore was important to the investigation of electrical proper- ties. Photoconductive effects have been observed for amorphous 8e’(105,107,124) GeTe,(125-128) (129) (130,131) AsZSe3, AszTea, and for (123) Photoconductive effects have 8180 been 2-1 investigated for switching glasses from the Ge-Si-As-Te system.(13 34) many other chalcogenides. Semiconducting samples illuminated with photons with energy beyond the absorption edge absorb much of the incident radiation near the surfaces generating a number of current carriers.* A thin sheet of carriers produced in this fashion has been used for measurements of drift mobility in amorphous semiconductors using the transit-time (107’135) The number of excess carriers generated and the technique. volume of sample in which they are generated can be controlled by adjustment of wavelength and intensity of the incident radiation. Since photoconductivity results from the generation of excess carriers it is * In amorphous semiconductors carrier-generation processes are complex. Excess carrier generation may be thought of as an activation of mobility for an additional number of current carriers. See Appendix A. 57 sensitive to wavelength and intensity of illumination. Conductivity measured for amorphous GelOAsone7o was not very sensitive to illumination at any wavelengths at room temperature. Very sensitive measurements were able to detect only slight photoconductive effects for moderate illumination with radiation at the absorption edge (A 8 1.7-um). Measurements were not made at lower temperatures.* Valuable information can be gained from the photoconductive relax- ation curve obtained upon interruption of sample illumination. Exponen- tial decay of photoconductivity with time can be related to lifetime of the excess carriers causing the photoconductive effects. Photoconductive-decay experiments have been performed for several amorphous chalcogenide semiconductors.(127’128’131'134) In general, the early part of the relaxation was very fast; the last part was very slow. The slow (hours at low temperature) decay was attributed to recombina- tion of trapped carriers.(131’133’134) Very-long post-illumination relaxation times have interfered with lifetime measurements. Estimates of lifetime for amorphous AszTe3 have been made as short as 50 ns.(131) Measurements of lifetime have not yet been made for amorphous GeloAsone7o. Equipment for measurement of photoconductivity versus temperature and wavelength is under construction. Equipment is also under construction for determin- (30) and ation of lifetime of carriers generated by photoconductive electron-beam injection. * Low-temperature conductivity measurements cited in this work.were determined with light-shielded apparatus. CHAPTER 5 OPTICAL CHARACTERISTICS 5.1 Introduction Reflectivity, absorption coefficient, and index of refraction were measured for glassy and crystallized GeloAsone7o. Initially, approximate reflectance and transmittance were determined. Later, careful measurements of absorption coefficient versus photon energy were also made. Only approximate reflection corrections could be made, and a region of the curve was undetermined due to difficulty obtaining suit— able sample thicknesses. Results in accordance with theory were generally obtained. Optical measurements, especially infrared absorption, have been very valuable for investigating semiconductor carrier energetics and other (1’3 e 8) solid-state properties. Calibrated optical measurements have been (4) used to determine impurity content of crystalline and vitreous (5-7) materials. Study of optical characteristics of chalcogenide glasses began with the discovery of infrared transmission in As283,(9’10) resulting in considerable effort to apply these materials to infrared (ll-14,52) optics. Surveys of optical characteristics for amorphous chalcogenide semiconductors have appeared recently.(14-18’48’S7) Many theoretical and experimental discussions of optical properties for amorphous semiconductors have been published. Early interest in arsenic-trisulfide-like glasses has continued.(19-26) (18,27-31) Amorphous (30-32) germanium and amorphous silicon have been studied, inpart to compare properties for the amorphous state to well known properties for the crystalline state. Selenium, readily rendered into 58 59 3“ amorph0ua condition,* has been studied extensively.(18t33'36) Also, optical pr0perties of amorphous GeTe have been investigated.(37’38) Relations between optical absorption and electrical conductivity (15.47.49) have been reported.(43) Electroabsorption experiments have been performed, showing only a slight effect. Reflection measurements (49—51) Index of reflection has been measured (1,14,52) have also been made. di (13,53) rectly, by using reflection data, and by analyzing interference effects.(55) The index of refraction frequently has a (53) this effect has been suggested for strong temperature dependence; light deflection applications. 5.2 Reflectance and Transmittance Data reported here was gathered using a spectrophotometer (Perkin- Elmer Model 450) with a 451-reflectance accessory (Perkin-Elmer 350 Specular Reflectance Accessory). Pieces of amorphous GeloAaone70 with smooth parallel sides were not available; attempts to cut and polish this fragile material were generally unsuccessful. All reflectance measurements were therefore made on flash-evaporated films (typically several umeters in thickness), and interference effects were present. Orthogonal reflectance, which provides equal optical-path lengths for both reflectance and transmittance tests would have been preferred, but equipment was not available. The total averaged 45° reflectance for a thin film of material on glass microscope coverslides, is shown in Figure 5.1. Reflectance included contributions from both front and back surfaces. The values were larger than values normally reported for reflectance of chalcogenide glasses. * Vitreous selenium is possibly the best known amorphous semicon- ductor. It is used as a photoconductive element in the electrophoto- graphic process, xerography. 60 03 Amorphous Gl'ohzo'rlfo O. Q 8 c” Q.“ a 30m- 3 -- Sash .- a: 0.2- GM- 0 0.! 3.2 g: 0' I5 I! II (I (Imflmq 15w.- Photon Energy (eV) Figure 5.I Average 45" Reflection Amorphous Be ”A: ”Te-,0 8 Transmittance 9 9 0| I l °s£ 0.2 0.4 as as to l.2 L4 LS is so 2.: efe ‘liw- Photon Energy (eV) Figure 5.2 Average Normal Transmission 61 Averaged normal transmittance is shown in Figure 5.2. Film thickness in both cases was about 4.5 um, and interference fringes in the transparent region (A > 1 um) were averaged. These results are qualitatively typical for all samples examined, but especially for reflectance were not quantitatively accurate. 5.3 Absorption Coefficient It was desired to compare the energy gap for optical activation As Te Careful with the energy gap for thermal activation in Ge10 20 70. measurements of optical absorption versus photon energy can provide useful information about the optical-energy gap. For crystalline semi- conductors, in the region of the spectrum where‘hm & W (the optical- 8 absorption edge), there is a rapid increase of optical absorption as a function of photon energy as electrons are excited across the energy gap, WS' In amorphous materials the same process is much more (18,39) (45) complex; the absorption edge is not usually well defined. The absorption coefficient increases at—first exponentially with photon (40,45,46) energy; no part of the curve can be identified positively with an Optical-energy gap. At higher energies, the absorption coefficient increases according to the relation: (1 n. (fun - Wg)In or in a more complex manner. Frequently, a small discrepancy exists between the energy gap determined optically and thermally.(42) Near-infrared absorbance was measured for amorphous GeloAsone70 film and bulk samples using the previously—cited spectrophotometer. Absorbance at longer wavelengths (2.5 umeters to 16 umeters) was measured using a simpler instrument (Perkin—Elmer Model 4373). The spectrophotometers provided data directly in absorbance units, A, which 62 were logarithmically related to transmittance,(59) T: A a Loglo( 1/T ) The double-beam technique was used to compensate for atmospheric and substrate absorptions. Exact correction for reflectance,(1t58) scatter— ing, and interference effects was not practical. Instead, approximate correction for reflection, by subtraction of equivalent absorbance, and averaging of interference fringes was done. Typical characteristics reported here resulted from many collections of data. Flash-evaporated films (0.1 umeters to 5 umeters in thickness) were used for short-wavelength measurements. Ground-down larger pieces, and freshly-broken chips of bulk glassy GeloAsone70 were used for measure— ments at longer wavelengths. Samples from 5 to about 100 pm in thickness were not obtainable, resulting in gaps in data gathered for absorption coefficient versus wavelength.(15) Absorbance as high as 6 was measured through the use of special neutral-density filters* in the reference- beam path. Careful adjustment of apparatus was necessary in order to maintain constant optical bandwidth to assure repeatability for data gathered in the absorption edge. Absorption coefficient, a , defined by Lambert's law for known thickness, d: T a e-ad (58) was computed according to the relation: a - Loge(lO)x-%- The natural log of absorption coefficient plotted versus photon energy, ‘fim, is shown in Figure 5.3. An exponential dependence of absorption was * Photo-etched thin-metal screens were used. These screens are available from Buckbee—Mears Co., St. Paul, Minnesota. They were individually calibrated for absorbance versus wavelength. 63 Ln(uo> >ua>wuusoaooouonm .au\>a ne.~ a cause censuses mansouasm .eanwmmom unanuuazm huoaoa use odonmauna .o.\>: oon+.e unusuammeou xoenoem mo.o3~b + on I o Hmoauuaoam “vuuuam 0. end. up aoneuosoaoo u¢ .aauu mawaoaass one ousueuonsuu ou aswuamaoo on: .>o n.c 1 N.o u m3 «au\nona mroaxm a o .unoan augusaaou oaaaaeum>uu >o son I b you menopause on sea amen seam“: >uw>auusvaou Aaxxszuv .uusumuomauu mo usooaonavsu >HHsUHuusum «vaus>auoe >HHmauenu on on .ao\mosa on a >uw>wuusvaoa Honduuuaao no swam consumes >uw>fiuusvaou Heuauuuuao av vHawmusoa .oo mud a unsusuamswu madnauuom .mxhx ~H.H I m< .09 can mason vauuuaowa was .Aoo av\a: ma I.maou doamaenxanamauana .mxnx «m.H I m< .oo New u>one unusuaoaa and .m\H Ha.o a m< .oo cad a>one aoaueuwadsum>uo .0. cos 9 Au. mv\a mw~.o . .u. can e As. mv\n ~wm.o 1 u. on 9 Au. mv\n sHm.o u use: oamaoosm u. on 9 Au. mv\a mmu.o 1 been unusuuam assumes .oo nme um maaau newsman .oo oqa um vexeaa «w\h um.~ was 0o ooa us mnoa now Nm.o omnu mood souusoaxo awuonomoau< mafiaeaase muse: 0H Ham caduceus casuanuooam .oo own I ensueuunauu mnauaaz .oo ONH I ausueuaoaau doausauomoseuelmmsau .ounum .anHon uo encompass menu nonwsou one mucus: .msouoo use ou wanedmmam .vammauum magma: .oaamsum Hmofimmsm assesses .oueum msonnuoam seau mama >uamaun .msoz.w.~ a mmaavuam .mao\m w.m was muumnun .muuaaamum>uo wanna use aswusaauu oawaamum>uu .sOfiuumummwo >nu1x oaueawoaa suauusn aoauumummao emuix .ofloama> uo >moomauuaa Hosanna mo uanm>uamoo ousuosuum ouaeusonn< mead a: OH 8 mouwaamum>uo .Sm«>ouw one Hana oz .xuno was guano .maamowonoaoo vuusuumuh oumum pauuaaoum>uo eunum msonouoad muuoooum ohmHONsaoa ac mo maauuonoum mo humaasm H.o manna 72 .a: c A 4 you m.m a sowuuaumuu mo xuoaH .moanamm Edam umuu 1Hnowumo you no on 9 >0 He.o usage was as .ms .huw>wuasvaoa you unease coaun>wuum 1anaua£u anu mafia» menu mama maunmwao on: as sets: . as: ._ a ”was easauow nouuao e ammo nowunuomoe emu QH meanness seamen u¢ Hosanna .u. on e >o Ed 1 as no on e >0 mno.o u 03 “muons 005 I 5 332m: . an: . .8: m.n v 4 v a: e vamp soauauomoe maouum ”amps aoaunuounm ea meanness 30H u< .an m.~ I x as amps sowumuomom endgame HH03 .ooaawao hauooo .3: m.H a 4 «was douumuome< ouaum wouwaaoum>uo oumum msonnuoa< humonoum Avuudouv H.o manna 73 and brisk; significant advancements have been reported recently. Published results have only served to generate additional questions and suggest subjects and opportunities for additional research. The following list summarizes suggested steps for improving experimental techniques and obtaining additional information about electrical-current-carrier type and transport mode: 1. 10. ll. 12. Improve methods for material synthesis. Determine sensitivity of properties to annealing and other preparational techniques. Examine procedures for cutting and polishing samples. Adapt four-point probe conductivity measurement apparatus for a wider range of temperatures. Obtain a better low-current source and dc-voltage preamplifier. Fabricate apparatus for ao—admittance measurement over wide temperature and frequency ranges. Fabricate apparatus for measurement of photoconductivity versus wavelength for a wide range of temperatures. Measure time response of photoconductivity to determine lifetime of carriers. Measure electric-field intensity dependence of dc conductance, ac admittance, and photoconductivity. Measure Seebeck coefficient versus temperature. Determine drift mobility as a function of temperature and electric-field intensity by transit-time measurements using photoconductive and electron-beam injection. Determine normal reflectance as a function of wavelength for bulk and film samples. Make additional measurements of absorption in the absorption edge. Measure energies of thermal reactions using DTA techniques. Investigate effects of annealing time and quenching rate. Determine property sensitivity to impurities and composition. Investigate switching effects and determine composition and structure of crystallized material. These were the most important suggestions which arose; certainly many more would occur with continued work. .n . 1 1.1-Ag 31’1“ .‘l‘li‘ ‘l.lllill 1|. APPENDIX APPENDIX A NON-CRYSTALLINE ELECTRICAL CONDUCTORS Part 1 -- Historical Introduction Identification and description of non-crystalline materials has become important in recent times. Several reviews of characteristics (1-4) of the amorphous state have been published. Considerable study of non-crystalline materials has provided information necessary for classi- fication and description of amorphous semiconductors.(5’6) Information about preparation and physical chemistry of amorphous semiconductors has become available.(7-ll) Much progress in the study of semiconducting 8138898 has been reported at conferences.(12-l7) Early research into characteristics and applications of non-crystal- line electrical conductors began in the Soviet Union with the work of (18—19) academician A. F. Ioffe. Much of the early theoretical work (20) was summarized by Gubanov. In the western world early work on disordered electrical conductors began as an extension of crystalline- semiconductor research; understanding of electrical properties of perfectly-crystallized solids prompted study of less-perfect materials. Many excellent general reviews of electrical properties of crystalline (21-25) Several short surveys of 6 26 27 work an amorphous semiconductors have been published.( ’ ’ ) solid-state materials are available. Professor N. F. Matt has contributed greatly to the understanding of electrical conduction in disordered materials.(28-3S) Non-crystalline conductors can be separated into about five categories: electrolytes, organics, liquid metals, oxide glasses, and chalcogenide glasses. Electrical conduction in electrolytes has been 74 75 studied extensively as a branch of chemistry.(36) Much of the work on organic semiconductors has also been done by chemists.(37) N. F. Mott initiated study of liquid metals in 1960, and considerable work has since been done.(38’39) Transition-metal oxides, studied initially as an extension of ordinary silicate-glass technology, are an important category of disordered solid-state electrical conductors.(4o’41) Perhaps the potentially most—important category is chalcogenide glass, a family of materials containing sulfur, selenium, or tellurium as a major constituent. Chalcogenides, investigated for some time for useful thermoelectric properties, became especially important with the recent development of conductivity switching effects* by S. R. Ovshinsky. Part 2 -- Theory of Electrical Conductivity Work on theory of electrical conduction in disordered materials began primarily in the Soviet Union.(20) Presence of short-range order in amorphous materials was suggested from atomic radial-distribution (43) Long-range order was functions derived from x-ray diffraction data. absent, confirmed by the absence of sharp peaks in the x-ray diffraction pattern. The familiar quantum-mechanical treatment for electrons in crystalline semiconductors was extended to disordered materials, by assuming a perturbed periodic arrangement of atoms in which long-range order was completely disrupted. It was determined that an energy-band (44-55’59) even for situations in which there was model was applicable no long-range order. However, it was determined that the mean-free path for electrons was very short, comparable to interatomic spacings, for only slight variations in nearest-neighbor distance. * A short review of phenomena and theory of switching can be found in Appendix D. 76 Studies of localized electrons have indicated that the electrons (56-62) can interact significantly. N. F. Matt has suggested that the electronic-wave functions may be considered localized only for energies below some particular value, and that transition to free-electron behavior is similar to phase or other thermodynamic changes.(63-68) The Mott nonmetal-metal transition has been studied extensively in an attempt to understand switching pehnomena.(69-72) Considerable study has been made of carrier dynamics in disordered materials. N. F. Matt has noted characteristics similar to compensated (34,35) impurity-level conduction for crystalline semiconductors. A form of carrier motion by thermally-activated hopping from one localized state to another has been described for polarons in transition-metal oxide glasses,(88’89) (90,99) and seems applicable to some amorphous chalcog- enides. Space-charge conduction at high fields has been observed (91-95) for insulators and may be important for amorphous semiconductors. Double-injection processes have been studied to explain switching (96-98) initiation. Chemical bonding and electrical conductivity have been related for amorphous conductors.(loo-102) An internal electric- field theory has been used to account for optical absorption in some amorphous chalcogenides.(103) K. W. Boer has studied disordered conductors and contributed (73-77) greatly to understanding of conduction mechanisms. Cohen, Fritzsche, and Ovshinsky have suggested a simple band model (CFO Model) (78) Together for electrical conduction in amorphous semiconductors. with theoretical contributions by N. F. Mott, these hypotheses have provided tentatively-satisfying explanations for properties of some disordered electrical conductors. The model assumes ambipolar 77 conductivity for a high density of carriers in self-compensated material.(75) An energy-band model for mobility is used, resulting in a mobility gap rather than a gap in the density of states to account for temperature activated conductivity. Exponential dependence of mobility, a frequently noted property for disordered electrical (79’80) results. P-type conduction occurs due to a higher P‘ (75) Much conductors, effective energy-level density in the "valence" band. progress understanding electrical conductivity for amorphous semiconduc- t l _ l tors has resulted from application of the CFO model.(38’81 86) i | APPENDIX B MATERIAL PREPARATION TECHNIQUES Part 1 -- Silica Ampul Method A popular technique for preparing toxic or reactive alloys is fusion in evacuated ampuls. This process was employed for the prepara- tion of amorphous GelOASZOTe70 from elemental starting constituents. Lumps, shot, and powders of germanium, arsenic, and tellurium were carefully weighed into quartz-glass ampuls. The selection and applica- tion of ampuls is described in this section. Initially, ineXpensive, non-transparent "satin-surface" silica tubing, 12 mm-inside diameter and 0.4 mm-wall thickness, was obtained.* However, this silica form was difficult to handle for most glassblowing Operations, and was suspected of contaminating the semiconductor. The more expensive and transparent variety of silica tubing was used successfully. Pieces 30 cm long were sealed at one end and constrictions were formed near the center. Considerable skill is required to work softened silica; considerable trial-and—error learning and practice was required. Quartz-glass fusing and forming operations were performed using an oxygen-hydrogen hand torch** with and without a holder. High-density welder's goggles were used for safety and eye protection. * All silica materials discussed here were obtained from Thermal American Fused Quartz Co., Montville, New Jersey. ** The glassblowing equipment was mostly obtained from National Welding division of Veriflo Corp., Richmond, California. 78 79 Subsequently, silica ampuls were procured already closed at one end* but without constrictions. Forming constrictions and sealing off the ampul under vacuum was difficult so a different sealing technique was adopted. A small dimple, formed in the wall of the ampuls before filling, supported a short close-fitting silica plug* dropped into the ampul just before evacuation. After evacuation, ampuls were sealed by fusing the walls to the plug. Seals formed in this manner never cracked or leaked. Evacuation of the ampuls was done using a small vacuum system specially fabricated for the purpose. A mechanical vacuum pump (Welch DuoSeal Model 1400B) was connected with flexible copper tubing to an all- glass manifold including valves and traps. One trap was chilled by liquid nitrogen to capture toxic vapors created during the sealing operation. Another was filled with glass wool to capture dust particles drawn from the ampul during evacuation. The input connection to the vacuum system was a ground-glass tapered joint. A mating joint was coupled to silica ampuls by a short length of rubber base. The rubber hose allowed easy inexpensive connections, and the ordinary-glass tapered joint allowed rotation of the ampul during sealing. Ampul pressure, measured by a Pirani gauge attached to the system, was reduced to below 10'3 torr with the aid of the cold trap. After evacuation and sealing, ampuls were disconnected from the vacuum system and a hole was drilled above the seal for a wire used to hang the ampuls in the rocking tube furnace. Also, excess tubing length was removed. After fusion, quench, and annealing ampul contents, the * Vitreosil tubing, 16 mm-bore, normal—wall and transparent, closed- one-end, and matching 15 mm-O.D. silica rod stock was obtained in one- foot lengths from the previously-mentioned manufacturer. 80 ampul was carefully opened in a disposable glove bag. Usually ampuls were fractured by a sharp blow; resulting in broken samples. Some ampuls were opened carefully using an ordinary glass-cutting saw, but attempts to slice amorphous material while still in ampuls were not successful. Pieces of glassy material, exhibiting conchoidal fracture, are shown in Figure B-l. Part 2 -- Rocking Tube Furnace A small 2000-watt tube furnace, of unknown manufacture, was avail- able for the research program. This furnace, about 3.5 cm bore and 0.5 meter long, was layer wound with nichrome wire and amply insulated for high-temperature operation. A steel outside casing completed construction. The furnace was suspended from a bracket and enclosed by a fume hood as illustrated in Figure 3-2. The furnace was rocked at approximately one cycle per second about the pivot by a small variable-speed motor and eccentric drive. The maximum displacement of the bottom of the furnace was only about 5 cm. The ampul, suspended from outside and above the furnace by a 0.5 mm- diameter nickel or stainless-steel wire, tapped the inside walls of the furnace lightly each stroke. This was considered advantageous, and abrasion of the ampuls caused no difficulties. Rocking was stopped upon completion of heat treatments, the thermocouple was removed, and an armored quench-bath was arranged directly below the bottom of the furnace. Quenching was accomplished simply and quickly by cutting the suspension wire, allowing the ampul to drop directly into the quench- bath. Furnace temperature was monitored by a cromel-alumel thermocouple inserted into the bore through the bottom opening. This thermocouple 81 Figure B—l Pieces of Amorphous GeloAsone7o Spud Control I 1 . . 2000 Watt Tube Furnac. / Support WI" _—’-” Figure B-Z Rocking Tube Furnace 83 was also used for temperature control by use of an on-off type temper- ature controller (Honeywell-Versatronic Model-R7161B, 0-1200 °C. type-K). A silicon-controlled rectifier-type power control was used to adjust heating rate by proportioning available power. Manual programming was used. The furnace arrangement and operation was versatile and easy to use; it proved satisfactory in every respect. Part 3 -- Flash Evaporation of Films Many techniques exist for preparation of thin films. Thermal evaporation from resistance-heated filaments or boat sources is a well established practice. Not so-well known are flash and quasi-flash- evaporation techniques. In the ordinary flash-evaporation method lumps of material to be vaporized are dropped piece-by-piece onto a preheated crucible and vaporize practically upon contact. However, some materials vaporize too violently or segregate greatly, and cannot be flash evaporated from an ordinary open crucible. An arrangement is available which provides an indirect path for travel of vapors and which allows easy mixing of vapors prior to emission. This arrangement, generally called a baffled source,* is illustrated in Figure B-3. Evaporation of materials which react with metals from which sources are constructed (molybdenum, tantalum), requires a thin coating of aluminum-oxide powder to be sintered onto the interior of the source. Very—rapid resistance heating of sources can sometimes be obtained. Occasionally, sufficiently-rapid evaporation rates can be obtained simply by switching on a very-high power source. This form of quasi-flash-evaporation was utilized to deposit thin films of GeloAsone7O. * "Baffled-Box" type evaporation sources (alumina-coated interiors if specified) are available from R. D. Mathis Co., Long Beach, Calif. 84 ’ Material Feed 1:57 i l y E 1 l Heater i f l + Power I I A :r H1 .L .' ' i ' | ' p --_-‘:.:.*.'.’-Ll:‘:;----- Radiation Shield Figure B-3a Flash Evaporation,Baffled Boat 10 Pre- Pre- Load Substrate Load 1 t 1 fl T—T -—|1—l .1.- , Heater 7. I l : l ' l _ Power Fl " ‘e ',h $1511 111 ' '1' A: (.9 '1) [VJ/.31 :1): g; ‘1 l1: 1: (y 1 s . l e “Pry/x ,‘1yi ' ”(fur-2’ <1”) "1"" *1111 | ”11 "awn. l L\ ,l 'vl.‘_-+_JJI tr \ ) (\2 *4 “W‘ 1 ”vii. Figure Bs-3b Quasi-Flash Evaporation; Baffled Boat 85 An alumina-coated "baffled-box" source was obtained (R. D. Mathis Types SMlO & SMll) and installed in a conventional vacuum coater (Norton Co. (N.R.C.) Model 3176) replacing an ordinary resistance-heated filament source. A holder for thin-glass* substrates was also installed. Very-thin film—glass substrates were used for minimum interference with optical-absorption measurements. The substrates were dc—glow-discharge cleaned prior to evaporation coating. The substrates were not heated during evaporation, but were allowed to come to equilib- rium with room temperatures prior to evaporation. Evaporation opera- tions occurred rapidly enough so that radiant heating from the source was not a problem. A few grains of glassy GeloAsone7O material were loaded into the boat source. A correct mixture of powdered elements might have been used instead, but this possibility was not investigated. The substrate holder was arranged about 30 cm from the source; shielding was used where necessary to simplify later toxic clean-up operations. The chamber was evacuated to less than 10"6 torr and heating current (greater than 400 amperes initially) was switched on abruptly. Vapor- ization of the entire load of material appeared to occur in less than twenty seconds. Film thickness was controlled by substrate—to—source distance and by initial-load size. Films thinner than 0.1 um were frequently pinholed; suitable films as thick as 1.0 um.were obtained in a single flash- evaporation step. Successive identical evaporations reduced difficulties with pinholes, and allowed buildup of films several um in * 48 mm x 50 mm Number-1 Exax cover glass[#l9l60, Kimble division of Owens-Illinois Inc., was used for this purpose. 86 thickness. No layer-interface problems were noticed, even though films were exposed to the atmosphere between evaporation steps. Adhesion to substrates was very good. Segregation difficulties, frequently observed when evaporating alloys, were not apparent. Evaporated films were not annealed. Structural differences from bulk material were possible and probable. Electrical and optical properties of quasi-flash-evaporated films differed only in an accountable manner from properties of bulk material quenched from a melt. However, careful analysis of data for differences was not performed, and exact composition of films was never determined. APPENDIX C DC CONDUCTIVITY MEASUREMENT Part 1 -- General Considerations The four—point probe method was used to measure electrical conduc- tivity of amorphous Ge 3 Te semiconductor. This technique over- 10A 20 70 came difficulties associated with high-resistance and rectifying contacts, frequently observed at metal-semiconductor connections. It also permitted measurement of conductivity for samples having a variety of sizes and shapes. These and other advantages proved useful for measuring conductivity in lump and film samples of GeloAsone7o. It was first anticipated that conductivity of amorphous GeloAsone70 would be comparable to or slightly less than that observed for familiar crystalline semiconductors. Measurement, however, indicated conductivity was as low as 10‘6 mhos/cm. This led to refine- ment of a conventional four-point probe system to provide low leakages, high sensitivity, and good isolation between current supplies and volt- meters. The system developed is described here. The basic technique for four-point probe conductivity measurement(1’2) is illustrated in Figure C-l. Four pointed probes were placed on a relatively-flat surface of material to be measured. A fixed current passed through the outer electrodes developed voltage drops in the material near the probes. Voltage drops at the current-supply contacts were absorbed by the current-supply system. The voltage drop between the two inside probes was determined without causing any current to flow in these electrodes through the use of a high-resistance voltmeter floating with respect to the current supply. Since no current flowed 87 I 1 f I I I l a ‘ S 1 ii Probes I I Semiconductor Material Constant Current .. isolated Source U High Resistance Voltmeter Semi- Infinite Sample Temperature Controlled i Heat -Sini1 : Figure C-lb Four-Point Conductivity Measurement 89 in the voltmeter circuit, voltage drops at these contacts were elimin- ated. Several arrangements of probes or contacts have been used.(4’5) Four in-line and equally-spaced probes were used. For an arbitrarily- shaped sample conductivity may be computed: a - I x 0 2n x 8 x V where 0 is a geometry-dependent correction factor. For flat-sided semi- F infinite samples frequently approximated by large lumps, this factor-is unity. 0 has been determined for many other cases and geometries such as samples with boundries near the measurement and very-thin (19225—8) samples. Many applications and variations of the four-point L probe conductivity-measurement technique have been developed and studied.(9-12’SO) Part 2 -- Four-Point Probe Head Purchase of a suitable four-point probe conductivity-measurement system was attempted, but apparatus was not generally available. Several systems have been offered recently, however, such as the Sylvania Model RM—l Surface-Resistivity Meter. Electronic systems for existing probe-head assemblies have also become available.* Probe—head assemblies were available from several sources.** An Abacus Model A4P-25 probe head was fastened to a discontinued resistivity-test stand (Model JP, Baird Associates Inc., Cambridge, Massachusetts) used to hold and position the probe head. The chosen probe head consisted of four adjustable-coil spring-loaded tempered-steel * Hewlett-Packard Model 4329A Resistance Meter and digital ohm- meters such as the Hewlett-Packard Model 3450A Multi-Function Meter. ** A. 8 M. Fell, LTD, London, England; Dumas Instrument Company, Costa Mesa, California; and Abacus Instruments division of Alessi Industries, El Segundo, California. 9O probe pins equally spaced 0.0635 cm i 1: by a cast-epoxy guide block. Maximum-available probe pressure, about 180 grams, was used. The manufacturer's technique and others(13) were used to maintain sharpness of the points. Part 3 -- Description of Electronic Subsystems Several factors were considered in the choice of an appropriate constant-current supply for the system. Measurement over a wide conductivity range was desired and since conductivity was fairly low, it was necessary to use a well-isolated current source having good regulation and high-voltage capability. Initially, the design of an isolated constant-current source was attempted, following the (14’15) of a Model LCD-078 Constant-Current Converter. example Argonaut Associates, Beaverton, Oregon. Good isolation was maintained first through the use of battery packs, and later by special isolation transformers.(16) Unfortunately however, a system with sufficiently- high isolation, regulation, and voltage could not be developed. Schematic diagrams of two trial systems are given in Figure C-2. Eventually, a suitable constant-current source (Hewlett-Packard COB-6181B Constant-Current DC Power Source) with an isolated and guarded output of O - 2.5/25/250 milliamperes at a maximum of 100 volts, was obtained. This instrument had quite satisfactory isolation and represented about the state-of-the—art for such systems.(17’18’19) * Also recently offered are other supplies. A Keithley Model 225 Constant-Current Supply, capable of supplying highly-regulated currents * Luchter Instruments, Walnut Creek, California; Electronic Measurements, Neptune, New Jersey; and Keithley Instrument Co., Cleveland, Ohio. The Hewlett-Packard COB-6186B with higher-resolution and higher- voltage capability has recently became available. 91 i l5K,2W : E 1511,2w i + E dL— : .2. 1 7:- "‘.75° 11120-11111. - 1 "I LO'O‘IO‘IA. . 'rI'-ss i 1.111 TI-56 i 111 L—c/c; - Figure C-Za Trial Constant-Current Supply Guarda Power {6----13+ Linc : 1113044 1. 5 "if 1’ i 113044 .000 , - i 2211 I toe-15 3 ' i ‘ 10‘ I. i011 i 2.3:: 2-5" ’e I 211.3739 l _so111= $ “2.0" I lN52lSS E ltd-5238 ' L I 211.3712 ' 39K 8e t i ' 1K : Case ,1: "i 311 410 2 I - f I I a1 + 112 = s Volts/Io," Figure C-Zb Trial Constant-Current Supply 92 as low as one nanoampere, would have been used had it been available when needed. For many measurements it was necessary to use a test current of 0.1 microampere. This was not attainable from the Model CCB—6181B Supply. A simple but effective method for obtaining this constant current is illustrated in Figure C-3. The dc constant-voltage supply used is a highly-stable and well-isolated model. It was necessary to adjust this system manually to achieve constant-current operation when very-low conductivities were measured. As will be described, most of the conductivity-test system operated almost automatically; manual adjustment proved practical for maintaining constant current. Power- dissipation limitations prohibited use of this system to provide higher currents. Fortunately, operation at a current above one microampere was achieved satisfactorily by using the Model COB-6181B Supply. Considerable time and effort was taken for development of a suitable voltage-measuring system with input resistance about 10 G0 (to prevent loading). A system of the form shown in Figure C-4 was developed, to measure and record conductivity over a wide range with accuracy and minimum operational manipulation. It was impractical to sufficiently isolate a voltmeter from the current source by separately floating these subsystems. Instead, the center of a differential amplifier was referenced to one-half the voltage developed across the current supply, (as supplied by a divider as in Figure C-3). Design of the preamplifier involved many considerations.(20-25) The preamplifier developed, shown in Figure C-5, was a differential dc type with high differential and common-mode input resistances. High common-mode rejection permitted moderate imbalances of electrode AC Line 1 53.1. 330 Harrison 6522A Power Supply 0-2000 F1 Volts Current Set 500 11119 3 . Partametric F— Voltmeter 500K 500K 1"”— ‘ jiuf Current Measure Floating Ground Figure C-S A Special 0111A Constant-Current Supply Floating Ground Current Pnampiifier m: 011111111111 J— Source Recorder Current ‘ y ‘ _.J. Mater f Probes hermocouple ' Figure C-4 DC Conductivity-Measurement System Diagram 94 5285 .326 5:35.... 3.2.255 on one 23.... d. 1 R: > m a. 2 93020 . «c. 3.53.. 82H Or| 2X). 6n 52:0 go. . . n C Aw A. 324 5.50 x8« «1 :8. . ad. a. a gas a .- au IZ_<0 - #5 > 6mm“ - . 5+ » ’H :8— 8°“ 4.2-(0 x00— x O 2 202 .118... p8. . 5.3.2. + . 0m! 0— «.smzu 2.32.1 ~~O ..--..-... . on: o. monza 95 voltages due possibly to inequal contact resistances. Usually, the voltage midway along the drop across the current supply was approx- imately the same as that found at the sample surface near the center of the probe array. Several features of the differential preamplifier require eXplan- ation. The device used to achieve differential to single-ended conversion was a Fairchild—Semiconductor uA725 amplifier-microcircuit With high stability. Adjustable feedback(26-28) was used to set gain to 1 i 0.12 with common-mode rejection greater than -80 decibels. Choice of unity-gain resulted from several trial measurements to deter- mine the size of voltages to be monitored. Risetime of the amplifier was set to about one second to reject noise and 60 Hz signals. High input resistance was obtained by addition of temperature-compensated field-effect-transistor voltage followers at the inputs of the feed- back amplifier. Protection against high transients was provided by RC networks in the input leads. Care was taken in the design of the differential preamplifier to assure maximum thermal stability. The microcircuit was a premium type and input FET's were matched pairs. The dc power supply for this amplifier was a balanced and stabilized commercially-available system. Special, highly-stable or precision components were used where appropriate. The output voltage offset could be nulled and was quite stable. A useful range from less than 1 mV to greater than 10 V, the maximum output of the microcircuit amplifier, was available. A commercially-available preamplifier could not be selected as during early stages of design of the conductivity-measurement system characteristics required for the amplifier were not firmly established. 96 (29-32) such as Properties characteristic to differential amplifiers gain, linearity, common-mode rejection, input resistance, dc stability, etc. were measured for the developed amplifier and found satisfactory. Any of several commercially—available data—acquisition systems would have been suitable; possibly the conductivity-measurement system could have been improved through application of one of them. Data was recorded with a Bryans Model 22000 X-Y Auto Plotter. Voltage measurements were recorded on the Y-axis over five or more orders of magnitude by range-switching sensitivity of the plotter. Accuracy and linearity of the voltage-measurement system were established using a Hewlett-Packard 3450A digital multimeter for calibration. Also, a panel digital voltmeter (Model 4304, API Instruments, Chesterland, Ohio), was used to continously monitor the output of the preamplifier. Temperature of the sample was measured<33’34) using a specially , 6 calibrated copper-constantan thermocoupleSBS 3 ) A reference thermo- couple was maintained at 0 °C by means of a water-ice bath. The difference between sample and reference-thermocouple voltages was recorded directly on the X-axis at the X-Y plotter. Also, a galvano- meter (Scalamp Model, Ealing Corp., Cambridge, Mass.) was available to continuously monitor temperatures. The sample thermocouple was placed adjacent to the sample during conductivity measurements, and prior (36-38) against several standardamelting—point to use was calibrated * references. A temperature versus thermocouple-voltage curve plotted by this process is shown in Figure C-6. A General-Electric Mark-I timesharing computer program, POLFIT***, was used to generate * TherMetric Standards, T-420 etc.; Fisher Chemical Company. 97 Dental c Asia ' ....L +30+ ('C) .l. O i ""9“” (o “C Water-ice Reference) Temperature -2 1 1L 4 1 1 L 1 1 -3 -4 -3 -2 -1 a +1 fl +4 +3 Copper-Constantan Thermocouple (miliiValts) Thermocouple Calibration Figure C-S F 43Volts Ac ——-e+C Volts Dc 2113713 _ A _ «‘- Tfifld 5 r—sox , 3 --"-' Reed Refs 3 u L 5.3 c 1 y 3 7° 2 >~ 2 1. g e - h a - . e- 8 y “'8 g 9 s 3 CD 3 '80'0'.‘ F8 8 TO Ac Power 11 9 Control 3 :1 Logic ' 2 ‘g a ' 1M ' a Floating Ground . . Probe-Reversing System Figure C-T 98 a polynomial equation fitting the data in a least-squares sense: T[°K] = 273. + 28.4th - 0.607ch2 - 0.176XVt3 - 0.0166ch” + 0.00696xvc5 where Vt was the thermocouple millivoltage. This equation was used instead of the curve to determine temperature from recorded data. Several additions were made to the simple conductivity measurement system described previously. A system for periodically reversing the direction of test—current flow(3’9’10’40) (and voltage measurement) was added to average sources of error such as unbalanced thermoelectric voltages generated at probe contacts. Four Magnecraft 103LMPCX9 Magnetic Latching-Reed Relays<41-43) were modified by removal of cases and terminal strips to reduce leakage currents, and used as periodic- reversing switches. Circuit diagrams of the switching system and associated power supply are shown in Figure C-7. A function sequence controller was added to complete the electronic system. The controller provided automatic periodic operation of the reversing switch described earlier, and automatic lifting of the X-Y plotter pen during switching operations. A schematic diagram of the control system is shown in Figure C-8. Logical operations performed here require no comment except to note that the pen was lifted about one-second prior to switching and was returned to operation three- seconds following switching (to prevent recording of switching transients). A complete period of operation required approximately seventeen seconds. The common reference for the differential preamplifier (and other subsystems) was returned to a mid-voltage point at the current source. This necessitated isolated power supplies for the subsystems. 3 Volts From-1101111111 111ce33111]_"< a A e c o s +CValtsDc DICAE DIME , - 3: " 5 W" ..., 23 LaglcPaeer '5 dual gate quadgate 21.102 Reset + MCISOOP scene at a - w . .1 iv v“ fiIHCMT'g F . w102x-3 Function Sequence Controller Figure C-S 3111.111 Probe m r"— Sample ,. - E Ma] y r|ml Heating Vacuum Element Mgi'tsurement 9”,.” ectran1cs 7 Copper Heat-Sink \\\\ LN Sample Holder a Temperature Control System Figure C-S. 100 An ultra-shielded power-line isolation transformer(16’2&’44'45) (United Transformer Company Model HIT-15, lSO-Watt) was used to supply the preamplifier power supply (Model 2Q15-100PC, Elasco Inc., Bloomfield, Connecticut), the six-volt relay-power supply, the digital panel meter, and the X-Y plotter. The secondary-winding box-shield was connected to the floating-ground system. Better operation was noted when the floating-ground system was allowed to remain isolated from earth ground. This probably reduced small unbalanced leakage currents to earth ground existing in the isolated current source and all shielding systems. Part 4p7- Sample-Temperature Control A simple technique was chosen to control temperature of samples. Early experiments indicated need to cover samples with dry gas to prevent condensation of atmospheric water vapor during low-temperature measurements. Accordingly, the four-point probe head and samples were enclosed by a bell jar which could be evacuated and refilled with dry nitrogen. Sample temperature was managed by remote control. Many (46,47,48) Three systems have been developed for temperature control. systems were considered for this application: thermoelectric modules, heat-sink conduction, and cold—gas-flow cooling of an electrically- heated sample-holding stage. The thermoelectric technique(49) was tried with limited success. Materials-Electronic-Products Corporation Model-CP7-l7-10 and Cambridge- Thermionic Corporation Model-801-3958—01 ceramic-insulated thermo- electric modules were used both separately and in cascade. A water- cooled heat-sink was used to establish the temperature of the bottom surface of the modules at about 10 °C. By changing the direction and magnitude of current flowing through the modules, the upper surface 101 could be maintained in the range from -50 °C. to +120 °C. with good stability. Unfortunately, early experiments indicated a need for substantially-lower sample temperatures. Attempts to lower temperatures further by the thermoelectric technique failed primarily due to a decrease in heat-pumping efficiency of the modules at low temperatures. It was noted that a minimum pressure in the bell jar of about 100 torr was required to ensure good thermal conductivity at the sample, module, and heat-sink interfaces. As seen in Figure C-9, the heat-sink—conduction technique was finally adopted. A capper rod was used as a stage to establish sample temperature; an electrically-heated element was attached. The bottom end of the rod could be immersed in liquid nitrogen. The high thermal- conductivity copper rod was insulated from the bell-jar baseplate by means of a stainless-steel sleeve brazed to the rod, and by means of an O-ring vacuum coupling. Samples were electrically insulated from the top of the rod by means of Thermalloy Company Type—B-1000-63 beryllium-oxide mounting plates. Wakefield—Engineering thermal compound was used to affix the platesto the rod in a removable fashion. Lumps of thermal compound were used to hold samples in place upon the mount- ing plates. Samples could be cooled to slightly lower than -150 °C. in this manner. Thermal losses due to rod-mounting technique, and assorted interfacial thermal drops seemed to prohibit lower-temper- ature operation. Another system was considered for control of temperature of a barrel-like sample stage by means of electrical-heating and flow of evaporating liquid nitrogen. Such a system could easily be directed by a simple on-off controller to establish the value and rate of change 102 of temperature to any reasonable value. The heat-sink conduction technique Operated quite satisfactorily, so effort required to implement the gas-flow-and—heater method was not expended. Electrical connection to the probe head was made through the bell- jar baseplate by means Of Amphonel UG-254A/U and UG-909/U type BNC bulkhead connectors and RG-59/U coaxial cable. Use of double-shielded and guarded coaxial interconnecting cables was considered; fortunately need for such measures was not indicated from tests Of the system. Part 5 —- System Operation and Data Analysis The previously described system for four-point probe conductivity measurements was Operated as follows. A large (compared to the probe spacing Of 25 mils) piece Of semiconducting material was fixed to a mounting plate and placed atop the copper rOd. Temperature was lowered by evaporation of liquid nitrogen at the lower end of the copper heat- sink rod. Suitable graph paper was put on the X-Y plotter. With sensitivities set properly, the heat-sink was allowed to warm while constant current was maintained through the sample (sometimes manually). Accelerated warm-up and maintenance Of elevated temperatures was accomplished with the electrical heater. A single run begun in this fashion required several hours for completion. Data was gathered in the form shown in Figure C-lO. Several features should be noted. The voltage drop, and hence electrical conductivity, appeared exponentially dependent on temperature. Data was gathered over about five orders Of magnitude of voltage by range- switching Y-axis plotter sensitivity. At higher temperatures, and higher conductivities, the voltage drop was small and subject to interference by thermoelectric effects and noise. These effects were 103 3:6 2313.5 32%... 0.10 953.... 3:223. .010 .2322? I. 32...) 22.32:...» V1.1 n... N... _+ O .11 Ni ml s . . 4 . . 4 .. a a a . J . 14 o 2 III I/ x. 1 . /” I coo . /. // .3 1 u , / .... da ’1’ / so. I. n a d / I; . woo .. W. lo . a; a. a. l V C s. . . A . . o If / aa a J n m ... . a . 1 a o a — 0 a . . L o . ’ o a .. . m o. — a — h m w. a. —— u m 2.2 mm u w .r . . N QE<1_.O "uguH . a a a l O . at cox—Ea: 2955 .. . . . . e , «2.05.8393 2.2.3064 .. . — _ 1 . g — . - . .6. 2”,, o. 2. 8.: 82: o. . w L . . 104 observed by automatic periodic reversal of sample-current direction as described earlier. Data interpretation required many difficult calculations so a computer program was developed to convert information tabulated from the curves to a more useful table of conductivity versus temperature. This Fortran computer program, illustrated in Table 0'1 , can be run on the General-Electric Mark II computer timesharing system, and produces output of the form shown in Table C-2 . Data was tabulated from curves by hand, but possibility exists for automatic digital-data acquisition in a form which can be fed directly to a computer. Logarithm of conductivity, inverse Kelvin temperature, and activation energy correSponding to eXponential temperature dependence were also determined and tabulated. IO 12 20 30 40 50 60 70 80 81 82 90 92 IOO I2 COND¢LINE)lXLN2¢XIl(PI#THIK*V(LINE))1 GO TO ID 110 15 PRINT:“ NUMBER OF DATA POINTS! ”pLINEI PRINTo” " I20 DO 20 18!:LINEI LNILINE+I-II TKICIIIlo/TKCLN) ISO 20 CONDLCI)IALOG(COND(LN)) 105 DIMENSION CONDCSO):CONDL(SO)aCONDLDCSO)oTK(SO)aTKI(SO) DIMENSION AtSO):V(50) FILENAME INaOUT PRINT3" INFILE:OUTFILE:TO"I INPUTpINaOUToTD XKII.38044E-233031.60206E-I9IPII30141593XLN28o693|47 LINEIO! LZEROIOI IERRGO) TKIOIlo/TDB NORDERI9 READ(IN:I) LL:S:XI.VOFF:TMIKI I FORMATCV) IO READthaloENDOIS) LL:TCMV:VV LINESLINE+II VCLINEI'VOLTCVVoVOFF) TK(LINE)I273. + 28.451TCMV - o6074¢TCMV$TCMV & - ol?S7*TCMV*TCMV#TCMV - .OI657*TCMVtTCMVtTCMV*TCMV & + .006961*TCMV*TCMVQTCMVITCMVtTCMV IFCIO.¢THIK.LT.S) GO TO 12 CONDCLINE)!.5*XI/(PI#S*V(LINE))3 GO TO ID _..I_. --—-. ‘_ m-s— 4‘_._A ‘I' nu... I70 LINELILINE-l} DO 30 JIZoLINEL g 180 30 CONDLD(J)8-((CONDL(J)-CONDL(J-l))/(TKI(J)-TKI(J-l)) 18] 182 183 & + (CONDL(J+I)-CONDL(J))/(TKI(JOII-TKI(J)) & + (CONDL(J+I)-CONDL(J-I))/(TKI(JOI)'TKI(J-I)) &)*XKI(0*3.) 185 CONDLD(I)ICONDLD(2)I CONDLDtLINE)ICONDLDCLINE-I) I90 DO 60 JIIaLINE 200 60 A(J)8EXP( CONDLCJ)+CONDLD(J)*TKI(J)*O/XK I 220 ETOI-FDRVULtTKIOoTKIaCONDL:LINE:NORDER)¢XKIO 230 AABTNTI(TKIOoLINEoTKIaAoNORDERaIERR) 240 PRINT:" CONDIAOEXPC -E*O/(KtT) ) O ”aTDo” KELVIN“ 24! PRINT:" E I ”aETO:” ELECTRON-VOLTS” 242 PRINT:" A l "0AA a" I/(OHM0CENTIHETERS)" 250 WRITECOUTa4) LZEROaIN 251 A FORMAT(IIa?Xo"COND"o6Xo"T"a3X:"LOG(COND)” 252 8a4X:"I/T":6X:”E"O6X9A8) 260 D9 40 K3139) KK'LINEvl-KI 6L30043429ASOCONDLCKK) 280 40 WRITECOUTaZ) KoCONDtK)pTK(K)aCLpTKI(KKIaCONDLDCKK) 290 2 FORMATCII:" ":EIZoSpF7olaFB-E:FlO-6JF703) 300 DO 50 K8102LINE3 KKILINE41-KJ 0L8004342945*CONDL(KK) 320 50 WRITECOUTca) K:COND¢K):TK(K):CL:TKI(KKIaCONDLDCKK) 330 3 FORMAT(I2:EI2.3:F7.IoFOoZoFIOoépFToO) 3A0 PRINTa” LIST FILE! "aOUTa”FOR MORE DATA OUTPUT” 350 PRINT." "3 999 STOP! END 400 FUNCTION VOLTCVV:VOFF)B VVBVV-VOFF 410 IF(VV.LT¢D.OOIZ)GO TO IOO! IF(VV.LT.I.2)GO TO 110 420 VOLTOCVV+oO3)/lo0031 RETURN 430 100 VOLT=VVlloDIJ RETURN 440 110 VOLTSVV/Io0073 RETURN 450 END Table C'1 Computer Program DATA 0CD~IOLMDJBND‘13 COND 001808-07 00214E-n7 002435-07 002698-07 003095-07 004105-07 00531E-07 00755E-07 001035-06 001648-06 00199E-06 00229E-06 002745-06 00324E-O6 00394E-06 005058‘06 00664E-06 0.934E-06 OolZOE-OS 001682-05 O019OE-05 002235-05 00260E-05 00304E-05 00376E-05 004845-05 006148-05 0.8388-05 00109E‘04 001325‘04 001575-04 001945-04 002215-04 002632-04 00323E-04 00420E-04 0056OE-04 00740E-04 00109E-O3 001578-03 001905-03 002208-03 002515-03 00301E-03 00378E.O3 004695‘03 006668-03 Table C-2 T 16004 16309 16507 16704 16901 17205 17508 18007 18505 19108 19409 19604 19905 20205 20506 21001 21406 22005 22500 23009 23308 23608 23907 24206 24700 25208 25806 26509 27106 27508 28001 28507 28804 29206 29800 30509 31307 32102 33007 33807 34503 34805 35106 35507 36008 36508 37307 106 LOG¢COND) .7014 '7067 -7061 ~7.57 -1.51 .7039 -7.27 ~7o12 -6099 -6078 -6070 96064 -6056 -6049 -6040 -6030 '6018 -6.03 “5092 ’5077 -5072 -5065 -5059 '5052 -5042 “5031 -5.2| ~5008 -4096 -4088 “4081 -4.11 “4066 '4058 ~4049 -4038 -4.25 -4013 -3096 -3.80 ‘3072 -3o66 -3060 -3052 ’3042 .3033 -3018 1/T 00006235 00006101 00006031 00005975 00005914 00005799 00005689 00005535 00005391 00005214 00005131 00005091 00005013 00004937 00004864 00004760 00004660 00004534 00004445 00004331 00004277 00004224 00004172 00004121 00004048 00003955 00003866 00003761 00003682 00003625 00003571 00003501 00003467 00003418 00003356 00003269 00003188 00003114 00003023 00002952 00002896 00002870 00002844 00002811 00002772 00002734 00002676 Output from DATA Program 8 00137 00137 00156 00169 00204 00207 00200 00193 00208 00215 00244 00242 00193 00211 00218 00220 00236 00238 00249 00226 00227 00259 00260 00259 00242 00232 00243 00271 00288 00278 00266 00296 00319 00295 00212 00282 00315 00349 00403 00368 00376 00460 00461 00490 00495 00508 00508 INAB APPENDIX D REVERSIBLE CONDUCTIVITY SWITCHING Part 1 -- Historical Introduction Knowledge of conductivity-switching effects has existed for more (1-3) than ten years. Early published reports of switching at point contacts to amorphous chalcogenides were available in 1962. Much of the early work was done on glasses developed primarily for encapsulation (4) of crystalline-semiconductor devices, diodes and transistors. Recognition of the utility of switching effects led to preliminary (5,6) investigations. Apparently, instability of early switching devices prevented immediate development of the phenomena; continued development and study was not done until recently by these early investigators.(7’8) An independent investigator, S. R. Ovshinsky* discovered useful switching phenomena in many materials, and in the early 1960's began -11 manufacture of relatively-stable devices utilizing the effect.(9 ) Publication of subsequent research into reversible switching for (12) generated considerable excitement in research, 1 -21 industrial, and business fields.( 3 ) Many excellent surveys of work disordered materials on these devices (referred to by the manufacturer as Ovonic switches) 22-26 have been published.( ) Part 2 -- The Switching Phenomena Reversible conductivity switching has been observed extensively (27-32) since early reports. The process occurs ”hen a piece, or film, of material exhibiting the effect is placed between two electrodes and * Energy Conversion Devices, Inc., Troy, Michigan, founded by‘S. R. Ovshinsky, has pioneered research on amorphous-semiconductor devices. 107 108 a suitably—high voltage is applied. Below some threshold electric- field intensity, typically about 10“ volts/cm, most materials exhibit low conductivity, less than about 10"5 mhos/cm. However, as the applied electric field exceeds the threshold value, the glassy material "switches" to a highly-conductive condition, remaining in this state until the electric field is reduced nearly to zero. Threshold switching can often be repeated indefinitely without degradation of switching material or electrodes. Switching occurs as a highly-conductive filament-like channel is grown from the positive electrode towards the negative electrode.(31) A short delay, about one microsecond, occurs before switching if the electric field is applied abruptly. After the short delay, switching takes place very rapidly, in less than 0.25 nanoseconds. Apparently the device remains in a low-conductance condition until a highly- conductive filamentary channel is completely formed between electrodes. Certain glasses remain highly conductive after switching, even if the applied electric field is removed, provided the device current was allowed to reach a highaenough value previously. This memory switching process seems to be associated with a modification or change (32) For some glasses, the of state for material forming the filament. low-conductance condition can be regained by passing a brief, high- current pulse through the device. Memory switching also appears to be indefinitely repeatable for some materials. While stable, repeatable switching at relatively-dow applied fields occurs primarily in amorphous chalcogenide glasses, switching (20,33) has been demonstrated for many materials. Threshold switching and memory switching have been reported for amorphous elemental 109 4-39) semiconductors,3 for transition-metal oxide semiconducting (11) glasses,(40-44) for many metal-oxide films, and for some polymers.(45’46) Practically any thin insulating film can be (11’20) a few times. switched The only commercially-available switching device employing thresh- old and memory-switching effects, the Ovonic threshold switch (Energy Conversion Devices, Inc.» is fabricated in thin-film form. Voltage is applied to electrodes on either side of a thin (about one nmeter in thickness) film of amorphous chalcogenide glass, basically a Ge-Si-As—Te composition. A highly-stable, radiation-resistant device results if the area of contact at the electrodes is restricted some- what to assure that the filament always is formed at the same location. Relatively-low switching voltages, about ten volts typically, are required. The threshold switching voltage is sensitive to tempera- (47,48) e9 tur about 1% per degree Celsius. Characteristics of similar devices and materials showing switching properties are generally available in literature cited in this chapter.(49-68) Part 3 -- Theory of Switching Theory for reversible conductivity switching in amorphous semi- conductors is not yet clear. Complete hypotheses for the mechanism of initiation and maintenance of threshold switching have not been reported, though excellent progress in the understanding of conduction processes for disordered materials is being made.* Theory satisfac- torily accounting for electric-field history and temperature dependence * A discussion of the theories related to electronic conduction in disordered materials, particularly amorphous chalcogenides, can be found in Appendix A as a literature review. 110 of switching delay has also not been developed. Electrothermal explanations for switching has proved satisfactory for some disordered materials, but this explanation seems not to account for particularly- abrupt switching observed for some other materials. Whatever the cause for initial formation of filamentary-conducting channels between electrodes, plausible eXplanations for memory behavior have been developed which assume phase transformations in the filament due to dissipated power. Some amorphous materials, notably boron, vanadium oxide, and several chalcogenides, show negative differential conductance<67) on application of suitably-high electric fields. This effect, possibly due to self heating, can lead to constriction of current into a (68’69) between electrodes. Filament formation: 70-73 observed for many materials,( ) is prominent in amorphous-semi- relatively—small filament conductor switching devices. Phase transformation or alteration of material properties within the filament is suggested. Perhaps negative differential admittance can account for apparent negative capacitance observed for amorphous chalcogenide glasses at high fields short of switching intensities.(74) Memory switching has been thoroughly examined. Determination of composition of conducting filaments has been done using differential 100 103 thermal analysis,(95'96’ ’ ) electrondmicroprobe analysis,(32) (97) Study of bulk amorphous semi- and x-ray diffraction analysis. conductors has generally indicated a tendency towards crystallization at temperatures substantially below the melting point. Tellurium crystallites have been identified in crystallized material,(97’99) as reported in Chapter 3 for GeloAsone7o. It is suspected that 111 filaments are melted and requenched to the glassy state upon applica- tion of a short high-current "turn—off" pulse. (75) Theoretical investigations have showed that electrothermal effects are a probable cause for switching in materials showing a tendency towards differential negative conductance. Examination of such hypotheses has led to reasonably satisfactory explanations(76’77) (78) (79,80) for time delays and switching, in a few cases. Switching voltage has been strongly correlated with thermal properties such as 101 102 glass-transformation temperature.( ’ ) Many excellent discussions of energy-controlled thermal-switching processes have been (Bl-84,104) published. Other suggested processes for initiation of switching in disordered materials include anomalous field emission,(85) (86) double injection,(87) and Mott transformation. zener-like tunneling, (88) Sophisticated quantumrtheoretically-based suggestions such as s new mobility-gap (89) may 90 91 lead to satisfactory explanations for initiation of switching.( ’ ) model for electronic conduction in amorphous semiconductors Excellent reviews of theory for switching have recently been (92,93) presented; many of these generally discount electrothermal initiation.(90’94) BIBLIOGRAPHY BIBLIOGRAPHY Part 1 —— References for Chapter 1 l. A. R. Hilton, New High Temperature InfraredTransmitting_class , Texas—Instruments, Inc., Dallas Texas, Report No. 08-65-121 September 30, 1965. 2. S. R. Ovshinsky, "Reversible Electrical Switching Phenomena in Disordered Structures", Phys Rev Letters, 21:20(1968) 1450-1453. 3. M. Hansen, Constitution of Binary Alloys. 2nd ed. New York: McGraw-Hill, 1958; R. P. Elliot, Constitution of Binary Alloys, First Supplement. New York: McGraw-Hill, 1965; F. A. Shunk, Constitution of Binary Alloys, Second Supplement. New York: McGrsw-Hill,l969. 4. K. W. Bagnall, The Chemistry of Selenium, Tellurium, and Polonium. New York: Elsevier Publishing Co., 1966. 5. H. Rawson, Inorganic Glass-Forming Systems. New York: Academic Press, 1967. 6. V. I. Davydov, Germanium. New York: Gordon & Breach Publishers, 1966, Ch IV-3. 7. C. S. Smith, "The Interconversion of Atomic weight and Volume Percentages in Binary and Ternary Systems", P. N. Rhines, ed., Phase Diagrams in Metallurgy, New York: McGrsw-Hill Book Co., 1956. 8. A. R. Hilton, C. E. Jones, M. Brau; "Non-Oxide IVA-VA-VIA Chalcogenide Glasses", Parts 1,2,3; Phys Chem Glasses, 7:4(1966)105-126. 9. A. R. Hilton, M. Brau; "New High Temperature Infrared Transmitting Glasses-III", Infrared Phys, 6(1966)l83-l94. 10. A. R. Hilton, "Optical Properties of Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)28-39. 11. R. C. Chittick, "Properties of Glow-Discharge Deposited Amorphous Germanium & Silicon", J Non-Cryst Solids, 3(1970)255-270. 112 12. 13. 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. K. A. D. 113 L. Chopra, 8. K. Bahl; "Amorphous vs. Crystalline GeTe Films", J Appl ths. Part I - 40;lO(1969)4l7l—4l78; Part II — 40:12 (1969)4940-4947; Part III - 41:5(1970)2196-2212. H. Clark, "A Review of Band Structure and Transport Mechanisms in Elemental Amorphous Semiconductors", J Non-Cryst Solids, 2(1970)52—65. . Adler et al, "Transport Properties of a Memory—Type Chalcogenide Glass", J Non-Cryst Solids, 4(1970)330—337. Bienenstock, F. Betts, S. R. Ovshinsky; "Structural Studies of Amorphous Semiconductors", J Non—Cryst Solids, 2(1970) 347-357. . Betts, A. Bienenstock, S. R. Ovshinsky; "Radial Distribution Studies of Amorphous Ge 4(1970)554-563. H xTe1_x Alloys , J Non Cryst Solids, . B. Dove et al, "Short—Range Order in Amorphous GeTe Films", Appl Phys Letters, l6:3(l970)l38-140. . Grigorovici, R. Manaila; "Short—Range Order in Amorphous Germanium", J Non-Cryst Solids, l(1969)37l-387. . L. Chopra, S. K. Bahl; "Structural, Electrical & Optical Properties of Amorphous Germanium Films", Phys Rev B, l:6(1970)2545-2556. . Takamori, R. Roy, G. J. McCarthy; "Structure of Memory-Switch- ing Glasses", Mat Res Bull, 5:7(1970)529-540. . E. Spicer, T. M. Donovan; "Electronic Structure of Amorphous Ge", Phys Rev Letters, 24:11(1970)595-598. . Senturia, C. R. Hewes, D. Adler; "NMR Study of a Memory-Type Amorphous Semiconductor", J Appl Phys, 41:1(1970)430-431. . V. Phillips, R. E. Booth, P. W. McMillan; "Structural Changes Related to Electrical Properties of Bulk Chalcogenide Glasses", J Non-Cryst Solids, 4(1970)510-517. Adler, M. H. Cohen, E. A. Fagen, J. C. Thompson; "Valence Electron Configuration of Te in Amorphous TeGe Alloys", J Non—Cryst Solids, 3(1970)402-406. . Shanefield, P. E. Lighty; "Instabilities in Semiconducting Glass Diodes", Appl Phys Letters, 16:5(1970)212-214. . H. Sie, "Electron MicrOprobe Analysis and Radiometric Microscopy of Electric Field Induced Filament Formation on the Surface of AsTeGe Glass", J Non-Cryst Solids, 4(1970)548-553. 27. H. 28. J. 29. H. 30. Mo 31. C. 32. E. 33. M. 34. R. 36. K0 37. K. 39. R. 40. A- 114 J. Stocker, "Phenomenology of Switching and Memory Effects in Semiconducting Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)371-381. W. Osmun, H. Fritzsche; "Electron Tunneling into Amorphous Germanium", Appl Phys Letters, l6:3(1970)87-89. J. Stocker, "Bulk and Thin Film Switching and Memory Effects in Semiconducting Chalcogenide Glasses", Appl Phys Letters, 15:2(1969)55-57. Kastner, H. Fritzsche; "Ambient Induced Changes of the Conductors of Amorphous Germanium", Mat Res Bull, 5:8(1970)631-640. Feldman, K. Moorjani; "Switching in Elemental Amorphous Semi- conductors", J Non-Cryst Solids, 2(1970)82-90. J. Evans, J. H. Helbers, S. R. Ovshinsky; "Reversible Conductivity Transformations in Chalcogenide Alloy Films", J Non-Cryst Solids, 2(1970)334-346. M. Cohen, M. Campi, J. D. Penar; "Electronic Switching Phenomena in Hot-Pressed Ge", J Non-Cryst Solids, 2(1970) 91-980 Uttecht et al, "Electric Field-Induced Filament Formation in As-Te-Ge Glass", J Non-Cryst Solids, 2(1970)3S8-370. . A. Walley, "Electrical Conduction in Amorphous Silicon and Germanium", Thin Solid Films, 2:4(1968)327-336. Weiser, M. H. Brodsky, G. D. Pettit; "Electrical and Optical PrOperties of Amorphous AszTe3 Films", J Non-Cryst Solids, 4(1970)43-44. Weiser, M. H. Brodsky; "DC Conductivity, Optical Absorption, and Photoconductivity of Amorphous Arsenic Telluride Films", Phys Rev B, 1:2(1970)791-799. . Hamacuchi, Y Sasaki, J. Nakai; "Electrical Conduction and Switching in Amorphous Semiconductors", Jap J Appl Phys, 9:10(1970)1195-1203. C. Guthmann, C. Hermann, J. M. Thurllier; "Recombination Phenomena in Tellurium", Phys Stat Sol (a), 3(1970)365-374. Goswami, R. H. Jog; "Study on Semiconducting Properties of Vacuum Deposited Tellurium Films", Indian J Pure Appl Phys, 6(1968)416-4l9. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. . Tamaka et al, "Electrical Nature of the Lock-0n Filament in . Iizima et a1, "Electrical and Thermal Properties of Semi- 115 S. Blakemore, "Recombination Processes in Tellurium", Proceedings of the Internationg1_Qggggggngg_gn_§gn1ggndngtgr Physics Prague 1269, New York: Academic Press, 1961, 981-983. . K. Bahl, K. L. Chopra; "Electrical and Optical Properties of Amorphous v.s. Crystalline GeTe Films", J Vac Sci & Technol, 6:4(1969)561—565. Albers, "Determination of Activation Energies in Polycrystalline Tellurium Films from Conductivity and Hall Measurements", Phys Stat Sol (a), 1(1970)K105-K107. Amorphous Semiconductors", Solid State Commun, 8(1970)75-78. W‘- m... l‘ 1._-- um-4i. n“? . conducting Glasses As-Te-Ge", Solid State Commun, 8(1970) 153-155. . Tanaka et a1, "Thermal Effect on Switching Phenomena in Chalcogenide Amorphous Semiconductors", Solid State Commun, 8(1970)387-389. Sugi et a1, "Evidence for Thermal Pinching Effect in Chalcogen- ide Amorphous Semiconductors", Solid State Commun, 8(1970)‘ 829-832. Tanaka et a1, "Electrical and Calorimetrical Analyses of Memory Phenomenon in Amorphous Semiconductors", Solid State Commun, 8(1970)1333-1336. Iizima et al, "Effect of Stabilization on Electrical Conduc- tivity in Chalcogenide Glass", Solid State Commun, 8(1970)1621-1623. H. Sie, "Memory Cell Using Bistable Resistivity in Amorphous As-Te-Ge Film", dissertation # 69-20 670, University Microfilms Inc., Ann Arbor, Michigan, 1969. Tsu, W. E. Howard, L. Esaki; "Photoconductivity and Optical Properties of Amorphous GeTe", J Non-Cryst Solids, 4(1970)323-327. E. Spicer, T. M. Donovan; "Photoemission and Optical Studies of Amorphous Germanium", J Non-Cryst Solids, 2(1970)66-80. J. Stiles et a1, "High-Field Photoconductivity of Amorphous GeTe and GeSe Films", Appl Phys Letters, 16:10(1970) 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 116 Tauc et al, "Infrared Absorption in Amorphous Germanium", J NonvCryst Solids, 4(1970)279e288. E. Howard, R. Tsu; "Photoconductivity and Density of States for Amorphous GeTe", Phys Rev B, 1:12(l970)4709-47l9. R. Hilton, "Non—Oxide Chalcogenide Glasses as Infrared Optical Materials", Appl Opt, 5:12(l966)1877-l882. Beaglehole, M Zavetova; "Fundamental Absorption of Amorphous Ge, Si and GeSi Alloys", J Non-Cryst Solids, 4(1970) 272-278. Feinleib, S. R. Ovshinsky; "Reflectivity Studies of the Te (Ge, As) Based Amorphous Semiconductor in the Conducting and Insulating States", J Non-Cryst Solids, 4(1970)564-572. Maschke, P. Thomas;"Ca1culation of ez-Spectra for Amorphous Se, Ge, and Si . . .", Phys Stat Sol, 41:2(1970)743-450. E. Lewis, "The Principal Energy Gap in GeTe: A Possible EXplanation for the Experimental Discrepancy Between E8(optical) and E8(thermal)", Phys Stat Sol, 39(1970)K5-K7. Herman, J. P. Van Dyke; "New Interpretation of the Electronic Structure and Optical Spectrum of Amorphous Germanium", Phys Rev Letters, 21:23(l968)1575-1578. B. Murphy, "Thermal Analysis", Anal Chem, 42:5(1970)268R-276R. P. Hunter, Handbook of Semiconductor Electronics, 3rd ed. New York: McGraw—Hill Book Co., 1970, Sections 2 and 20. . E. Spear, "Drift Mobility Techniques for the Study of Electrical Transport Properties in Insulating Solids", J Non-Cryst Solids, 1(1969)197-214. ASTM, "Standard Method of Test for Conductivity Type of Extrinsic R. A. P. A. Materials" F—42-69, Annual Book of ASTM Standards, 8(1970) Roy, "Classification of Non—Crystalline Solids", J Non-Cryst Solids, 3(1970)33-40. R. Hilton, "Chemical Composition and Glass Formation in Chalcogenide Systems", Phys Chem Glasses, 9:5(1968)148-152. F. Weller, "An Introduction to Principles of the Solid State", J Chem Educ, 47:7(1970)501-507. R. Hilton, "Chemical Composition and Glass Formation in Chalcogenide Systems", Phys Chem Glasses, 9:5(1968)148—152. .‘L'Y‘: ”up". D __L '41: .. 117 Part 2 -- References for Chapter 2 1, A. I. Gubanov, Quantum Electron Theory of AmorphOus Conductors. New York: Consultants Bureau, 1965, 24-25 & 213-214. 2, B. D. Wedlock, F. M Norton; "Purification of Tellurium by Distillation", UltrapurificatiOn of SemiConductor Materials, Proceedings of the April 1961 Conference in Boston. New York: MacMillan Co., 1961. Pt, 3, L. D. Pye, "Semiconductivity in Some Chalcogenide Glasses", ?= dissertation #68—16619, University Microfilms, Inc., E Ann Arbor, Michigan, 1968. f 4, W. C. Lacoruse, V. A. Twaddell, "Effects of Impurities on the 1 Electrical Conductivity of Glassy Selenium", J Non-Cryst g, Solids , 3(1970)234-236. g 5, N. I. Sax, Dangerous Properties of Industrial Materials. 2nd ed. New York: Reinhold Publishing Corp., 1963. 6. R. E. Honig, D. A. Kramer, "Vapor Pressure Data for the Solid and Liquid Elements", RCA Review, 30:2(1969)285-305. 7, C. J. Smithells, Metals Reference Book. Vol II, 3rd. ed. Washington: Butterworth & Co., 1962. 3, R. C. Weast, ed., Handbook of Chemistry and Physics. 46th ed. Cleveland: The Chemical Rubber Co., 1965. 9, C. H. Sie, "Memory Cell Using Bistable Resistivity in Amorphous As-Te-Ge Film", dissertation #69-20670, University Microfilms Inc., Ann Arbor, Michigan, 1969. 10, A. R. Hilton et al, "New High Temperature Infrared Transmitting Glasses - III", Infrared Phys, 6(1966)183-194. 11, M. Hansen, Constitution of Binary Alloys. 2nd ed. New York: McGraw—Hill, 1958; R. P. Elliot, Constitution of Binary Alloys, First Supplement. New York: McGraw-Hill, 1965; F. A. Shunk, Constitution of Binary Alloys, Second Supplement. New York: McGraw-Hill, 1969. 12, E. D. Dietz, "The Glassy State", Sci & Technol, 82(Nov 1968) 13, R. J. Charles, "The Nature of Glass", General Electric Corp., Report No. 67—C-3l7, Aug 1967. Also: Sci Am, 217:3 (Sept 1967)127-156. 14, R. W. Douglas, "The Chemical Approach to Problems of the Glassy State", V.D. Frechette, ed., Non-Crystalline Solids. New York: John Wiley & Sons, 1960, 374—411. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 118 R. Secrist, J. D. Mackenzie; Modern Aspects of the Vitreous State. Vol 3. Washington: Butterworth & Co., 1964. Ch 6. . T. Sarjeant, R. Roy; "New Glasses and Polymorphic Oxide Phases Using Rapid Quenching Techniques", J Am Ceram Soc, 50:10(1967)500-503. Iizima et al, "Electrical and Thermal Properties of Semi- ducting Glasses As-Te-Ge", Solid State Commun, 8(1970) 153-155. Iizima et al, "Effect of Stabilization on Electrical Conduc- { tivity in Chalcogenide Glass", Solid State Commun, ! 8(1970)1621-1623. § A. Fagen, H. Fritzche, "Electrical Conductivity of Amorphous 1 Chalcogenide Alloy Films", J Non-Cryst Solids, 2(1970) 3 170-179. g . A. Dillon, "Purification Methods in the Study of Semiconductor Surfaces", M. S. Brooks, J. K. Kennedy, ed's., Ultrapurification of Semiconductor Materials, Proceedings of the April 1961 Conference in Boston. New York: MacMillian Co., 1961. D. Pearson et al, "Chemical, Physical and Electrical Properties of Some Unusual Inorganic Glasses", Advances in Glass Technology. New York: Plenum Press, 1962, 357-365. . Takamori, R. Roy, G. J. McCarthy, "Structure of Memory-Switching Glasses, I - Crystallization Temperature and Its Control in Ge-te Glasses", Mat Res Bull, 5:7(1970)529-540. D. Mackenzie, "Preparation and Properties of Noncrystalline Films", J Vac Sci Technol, 6:2(1969)658-662. C. Schottmiller, D. L. Bowman, C. WOod; "New Vitreous Semi- conductors", J Appl Phys, 39:3(1968)1663-1669. . Freller, K. G. Gunther, "Das Aufdampfen von Verbindungshalblei- tern im Hochvakuum", Thin Solid Films, 3(1969)417-438. C. Chittick, "Properties of Glow-Discharge Deposited Amorphous Germanium and Silicon", J Non-Cryst Solids, 3(1970)255-270. R. Olson, "Evaporation of Metals from a Covered Boat Source", R. D. Mathis Co., Long Beach, California, Oct 1963. Tauc et al, "Infrared Absorption in Amorphous Germanium", J Non—Cryst Solids, 4(1970)279-288. Goswami, R. H. Jog, "Study on Semiconducting Properties of Vacuum Deposited Tellurium Films", Indian J Pure Appl Phys, 6(1968)416-419. 119 30. R. Roy, "Classification of Non-Crystalline Solids", J Non-Cryst Solids, 3(1970)33-40. 31. ASTM, "Standard Method for Analyzing Stress in Glass Designation", F 218-68, Annual BOok of ASTM Standards, 8(1970)968-974. 32. B. Schwartz, N. Schwartz; Measurement TechniQues'for Thin Films. New York: The Electrochemical Society, May 1967. 33. J. D. MacKenzie, "Electronic Conduction in Non—Crystalline Solids", J Non—Cryst Solids, 2(1970)16—26. Also available #AD 69 0202, May 1962, Clearinghouse for Federal Scientific 5 Technical Information. 34. E. A. Fagen, H. Fritzsche, "Electrical Conductivity of Amorphous Chalcogenide Films", J Non-Cryst Solids, 2(1970)170-179. 35. C. Wood, T. H. Jones, "A Simple Apparatus for the Preparation of Thin Films by Co-evaporation", Rev Sci Instr, 41:6(1970) 874-875. * Reprints of certain technical reports are available from the Clearinghouse for Federal Scientific and Technical Information, Springfield, Virginia, 22151, by ordering the appropriate catalog number and including $3.00 per report. Some governmental department reports are also available directly from the Superintendent of Documents, Government Printing Office, Washington, D. C., 20402. _ “-m—‘m .‘—.—‘ u : :u A - t « 120 Part 3 -- References for Chapter 3 l. 10. 11. 12. 13. 14. 15. D. E. A. R. S. D. Gray, ed. American Institute of Physics Handbook. 2nd ed. New York: McGraw—Hill Book Co., 1963, Ch 4-j. Hiton et al, "NonOxide IVA-VA-VIA Chalcogenide Glasses: Part 3 - Structural Studies", Phys Chem Glasses, 7:4(1970) 116-126. Senturia, C. R. Hewes, D. Adler; "NMR Study of a Memory Type Amorphous Semiconductor", J Appl Phys, 41:1(1970)430-431. D. Adler, M. H. Cohen, E. A. Fagen, J. C. Thompson; "Valence Electron Configuration of Te in Amorphous TeGe Alloys", J Non-Cryst Solids, 3(1970)402-406. T. Takamori, R. Roy, G. J. McCarthy; "Structure of Memory Switching A. R. Glasses: I - Crystalization Temperature and Its Control in Ge-Te Glasses", Mat Res Bull, 5:7(1970)529-S40. Phillips, R. E. Booth, P. W. McMillan; "Structural Changes Related to Electrical Properties of Bulk Chalcogenide Glasses", J Non-Cryst Solids, 4(1970)510-517. Sie, "Electron Microprobe Analysis and Radiometric Micro- scopy of Electric Field Induced Filament Formation on the Surface of As-Te-Ge Glass", J Non-Cryst Solids, 4(1970) Westman, ed., Reference Data for Radio Engineers. 5th ed. New York: H. W. Sams & Co., 1968, section 4. Hilton, New High Temperature Infrared Transmitting:Glasses. Texas-Instruments, Inc., Dallas, Texas, Report No. 08-65-121, September 30, 1965. Hilton et a1, "Non-Oxide IVA-VA-VIA Chalcogenide Glasses: Part 1 - Glass-Forming Regions and Variations in Physical Properties", Phys Chem Glasses, 7:4(1966)105-112. Hilton, "New High Temperature Infrared Transmitting Glasses- III", Infrared Phys, 6(1966)183-194. R. Roy, V. Caslavska; "Di-Phasic Structure of Switching and Memory C. B. W. W. W. J. Devices 'Glasses'", Solid State Commun, 7(1969)l467-1473. Murphy, "Thermal Analysis", Anal Chem, 42:5(1970)268R—276R. Wendlandt, Thermal Methods of Analysis. New York: John Wiley & Sons, Interscience, 1964. Smothers, Y. Chiang, Handbook of Differential Thermal Analysis. New York: Chemical Publishing Co., 1966. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 121 P. D. Garn, Thermoanalytical Methods of Investigation. New York: Academic Press, 1965. M. Kastner, H. Fritzsche; "Ambient Induced Changes of the Conductance of Amorphous Germanium", Mat Res Bull, 5:8(1970)631-640. H. Fritzsche, S.R. Ovshinsky; "Calorimetric and Dilatometric Studies of Chalcogenide Alloy Glasses", J Non-Cryst Solids, 2(1970)148-154. S. A. Dembovsky, "The Elastic Constants, Softening Temperature, and Structure of Chalcogenide Glasses", Phys Chem Glasses, 10:2(1969)73-74. B. Eckstein, "The Inner Parameters of Glassy Systems", J Non-Cryst Solids, 4(1970)366-369. B. G. Bagley, H. E. Bair; "Thermally Induced Transformation in Glassy Chalcogenides", J Non-Cryst Solids, 2(1970)155-160. N. Onodera, H. Suga, S. Seki; "Glass Transition in Amorphous Preceipitates", J Non-Cryst Solids, l(1969)331-334. DuPont, "Detection of Glass Transitions by DTA and TMA", App Brief #14, Jan 1968, E. I. DuPont de Nemours 8 Co., Instrument Products Division, Wilmington, Delaware. 8. Iizima et al, "Effect of Stabilization on Electrical Conduc- tivity in Chalcogenide Glass", Solid State Commun, 8(1970) 1621-1623. DuPont, "Specific Heat", App Brief #11, Jan 1968, E. I. DuPont de Nemours & Co., Instrument Products Division, Wilmington, Delaware. DuPont, DuPont 900 Differential Thermal Analyzer, Instruction Manual. E. I. DuPont de Nemours & Co., Instrument Products Division, Wilimington, Delaware, Section 15. M. Hattori et a1, "Heat Capacities of As—S Glasses", J Non-Cryst Solids, 3(1970)195-204. S. Iizima et a1, "Electrical and Thermal Properties of Semi- conducting Glasses As-Te-Ge", Solid State Commun, 8(1970) 153-155. K. Tanaka et al, "Electrical and Calorimetrical Analyses of Memory Phenomenon in Amorphous Semiconductors", Solid State Commun, 8(1970)1333«1336. S. V. Phillips, R. E. Booth, P. W. McMillan; "Structural Changes Related to Electrical Properties of Bulk Chalcogenide Glasses", J Non—Cryst Solids, 4(1970)SlO-Sl7. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 122 G. Bagley, W. R. Northover; "Electron Microscopic Observation of Thermally Induced Transformations in Amorphous Chalcogenide Thin Films", J Non-Cryst Solids, 2(1970)161-169. . E. Hall, "The Specific Heat of Amorphous Chalcogenide Semi- conducting Material", Phys Rev, 178:3(1969) Guinier, D. L. Dexter;"EgggypStudiesvgf;Materials. New York: John Wiley 5 Sons, Interscience, 1963. . P. Klug, L. E. Alexander; XéRay_DiffraCtion Progedures. New York: John Wiley & Sons, 1954. D. Cullity, Elements of X-Ray Diffraction. Reading, Mass.: Addison—Wesley Publishing Co., 1959. S. Peiser, H. P. Rooksby, A. Wilson; X—Ray Diffraction by Polycrystalline Materials. New York: Reinhold Publishing Co., 1960. . H. Brodsky et a1, "Structural, Optical, & Electrical Properties of Amorphous Silicon Films", Phys Rev B, 1:6(1970)2632-264l. Uttecht et a1, "Electric Field-Induced Filament Formation In As-Te-Ge Glass", J Non-Cryst Solids, 2(1970)358-370. B. Dove et al, "Short-Range Order In Amorphous GeTe Films", Appl Phys Letters, 16:3(1970)138-140. . Bienenstock, F. Betts, S. R. Ovshinsky; "Structural Studies of Amorphous Semiconductors", J Non-Cryst Solids, 2(1970) 347-357. Betts, A. Bienenstock, S. R. Ovshinsky; "Radial Distribution Studies of Amorphous GexTe1_xAlloys", J Non-Cryst Solids, 4(1970)554-563. R. Shappirio, D. W. Eckart, C. F. Cook; "Chemical and Structural Characterization of Amorphous Semiconducting Materials in the System Ga-As-Te-Ge", J Non-Cryst Solids, 2(1970)217-228. Grigorovici, "Short-Range Order in Amorphous Semiconductors", J Non-Cryst Solids, l:14(l969)303-325. F. Ioffe, A. R. Regel; "Non-Crystalline, Amorphous and Liquid Electronic Semiconductors", A. F. Gibson, ed., Prpgress in Semiconductors. Vol 4. London: Heywood & Co., 1960. H. Bauer, "Electron Diffraction Techniques for the Study of Amorphous Systems", V. D. Frechette, ed.,‘Nonacrystalline Solids. New York: John Wiley & Sons, 1960. Roy, "Classification of Non-Crystalline Solids", J Non-Cryst Solids, 3(1970)33-40. 47. 48. 49. 50. 51. 123 K. Jonscher, P. A. Walley; "Electrical Conduction in Non- Metallic Amorphous Films", J Vac Sci Technol, 6:4(1969) 662-669. . Cervinka, R. Hosemann, W. Vogel; "Relations Between the Structure of Crystalline and Amorphous CdGexAsz", J Non- Cryst Solids, 3(1970)294-304. Debye, "Scattering of Radiation by Non-Crystalline Media", V. D. Frechett, ed.,'Nonécrystalline‘Solids. John Wiley & Sons, 1960. Wetanabe, C. Munakata; "Measurement of Resistance by Mbans New York: Electron BeamsI", Jap J Appl Phys, 4:4(1965)250-258. C. Weast, ed., Handbook of Chemistry and Physics. 46th ed. Cleveland: The Chemical Rubber Co. 1965. of 124 Part 4 -- References for Chapter 4 1. 10. 11. 12. 13. T. Peters, "Physical Characterization of Electronic Materials, Devices and Thin Films", Available #AD 66 7547 or N68 23849, 1968, and #AD 70 3275 or N70 32505, 1970, Clearinghouse for Federal Scientific and Technical Information. M. Bullis, "Measurement Methods for the Semiconductor Device Industry -- A Summary of NBS Activity", NBS Technical Note 511, Dec 1969, available U. S. Government Printing Office. M. Bullis, "Methods of Measurement for Semiconductor Materials, Process Control, and Devices", National Bureau of Standards Technical Notes #: 472,475,488,495,520,527,555, etc. Washington D.C.: Superintendent of Documents, Government Printing Office. . L. Horovitz, V. A. Johnson, ed's., Solid State Physics. Vol 6. New York: Academic Press, 1959. Ch 7. . P. Hunter, ed., Handbook of Semiconductor Electronics. 3rd ed. New York: McGraw-Hill Book Co., 1960. Section 20. ASTM, Annual Book of ASTM Standards. 33 vols. Philadelphia: . Cimpl, F. Kosek, M. Matyas; "Magnetic Properties of A328 American Society for Testing and Materials. 1970. F. Ioffe, Physics of Semiconductors. New York: Academic Press, 1960. Ch 6. J. Welles, "The Electronic Properties Information Center (EPIC)" Available #AD 70 8737, May 1970, Clearinghouse for Federal Scientific and Technical Information. Neuberger, "II-VI Semiconducting Compounds Data Tables", #AD 69 8341, October 1969; "IV-VI Semiconducting Compounds Data Tables", #AD 69 9260, October 1969; "Bibliography of III-V Semiconducting Films", #AD 70 1074, December 1969; Clearinghouse for Federal Scientific and Technical Information. . J. Walsh, P. Calella; "Pressure Effects on Thin Films of Amorphous Chalcogenide Semiconductors", Phys Rev, 178:3(1969) Fuhs, J. Stuke; "Elastoresistivity of Amorphous Semiconductors", Mat Res Bull, 5:8(1970)611-620. Devenyi, A. Belu, G. Korony; "Piezoresistance in Amorphous and Polycrystalline Germanium", J Non-Cryst Solids, 4(1970) 380-390. X Glasses", Phys Stat Sol, 41:2(1970)535-538. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 125 H. Mell, J. Stuke; "Negative Magnetoresistance of Amorphous Semi- conductors", J Non-Cryst Solids, 4(1970)304-310. L. Reimer, J. Seuss; "Change of Conductivity in Amorphous Selenium Films by Electron Bombardment (10-100 KeV)", Phys Stat Sol (a), 2(1970)809-816. J. W. Osmun, H. Fritzsche; "Electron Tunneling into Amorphous Germanium", Appl Phys Letters, 16:3(1970)87-89. I. Giaeva, H. R. Zeller; "Tunneling into and Through Evaporated Semiconducting Films", J Vac Sci Technol, 6:4(1969)502-508. C. Guthmann, C. Hermann, J. M. Thullier; "Recombination Phenomena J. S. ASTM, ASTM, in Tellurium", Phys Stat Sol (a), 3(1970)365-374. Blakemore, "Recombination Processes in Tellurium", Proceedings, Int. Conf. on Semiconducting Physics, Prague, 1960. New York: Academic Press, 1960, 981-983. "Measuring the Minority-Carrier Lifetime in Bulk Germanium and Silicon", F28-66, Annual Book of ASTM Standards, 8(1970)515-521. Stretchberry, "A Summary of the Measurement and Interpretation of the Hall Coefficient and Resistivity of Semiconductors", NASA TM-X—1711, Dec 1968, available Clearinghouse for Federal Scientific and Technical Information. "Standard Methods of Test for Resistivity of Semiconductor Materials", F43-70, Annual Book of ASTM Standards, 8(1970) 557-564. . Lisher, "Variational Methods for Experimental Investigations of the Thermal and Electrical Properties of Semiconducting Materials", (translated from Russian), available #AD 70 10759, Clearinghouse for Federal Scientific and Technical Information. Wagstaff, R. J. Charles; "Device for Electrical Conductivity Measurements of Glasses During Heating and Cooling", Rev Sci Instr, 40:5(1969)709. E. King, "A Compact Design for Measuring Electrical Conductivity at High Pressures and Varying Temperatures", J Sci Instr(J Phys E) 2(1969)59-62. T. M. Dauphinee, E. Mooser; "Apparatus for Measuring Resistivity and Hall Coefficient of Semiconductors", Rev Sci Instr, 26:7(1955)660—664. J. Hennephof, G. Hoekstra, J. Van Overbeeke; "Thermostat for Resistivity Measurements from 4 °K to 350 °K", J Sci Instr (J Phys E), 2:1(1969)96-97. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 126 T. Elliot, D. J. Wilson; "Apparatus for Automatic Electrical Measurements on Semiconductors from Liquid Helium Temperature to 400 °K", J Sci Instr(J Phys E), 2(1969) 956-958. C. Pack, G. G. Libowitz; "Low Temperatue Electrical Measurements on Reactive Samples", Rev Sci Instr, 40:3(1969)420—422. . R. Drabble, T. D. Whyte;"The Direct Current Measurement of Very High Resistance at Low Temperatures", J Sci Instr(J Phys E) 3(1970)515-517. Fischer, D. Grieg, E. Mooser; "Apparatus for the Measurement of Galvonomagnetic Effects in High Resistance Semiconductors", Rev Sci Instr, 32:7(1961)842-846. . Watanabe, C. Munakata; "Measurement of Resistance by Means of Electron Beam", Jap J Appl Phys, Part 1 - 4:4(1965)250-258; Part 2 - 5:12(1966)157-1l60; Part 3 - 8:11(1969)1307-1309. . N. Chester, B. B. Kosicki; "An Electron Beam Method for Measuring High Sheet Resistances of Thin Films", Rev Sci Instr, 41:12(l970)1817-1824. . R. Ovshinsky, "Reversible Electrical Switching Phenomena in Disordered Structures", Phys Rev Letters, 21:2(1968) 1450-1453. Iizima et a1, "Electrical and Thermal Properties of Semiconducting Glasses As-Te-Ge", Solid State Commun, 8(1970)153-155. Tanaka et al, "Thermal Effect on Switching Phenomenon in Chalcogenide Amorphous Semiconductors", Solid State Commun, 8(1970)387-389. . L. Chopra, S. K. Bahl; "Structural, Electrical, and Optical Properties of Amorphous Germanium Films", Phys Rev B, 1:6(1970)2545-2556. . A. Walley; "Electrical Conduction in Amorphous Silicon and Germanium", Thin Solid Films, 2:4(1968)327-336. . Muller, M. Muller; "Current-Voltage Characteristics at High Fields in Amorphous Selenium Thin Films", J Non-Cryst Solids, 4(1970)504-509. . Croitoru, L. Vescan, C. POpescu, M. Lazarescu; "Non-Ohmic Properties of Some Amorphous Semiconductors", J Non—Cryst Solids, 4(1970)493-503. E. Hall, "Transient Electrical Measurements of Amorphous Chalcogenide Thin Films", J Non-Cryst Solids, 2(1970) 125—132. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 127 Boer, R. Haislip; "Semiconductivity of Classes", Phys Rev Letters, 24:5(1970)230-233. . C. Crassie, D. B. Green; "Current-Voltage Characteristics of Disordered Films", Solid State Commun, 8(1970)247-250. Kerr, "Current Voltage Characteristics of Chalcogenide Glass Films", J Non-Cryst Solids, 2(1970)203-204. Walsh et a1, "EXperimental Results in Amorphous Semiconductor Switching Behavior", J Non—Cryst Solids, 2(1970)107-124. Thomas, "Preswitching Behavior of Amorphous Chalcogenide Semiconductor Films", Electron Lett, 6:3(5 Feb 1970)62-64. Walsh, R. Vogel, E. J. Evans; "Conduction and Electrical Switching in Amorphous Chalcogenide Semiconductor Films", Phys Rev, 178:3(1969)1274-1279. Also available #AD 68 3465, Nov 1968, Clearinghouse for Federal Scientific and Technical Information. Fagen, H. Fritzsche; "Electrical Conductivity of Amorphous Chalcogenide Alloy Films", J Non—Cryst Solids, 2(1970) 170-179. Davis, R. F. Shaw; "Characteristic Phenomena in Amorphous Semiconductors", J Non-Cryst Solids, 2(1970)406-431. Owen, "Semiconducting Glasses", Contemp Phys, 11:3(1970) 227-286. Stevels, Progress in the Theory of the Physical Properties of Class. New York: Elsevier Publishing Co., Inc., 1948, Ch 111. Pearson, Modern Aspects of the Vitreous State. Vol 3. J. D. MacKenzie, ed. London: Butterworth Scientific Publications. 1964. Fritzsche, S. R. Ovshinsky; "Conduction and Switching Phenomena T. in Covalent Alloy Semiconductors", J Non-Cryst Solids, 4(1970)464-479. Edmond, "Measurements of Electrical Conductivity and Optical Absorption in Chalcogenide Glasses", J Non-Cryst Solids, 1(1968)39-48. Hamaguchi, Y Sasaki, J. Nakai; "Electrical Conduction and E. Switching in Amorphous Semiconductors", Jap J Appl Phys, 9:10(1970)1195-1203. Information Systems, Mark II Numerical Analysis Routines. Publication #807231A (7M). General Electric Co., Information Service Department, Bethesda, Maryland. May 1969. 53. 57.4 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 128 Iizima et al, "Effect of Stabilization on Electrical Conductivity in Chalcogenide Glass", Solid State Commun, 8(1970)1621—1623. . Sugi et al, "Evidence for Thermal Pinching Effects in Chalcogenide Amorphous Semiconductors", Solid State Commun, 8(1970)829-832. . V. Phillips, R. E. Booth, P. W. McMillan; "Structural Changes Related to Electrical Properties of Bulk Chalcogenide Glasses", J Non-Cryst Solids, 4(1970)510-517. K. Bahl, K. L. Chopra; "Amorphous versus Crystalline GeTe Films: III - Electrical Properties and Band Structure", J Appl Phys, 41:5(1970)2196-2212. . K. Bahl, K. L. Chopra; "Electrical and Optical Properties of Amorphous versus Crystalline GeTe Films", J Vac Sci Technol, 6:4(1969)561-563. . Tsu, W. E. Howard, L. Esaki;"Photoconductivity and Optical Properties of Amorphous GeTe", J Non-Cryst Solids, 4(1970) 322-327. . Weiser, M. H. Brodsky, G. D. Pettit; "Electrical and Optical Properties of Amorphous AsZTe3 Films", J Non-Cryst Solids, 4(1970)43-44. Weiser, M. H. Brodsky; "Dc. Conductivity, Optical Absorption, and Photoconductivity of Amorphous Arsenic-Telluride Films", Phys Rev B, 1:2(1970)791-799. ASTM, "Standard Method of Test for Thermoelectric Power of Electrical-Resistance Alloys", B77-33; and "Conductiviy Type of Extrinsic Semiconducting Materials" F42-69, Annual Book of ASTM Standards, 8(1970). . W. Gray, Americag Igstigutg pf Pbyaiga Handbook, 2nd. ed. New York: McGraw—Hill Book Co., 1963. Section 4a. K. Henisch, M. Francois; "On the Thermo-electric Properties of Selenium", Semi-Conducting Materials. London: Butterworth Scientific Publications. 1951. E. Owen, J. M. Robertson; "Electronic Properties of Some Simple Chalcogenide Glasses", J Non—Cryst Solids, 2(1970)40-51. J. Curry, S. R. Pollack; "The Thermoelectric Properties of Quarternary Chalcogenides", Radiation Effects, 4(1970)115- 120. N. Krestovnikov et a1, "Thermoelectric Effect at the Boundary of Solid and Liquid Phases of Semiconductors from Various Groups", Soviet Physics-Doklady, 13:2(1968)l65-167. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 129 N. Gordiakova, G. V. Kokosh, S. S. Sinani; "Investigation of Thermoelectric Properties of Solid Solutions BizTe3 - BiZSeBN’ Soviet Physics-Technical Physics, 3:1(1958)1-14. . I. Gubanov, Qpantum Electron Theory of Amorphous Conductors. New York: Consultants Bureau, 1965. . S. Davis, Nomography and Empirical Equations. 2nd ed. New York: Reinhold Publishing Co., 1962. J. Wang, Solid State Electronics. New York: McGraw-Hill Book Co., E} 1966. Ch 4,7. 53 C. Feldman, K. Moorjani; "Switching in Elemental Amorphous Semi- ' conductors", J Non-Cryst Solids, 2(1970)82-90. . M. Cohen, M. Campi, J. D. Penar; "Electronic Switching i Phenomena in Hot-Pressed Ge", J Non-Cryst Solids, 2(1970) E 91-980 . Chapman, "Bistable Electrical - Switching Behaviour in Glasses and Other Media", Electron Lett, 5:11(29 May l969)246-247. H. Sie, "Electron Microprobe Analysis and Radiometric Micro- scOpy of Electric Field Induced Filament Formation on the Surface of As-Te-Ge Glass", J Non-Cryst Solids, 4(1970) 548-563. Uttecht et a1, "Electric Field-Induced Filament in As-Te-Ge Glass", J Non-Cryst Solids, 4(1970)358-370. Tanaka et al, "Electrical Nature of the Lock-ON Filament in Amorphous Semiconductors", Solid State Commun, 8(1970) 75-780 Tanaka et al, "Electrical and Calorimetrical Analyses of Memory Phenomenon in Amorphous Semiconductors", Solid State Commun, 8(1970)1333-1336. Tanaka et a1, "Thermal Effect on Switching Phenomenon in Chalcogenide Amorphous Semiconductors", Solid State Commun, 8(1970)387-389. J. Stocker, "Bulk and Thin Film Switching and Memory Effects in Semiconducting Chalcogenide Glasses", Appl Phys Letters, '15:2(l969)55-57. Shanefield, P. E. Lighty: "Instabilities in Semiconducting Glass Diodes", Appl Phys Letters, 16:5(1970)212-214. . F. Weirauch, "Threshold Switching and Thermal Filaments in Amorphous Semiconductors", Appl Phys Letters, 16:2(1970) 72-730 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. H. J. 130 Stocker, "Phenomenology of Switching and Memory Effects on Semiconducting Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)371-381. Ure, R. Bowers, R. C. Miller; "Materials for Thermoelectric Applications", H. C. Gatos, ed., PrOperties of Elemental and Compound Semiconductors. New York: Interscience Publishers, Inc., 1960. 245-259. T. Minami, "Bond-Structural Approach to Electrical Conduction in Vitreous Semiconductors in the System As—S-Te", J Non-Cryst Solids, 3(1970)327-343. J. Stuke, "Review of Optical and Electrical Properties of Amorphous ASTM, Semiconductors", J Non-Cryst Solids, 4(1970)1-26. "Standard Method for Measuring Hall Mobility in Extrinsic Semiconductor Single Crystals", F76-68, Annual Book of ASTM Standards, 8(1970). McKimzie, ”Systematic Errors in Alternating-Current Hall- Effect Measurements", J Appl Phys, 40:12(1969)4954-4958. Lupu, N. M. Tallan, D. S. Tannhauser; "Apparatus for Measur- ing the Hall Effect of Low-Mobility Samples at High Temperatures", Rev Sci Instr, 38:11(1967)1658-1661. Boer, "The Electrical Conduction Mechanism in Highly Disordered Semiconductors", J Non-Cryst Solids, 2(1970)444-453. Kurtz, "A Study of the Anomalous Hall Effect in the Chalcogenide Semiconductor", Available #AD 71 0600, Feb 1970, Clearinghouse for Federal Scientific and Technical Information. D. Adler et a1, "Transport Properties of a Memory-Type Chalcogenide K. We Glass", J Non-Cryst Solids, 4(1970)330-337. Boer, "Carrier Mobility in Highly Disordered Semiconductors", J Non-Cryst Solids, 4(1970)583-592. J. Bednarczyk, D. Bednarczyk; "Effective Mobility in Polycrystalline M. H. M. H. Semiconductors", Acta Physics Polonica Fasc, 36(1969) 1011-1016. Cohen, H. Fritzsche, S. R. Ovshinsky; "Simple Band Model for Amorphous Semiconducting Alloys", Phys Rev Letters, 22:20(l969)1065-1068. Cohen, D. L. Johnson; "Carrier Kinetics in Amorphous Semi- conducting Alloys", J Non-Cryst Solids, 3(1970)271—278. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. . E. Spear, "Drift Mobility Techniques for the Study of Electrical . R. Haynes, W. Shockley; "The Mobility and Life of Injected Holes . M. Pai, S. W. Ing; "Photogeneration of Carriers in Vitreous 131 H. Cohen, "Electronic Structure and Transport in Covalent Amorphous Semiconducting Alloys", J Non-Cryst Solids, 2(1970)432-443. . E. Spear, "Transit - Time Measurements of Charge Carriers in Amorphous Selenium Films", Proc Phys Soc, B70(1957)669-675. E. Spear, "The Hole Mobility in Selenium", Proc Phys Soc, 76(1960)826-832. -7 .'..." 2,0 I Transport Properties in Insulating Solids", J Non-Cryst Solids, 1:3(1969)197-214. and Electrons in Germanium", Phys Rev, 81:5(1951)835-843. . #< “Hn—J-t‘ -. -_—- Selenium", Phys Rev, 173:3(1969)729-734. . G. Kepler, "Charge Carrier Production and Mobility in Anthracene Crystals", Phys Rev, 119:4(1960)1226-1229. . D. Tabak, P. J. Warter; "Field-Controlled Photogeneration and Free-Carrier Transport in Amorphous Selenium Films", Phys Rev, 173:3(1968)899-907. . Shockley, "Currents to Conductors Induced by a Moving Point Charge", J Appl Phys, 9:10(1938)635-636. . E. Spear, H. P. D. Lanyun, J. Mort; "Some Applications of Mercury Wetted Contact Relays", J Sci Instr, 39(1962)81-83. Alberigi et a1,"A 40-keV Pulsed Electron Accelerator", Rev Sci Instr, 41:8(1970)1205-1213. . P. Wagner, K. H. Besocke; "Alternating-Current Method for Separating the Contact Influence from Bulk Properties of Semiconductors", J Appl Phys, 40:7(1969)2916-2922. . A. Fagen, S. R. Ovshinsky, H. Fritzsche; "Hopping Conduction in an Amorphous Chalcogenide Alloy Film", Bull Am Phys Soc (II), 14:311(1969). . K. Rockstad, "Evidence for Hopping Conduction in Amorphous Chalcogenide Films", Solid State Commun, 7(1969)lSO7-1509. E. Owen, J. M. Robertson; "Electronic Properties of Some Simple Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)40-51. K. Jonscher, P. A. Walley; "Electrical Conduction in Non- Metallic Amorphous Films", J Vac Sci Technol, 6:4(1969) 116. 117. 118. 119. 120. 121. 122 O 123. 124. 125. 126. 127. 128. 129. 132 E. B. Ivkin, B. T. Kolomiets; "Hi-Frequency Conductivity in Arsenic Selenide", J Non-Cryst Solids, 3(1970)41-45. R. Vogel, P. J. Walsh; "Negative Capacitance in Amorphous Semi- conductor Chalcogenide Thin Films", Appl Phys Letters, 14:7(1969)216-218. Also available #AD 68 3467, Dec 1968, Clearinghouse for Federal Scientific and Technical Information. A. E. Owen, "Electronic Conduction Processes in Glass", Available #N69 13435, Feb 1968, Clearinghouse for Federal Scientific and Technical Information. .2! I“ D. L. Eaton, "Electrical Conduction Anomaly of Semiconducting Classes in the System As-Te-I", J Am Ceram Soc, 47:11(1964) -. —“ ,..—. A; s- “an-.- ... . - . J. Stuke, "Review of Optical and Electrical Properties of Amorphous Semiconductors", J Non-Cryst Solids, 4(1970)1-26. D. W. Deis, E. A. Dancy, A. Patterson; "On The Switching Character- istics and the Bulk Properties of Chalcogenide Glasses", 2(1970)141-147. M. H. Cohen, "Review of the Theory of Amorphous Semiconductors", J Non-Cryst Solids, 4(1970)391-409. J. D. MacKenzie, "Photoconductivity and Phototrophy in Non- Crystalline Solids", Adli Bishay, ed. Proceedings, Conference on Interaction of Radiation with Solids. 1966. 133-154. E. M. Pell, "Properties of Amorphous Photoconductors" Dessauer, Clark, ed's. Xerography,and Related Processes. New York: Focal Press, 1965. S. K. Bahl, K. L. Chopra; "Amorphous versus Crystalline GeTe Films: II - Optical Properties", J Appl Phys, 40:12(1969) 4940-4947. S. K. Bahl, K. L. Chopra; "Electrical and Optical Properties of Amorphous v.s. Crystalline GeTe Films", J Vac Sci Technol, 6:4(1969)561-565. R. Tsu, W. E. Howard, L. Esaki; "Photoconductivity and Optical Properties of Amorphous GeTe", J Non-Cryst Solids, 4(1970)322-327. W. E. Howard, R. Tsu; "Photoconductivity and Density of States for Amorphous GeTe", Phys Rev B, 1:12(1970)4709-4719. B. T. Kolomiets, V. H. Ljubin, V. L. Averjanov; "Investigation of Local States in Vitreous Semiconductors by Photo- conductivity and Thermally Stimulated Depolarization Methods", Mat Res Bull, 5:8(1970)655-664. 130. 131. 132. 133. 134. 135. 136. 133 K. Weiser, M. H. Brodsky, G. D. Pettit; "Electrical and Optical Properties of Amorphous AszTe3 Films", J Non-Cryst Solids, 4(1970)43-44. K. Weiser, M. H. Brodsky; "Dc Conductivity, Optical Absorption, and Photoconductivity of Amorphous Arsenic-Telluride Films", Phys Rev B, 1:2(1970)791-799. T. Botila, A Vancu; "Photoelectric Effects in Semiconducting Amorphous Chalcogenides", Mat Res Bull, 5:11(1970)925-932. E. A. Fagen, H. Fritzsche; "Optical Properties of Amorphous Chalcogenide Alloy Films", J Non-Cryst Solids, 2(1970) 180-191. E. A. Fagen, H. Fritzsche; "Photoconductivity of Amorphous Chalcogenide Alloy Films", J Non-Cryst Solids, 4(1970) W...“ ‘9‘“: “Id-5' a S. W. Ing, J. H. Newhart, F. Schmidlin; "Charge Transport and Photoconductivity in Amorphous Arsenic TriSulfide Films", J Appl Phys, 42:2(1971)696-703. D. Halliday, R. Resnick; Physics, Part I and II. New York: John Wiley & Sons, Inc., 1966. 698-699. 134 Part 5 -- References for Chapter 5 1. 10. 11. 12. 13. T. 8. Moss, Optical Properties of Semi-Conductors. London: Butterworth & Co., 1961. J. F. Black, E. Lanning, S. Perkowitz; "Infrared Techniques for Semiconductor Characterization", Infrared Phys, 10:2(1970) 125-139. W. Paul, "Use of Thin Films in Determining the Fundamental Optical ASTM, Properties of Semiconductors", J Vac Sci Technol, 6:4(1969) 483-493. Also available #AD 70 7476, Clearinghouse for Federal Scientific & Technical Information. "Infrared Absorption Analysis of Impurities in Single Crystal Semiconductor Materials", F120-70T;"Interstitial Atomic Oxygen Content of Silicon by Infrared Absorption", F121—7OT; "Interstitial Atomic Oxygen Content of Germanium by Infrared Absorption",F122-7OT; "Substitutional Atomic Carbon Content of Silicon by Infrared Absorption", F123-70T, Annual Book of ASTM Standards, 8(1970). Lacourse, V. A. Twaddell, J. D. Mackenzie; "Effects of Impurities on the Electrical Conductivity of Glassy Selenium", J Non-Cryst Solids, 3(1970)234-236. Vasko, D. Lezal, I. Srb; "Oxygen Impurities and Defects in T. O J. A. Chalcogenide Glasses", J Non-Cryst Solids, 4(1970)311—321. Kolomiets, T. N. Mamontova, G. I. Stepanov; "Influence of Impurities on the Energy Spectrum of Charge Carriers in Vitreous T1 Se - As Te ", Phys Stat Sol (a), 3(1970) 309-314. 2 2 3 Kreidl, "Semiconductors and Optical Properties", Glass Ind, 51:6(1970)264-267. Fraser, J. Juger; "Arsenic Trisulfide. A New Infrared Trans- mitting Glass", J Opt Soc Am, 43:12(1953)332A. Frericks, "New Optical Glasses with Good Transparency in the . R. Infrared", J Opt Soc Am, 43:12(1953)1153-1157. Hilton, M. Brau; "New High Temperature Infrared Transmitting Glasses", Infrared Phys, Part I - 3(1963)69-76; Part II - 4 (l964)213-221; Part III - 6(1966)183-194. Hilton, "Nonoxide Chalcogenide Glasses as Infrared Optical Materials", Appl Opt, 5:12(1966)1877-1882. Hilton, New High Temperature Infrared Transmitting Glasses. TeXas Instrument, Inc., Dallas, Texas, Report No. 08-65-121, September 30, 1965. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. . A. Davis, R. F. Shaw; "Characteristic Phenomena in Amorphous . T. Kolomiets, T. F. Mazets, A. M. Andriesh,Sh. M. Efendiev; 135 . R. Hilton, "Optical Properties of Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)28-39. . A. Fagen, "Optical Properties of Amorphous Chalcogenide Alloy Films", J Non-Cryst Solids, 2(1970)180-191. Stuke, "Review of Optical and Electrical Properties of Amorphous Semiconductors", J Non-Cryst Solids, 4(1970)1—26. Semiconductors", J Non-Cryst Solids, 2(1970)406-431. E. Owen, "Semiconducting Glasses", Contem Phys, 2:3(1970) 227-286. I '.. 17. ' Ti ‘57." "On the Energy Spectrum of Vitreous Arsenic Sulphide", J Non-Cryst Solids, 4(1970)45-56. View“-"- T. Edmond, "Far Infra-red Study of As28e3- Type Glasses", Proc Phys Soc, 81(1963)378-379. . F. Shaw, W. Y. Llang, A. D. Yoffee;"0ptica1 Properties, Photoconductivity, and Energy Levels in Crystalline and Amorphous Arsenic Triselenide", J Non-Cryst Solids, 4(1970)29-42. C. Taylor, S. G. Bishop, D. L. Mitchell; "Far Infrared and Microwave Optical Properties of Vitreous As28e3", Solid State Commun, 8(1970)1783-1787. Nagels, R. Callaerts, M. Denayer, R. DeConinck; "Electrical Properties of Vitreous T1 Te - As Te ", J Non-Cryst Solids, 2 2 3 4(1970)295-303. . T. Edmond, "Measurements of Electrical Conductivity and Optical Absorption in Chalcogenide Glasses", J Non-Cryst Solids, 1(1968)39-48. . Weiser, M. H. Brodsky; "dc Conductivity, Optical Absorption, and Photoconductivity of Amorphous Arsenic-Telluride Films", Phys Rev B, 1:2(1970)791-799. Procarione, C. Wood; "The Optical Properties of Sb28e3 - szTe3", Phys Stat Sol, 39(1970). . E. Spicer, T. M. Donovan; "Photoemission and Optical Studies of Amorphous Germanium", J Non-Cryst Solids, 2(1970)66-80. . Tauc, A. Abraham; "Infrared Absorption in Amorphous Germanium", J Non-Cryst Solids, 4(1970)279-288. L. Chopra, S. K. Bahl, "Structural, Electrical and Optical Properties of Amorphous Germanium Films", Phys Rev B, 1:6(1970)2545-2556. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 136 . Beaglehole, M. Zavetova; "The Fundamental Absorption of Amorphous Ge,Si and GeSi Alloys", J Non-Cryst Solids, 4(1970)272-278. . C. Chittick, "Properties of Glow-Discharge Deposited Amorphous Germanium and Silicon", J Non-Cryst Solids, 3(1970)255-270. . H. Brodsky et a1, "Structural, Optical, and Electrical Properties of Amorphous Silicon Films", Phys Rev B, 1:6(1970)2632-264l. . A. Davis, "Optical Absorption, Transport and Photoconductivity in Amorphous Selenium", J Non-Cryst Solids, 4(1970)107-116. . Vasko, "Optical Constants of Amorphous Selenium from X-Rays to the Far Infrared", J Non-Cryst Solids, 3(1970)225-233. Aoki, S. Okano; "Optical Absorption in Amorphous Selenium in High Electric Fields", Jap J Appl Phys, 5(1966)980-981. . Kramer et a1, "Optical Properties of Amorphous Selenium", Phys Rev Letters, 25:15(1970)1020-1022. K. Bahl, K. L. Chopra; "Electrical and Optical Properties of Amorphous vs Crystalline GeTe Films", J Vac Sci Technol, 6:4(1969)561—565. K. Bahl, K. L. Chopra, "Amorphous Versus Crystalline Gete Films: II - Optical Properties", J Appl Phys, 40:12(1969)4940-4947. Tauc, "Optical Properties and Electronic Structure of Amorphous Semiconductors", S. Nudelman, S. S. Metra, F. Abeles, ed's. Optical Properties of Solids. New'Ybrk: Plenum Press, 1969. Ch 5. Urbach, "The Long Wavelength Edge of Photographic Sensitivity and of the Electronic Absorption of Solids", Phys Rev, 92:4(1953)1324. K. Hindley, "Random Phase Model of Amorphous Semiconductors", J Non-Cryst Solids, 5(1970)17-30. E. Lewis, "The Principal Energy Gap in GeTe: A Possible Explanation for the Experimental Discrepancy Between E (optical) and Eg(therma1)", Phys Stat Sol, 39(1970) K5-K7. T. Edmond, "A Relationship Between Optical Absorption and Electrical Conductivity in Glassy Semiconductors H AsZSe1_xTe2+x , Phys Stat Sol (a), 3(1970)K129. Tauc, "Electronic Properties of Amorphous Materials", Science, 158(22 Dec 1967)1543-1548. L45: affirm. " ‘73“, ‘HF¥”“ 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 137 J. Tauc, "Absorption Edge and Internal Electric Fields in J. T. C. V. Amorphous Semiconductors", Mat Res Bull, 5:8(1970)721-730. Edmond, "Electronic Conduction In A828e , As Se Te and Similar Materials", Brit J Appl Phys, 19:8(13661979-989. Seraphin, Semiconductors and Semimetals. Vol VI. New York: Academic Press, 1968. Owen, J. M. Robertson; "Electron Properties of Some Simple Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)40-51. Hodge, C. A. Baumgardner; "Current-Induced Reflectivity Effects in a Semiconductor", Phys Rev B, 1:8(1970)3347-3350. J. Feinleib, S. R. Ovshinsky; "Reflectivity Studies of the Te(Ge,As)-Based Amorphous Semiconductor in the Conducting and Insulating States", J Non-Cryst Solids, 4(1970)564-572. Goryunova et a1, "Electronic and Vibrational Reflection Spectra of CdGeAsz in Crystalline and Vitreous States", J Non-Cryst Solids, 4(1970)57-65. Hilton, C. E. Jones, M. Brau; "Non-Oxide IVA-VA-VIA Chalcogenide Glasses", Phys Chem Glasses, 7:4(1966)105-126. Hilton, C. E. Jones, "The Thermal Change in the Nondispersive Infrared Refractive Index of Optical Material", Appl Opt, 6:9(1967)1513-1517. Boer, "Deflecting of Light Beams Using a Gradient of the Index of Refraction in Semiconducting Glasses", Appl Phys Letters, 16:6(1970)228-230. Perkin-Elmer, Model 450 Instruction Manual. July 1965. J. Tauc, "Optical Properties and Electrical Structure of A. H. Amorphous Germanium and Silicon", Mat Res Bull, 3:1(1968) 37-46. Clark, "A Review of Band Structure and Transport Mechanisms in Elemental Amorphous Semiconductors", J Non-Cryst Solids, 2(1970)52-65. R. Tsu, W. E. Howard, L. Esaki; "Optical and Electronic Properties D. E. and Band Structure of GeTe and SnTe", Phys Rev, 172:3(1968) 779-788. Gray, American Institute of Physics Handbook. 2nd ed. New York: McGraw-Hill Book Co., 1963. Section 6. 138 Part 6 -- References for Appendix A 1. 10. 11. 12. 13. 14. 15. 16. R. J. Charles, "The Nature of Glass", Sci Am, 217:3(Sept 1967) 127-136. Also available: G.E. Report #67-C-317, General Electric, Inc., R. W. Douglas, "The Chemical Approach to Problems of the Glassy State", V. D. Frechett, ed., Non-Crystalline Solids. New York: John Wiley & Sons, Inc., 1960, 374-411. E. D. Dietz, "The Glassy State", Sci & Technol, 82(Nov 1968). B. Eckstein, "The Inner Parameters of Glassy Systems", J Non-Cryst Solids, 4(1970)366-369. R. Roy, "Classification of Non-Crystalline Solids", J Non-Cryst Solids, 2(1970)33-40. wxrrfia‘zm, A. E. Owen, "Semiconducting Glasses", Contemp Phys, 11:3(1970) 227-286. A. R. Hilton, New High Temperature Infrared Transmitting Glasses. Texas Instruments, Dallas, Texas, Report No. 08-65-121, September 30, 1965. H. Rawson, Inorganic Glass-Forming:Systems. New York: Academic Press, 1967. J. D. Mackenzie, ed., Modern Appects of the Vitreous State. 3 Vols. Washington: Butterworth Scientific Publications, 1964. A. R. Hilton, "Chemical Composition and Glass Formation in Chalcogenide Systems", Phys Chem Glasses, 9:5(1968)l48-152. J. D. Mackenzie, "Preparation and Properties of Noncrystalline Films", J Vac Sci Technol, 6:4(1969)658—662. Proceedings, "Symposium on Instabilities in Semi-Conductors, March 1969, New York", IBM J Res Develop, 13:5(1969). Proceedings, "American Physical Society Meeting, March 1969, Philadelphia", Bull Phys Soc, 14:3(1969). Proceedings, "Symposium on Semiconductor Effects in Amorphous Solids, May 1969", J Non-Cryst Solids, 2(1970). Gordon Research Conference, Chemistry and Metallurgy of Semiconduc- tors, July 1969, Andover, New Hampshire. Proceedings, "International Conference on Amorphous and Liquid Semiconductors, Sept 1969, Cambridge, England", J Non- Cryst Solids, 4(1970); Also available #AD 70 1192, Jan 1970, Clearinghouse for Federal Scientific & Technical Information. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 139 Proceedings, "American Physical Society Meeting, Solid State Physics, Mar 1970, Dallas, Texas", Bull Am Phys Soc, 15:3(1970). Ioffe, Physics of Semiconductors. New York: Academic Press, Inc., 1960. Ioffe, A. R. Regel; "Non-Crystalline, Amorphous, and Liquid Electronic Semiconductors", A. F. Gibson, ed., Progress in Semiconductors. Vol 4. London: Heywood & Co., 1960. Gubanov, Quantum Electron Theory of Amorphous Conductors. Translated by A. Tybulewicz. New York: Consultants Bureau, 1965. S. Wang, Solid State Electronics. New York: McGraw—Hill Book Co., J. C. E. F. 1966. Phillips, "Bonds and Bands in Semiconductors", Science, 169 (11 Sept 1970)1035-1042. Bass, "Chemistry of Semiconductors", Chem Brit, 5:3(1969) 100-103. Weller, "An Introduction to Principles of the Solid State", J Chem Educ, 47:7(1970)501-507. Gurnee, "Fundamental Principles of Semiconductors", J Chem Educ, 46:2(1969)80-85. J. Tauc, "Electronic Properties of Amorphous Materials", Science, E. A. 158(22 Dec 1967)1543-1548. Davis, R. F. Shaw; "Characteristic Phenomena in Amorphous Semiconductors", J Non-Cryst Solids, 2(1970)406-431. Mott, "Electrons In Disordered Structures", Advan Phys, l6(1967),49-144. Mott, "Conduction in Non-crystalline Systems: Part I - Localized Electronic States In Disordered Systems", Phil Mag, 17:150(June, 1968)1259-1268. Mott, E. A. Davis; "Conduction in Non-crystalline Systems: Part II - The Metal-insulator Transition in a Random Array of Centres", Phil Mag, 17:150(June 1968)1269-1284. Mott, "Conduction in Non-crystalline Materials: Part III - Localized States in a Pseudogap and Near Extremities of Conduction and Valence Bands", Phil Mag, 19:160(1969) 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 140 . F. Mott, "Conduction in Non-crystalline Systems: Part IV - Anderson Localization in a Disordered Lattice", Phil Mag, 22:175(1970)7-29. . A. Davis, N. F. Mott; "Conduction in Non-crystalline Systems: Part V - Conductivity, Optical Absorption and Photo- conductivity in Amorphous Semiconductors", Phil Mag, 22:179 (1970)903-922. F. Mott, "Charge Transport in Non-crystalline Semiconductors", Festkoerperprobleme, 9(1969)22-45. F. Mott, "Conduction and Switching in Non-crystalline Materials", Contemp Phys, 10:2(1969)125-138. Kummer, M. E. Milberg; "Ionic Conduction and Diffusion in Solids", Chem Eng News, 47(12 May 1969)90-99. W, -r:—‘-.-“‘+":-z-**1—rn . Gutmann, L. E. Lyons; Organic Semiconductors. New York: John Wiley & Sons, Inc., 1967. . M. Ziman, "Electrons in Liquid Metals and Other Disordered Systems", Proc Roy Soc Lond, A318(Sept 1970)401-420. . F. Mott, "The Electrical Properties of Liquid Mercury", Phil Mag, 13:125(1966)989-1014. Adler, "Electrical and Optical Properties of Transition-Metal Oxides", Radiation Effects, 4(1970)123-131. N. R. Rao, G. V. Subba Rao; "Electrical Conduction in Metal Oxides", Phys Stat Sol (a), 1(1970)597-652. R. Ovshinsky, "Reversible Electrical Switching Phenomena in Disordered Structures", Phys Rev Letters, 21:20(1968) 1450—1453. Grigorovici, "Short-Range Order in Amorphous Semiconductors", J Non-Cryst Solids, 1:4(1969)303-325. . Lax, H. L. Frisch, "Electronic Band Structure of One-Dimensional Disordered Arrays", V. D. Frechette, ed., Non—Crystalline Solids. New York: John Wiley & Sons, Inc., 1960, 144-167. Y. Tong, "Electronic Structures of One-Dimensional Systems", J Non-Cryst Solids, 4(1970)455-458. Landauer, "Electrical Resistance of Disordered One—Dimensional Lattices", Advan Phys, 18(1969)863-867. . Moorjani, C. Feldman; "Electrical Properties of Amorphous Semiconductors", Rev Mod Phys, 36(1964)1042-1049. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 141 . Gosar, "Narrow Band Conduction in a Quasi-Amorphous Lattice", J Non-Cryst Solids, 4(1970)436-441. I. Gubanov, "On the Theory of AIIIBv-Type Semiconductor", Soviet Physics-Technical Physics, 1:2(1956) . V. Gordeev, "The Electric Conductivity of Amorphous Semi- conductors", Soviet Physics-Technical Physics, 3:1(1958) . I. Gubanov, "Length of the Free Path of the Electrons in Liquid and Amorphous Conductors", Soviet Physics-Technical Physics, 1:2(1956)1605-1611. F. Edward, "The Electronic Structure of Disordered Systems", Phil Mag, 6:65(1961)617-637. . Saven,"Contribution to the Theory of Disordered Alloys", Phys Rev, 178:3(1969)1136-1144. . K. Hindley, "Random Phase Model of Amorphous Semiconductors", J Non-Cryst Solids, 5(1970)17-40. Wilma, P. Phariseau; "On A Formal Theory of Conductivity In Amorphous Media", J Non-Cryst Solids, 4(1970)442-451. . W. Anderson, "Absence of Diffusion Certain Random Lattices", Phys Rev, 109(1958)1492-1505. W. Anderson, "Self-Consistent Pseudopotentials and Ultralocal— ized Functions for Energy Bands", Phys Rev Letters, 21:1(1968)13-16. W. Anderson, "The Fermi Glass: Theory and Experiment", Comm Solid State Phys, 11:6(1970)193-197. Brouers, "Some Questions About Electron Localization in Disordered Systems", J Non-Cryst Solids, 4(1970)428-432. . L. Bonch-Breuevich, "Behavior of the Charge Carriers in Slowly Varying Random Field and Semiphenomenological Approach to the Theory of Electronic Processes in Disordered Systems", J Non-Cryst Solids, 4(1970)410-416. Velicky, "Theory of Electronic Transport in Disordered Binary Alloys: Coherent-Potential Approximation", Phys Rev, 184:3(1969)614-627. Fischer, S. A. Rice; "Theory of Electron Transport in Disordered Systems: The Nearly Free-Electron and Nearly Bound-Electron Limits", J Chem Phys, 51:8(1969)3352-3360. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 142 . F. Mott, "The Transition to the Metallic State", Phil Mag, 6:62(1961)287-309. . F. Mott, "Metal-Insulator Transition", Rev Mod Phys, 40:4(1968) 677-683. . I. Halperin, T. M. Rice; "The Excitonic State at the Semi- conductor—Semimetal Transition", Frederick Seitz, David Turnbull, Henry Ehrenreich, ed's., Solid State Physics. New York: Academic Press, 1968, 115-192. Doniach, "The Insulator-Metal Transition", Advan Phys, 18:76 (1969)819-848. F. Mott, "The Metal-Non-Metal Transition", Comm Solid State Phys, 11:6(1970)183-192. Cutler, N. F. Mott; "Observation of Anderson Localization in an Electron Gas", Phys Rev, 181:3(1969)1336-1340. . F. Mott, "Conduction in Solids", Rivista Del Nuovo Cimento, Numero Specia1e(l969)174-183. . K. Chakraverty, "Metal-Insulator Transition; Nucleation of a Conducting Phase in Amorphous Semiconductors", J Non-Cryst Solids, 3(1970)317-326. . H. Lieb, F. Y. Wu; "Absence of Mott Transition in an Exact Solution of the Short-Range, One-Band Model in One Dimension", Phys Rev Letters, 20:25(1968)1445-1448. F. Edwards, "The Localization of Electrons in Disordered Systems", J Non-Cryst Solids, 4(1970)417-425. W. Boer, "Remarks to the Ovshinsky Effect", Phys Stat Sol (a), 1(1970) K21—K24. Also available #AD 70 8820, Clearinghouse for Federal Scientific & Technical Information. W. Boer, R. Haislip; "Semiconductivity of Classes", Phys Rev Letters, 24:5(1970)230-233. . W. Boer, "The Conduction Mechanism of Self-Compensated Highly Disordered Semiconductors: Parts I 8 II", Phys Stat Sol, 34(1969)721-741. . W. Boer, "Carrier Mobility in Highly Disordered Structures", J Non-Cryst Solids, 4(1970)583-592. W. Boer, "The Electrical Conduction Mechanism in Highly Disordered Semiconductors", J Non-Cryst Solids, 2(1970) 444-453. a-“ is T_‘____y.—__~ n. Iv, "‘2. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 143 H. Cohen, H. Fritzsche, S. R. Ovshinsky; "Simple Band Model for Amorphous Semiconducting Alloys", Phys Rev Letters, 22:20 (1969)1065-1068. . Bednarczyk, D. Bednarczyk; "Effective Mobility in Polycrystal- line Semiconductors", Acta Physica Polonica, 36(1969) 1011-1016. M. Stevels, Progress in the Theory of the Physical Properties of Class. New York: Elsevier Publishing Co., Inc., 1948. . Fritzsche, S. R. Ovshinsky; "Electronic Conduction in Amorphous Semiconductors and the Physics of the Switching Phenomena", J Non-Cryst Solids, 2(1970)393-405. . H. Cohen, "Review of the Theory of Amorphous Semiconductors", J Non—Cryst Solids, 4(1970)391-409. N. Economou et al, "Localization in Disordered Materials: Binary Alloys", Phys Rev Letters, 25:8(1970)520-523. . H. Cohen, D. L. Johnson; "Carrier Kinetics in Amorphous Semi- conducting Alloys", J Non-Cryst Solids, 3(1970)271-278. H. Cohen, "Electronic Structure and Transport in Covalent Amorphous Semiconducting Alloys", J Non-Cryst Solids, 2(1970)432-443. N. Economou, M. H. Cohen; "Anderson's Theory of Localization and the Mott-CFO Model", Mat Res Bull, 5:8(1970)577-590. . F. Mott, "Conduction in Glasses Containing Transition-Metal Ions", J Non-Cryst Solids, 1(1968)1-17. . F. Mott, "Polarons in Transition-Metal Oxides", Comm Solid State Phys, 1:4(1968)105-111. I. Klinger, "Some Problems Concerning Polarons in Disordered Systems", J Non—Cryst Solids, 5(1970)78-102. . G. Austin, N. F. Mott; "Polarons in Crystalline and Non-Crystal- line Materials", Advan Phys, l8:71(1969)41-102. . J. O'Dwyer, "Theory of High-Field Conduction in 3 Dielectric", J Appl Phys, 40:10(l969)3887-3890. . J. O'Dwyer, "Theory of Dielectric Breakdown in Solids", Solid State Sci, 116:2(1969)239-242. . C. Lilly, Jr; D. A. Lowitz; John C. Schug; "Space-Charge Conduction in Insulators", J Appl Phys, 39:9(1968)4360-4364. ‘IIE¢T**“”*_TZQDW 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. . Lampert, "Double Injection in Insulators", Phys Rev, 125:1 144 Tredgold, "Space-Charge Conduction", Contemp Phys, 8:4(1967) 401-411. Pulfrey, A. H. M. Shouska, L. Young; "Electronic Conduction and Space Charge in Amorphous Insulating Films", J Appl Phys, 41:7(1970)2838-2843. O'Dwyer, "Theory of Double-Charge Injection From a Dielectric", J Appl Phys, 39:9(1968)4356-4359. (1962)126-141. .. on- *I J"" Rosenburg, M. A. Lampert; "Double Injection in the Perfect Insulator: Further Analytic Results", J Appl Phys, 41:2(1970)508-521. Bonch—Bruevich, "On the Theory of Absorption of Low Frequency Electro-Magnetic Waves in Disordered Semiconductors", Mat Res Bull, 5:8(1970)555-566. ‘- “flu; A ‘.A_'- 4": --.‘I . Krebs, "Electrical Conductivity and Chemical Bonding in Crystalline, Glassy and Liquid Phases", J Non-Cryst Solids, 0(1969)455-473. ' . Drake, I. F. Scanlan; "Bond Type, Polarisation and Electrical Conduction in Some Oxide and Chalcogenide Glasses", J Non-Cryst Solids, 4(1970)234-247. Brodsky, P. J. Stiles; "Electronic Dielectric Constant of Amorphous Semiconductors", Phys Rev Letters, 125:12 (1970)798-801. J. Tauc, "Absorption Edge and Internal Electric Fields in Amorphous Semiconductors", Mat Res Bull, 5:8(1970)721-730. 145 Part 7 -- References for Appendix C 1. L. B. Valdes, "Resistivity Measurements on Germanium for Transistors", Proc IRE, 42(1954)420-427. 2. F. M. Smits, "Measurement of Sheet Resistivities with the Four- Point Probe", Bell System Tech J, 37:3(1958)711—718. 3. ASTM, "Measuring Resistivity of Silicon Silica with a Collinear Four-Probe Array", F84—70, Annual Book of ASTM Standards, 8(1970). 4. J. Hornstra, L. J. Van der Pauw; "Measurement of the Resistivity Constants of Anisotropic Conductors by Means of Plane- Parallel Discs of Arbitrary Shape", Phillips Research Reports, 13:1(1958)169-17l. I'..,A.C-Ai \‘4 Q 1' m _ . 5. L. J. Swartzendruber, "Correction Factor Tables for Four-Point ' Resistivity Measurement on Thin, Circular Semiconductor Samples", NBS Technical Note No. 199. Available #AD 68 3408, Clearinghouse for Federal Scientific & Technical Information. 6. M. A. Logan, "Sheet Resistivity Measurements on Rectangular Surfaces-General Solution for Four—Point Probe Conversion Factors", Bell System Tech J, 46:10(1967)2277-2322. 7. D. O. Townley, "Four-Point Probe Resistivity Measurements Demonstrating the Effect of Probe Position Correction Factors for Thin Circular Slices", also: Qgrrgggign, Eggpors for Slice Resistivitngeasurements Using 62.7 mil - —P P .. Dow Corning Corp., Electronic Product Division, Hemlock, Michigan. 8. S. Murashima, H. Kanamori, F. Ishibashi; "Correction Divisors for the Four-Point Probe Resistivity Measurement on Cylindrical Semiconductors", Jap J Appl Phys, 9:1(1970) 58-67. 9. L. J. Swartzendruber, F. H. Ulmer, J. A. Coleman; "Direct-Reading Instrument for Silicon and Germanium Four-Probe Resistivity Measurements", Rev Sci Instr, 39:12(1968)l858—1863. 10. J. F. Hallenback, D. Piscitelli; "Precision Four-Probe Resistivity Instrumentation", R. R. Haberecht, ed., Semiconductor Silicon. l'st International Symposium on Silicon Materials, Science & Technology, New York, 1969, 736-746. 11. R. A. Cooper, E. Lerner; "Four-Point Mercury Contact Probe for Electrical Resistivity Measurements of Thin Films", Rev Sci Instr, 39:8(1968)1207-1208. 146 12. A. Tong, A. Dupnock; "The Two-Level Effect in Sheet Resistance Measurements Made with a Four-Point Probe", J Electrochem Soc, 118:2(1971)390-394. 13. W. G. Pfann, "An Electrolytic Method for Pointing Tungsten Wires", Metal Technology, 14(1947)1-4. 14. A. Associates, "The Parameters of Operation of Constant Current Devices", Insight, Argonaut Associates, Beaverton, Oregon, (Fall 1968). .- A . 15. J. T. Stock, "Transistorized Power Sources for Constant Current 8 Coulometric Titration", J Chem Educ, 46:12(l969)858-860. j 16. Editorial Staff, "Applying and Synthesizing Isolated Power Supplies i in Instrumentation Systems", Electronic Instr Digest, , 4:8(1968)8-11. E 17. J. C. Perkinson, W. C. Pierce; "Precision DC Current Sources", Hewlett-Packard J, 21:1(1969)15-20. 18. G. F. Walker, "Applications of a DC Constant-Current Source", IEEE Spectrum, 7:9(Sept 1970)47-58. 19. W. C. Pierce, "Constant Current Sources Gain New Capabilities", EDN Mag, 15(1 Sept 1970)47-51. 20. Editorial Staff, "Operational and Instrument Amplifiers for Signal Conditioning", Electronic Instr Digest, 3:12(1967) 104-111. 21. J. Rose, "Straight Talk About Data Amplifiers", EDN Mag, 11:14 (Nov 1966)26-33. 22. R. Y. Moss, "Errors in Data Amplifier Systems", Hewlett-Packard J, 18:11(1967)17-20. 23. T. Helstern, "Beware Common Mode Voltages", Electromechanical Design, 13:ll(l969)16-17. 24. Editorial Staff, "Rules Governing Design of Shielding and Grounding for Instrumentation Systems" Electromechanical Design, 10:2(1966)52-59. 25. D. H. Sheingold, L. Smith; "Operational-Amplifier-Circuit Noise Characteristics", Electronic Instr Digest, 5:9(1969)50-65. 26. Burr—Brown, Handbook of Operational Amplifier Applications. lst ed. Burr-Brown Research Corp., Tucson, Arizona, 1963. 27. J. N. Giles, Linear Integrated Circuits Applications Handbook. Fairchild Semiconductor, Mountain View, California, 1967. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 147 Philbrick-Nexus, Applications Manual for Operational Amplifiers. 2nd ed. Philbrick-Nexus Research, Dedham, Mass., 1968. R. Stata, "Operational Amplifiers", Parts I-III, Electromechanical Design, 9:9(1965), 9:11(1965), 10:2(1966); Part IV- Analog Devices, Cambridge, Massachusetts. W. Kaelin, C. V. Weden; "Review of Operational Amplifier Character- istics", Fairchild Controls Application Bulletin, Dec 1968. L. J. Giacoletto, Differential Amplifiers. New York: John Wiley & Sons, Interscience, 1970. I -_'_ «1 .7 "T R. D. Middlebrook, Differential Amplifiers. New York: John Wiley & Sons, 1963. J. F. Swindells, ed., Preceision Measurement & Calibration, “ Temperature. National Bureau of Standards Special g Publication 300, Vol 2, Washington: U. S. Government Printing Office, Aug 1968. W. A. Macan, "Electrical Sensors of Temperature", Res Develop, 21:6(1970)22-29. P. H. Dike, Thermoelectric Thermometry. Leeds & Northrup Co. Publication, Sept 1954. Omega Engineering, Thermocouple Tables and Alloy Data, #CT-Z. Omega Engineering Inc., Stamford, Conn. J. A. Hall, The Measurement of Temperature. London: Chapman 8 Hall Ltd., 1966, Ch 6. D. E. Gray, ed., American Institute of Physics Handbook. 2nd ed. New York: McGraw-Hill, 1963. Ch 4a. General Electric, Program Library, Mark-I TimesharingpService, #228116A (5M). General Electric Information Service Department, Bethesda, Maryland, July 1968, P.53. T. M. Dauphinee, E. Mooser; "Apparatus for Measuring Resistivity and Hall Coefficient of Semiconductors", Rev Sci Instr, 26:7(1955)660-664. ’ General Electric, Use and Operation of DryiReed Switches, #FTD-4409. General Electric Tube Department, Owensboro, Kentucky, Dec 1965. Editorial, "Reeds: Switches & Relays", Electromechanical Design, 12:4(1968)87-97. 148 43. A. W. Ayers, "How to Use Dry Reed Relays", Electromechanical Design, 14:3(1970)30-39. 44. B. I. Sommer, G. W. Plice; "Specification and Testing of Shielded Transformers", Electro-Technology, 71:5(May 1963)102-105. 45. Elcor, "Reduction of Noise in Low-Level Measurement and Control Systems Through Proper Use of Isolation Devices", Bulletin 95-765 & 79-364. Elcor Division, Halliburton Co., Silver Spring, Maryland. 46. W. G. Holzbock, Instruments for Measurement and Control. 2nd ed. New York: Reinhold Publishing Co., 1962, Ch 1. 47. M. Kutz, Temperature Control. New York: John Wiley 8 Sons, 1970. 48. M. Van Swaay, "The Control of Temperature", J Chem Educ, Part I - r 46:8(1969)A515-A528; Part II - 46:9(1969)A565-A575. “in 3'." .' 49. G. Fischer, D. Greig, E. Mooser; "Apparatus for the Measurement of Galvanomagnetic Effects in High Resistance Semiconductors", Rev Sci Instr, 32:7(1961)842-846. 149 Part 8 -- References for Appendix D l. J. F. Dewald et al, "Semiconducting Glasses". Oral Presentation, Electrochemical Society Meeting, Los Angles, May 1962. Abstracted: J Electrochem Soc,109(1962)243C. 2. A. D. Pearson et a1, "Chemical, Physical, and Electrical Properties of Some Unusual Inorganic Glasses", Advances In Glass Technology. New York: Plenum Press, 1962, 357-369. 3. A. D. Pearson, "Chemical, Physical and Electrical Properties of Some Unusual Inorganic Glasses: Part 2 ", Advances in Glass Technology. New York: Plenum Press, 1963, 145. 4. S. S. Flaschen et al, "Glass Composition and Coated Article", U. S. Pat. #3,034,119. Granted March 6, 1962. —-.:r.r.--: =.-":-.-.-r_.-‘-"ri 5. W. R. Northover, "Glass Composition", U. S. Pat. #3,117,013. Granted January 7, 1964. Q? 6. D. L. Eaton, "Electrical Conduction Anomaly of Semiconducting Classes in the System As-Te-I", J Am Ceram Soc, 147:11 (1964)554-558. 7. A. D. Pearson, "Characteristics of Semiconducting Glass Switching/ Memory Diodes", IBM J Res Develop, 13(1969)510-514. 8. A. D. Pearson, "Memory and Switching in Semiconducting Glasses: A Review", J Non-Cryst Solids, 2(1970)1-15. 9. M. P. Southworth, "The Threshold Switch: A New Component for AC Control", Control Engr, 11:4(April 1964)69-72. 10. S. R. Ovshinsky, "Thin-Film Semis, At Last", EEE Mag, (May 1966). 11. S. R. Ovshinsky, "Symmetrical Current Controlling Device", U. 8. Pat. #3,271,591. Granted September 6, 1966. 12. S. R. Ovshinsky, "Reversible Electrical Switching Phenomena in Disordered Structures", Phys Rev Letters, 21:20(1968) 1450-1453. 13. G. Sideris, "Transistors Face an Invisible Foe", Electronics, 39(19 May l966)19l-195. 14. L. Lydon, "The Tumult and the Shouting Die". Electronic News, 13:682(18 Nov 1968)1 15. Editorial Staff, "Companies Making It", Electronics, 41(25 Nov 1968). 16. Editorial Staff, "Glassy Semiconductors--Embryonic Enigma", Ind Res Mag, (Dec 1968)19-20. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 150 Editorial Staff, "Glassy Semiconductors Show Switching and Memory Effects", Phys Today, 22:1(1969)62-64. Editorial Staff, "Support for Ovshinsky", Electronics, 42(3 March K. L. 1969)58. Chopra, "The Ovshinsky Effect", Phys Today, 22:3(March 1969) 10. Chapman, "Bistable Electrical - Switching Behavior in Glasses and Other Media", Electronics Letters, 5:11(29 May 1969) 246-247. Paul, "Amorphous Semiconductors Stimulate Fundamental and J. Applied Research", Phys Today, 22:10(Oct 1969)97-101. . Ovshinsky, "Amorphous Semiconductors", Sci J, 5A:2(Aug 1969) 73-78 0 Ovshinsky, "An Introduction to Ovonic Research", J Non-Cryst Solids, 2(1970)99-106. Shanks, "Ovonic Threshold Switching Characteristics", J Non- Cryst Solids, 2(1970)504-514. Neale, "Device Structures and Fabrication Techniques for Amorphous Semiconductor Switching Devices", J Non-Cryst Solids, 2(1970)558-574. Nelson, "Ovonic Device Applications", J Non-Cryst Solids, 2(1970)528-539. Walsh et al, "Conduction and Electrical Switching in Amorphous Chalcogenide Semiconductor Film", Phys Rev,, 178:3(1969)1274-1279. Also available #AD 68 3465, Clearing- house for Federal Scientific & Technical Information. Stocker, "Bulk and Thin Film Switching and Memory Effects for Semiconducting Chalcogenide Glasses", Appl Phys Letters, 15:2(1969)55-57. Stocker, "Phenomenology of Switching and Memory Effects in Semiconducting Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)371-381. Walsh et al, "Experimental Results in Amorphous Semiconductor Switching Behavior", J Non-Cryst Solids, 2(1970)107-124. Uttecht et al, "Electric Field-Induced Filament Formation in As-Te-Ge Glass", J Non-Cryst Solids, 2(1970)358—370. 32. C. 34. C. 35. P. 36. C. 37. M. 38. K. 39. C. 40. R. 41. G. 42. F. 43. C. 44. T. 45. C. 46. A. 151 H. Sie, "Electron Microprobe Analysis and Radiometric Micro- scopy of Electric Field Induced Filament Formation on the Surface of As-Te-Ge Glass", J Non-Cryst Solids, 4(1970) 548-533. . O. Silva, G. Dir, C. Griffiths; "Bistable Switching and Memory Devices", J Non-Cryst Solids, 2(1970)316-333. H. Champness, R. H. Hoffman, "Conductivity Changes Associated with the Crystallization of Amorphous Selenium", J Non- Cryst Solids, 4(1970)138-148. Tronc, "A Method of Realisation of Thin Films Presenting the Memory Effect", Thin Solid Films, 5:1(1970)R29-R30. Feldman, K. Moorjani, "Observation of Filament Formation in Amorphous Films During Switching", Thin Solid Films, 5:1(1970)Rl-R4. M. Cohen, M. Campi, J. D. Penar; "Electronic Switching Phenomena in Hot-Pressed Ge", J Non-Cryst Solids, 2(1970)91-98. Moorjani, C. Feldman; "Electrical Conduction in Amorphous and Silicon", J Non-Cryst Solids, 4(1970)248—255. Feldman, K. Moorjani; "Switching in Elemental Amorphous Semi- conductors", J Non-Cryst Solids, 2(1970)82-90. H. Walden, "Two Switching Devices Utilizing VO ", IEEE Trans ED, l7:8(1970)603-612. Dearnaley, D. V. Morgan, A. M. Stoneham; "A Model for Filament Growth and Switching in Amorphous Oxide Films", J Non-Cryst Solids, 4(1970)593—612. Argall, "Switching Phenomena in Titanium Oxide Thin Films", Solid State Electron, 11(1968)535-541. F. Drake, I. F. Scanlan, J. H. Alexander; "Electrical Switching and Memory Devices", British Patent #1,l41,644. W. Hickmott, W. R. Hiatt; "Electrode Effects and Bistable Switching of Amorphous Nb205 Diodes", Solid State Electron, 13(1970)1033-1047. H. Sie, "Memory Cell Using Bistable Resistivity in Amorphous As-Te-Ge Film", Dissertation #69-20-670. University Microfilms, Inc., Ann Arbor, Michigan, 1969. Szymanski, D- C- Larson, M. M. Labes; "A Temperature-Independent Conducting State in Tetracene Thin Film", Appl Phys Letters, 14:3(1969)88-90. 152 47. N. Croitoru, L. Vescan; "Temperature Dependence of Switching in TlZSeAszTe3 Glass", Rev Roum Phys, 15:1(1970)107-108. 48. O. E. Wagner, "Temperature Dependence of Ovshinsky-Type Devices", J Appl Phys, 40:10(l969)4212-4213. 49. A. Kinoshita, T. Aono, T. Nakano; "Experiment on Memory Diode in Amorphous Semiconductor", Jap J Appl Phys, 9(1970)411. 50. D. L. Thomas, "Preswitching Behavior of Amorphous Chalcogenide Semiconductor Films", Electron Lett, 6:3(1970)62-64. 51. R. Holstrom, "Switching and Conduction Behavior of Amorphous Semiconductor Diodes", Proc IEEE, 57:8(Aug 1969)1451-1453. j 52. J. T. Kerr, "Current-Voltage Characteristics of Chalcogenide E Glass Films", J Non-Cryst Solids, 2(1970)203-204. " 53. H. J. Hovel, "Switching and Memory in ZnSe-Ge Heterojunctions", 1 Appl Phys Letters, 17:4(1970)141-143. 54. M. Kikuchi, S. Iizima;"'Memory Exchange'in Amorphous Semi- conductors", Appl Phys Letters, 15:10(1969)323-325. 55. Minnesota Mining and Manufacturing Co., "Solid State Switching Device". British Patent #1,117,001, 12 June 1963- 56. A. D. Pearson, C. E. Miller; "Filamentary Conduction in Semi- conducting Glass Diodes", Appl Phys Letters, 14:9(1969) 57. D. F. Weirauch, "Threshold Switching and Thermal Filaments in Amorphous Semiconductors", Appl Phys Letters, 16:2(1970) 72-73 0 58. D. Shanefield, P. E. Lighty; "Instabilities in Semiconducting Glass Diodes", Appl Phys Letters, 16:5(1970)212-214. 59. M. Sugi et al, "Evidence for Thermal Pinching Effect in Chalcogenide Amorphous Semiconductors", Solid State Commun, 8(1970) 829-832. 60. K. Tanaka et al, "Electrical Nature of the Lock-On Filament in Amorphous Semiconductors", Solid State Commun, 8(1970) 75-78 0 61 P. Calella et a1, "Some Experiments on Switching-Type Amorphous Semiconductors", #AD 70 3858, Jan 1970, Clearinghouse for Federal Scientific & Technical Information. 62. D. W. Deis, E. A. Dancy, A. Patterson; "On the Switching Character- istics and the Bulk Properties of Chalcogenide Glasses", J Non-Cryst Solids, 2(1970)141—147. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 153 N. Croitoru et a1, "Non-Ohmic Properties of Some Amorphous Semiconductors", J Non-Cryst Solids, 4(1970)493-503. C. Hamoguchi, Y. Sasaki, J. Nakai; "Electrical Conduction and A. M. Switching in Amorphous Semiconductors", Jap J Appl Phys, 9:10(1970)1195-1203. Stone et al, "An Amorphous Semiconductor R. F. Switch", Proc IEEE, 59:2(197l)323-324. Evans, J. H. Helbers, S. R. Ovshinsky; "Reversible Conductivity Transformations in Chalcogenide Alloy Films", J Non-Cryst Solids, 2(1970)334-346. Pamplin, "Negative Differential Conductivity Effects in Semi- conductors", Contemp Phys, 11:1(1970)l-19. Ridley, "Specific Negative Resistance in Solids", Proc Phys Soc, 82(1963)954-966. Barnett, "Current Filaments in Semiconductors", IBM J Res Dev, 13(1969)522-528. Ferro, S. K. Ghandi; "Observations of Current Filaments in Chromium-Doped GaAs", Appl Phys Letters, 16:5(1970)196-198. Barnett, H. A. Jensen; "Observation of Current Filaments in Semi-Insulating GaAs", Appl Phys Letters, 12:10(1968)341-342. G. Nimtz, K. Seeger; "Current Controlled Negative Differential Resistivity of P-Type Tellurium", Appl Phys Letters, 14:1 (1969)19-21. E. Harnik, A. Lepek; "Current Filaments in Acoustically Amplifying CdS", Solid State Commun, 8(1970)1287-1289. R. Vogel, P. J. Walsh; "Negative Capacitance in Amorphous Semi- A. C. conductor Chalcogenide Thin Films", Appl Phys Letters, 14:7(1969)216-218. Also Available #AD 68 3467, Clearing- house for Federal Scientific and Technical Informtion. Warren, "Switching Mechanism in Chalcogenide Glasses", Electron Lett, 5:19(l969)461-462. A. Csillag, "Investigation of the Switching Characteristics of the Te-As-Ge-Si Glass-Films", J Non-Cryst Solids, 4(1970)518- 522. 77. A. Csillag, H. Jager; "Energy—Controlled Switching Process in the 78. K. W. Amorphous System Te-As-Ge-Si", J Non-Cryst Solids, 2(1970)133-140. Boer, G. Dohler, S. R. Ovshinsky; "Time Delay for Reversible Electric Switching in Semiconducting Glasses", J Non-Cryst Solids, 4(1970)573-582. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 154 Fritzsche, S. R. Ovshinsky; "Conduction and Switching Phenomena in Covalent Alloy Semiconductors", J Non-Cryst Solids, 4(1970)464-479. . W. Boer, S. R. Ovshinsky; "Electothermal Initiation of an Electronic Switching Mechanism in Semiconducting Glasses", J Appl Phys, 41:6(1970)2675-2681. J. Stocker; C. A. Barlow, Jr.; D. F. Weiraugh; "Mechanism of Threshold Switching in Semiconducting Glasses", J Non-Cryst Solids, 4(1970)523-535. W. Boer, "Remarks to the Ovshinskyfofect", Phys Stat Sol (a), 1(1970)K21-K24. Also available #AD 70 8820, Clearinghouse for Federal Scientific & Technical Information. . W. Boer, "Electra-Thermal Analysis of a Thin Semiconducting Sandwich with Field-Independent Conductivity", Phys Stat Sol (8), 2:4(1970)817-823. . S. Chen, T. T. Wang; "On the Theory of Switching Phenomena in Semiconducting Glasses", Phys Stat Sol (a), 2(1970)79-84. G. Zhdan et al, "Current-Voltage Characteristics of S-Type in Some Film Systems and Its Interpretations", J Vac Sci Technol, 6:4(1969)673. . C. Mattes, "Theory of Electronic Switching Effect As A Coopera- tive Phenomenon", Phys Rev Letters, 22:18(1969)936-939. Balberg, "Simple Test for Double Injection Initiation of Switching", Appl Phys Letters, 16:12(1970)491-493. K. Chakraverty, "Metal-Insulator Transition: Nucleation of a Conducting Phase in Amorphous Semiconductors", J Non-Cryst Solids, 3(1970)317-326. H. Cohen, H. Fritzsche, S. R. Ovshinsky; "Simple Band Model for Amorphous Semiconducting Alloys", Phys Rev Letters, 22:20(1969)1065-1068. Fritzsche, S. R. Ovshinsky; "Electronic Conduction in Amorphous Semiconductors and the Physics of the Switching Phenomena", J Non-Cryst Solids, 2(1970)393-405. K. Henisch, E. A. Fagen, S. R. Ovshinsky; "A Qualitative Theory of Electrical Switching Processes in Monostable Amorphous Structures", J Non-Cryst Solids, 4(1970)538-547. P. Gupta, "On Electrical Switching and Memory Effects in Amorphous Chalcogenides", J Non-Cryst Solids, 3(1970)148-154. H. Cohen, "Review of the Theory of Amorphous Semiconductors", J Non-Cryst Solids, 4(1970)391-409. W37 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 155 . Fritzsche, "Physics of Instabilities In Amorphous Semi- conductors", IBM J Res Develop, 13(1969)515-521. . Fritzsche, S. R. Ovshinsky; "Calorimetric and Dilatometric Studies on Chalcogenide Alloy Glasses", J Non-Cryst Solids, 2(1970)148-154. . G. Bagley, H. E. Bair; "Thermally Induced Transformations in Glassy Chalcogenides", J Non-Cryst Solids, 2(1970)155-160. . Bienenstock, F. Betts, S. R. Ovshinsky; "Structural Studies of Amorphous Semiconductors", J Non-Cryst Solids, 2(1970) 347-357. . L. Cook, "Model for the Resistive-Conductive Transition in Reversible Resistance-Switching Solids", J Appl Phys, 41:2(1970)551-554. . D. Senturia, C. R. Hewes, D. Adler; "NMR Study of a Memory—Type Amorphous Semiconductor", J Appl Phys, 41:1(1970)430-431. . Tanaka et a1, "Electrical and Calorimetrical Analyses of Memory Phenomenon in Amorphous Semiconductors", Solid State Commun, 8(1970)1333-1336. . Iizima et a1, "Electrical and Thermal Properties of Semi- conducting Glasses As-Te-Ge", Solid State Commun, 8(1970) 153-155. . Tanaka et a1, "Thermal Effect on Switching Phenomenon in Chalcogenide Amorphous Semiconductors", Solid State Commun, 8(1970)387-389. . V. Phillips, R. E. Booth, P. W. McMillan; "Structural Changes Related to Electrical Properties of Bulk Chalcogenide Glasses", J Non-Cryst Solids, 4(1970)510-517. A. C. Warren, "Thermal Switching in Semiconducting Glasses", J Non-Cryst Solids, 4(1970)613-616. W “'6 Tim mun- ”'1711111111111117111111111“