LBL-425 UC-2 Genera!,Miscellaneous, and Progress Reports TID-4500 (59th Ed.)
d
INORGANIC MATERIALS RESEARCH DIVISION ANNUAL REPORT 1971
April 1972
-NOTICE- This report was prepared as an account of work sponsored by the United States Government. Neither the United States nor the United States Atomic Energy Commission, nor any of their employees, nor any of their contractors, subcontractors, or their employees, mBkes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, com pleteness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights.
Lawrence Berkeley Laboratory University of California Berkeley, California
MOTflWJTW* Of THIS BOCUWtMT IS OtH-WTM Work done under U.S. Atomic Energy Commission Contract No. W-7405-eng-48
Printed in the United States of America Available from National Technical Information Service U.S. Department of Commerce 5285 Port Royal Road Springfield, Virginia 22151 Price: Printed Copy $3.00; Microfiche $0.95 -iii-
INORGANIC MATERIALS RESEARCH DIVISION ANNUAL REPORT. 1971*
Contents
GENERAL INTRODUCTION xi
I. CHEMISTRY A. Inorganic Chemistry William L. Jolly, Principal Investigator 1. X-Ray Photoelectron Spectroscopy 1 2. Studies of the Hydrides of Groups TV and V 2 3. Studies in Liquid Ammonia 5 4. Boron Hydride Chemistry 6 5. Sulfur-Nitrogen Chemistry 7 f. Research Plans for Calendar Year 1972 7 7. 1971 Publications and Reports 8 Robert E. Conniak, Prinoipal Investigator 1. Water Exchange from the First Coordination Sphere of Some Diamagnetic Ions 10 2. Oxygen-17 NMR Studies of Waters in the Second Coordination Sphere of Cr3+ ..... 10 3. Oxygen-17 NMR Study of the Rate of Water Exchange from Partially Complexed Nickel Ion 11 4. Research Plans for Calendar Year 1972 11 5. 1971 Publications and Reports 11 Neil Bartlett, Prinoipal Investigator 1. Noble Gas and Halogen Chemistry 12 2. Structural Investigations of High Valence Cations . 13 3. Research Plans for Calendar Year 1972 14 4. 1971 Publications and Reports 15 B. Solid State Chemistry and Physics Norman E. Phillips, Principal Investigator 1. Calorimetric Evidence for a Singlet Ground State in CuFe 16 2. Low Temperature Heat Capacity of Cerium 17 3. Heat Capacity of AuGd Alloys 17 4. Pressure Dependence of the Low Temperature Heat Capacity of V2O3 ...... 18 5. Temperature Measurements 18 6. Dilution Refrigerator 18 7. Low Temperature, High Pressure Calorimetry 18 8. New Adiabatic Demagnetization Cryostat and Calorimeter 19 9. Research Plans for Calendar Year 1972 19 10. 1971 Publications and Reports 19 George Jura, Prinoipal Investigator 1. Pressure (Up to 40 Kbar) as a Parameter in the Study of Vibrational Excition Levels and Crystalline Potentials of Molecular Crystals: Naphthalene and N-Paraffins ..... 23 2. Heat Capacity of Iron as a Function of Pressure and Temperature 21 3. Specific Heats and Heats of Transition of Bismuth as a Function of . Temperature and Pressure 21 4. Research Plans for Calendar Year 1972 22 5. 1971 Publications and Reports 22 Gabor A. Somorjai, Prinoipal Investigatcr 1. Low Energy Electron Diffraction Studies of Surfaces . . , 24 2. Auger Electron Spectroscopy of Surfaces 26 3. Molecular Beam Scattering From Surfaces 27 4. Studies of the Vaporization Kinetics of Hydrogen Bonded Liquids 27 5. Research Plans for Calendar Year 1972 -• 28 6. 1971 Publications and Reports 28 Lee F. Donaghey, Principal Investigator 1. Czochralski Crystal Growth 30 2. Chemical Vapor Deposition 31 3. Chalcogenide Spinel 31
"JU' "Preceding Annual Reports: UCRL-19155, UCRL-20500 -iv-
4. Transport in Solids 31 5. Magnetic Susceptibility Measurements 31 6. Computer Simulation 32 7. Research Plans for Calendar Year 1972 32 8. 1971 Publications and Reports 32 C. Electrochemistry Charles W. Tobias, Principal Investigator 1. Fundamental Studies of Ionic Mass Transport 34 2. Electrolysis at High Current Densities 35 3. Electrochemistry in Nonaqueous Solvents 37 4. Research Plans for Calendar Year 1972 37 5. 1971 Publications and Reports 38 Rolf H. Muller, Principal Investigator 1. Optical Studies of Interfacial Phenomena 39 2. Research Plans for Calendar Year 1972 41 3. 1971 Publications and Reports 42 John Newman, Principal Investigator 1. Diffusion in Aqueous KOH Solutions 43 2. Heat Transfer to Turbulent Gas-Solids Suspension Flow In Pipes 43 3. Mass Transfer to a Plane Below a Rotating Disk 43 4. Diffusion of a Binary Fluid Across sin Initially Sharp Boundary; Bouyancy-Driven Hydrodynamic Instability Due to Heat of Mixing 44 5. Empirical Correlation for Turbulent Couette Flow 44 6. Corrosion of Rotating Disks 44 7. Mass Transfer Coefficient in Porous Flow-Through Electrodes 45 8. Warburg Impedance of a Rotating Jisk Electrode 45 9. Fluid Flow in a Propagating Crack 45 10. Mass Transfer to a Rotating Sphere at High Schmidt Numbers 45 11. Mass Transfer in Electrochemical Systems 45 12. Research Plans foT Calendar Year 1972 46 13. 1971 Publications and Reports 46 D. Physical Chemistry Leo Brewer, Principal Investigator 1. Energies of the Electronic Configurations of the Singly, Doubly, and Triply Ionized Lanthanides and Actinides 48 2. Radiative Lifetimes of TiO and VO 48 3. Thermodynamic Properties of Solutions of Third and Fourth Group Transition Metals with Platinum Group Metals 48 4. Photochemistry of Iodine and Ultraviolet Spectrum of I2 ... 48 5. Research Plans for Calendar Year 1972 48 6. 1971 Publications and Reports 49 Charles B. Harris, Principal Investigator 1. Introduction 50 2. The Dynamics of Coherent Exciton Migration 50 3. The Dynamics of Coherent Triplet Exciton Trapping .... 51 4. Optica.'ly Detected Magnetic Resonance and the Measurement of Intramolecular Energy Transfer Processes in Triplets 51 5. Properties of Molecules in Excited Triplet States 52 6. Research Plans for Calendar Year 1972 53 7. 1971 Publications and Reports 53 Ronald R. Herm, Principal Investigator 1. Photodissociation of Nal Vapor and the Energy Dependence of the Quenching of Na*(^p 2p) by Foreign Gases 55 2. Molecular Beam Kinetics: Reactions of Mg, Ca, and Ba with CI2, Brj, and IC1 55 3. Molecular Beam Kinetics: Reactions of Ca and Ba with HI, NO2, SOj, PCI3, SnCl4, CCI4, CH3I, CH2I2, and 2-Nitropropane ,56 4. Molecular Beam Kinetics: Magnetic Deflection and Product Velocity , Analysis of the Li + Br2 Reaction : 57 5. Research Plans for Calendar Year 1972 1 58 6. 1971 Publications and Reports '59 -v-
Harold S. Johnston, Principal Investigator 1. Introduction 60 2. Catalytic Reduction of Stratospheric Ozone 60 3. Dissociation of Homonuclear Diatomic Molecules at High Temperatures .... 61 4. Molecular Modulation Spectrometry 61 5. Fluorescence of Sulfur Dioxide . 62 6. Research Plans for Calendar Year 1972 62 7. 1971 Publications and Reports 63 Bruce ,'!. Mohan, Principal Investigator 1. Molecular Orbital Correlation Diagrams and Ion-Molecule Reaction Dynamics 64 2. Determination of Intermolecular Potentials from Large Angle Elastic Scattering 64 3. Deconvolution of Molecular Beam Inelastic Scattering Data 65 4. Nonreactive Scattering in the Ne* - H2 and Ne+ - He Systems 65 5. Dynamics of the Reactions of O2 with H2, D2, and Iffl 66 6. Thermal Ion-Molecule Reactions in Gaseous Methane-Oxygen Mixtures 67 7. Research Plans for Calendar Year 1972 67 8. 1971 Publications and Reports 67 William H. Miller, Prinaival Investigator 1. The Classical S-Matrix for Rotational Excitation; Quenching of Quantum Effects in Molecular Collisions . . 68 2. Collisional Quenching of Metastable Hydrogen Atoms 68 3. Molecular Ramsauer-Townsend Effect in Very Low Energy He^-He'1 Scattering 69 4. Classical Path Approximation for the Boltzmann Density Matrix 69 5. The Classical S-Matrix for Linear Reactive Collisions of H + CI2 70 6. Predissociation in the Schumann-Runge Bands of O2 71 7. Research Plans for Calendar Year 1972 71 8. 1971 Publications and Reports 73 Rollie J. Myers, Principal Investigator 1. Spin Relaxation in a-NiS04-6H2C' 74 2. Electron-Electron Double Resonance in Solutions 74 3. Measurement of Transition Metal Ion Relaxation in Solution 75 4. Research Plans for Calendar Year 1972 76 5. 1971 Publications and Reports 76 Kenneth S. Pitser, Principal Investigator 1. Research Plans for Calendar Year 1972 78 2. 1971 Publication 78 E. Nuclear Science Donald R. O'Umder, Principal Investigator 1. An Apparatus for Investigation of Gas-Solid Reactions by Modulated Molecular Beam Mass Spectronetiy 79 2. Reactions of Modulated Molecular Beams with Pyrolytic Graphite 79 3. The Rate of Oxidation of the Basal and Prismatic Surfaces of Pyrolytic Graphite in the Transition Regime Between Chemical and Diffusional Control 80 4. Theory of lie Gas Centrifuge 80 5. Laser Vaporization of Solids 81 6. The Viscosity of Molten Uranium Dioxide 81 1. Ressarch Plans for Calendar Year 1972 82 8. 1971 Publications and Reports 82
METALLURGY A. Crystal Imperfections Jack Washburn, Principal Investigator 1. Field Ion Electron Microscope Observations on Peritectoid Growth of Ni4W 83 2. Transmission Electron Microscopic Studies on Magnesium Oxide 84 3. Atom Probe Field Ion Microscope 86 4. Research Plans for Calendar Year 1972 86 5. 1971 Publications and Reports 87 -vi-
B. Mechanisms of Plastic Deformation John E. Dorn (deceased Sept. 24, 1971) and John W. Morris, Jr., Principal Investigators 1. Research on Mechanisms of Dislocation Glide 88 2. Research on Plastic Deformation 88 3. Research Plans for Calendar Year 1972 90 1. 1971 Publications and Reports 90 C. Thermodynamics of Inhomogeneous Materials John W. Morris, Jr., Principal Investigator 1. Research Plans for Calendar Year 1972 92 2. 1971 Publications and Reports 92 D. Relation of Structure to Properties Gareth Thomte, Principal Investigator Steels 93 1. The Enhancement of Strengthening Dislocated Martensite 94 2. Structure and Mechanical Properties of Fe-Cr-C-Co Steels 95 3. Orientation Dependence on Deformation Induced z and a Martensite in Fe-Ni-Cr Single Crystals 95 Phase Transformations 96 4. Early Stages of Ordering in NijMo 96 5. Field Ion Microscopic Study of Antiparallel Twins in Ni,)Mo 97 6. Electron Metallography of Spinodally Decomposed (Cu-Mn)3 Al Ailovs 99 7. Interphase Interfaces in Spinodal Alloys 102 E. Electron Microscopy Gareth Thonus, Principal Investigator 1. Dislocation Contrast at High Voltages 104 2. Disappearance Voltage Determinations Using a Convergent Beam 107 3. An C/rthogonal Drive Tilting Stage for the Hitachi 650 kV Microscope .... 108 4. High Voltage Electron Microscopy of Interfacial Defects in GaAs 108 5. High Voltage Electron Microscopic Studies of Minerals 110 6. Research Plans for Calendar Year 1972 Ill '!. 1971 Publications and Reports Ill F. High Strength Materials Victor F. Zaokay and Earl Ft. Parker, Principal Investigators Introduction 114 1. Precipitation Studies Using Nuclear Magnetic Resonance 114 2. Prediction of the Stress-Strain Behavior of Polycrystalline a-Iron Containing Hard Spherical Particles 115 3. Tr Design of Low Alloy Carbon-Free Iron-Base Alloys 117 4. Relationships Between Microstructure and Fracture Toughness in a Secondary Hardening Steel 118 5. Untempered Martensite Steels of High Strength and Toughness 119 6. Stability and Mechanical Properties of Metast.able Austenite Steels 120 7. The Effects of High Prior-Deformation Temperatures on TRIP Steels 121 8. Hydrogen Embrittlement of TRIP Steels 122 9. Research Plans for Calendar Year 1972 124 10. 1971 Publications and Reports 124 G. High Field Superconductivity Leo Breuer, Earl R. Parker, Victor F. Zaakay, and Milton R. Piakus, Principal Investigators 1. "Synthesis of Metastable Superconducting A-15 Compounds 125 2. High Field Superconducting Composites by Powder Metallurgy 128 3. Superconductivity in Amorphous Transition Metal Alloys 132 4. Research Plans for Calendar Year 1972 137 5. 1971 Publications and Reports 137 H. Thermodynamics of Metal Systems Ralph R. Hultgren, Principal Investigator 1. Evaluation of Thermodynamic Data 138 2. Low Temperature Heat Capacity of Binary Alloys 138 3. Vapor Pressures of Pb-Pt Alloys 138 4. Heat Contents of Sodium Nitrate and Sodium Nitrite 138 5. Research Plans for Calendar Year 1972 138 6. 1971 Publications and Reports 138 -vii-
I. Advanced Materials Robert H. Bragg, Principal Investigator 1. Structure and Properties of Some Advanced Materials 139 2. Effect of Polarized X-Ray Beam in Small-Angle Scattering by Pyrolytic Graphite 139 3. Structure of Glassy Carbon 140 4. Brittle Fracture in Filament-Reinforced Materials 140 5. Eutectic Solidfication 142 6. Research Plans for Calendar Year 1972 142 7. 1971 Publications and Reports 142 III. Ceramic Science A. High Temperature Reactions Alan W. Searcy, Principal Investigator 1. The Kinetics of Vaporization and of Endothermic Decomposition Reactions 145 2. The Kinetics of Decomposition of Calcite Single Crystals 145 3. Anisotropic Vaporization Rates of Opposite Faces of Zinc Oxide Single Crystals: Experiments and Theory 146 4. The Stabilities of Ce2F6(g) and LajFgCg) 146 5. The Formation of Vapor Phase Nuclei in Nucleate Boiling 148 6. The Vapor Pressures of Gallium Trifluoride Monomer and Dimer 149 7. Research Plans for Calendar Year 1972 150 8. 1971 Publications 151 B. Mlcrostructure and Behavior of Ceramic Materials: Glass and Ceramic- Metal Systems Joseph A. Pask, Principal Investigator 1. Diffusion Kinetics in Mullite (3Al203-2SiO2) and Alumina-Silica Liquids . . 152 2. Decomposition of Mullite 152 3. Capillary Penetration of Silicate Liquids into MgO Compacts . . 152 4. Permeation of Silicates in Magnesia and Forsterite Compacts 153 5. Microstructure Study of Polycrystalline Spinel Developed from Powders Prepared by the Freeze-Drying Technique 153 6. Liquid Phase Sintering Studies 153 7. Effect of LiF on Creep of MgO 154 8. High Temperature Mechanical Behavior of MgO-CaO-MgO-Si02 Materials 154 9. Mechanical Behavior of MgO Solid Solutions 155 10. Fracture of Brittle Transparent Materials 156 11. Interdiffusion and Moving Boundaries in NiO-CaO and NiO-MgO Single-Crystal Couples 156 12. The Effect of Atmosphere on Interfacial Reactions and tfetting Behavior of Glass-Metal Systems . 156
13. Electrochemical Study of the CoO-Na2Si2Os Glass System 157 14. Electrochemical Studies of the Fen Q5O-Na20-Si02 Glass System 157 15. Research Plans for Calendar Year 1972 157 16. 1971 Publications and Reports 158 C. Relation of Microstructure to Properties in Ceramics Richard M. Fulrath, Principal Investigator 1. Relation of Microstructure to Properties in Ceramics 159 2. Non-Stoichiometry of Lead Zirconate Titanate 159 3. Intrinsic Non-Stoichiometry in Single-Phase Lead ZiTconate-Titanate 159 4. Processing of Sodium-Potassium Niobate Ceramics 161 5. The Solubility of Gases in Glass 162 6. Diffusion of Gases in Glass 162 7. Influence of Surface Damage on the Strength of a Brittle Matrix Composite 164 8. Characterization and Sintering Behavior of Barium and Strontium Ferrites 164 9. High Temperature Scanning Electron Microscopy 165 10. Research Plans for Calendar Year 1972 165 11. 1971 Publications and Reports 165 -viii-
IV. SOLID STATE PHYSICS A. Theoretical Solid State Physics Marvin L. Cohen, Principal Investigator 1. Electronic Structure of Semiconductors 167 2. Electronic Structure of Noble and Transition Metals and Their Compounds 167 3. Superconductivity 167 4. Narrow Gap Semiconductors 167 5. Research Plans for Calendar Year 1972 168 6. 1971 Publications and Reports 168 B. Magnetic Properties of Solids Alan M. Portie, Principal Investigator 1. Electron Resonance in Metallic Particles 170 2. Research Plans for Calendar Year 1972 170 3. 1971 Publication 170 C. Far Infrared Properties of Solids Paul L, Richards, Principal Investigator 1. The Josephson Junction as a Phase-Locked Heterodyne Detector 171 2. Generation of Far Infrared Picosecond Light Pulses . 172 3. Theoretical Study of Coherent Phonon Generation by Optical Mixing in a One-Dimensional Superlattice 172 4. Construction of a Polarizing Interferometer for Fourier Spectroscopy 173 5. Ground Based Observations of the Cosmic Background Radiation 173 6. Balloon-Borne Far Infrared Spectrometer 174 7. Vorticity in Superfluid Helium 17S 8. Research Plans for Calendar Year 1972 175 9. 1971 Publications and Reports 175 D. Superconductivity Gene I. Rochlin, Principal Investigator 1. Externally Shunted Oxide Barrier Josephson Junctions: Generalized Weak Links 176 2. Observation of Phonon Structure in the Tunneling Density of States of Superconducting Vanadium 177 3. Thin Film Vanadium Oxides 178 4. Transition from Bulk to Josephson-Junction Behavior in Weak Links 178 5. Scratched Film Quantum Interferometers 179 6. Tunneling Into Very Small Superconducting Normal Sandwiches 179 7. Tunneling in Normal Metal Junctions 179 8. Noise 180 9. Research Plans for Calendar Year 1972 180 10. 1971 Publications and Reports 180 John Clarke, Principal Investigator 1. A New Method for Determining Very Low Superconducting Transition Temperatures 181 2. A Search for Superconductivity in Degenerate Semiconductors 182 3. The Superconducting Fountain Effect , 182 4. A Josephson Junction Bolometer 183 5. Josephson Junction Amplifier 184 6. A New Type of Fermi Surface Measurement 184 7. Research Plans for Calendar Year 1972 185 8. 1971 Publications and Reports 185 E. Experimental Solid State Physics and Quantum Electronics Y. Ron Shen, Principal Investigator 1. Self-Focusing and Small-Scale Filaments of Light 187 2. Phase-Matched Third-Harmonic Generation in Cholesteric Liquid Crystals .... 187 3. Brillouin Scattering in Liquid Crystals 187 4. Raman Study of Smetic Liquid Crystals 188 5. Far Infrared Generation by Optical Mixing 188 6. Wavelength-Modulation Spectra of Solids 188 7. Dispersion of Nonlinear Optical Susceptibilities 188 8. Theoretical Study of Coherent Optical Phonon Generation by Optical Mixing in a One-Dimensional Superlattice 188 9. Research Plans for Calendar Year 1972 189 10. 1971 Publications and Reports 189 -ix-
F. Solid State Spin Resonance and Coherent Laser Pulse Studies E. I. Hahn, Prinoipal Investigator 1. Picosecond Optical Pulse Nonlinear Propagation Effects 190 2. 1971 Publications and Reports 190
APPENDICES INiRD STAFF FOR 1971 191 VISITING SCIENTISTS 197 DEGREES AWARDED IN 1971 198 1971 RESEARCH SEMINARS 201 GENERAL INTRODUCTION
Leo Brewer, Division Head Viator F. Zaokay, Associate Head
The vigor and productivity of the Division and a true scholar who was renowned throughout have remained at the same high level in spite the world. All of his colleagues will sorely of the tightening budgetary situation. One of miss him. the significant departures from the past has hcen the active solicitation of contracts from Two new principal investigators have joined other agencies by the principal investigators the Division during the past year. Professor in an effort to maintain the quality and J. IV. Morris comes to the Division from Dell quantity characteristic of the Division. Dur Aerospace Co., Buffalo, New York where he ing 1971, 36 Ph.D.'s, 24 M.S.'s and two spent several years as u materials scientist Doctor of lingincoring degrees were awarded to and engineer, lie received his doctorate from II WD students, and 140 journal papers or book Massachusetts Institute of Technology and is chapters were published. The total number of regarded as one of the most promising young oust doctorals remained about the same; the theoreticians in the nation in the fields of fi'action caning with their own salary support nucleation, phase transformations, and dis continues to be high. During the year the location interactions. Professor Kenneth aiCTagc number of graduate students was 180. Pitzer returned to the Division this year. He Tie growth of the Division over the past was active in the establishment of the In dh'Zcn years with respect to principal inves organic Materials Research Division and was tigators, students and staff is shown in the tvo graphs below. its first Division Head. He left in 1961 to become President of Rice University and just prior to returning to the Inorganic Materials In spite of the current national economic Research Division he was President of Stanford outlook, there was a continued demand by em University. ployers for qualified students at both the M.S. and Ph.D. levels. In fact, with the Professor Marshall Merriam left the Divi possible exception of several foreign students sion for the post of Assistant Director, at the Ph.D. level, virtually all students ob Technical and Development Institute, Honolulu, tained employment. In some cases, there were Hawaii. not enough students with advanced degrees to supply the requests of potential employers. A number of Inorganic Materials Research Division principal investigators are making The Division lost one of its most prestig- substantial contributions to the University ioxs and productive scientists this past year administration. Professor Charles Tobias is through the death of Professor John E. Dorn. Chairman of the Chemical Engineering Depart Professor Dorn was a major contributor to the ment, Professor Victor Zackay is Associate field of metal plasticity, a revered teacher, Dean and Professor Joseph Pask is Assistant
IMRD GRADUATE STUDENTS AND SUPPORT PERSONNEL IMRO PRINCIPAL INVESTIGATORS
I960 1965 1970 I960 1965 1970 s
Dean of the College of Engineering. Profes SOME HIGHLIGHTS OF THE 1971 RESEARCH PROGRAM sors Gareth Thomas and Robert Connick served on the Chancellor's Staff, the former as The rate of water exchange on Al has been Special Assistant to the Chancellor-Academic redetermined with significant improvement in Affairs and the latter as the Vice Chancellor accuracy. The very high entropy of activation until July 1, 1971. Professor Alan Portis is found previously was shown to be spurious. the Director of the Lawrence Hall of Science. (R. E. Connick, p. 10) Dr. Rolf Muller is Chairman of the Lawrence Berkeley Laboratory Safety Kbview Committee. The catalytic dehydrocyclization of n-heptane to toluene showed remarkable sensi Several Inorganic Materials Research Divi tivity to the surface structure of platinum sion investigators received awards and honors crystals. Stepped surfaces with (111) during the year: crystal orientation are much more reactive Professor Gareth Thomas--a Guggenheim Fellow than stepped surfaces with (100) crystal ship. orientation. Professor Earl Parker--a Fellow of the (G. A. Somorjai, p. 24) American Society of Metals (ASM). Professor Victor Zackay--the Albert Sauveur Based on the reversible electrode behavior Achievement Award of the ASM, the highest of the heavy alkali metals (K, Rb, Cs) found research award in the Society. in propylene-carbonate-AlCl3, a patent dis Professor John Dorn--Distinguished Teaching closure has been made on electrowinning Award, awarded posthumously by the University alkali metals by preparing the amalgams from of California at Berkeley. aqueous chloride solutions and subsequently Professor Leo Brewer--1971 Palladium Medalist "refining" them to the pure state in this from the Electrochemical Society. solvent. Professor Neil Bartlett--Tour speaker of the (C. W. Tobias, p. 34) year award from the Central South Circuit of the American Chemical Society and the Daraie In the ion-molecule reaction dynamics pro Heineman prize for excellence in chemistry. gram, a test of statistical theories of chem Dr. Rolf Muller--Chairman of the San Francis ical reactions, has shown the influence of co Section of the Electrochemical Society. the angular momentum of the reactants on the relative yields of several possible products. Professors Norman Phillips and Harold Wit>: this information, the statistical theory Johnston were on Sabbatical leaves until June -A reactions can become a facile, predictive 30, 1971. Dr. Phillips in Helsinki, Finland tool for the chemist. and Dr. Johnston in residence. (B. H. Mahan, p. 64, see figure)
The Fifth International Materials Symposium Ne*(H»,Ht>m«*l7S.OeVI on The Structure and Properties of Materials-- 12.5 eV Relative Energy Techniques and Applications of Electron Micros copy sponsored by the Inorganic Materials Re search Division and the University of Califor nia at Berkeley, Department of Materials Sci ence and Engineering was held in Berkeley, September 13-17, 1971. An active group of scientists, from this country and abroad, presented papers and held stimulating discus sions. The proceedings of the symposium have been published by the University of California Press.
The Fifth Annual Program Review of the In organic Materials Research Division was held January 26 and 27, 1971 and the Sixth is scheduled for February 2 and 3, 1972. Review ers at the January 1971 Review were: Professor Norman Nachtrieb, University of Contour map of the specific intensity of lie+ Chicago, Professors Oleg Sherby and A. L. scattered from He. Schawlow, both from Stanford University, Professor John Waugh from Massachusetts Insti Available late constant data have been tute of Technology and Professor John Chipman, used to examine the ecological aspects of Emeritus from Massachusetts Institute of wide-spread use of supersonic transports. Technology. In the stratosphere, oxides of nitrogen from SST exhaust pose a threat to the ozone shield. The mechanical behavior of polycrystalline The projected increase in stratospheric oxides magnesia at high temperatures is strongly of nitrogen could reduce the ozone shield hy affected by the distribution of a silicate about a factor of 2, thus permitting the phase along MgO grain boundaries. harsh radiation below 300 nanometers to (J. A. Pask, p. 152, see figure.) permeate the lower atmosphere. (H. S. Johnston, p. 60)
For studies of the mechanisms of disloca tion glide, a numerical code has been written to model the thermally activated motion of an idealized dislocation across a plane contain ing a random array of obstacles. (J. E. Dorn and J. W. Morris, Jr., p. 88)
A new superconducting compound, NbjSi, with the A15 structure, has been synthesized by co-condensation of the elements. Preparation of Nb3Ge and NbjGa by similar means has re sulted in an extension of the solubility of Ge and Ga to the stoichiometric composition (3:1) and a confirmation of the dependence of Tc on the increased order derived from stoi- chiometry. (Brewer, Parker, Zackay, Pickus, p. 125) Strain rate vs. normalized stress for areep A procedure has been developed for the specimens of MgO containing 5% CaO-MgO- Si02 preparation of continuous M^Sn superconduct tested in the temperature range 1200-1450"C. ing tape by use of powder metallurgy techni ques. Mountain-top far infrared measur. Tients (M. R. Pickus, p. 130, see figure.) show that the cosmic background radiation remaining from the creation of the universe does not exist in the form of narrow spectral lines. Far infrared radiation has been generated with a picosecond laser pulse in a nonlinear crystal as a result of beating between its various frequency components. (P. L. Richards, Y. Shen, p. 171, see figure.
Superconducting porous Niobium tape. White: Tin; tight gray: niobium matrix; dark gray: Nbjin. PLATE SEPARATION, mm
The nature of brittle fracture in filament- Fabry-Perot fringes of far infrared radiation reinforced materials has been explored using generated in LiNbO^ by picosecond pulses both an analytical and an experimental ap simultaneously phase-matched at 23.5 arrrl proach . and 6. 7 am-1. (R. H. Bragg, p. 139) -xiv-
Electronic charge density calculations were A new technique has been developed for made fcr Si, Ge, Sn, GaAs, InAs, InSb, ZnSe, measuring the transition temperature of super Cdle, snd PbTe and compared with experimental conductors, even though this temperature may measurements. be far below the working temperature of the (M. L. Cohen, p. 167) experiment. The technique involves measuring the resistance of superconductor-normal- A new type of Josephson effect device--the superconductor junctions, where the normal externally shunted Josephson junction has been metal is the material being studied. The discovered. This device, consisting of a thin theory suggests that transition temperatures f i Ira superconductor - insulator - superconductor as low as 10'400K may be detected. (S-I-S) Jorephson junction of the classical (J. Clarke, p. 181) type shunted with a vacuum deposited silver strip, can be tuned with temperature and/or magnetic field to duplicate the current- voltage characteristics of a host of weak- link Josephson devices. (G. I. Rochlin, p. 176) «*5w=w>)a.wi.''ii,"!.
4 /
1 '
I. Chemistry
*«r 1
INORGANIC CHEMISTRY
William L. Jolly, Principal Investigator energy range. We chose to study compounds of nitrogen because of the wide variety of 1. X-RAY PH0TOELECTR0N SPECTROSCOPY bonding types that they possess and the probability that they would yield a wide a. Chemical Shifts in Core Electron Binding range of chemical shifts. We found that the r^ergies for Some Gaseous Nitrogen Compounds* estimation methods that were most readily tested with our data were the atomic charge Patricia A. Finn, Richard K. Pearson, Jack M. correlation method and the thermodynamic Hollander, and William L. Jolly method.1 In this investigation we studied only gaseous nitrogen compounds. Although X-ray photoeleo. -on spectroscopy is of many successful correlations and estimations great interest to chemists because the meas have been made by using chemical shift data ured core electron binding energies are a for solid compounds, there are both experi function of the chemical environment of the mental and theoretical reasons for preferring atoms--that Is, the binding energies show the study of gaseous compounds. For example, chemical shifts. In order to appraise the by use of gaseous mixturos, it is a simple relative usefulness of various methods of matter to measure chemical shifts between estimating chemical shifts, it is important compounds of the same element from a single to have chemical shifts that span a wide spectrum. Chemical shifts foT gaseous corn-
Table I. Nitrogen Is binding energies for some gaseous compounds.
No. Molecule Relative Pauling CNDO Thermo. binding atomic atomic estimated energy charge charge rel. binding (eV) (eV)
1 CNF, 7.1 1.24 0.70
2 NF3 4.3 0.45 0.30
3 N02 3.0 0.53 0.41 3.3
4 NNO 2.6 0.51 0.52
5 2.4 0.32 0.23 2.8 N2F4 6 0NC1 1.5 0.05 0.17 7 NO 0.8 -0.37 0.05 0.9
8 N 0.0 0.00 0.00 0.0 2 9 NNO -1.5 -0.34 -0.24 -0.9 10 HCN -3.1 -0.06 -0.10 -2.6
N H -3.8 -0.14 1J 2 4 -0.26 12 Nllj -4.3 -0.39 -0.26 -3.5
13 oy«2 -4.8 -0.32 -0.21 -3.9
14 (CH3)2NH -5.0 -0.23 -0.17 -4.2
IS (CH3)3N -5.2 -0.15 -0.14 pounds do not suffer from uncertainties of can be estimated. A plot of experimental vs. work functions which arise in the case of thtiTncdynamically estimated binding energies solid compounds. Both theoretical and empir is given in Fig. 1. ical calculations of core electron binding energies are much simpler for gaseous mole cules than for solid compounds. Abstracted from Inorg. Chem. 1£, 378 (1971). 1. See I. M. Hollander and W. L. Jolly, The measured nitrogen Is binding energies, Accounts Chem. Res. 3, 193 (1970). relative to molecular nitrogen, are given in Table I. The values span a total range of 12.3 eV. b. X-ray Photoelectron Spectra of Nitrosyl The Pauling atomic charge method, although and Nitrogen Complexes" it gives very rough correlations with binding energy, has the advantage that it can be ap Patricia A. Finn, Jack M. Hollander, William plied to any set of molecules for which com I. Jolly, Philip C. Douglas,* Robert D, plete octet structures can be written. Fcltham,* C, Dolcctti,+ P. Fan>ham,+ and Although the CNDO atomic charge method gives James Collman* a slightly better correlation than the Pauling method, many chemists may consider this The nitrogen Is binding energies of some advantage to be outweighed by the relative transition metal nitrosyl complexes and difficulty and expense of the CNDO calcula dinitrogen complexes were determined by x-ray tions. The thermodynamic method gives more photoelectron spectrometry. Most of the accurate predictions than the atomic charge nitrosyl binding energies were in the range method, and it is completely nonempirical, at 399.6 - 400.7 eV. Huwever, higher binding least with respect to binding energies. energies were observed for the following com However, it is applicable only when the pounds: [Co(das)2NO](C104)2, 407.3 eV; appropriate heats of formation are known or [Rh(PPh3),(NO)2]PF6, 401.3 eV;
Ru(PPh3j2Cl2(NO)2, 400.2 and 402.6 eV;
Na2[Fe(CN)5NO]- 2H20, 403.3 eV. The higher binding energies probably correspond to coordinated NO* groups; the lower Hnding energies probably correspond either to NO* groups with extensive back-bonding from the metal atoms or to NO" and NO groups. The two arc con peaks observed for Ru (PPI13) ;>C12 0*>) 2 " sisteut with its structure, which has non- UJ equivalent NO groups. Two peaks of equal intensity at 400.7 and 402.3 eV were observed
for nans-[RuCl(das)2N,]PF0, and a single peak at 400.1 eV was 'observed for N {[RuCl(das)2]2 2}CBPh4)2- These data indicate that the lower-energy peak in the former com plex corresponds to the directly coordinated N atom. The lower binding eneTgy corresponds to a higher electron density, probably a consequence of back-bonding from the ruthenium atom.
-5 -4-3-2-1 0 I 2 3 Department of Chemistry, University of Arizona, Tucson, Arizona. Thermo. Estd. Rel. Binding Energy, eV Department of Chemistry, Stanford Univ.-sity, Stanford, California. Fig. 1. Comparison of experimental and thermodynamically estimated binding energies. The line has a slope of unity and passes through the origin. The numbers refer to 2. ST (DIES OF TIE HYDRIDES OF CROUPS IV AND V the compounds in Table I. The line would fit the points much more closely if point 9 or a. Th Synthesis and Characterization of the 10 (rather than 8) had been arbitrarily Halost Wanes* placed at the origin. It is significant that most cf the points fit a line of unit slope. John R Webster, Merle M. Millard, and William XBL 706-10% L. Jol y Stannane, S11H4, reacts with HC1, HBr, and in >x> different ways. The reaction whi.rh HI at -112° to give the monohalostannanes produces an organogemane undoubtedly Involves and hydrogen. The monohnlostannanes are nucleophilic attack by a gernfl anion on a colorless volatile compounds that decompose as carbn atom, with simultaneous halide ior solids above about -40 . They have been displacement. The reaction which produces characterized by the study of their decomposi a hydrocarbon and GeHj polymer probably tion reactions (which afforded chemical involves nucleophilic attack by a gernyl uiion analyses), by mass spectrometry, by infrared on a halogen atom, with simultaneous or spectrometry, and by Raman spectrometry. The subseq.ient formation of a C-H bond tine' cleav Raman spectra of the solids suggest structures age of a Ge-H bond. A possible mechanism for involving relatively independent SnHj* the latter, hydrogen-substitution, reaction cation?. is
RX + Gellv R X'Gellj* RH • X >• (1/n) (GelL) Abstracted from Inorg. Chem. j£, 879 0971). 2Jrf Halogen worn attack riechanisms have been pro posed for muiy '/suctions of organcqihosphincs
1 and for tin' reaction of organic hulides with b. The SnH- Ion in Acid Sc ions* it potassium triphonylstlyl and socttum triphenylgci-nryl. John R. Webster and William L. Jolly The hydrogen-substitution reaction pre Stannane reacts slowly with strong aqueous dominates in -.ases where one would expect acids below -SO" to give one mole of hydrogen halidc ion di.«,pljtcesnent by the germyl anion per mole of stannane. Ihe resulting solutions to be slow. Consequently, high yields of the evolve an additional two moles of hydrogen hydrogen-subptiluted products are obtained in as they are warmed to room temperature. the reactions with aromatic halides, neopentyl Stannane undergoes a similar, but relatively and neophyl halides, and dihalomethanes. rapid, reaction with anhydrous fluorosulfuric However, even in the case of iodomethane, a acid at -78°; when the resulting solution is trace of the hydrogen-substituted product, warmed to room temperature, the solvent is methane, is obtained. It appears that the two reduced but no hydrogen is evolved. Con types of reaction always occur simultaneously. ductivity studies of the cold undecomposed Perhaps a four-center activated complex is fluorosulfuric acid solutions indicate the formed; such a mechanism is almost equivalent formation of one male of fluorosulfate ion to a cage reaction in which the phenyl anion per male of stannane reacted, and H9sn NMl reacts with broraofjermane before it can be spectra of such solutions show a 1:3:3:1 dispersed and reict with the solvent. This quartet. These results are interpreted in type of nEchanisai is terms of the solvated stannonium ion, SnHj*.
Abstracted from Inorg. Chem. 10.. 877 (1971). RX • GeH, a/' ;>"2
RH • X" + (l/n)(GeH )n c. The Reaction of Potassium Gerroyl with 2 Organic Halides*
Robert M. Dreyfuss and William L. Jolly Abstracter from Inorg. Chem. 10_, 2567 (1971).
Many organic halides react with potassium gerayl in diglyme to give the corresponding organogeimanes. However, halides for which d. The Reaction of Potassium Dihydrogen nucleophilic substitutions are slow react Arsenide wit'n'RaTobenzenes*' with potassium germyl to give principally the corresponding hydrocarbons. Data relevant to Harng-Shen 5.'ito and William L. Jolly the mechanism of the latter type of reaction are presented. A method for the preparation Nucleophiiic attack on the halogen atoms of and standardization of s'cock solutions of organic halides constitutes a ••all-documented potassium gemyl is described. A new com class of rjctctions. Evidence for such attack pound, digermylmethane (1,3-digermapropane), by various nucleophilic reagents, including is reported, and its f*«, mass, and infrared the triphenylsilyl anion, germyl anions, and spectra are discussed. phosphines, has been cite-.!. In this paper we report the results of a study of the Potassium germyl reacts with organic halides reaction of diglyme solutions of potassium 4
dihydrogen arsenide (KAsiy with iodo-, bromo- The acid-strengthening effect of the phenyl
and chlorobenzcne. Of particular interest, group in CgllsMHx is due (in addition to a from a synthetic point of view, were the relatively weak a inductive effect) to a relative yields of benzene (from attack on stabilization of the anion by interaction of nulogcn atoms) and phenylarsines ffrom cou an empty pir molecular orbital of the phenyl pling reactions). ring with a filled pir "nonbonding" orbital of the M atom. The effect is operative in the Potassium dihydrogen arsenide reacts with phenyl derivatives of methane and of the arylhalides at .jnbient temperature to give hydrides of elements of groups V and VI. For benzene, tetraphenyldiarsine, and example, in the case of phenylarsine, triphenylarsinc. The yield of benzene ranged resonance stiuctures of the following type from 74$ in the case of iodobenzene to 0% in can be written for the anion the case of chlorobenzene. The normal cou pling product, triphcnylarsinc, was obtained (in low yields) only from chlorobcnzenc and bfomobciizene.
The acid-strengthening effect is inoperative Abstracted from J. Organomctal. Chem, 3_0, in phcnylgermane because the lone-pair orbital 361 (1971). on the germanium atom of the CfjUcGelb ion (like that on the arsenic atom of the isoelec- tronic CgHjAsH?) has little p character. Consequently there can be no significant pn- e. The Acidities of Germane and the piT interaction between the phenyl ring and the Phenylgermanes* gc-manium atom in the anion without an energetically unfavorable hybridization of the William L. Jolly germanium orbitals. Generally, a phenyl derivative of a nonmetal Abstracted from inorg. Chem. 1£, 2364 (1971). hydride, Cf,IW3lx.j, is more acidic than the
Darent hydride, MHX. For example, phenylarsine is a stronger acid than arsine, and diphenyl- arsine is a stronger acid than phenylarsine. However, germane and the phenylgermanes con f. Reactions of Potassium Germyl stitute a puzzling exception to this rule. The phenylgermanes are weaker acids than germane-- Chris Riddle, William L. Jolly, Esther A. the weakness increasing with increasing phenyl Staricco, and Eugene C. Jekel substitution. The purpose of this communica Solutions of the geraiyl anion, GeH ", in tion is to show that the data may be explained 7 in terras of two opposing effects of phenyl mono- and diglyme have continued to Ee used substitution: an acid-weakening effect (which extensively in our study of the chemistry we ascribe to pTt-dr bonding) and an acid- associated with Group IV-A hydride species. strengthening effect (mainly due to pu-pir Attempts to prepare specific germyl-containing bonding). compounds have been frustrated by the multi plicity of reaction pathways available to the germyl ion. The phenyl group in the molecule CAHgMHg can exert an acid-weakening effect if the atom For example, germyl iodide, Gef^I, was not M has empty valence dir orbitals which interact with the filled pir molecular orbitals of the isolated as a product of the reaction between phenyl ring. Such dn orbitals are available iodine and potassium germyl. We postulate when M is from the second or third row of the the intermediate formation of GeHjI. periodic table. The acid-weakening effect This reacts further with gerrayl ion, predomi raay be ascribed to a stabilization of the nantly according to equation, to form germane acid, relative to its conjugate base. There and the polymeric hydride (GeH2)x; some is little such stabilization of the conjugate digermane is also formed. base because the pir-dir interaction would cause a shift o^ electron density toward an GeH3" + I2 - GeHjI + i" , atom with a forr,\l charge of 1-. For the cases of phenylgermane and phenylarsine, the effect probably has a similar magnitude, and GeHjI + GeH3" - GeH4 + 1/x (GeH2)x + i" . contributing resonance structures of the fol lowing type can be written The strongly basic nature of the germyl anion is further demonstrated in its reactions 0G5ll3 +G-4iH2 with simple tiivalent boron compounds. With BX3 (X = I, CI, OCH3) we believe the initial acidity in aqueous solution exhibit acidic product to be an etnerate of JC* (GeH3BX3)". behavior in liquid ammonia and can be titrated Polymerization resulted when we attempted to with strong bases such as potassium amide. isolate these species or their neutral However, in spite of its many potential derivatives. applications, acid-base titrimetry in liquid ammonia has been neglected--mainly because of Hydroborate, Bfy", forms diborane when the lack of a simple method of end-point added to 854 phosphoric acid. When detection. End points of acid-base titrations germyltrihydroborate, GeHjBHj", was added to in liquid ammonia have been detected using phosphoric acid, sulphuric acid, or boron the glass electrode, the hydrogen electrode, trichloride there was no evidence to suggest other potentiometric methods, conductivity similar loss of hydride ion; instead, the measurements, and acid-base indicators. We Ge-B bond was cleaved. believe that, of these methods, the use of acid-base indicators is the simplest. How Hie tris(pyridine)hydroboron(III) dication, ever, the intelligent use of indicators 2 pyjBH *, proved stable to attack by GeH3" requires a knowledge of their pK values (i.e., while trimethylaraine dibromoborane, the p!l values corresponding to their color MejN:BHBr2, gave a complex reaction involving changes), and pK values are known for only a the formation of germanium chains. few indicators in liquid ammonia. Thus our main objective in this study was to evaluate We believe diethylchloroborane, EtjBCl, approximate pK values for a group of indica reacts with GeH?" to form IC(GeH3B(CI)Et2)": tors spanning, as far as possible, the pH etherate. Similar products result when the scale available in ammonia. pyridine adducts, py:Et2BCl and py:BCl3, are used. These reactions will be studied Other problems to be solved were those of further in an attempt to isolate devising simple methods for obtaining pure diethylgermylboron, Et2BGeH3, its pyridine liquid ammonia and for gradually mixing the adduct, or that of the trigermylboron species. acidic and basic reagents. Many published methods for carrying out titrations in liquid Reaction of potassium germyl with a series ammonia require special equipment such as of organo-fluoro-compounds is being investi high-pressure apparatus, a vacuum line, or a gated. Potassium germyl reacts 1:1 with buret for use with cold liquid ammonia solu trifluoroacetylchloride, CFjCOCl, to give tions. In the present study, we purposely potassium chloride and a clear, involatile avoided the use of anything more complicated mixture of polymers. The initial product is than an unsilvered dewar flask fitted with a presumed to be CF3COGeH3 which polymerizes dry ice-cooled condenser. With this simple to (CF3CGeH3)C-0)n. equipment we have carried out a variety of practical acid-base titrations tliat illustrate We have found that K+GeH," and boron the usefulness of a set of indicators of trimethyl react rapidly to form a 1:1 adduct, known pK values. GeH " + B(CH ) + GeH B(CH ) " . 3 3 3 3 3 3 Abstracted from J. Inorg. Chem. 33, 783 Disproportionation then occurs to form the (1971). ~ analogous digermanyl species and the trimethylhydroborate ion,
2 GeH3B(CH3;)3' - Ge^B(CHJ ' + HB(CH3)3". b. Coulometric Potentiometric Titrations of Weak Acids in Liquid Ammonia*' Further work on this reaction is in progress. William L. Jolly and Edward A. Boyle
Liquid ammonia is a strongly basic solvent in which many substances chat show no ap 3. STUDIES IN LIQUID AMWJIA preciable acidity in aqueous solution behave as moderately strong acids. Thus liquid a. The Use of Indicators in Acid-Base ammonia potentially is a useful solvent for Titrations in Lfquid Ammonia* carrying out acid-base titrations for the quantitative estimation of such substances. William L. Jolly, Edward A. Boyle, and However, this analytical technique has been Maurice L. Javet neglected because of difficulties inherent in the preparation and manipulation of Because ammonia is much more basic than standard solutions which boil below room water, many weak acids that have no detectable temperature. (Liquid ammonia boils at 6
-33.4°C.) We have avoided these volumetric Our experimental results can be explained difficulties by employing a coulometric in terms of a mechanism involving the follow method for generating the titrant. The method ing initial steps. (Here M denotes an active has the added advantage of applicability to site on the catalyst surface.) very small amounts of reagents.
2M + H2 * 2MH A technique has been developed for deter mining very weak protonic acids [in 0.05 - NH =» M 2.0 meq amounts) with an accuracy and preci MH 2 NH, sion of ± 1.5%. The acid is dissolved in liquid ammonia and is coulometrically titrated M ••= M + e with ammoniated electrons, generated at a platinum cathode. The endpoint is detected Subsequent steps are uncertain; they involve potentiometrically, using a platinum wiTe the substrateaccepting either electrons electrode, which serves as a crude hydrogen (from M ore J or hydride ions (from MH ). electrode, and a cationic glass reference Although the details of the overall hydrogona- electrode. The technique has been successful tion mechanism are uncertain, it seems very ly applied to amnonium chloride, boron tri- likely that the first two steps (the bonding fluoride ammine, guanidinium thiocyanate, of hydrogen atoms to the catalyst and the sufaraide, water, urea, and diphenylamine. formation of a low oxidation state form of the catalyst) are involved. The facts that these Abstracted from Anal• Chem., 43, 514 (1971). steps fit the data for both reaction (1) and the hydrogenations and that they are common features of hydrogenations are strong support for their involvement. c. The Catalyzed Hydrogenation of Tolane and Stilbene in Liquid Ammonia* Abstracted from J. Org. Chem. 36, 3649 (1971). Kenneth A. Strom and William L. Jolly
The reaction between the ammoniated electron d. The Preparation of Copper Hydride Species and liquid ammonia 1 Kenneth A. Strom and William L. Joll) NH, NH, 2 H2 (1) We are continuing our study ol the compound obtained from the reaction of hydrogen with has been shown to be reversible in the pre x Cu(NH2)x in liquid ammonia. The compound sence of certain solid catalysts and to have burns in air and slowly decomposes in vacuo an equilibrium constant of 5 x 10^ atrrr'' t a at -40°. Analysis gives the approximate room temperature. There are many compounds composition KCujHj-xNHj, where x ranges be- which are highly reactive toward ammoniated tween0.2and 0.6. The compound is diamagnetic, electrons, but which are essentially inert as expected for cuprous compounds. It does toward molecular hydrogen. We therefore felt not dissolve in cyclohexene or pyridine. that a liquid ammonia system containing Reaction of the compound with D 0 or DC1 potassium amide and a catalyst for reaction 2 yields H2(55-60t), HD(40-4S%), and D,(l-5%). (1) might serve as a useful medium for The hydrogens appear to be weakly hydridic. "activating" molecular hydrogen. We have Partial thermal decomposition of the compound studied the reactions of stilbene produces a residue which reacts with D 0 to (1,2-diphenylethylene) and tolane (diphenyl- 2 yield H, (32%), HD (524), and D2 (16*). The acetylene) in this system at -45° and at room higher HD percentage indicates that the temperature. Both compounds are Known to residue is more hydridic than the original undergo reduction by the ammoniated electron, material. and we hoped that reaction sequences such as the following (illustrated by stilbene) would take place: 4. BORON HYDRIDE CHEMISTRY H, + 2NH, * 2e , + 2NH, 2 2 amm 3 Francis T. Wang, Patricia A. Finn, John D. Illige, and William L. Jolly 2e" + C,H CHCHC,H + 2NH, •> am 6 C5 6 5C 3 Our studies of the intermediates in the hy-
C H CH CH C H • 2NH drolyses of BH4" and B2H(j are near completion. 6 5 2 2 6 5 2 We have identified, and studied the kinetic + reactivities of, H2OBH3, BH3OH", BH2(H20)2 , mA H2OBH2OH, BH2(OH)2". H2OBH(OH)2, BH(ai)3". b. Reactions of Heptasulfur Imide The asymmetric cleavage of diborane by water, + to yield BH2(H20)2 and BH4", has been proved. Marshall H. Mendelsohn and William L. Jolly
We have determined the maximum yield of Investigations of the reaction of Na[BHjN3] with N0+ are continuing, utilizing NOC1 as a S7N" from the reaction of S7NH with potassium source of nitrosyl cation. A product corre flnorenyl to be 80%. This was found by quantitatively determining the amount of sponding to BH3N2 or BH3N4O may exist at low temperatures. Studies of the aqueous S7NCH3 formed upon addition of CH-,1 to the chemistry of Na[BHjNj] are precluded by the solutions. By the same analytical technique imnediate rapid evolution of hydrogen upon we have found that solutions of S7N" can be the addition of BH3N3" to aqueous solutions quantitatively prepared at -63° from KOH in (even 10 M sodium hydroxide]. THF by stirring the reaction mixture for IS ± S min. The same procedure followed with LiOII does not give quantitative production of the heptasulfur imide anion. 5. SULFUR-NITROGEN CHEMISTRY K+S7N" solutions in THF react with carbon a. The Reaction of Tetrasulfur Tetranitride dioxide to give a white solid which pre with Sulfuric Acid* cipitates from solution. The solid, pre + sumably K S7NC02", gives off C02 when treated Steven A. Lipp and William L. Jolly with 6 M HC1, but we were unable to character ize the solid further because of its The stable products of the reaction of instability. We were able to prepare and tetrasulfur tetranitride with 100* sulfuric characterize two derivatives of the general acid are sulfur dioxide, sulfamic acid, formula S7NCC7R by the reaction of S7N" with bisulfate ion, disulfate ion, ammonium ion, C1C02R, where R = CH3CH2" and C«HcCH2"- These and an inidentified cation which has an new compounds were characterized by chemical intense UV and a complex visible absorption analysis, *H NMR, infrared, and mass spectros spectrum. The approximate stoichiometry of copy. the overall reaction has been determined by a combination of analytical techniques. The We have also made a preliminary investiga analytical data can be summarized in the tion of solutions of S4N4 and aluminum form of the following equation for the net trichloride in nitromethane. We have observed reaction of S4N4 with sulfuric acid: the species S^* in an ESR experiment and the + species NH4 in an NMR experiment. The UV-vis S N + 6.8H S0 ->• 2.0SO + I.SNHJSOJ + spectrum of the solution shows peaks at 4650A, 4 4 2 4 2 3900A, and 3260A. Possibly the unidentified cationic species mentioned in part A (above) + l.SNH4 + 0.1S + ~1.6HS20?" + ~0.4HSO4" + is present in these solutions.
+ ~0.3H2S2O7 + [O.SS6N2OgH J. 6. RESEARCH PLANS FOR CALENDAR YEAR 1972 It should be emphasized that the coefficients for HS207", HSO4", and H2S2O7 are very inaccurately known. The last species in the William L. Jolly equation was empirically formulated to balance the atoms and charges. In view of a. X-Ray Photoelectron Spectroscopy the uncertainties in the analytical data, particularly in the yields of HS2O7", HSO4", We shall complete current studies of core and H2S2O7, the formula o'f this species has electron binding energy shifts in (a) gaseous practically no significance. We do know boron compounds, (b) transition metal nitrosyl that an unidentified cationic species, pro complexes, and (c) azide, dinitrogen, and bably containing at least one SN linkage, is nitride complexes of transition metals. formed in high yield. (If less than 0.5 mol Methods for predicting chemical shifts are of this species forms per mole of S4N4, its being devised and perfected. In this con UV molar extinction coefficient exceeds nection, we are working on a method for the 54,000 M"*--a value which is unusually high.) prediction of gas-phase Lewis acid-basv Clearly the reaction of S4N4 with sulfuric dissociation heats; this will be useful in acid cannot be said to be understood until applications of the "thermodynamic" method this important species is identified. of chemical shift prediction. We are also working on a scheme for estimating the charges on atoms in compounds; these charges can be Abstracted from Inorg. Chem. 10_, 33 (1971). used to predict chemical shifts. 8
b. Studies of the Hydrides of Groups IV and V 2. P. A. Finn, R. Pearson, J. M. Hollander, and W. L. Jolly, Chemical Shifts in Core Attempts tj prepare fluorocarbon derivatives Electron Binding Energies for Some Gaseous of germane will be continued. The possibility Nitrogen Compounds, Inorg. Chem. 10, 378 of using transition metal complexes as pre (1971). cursors in the formation of GeH3" substituted ketones will be studied. To avoid some of the 3. W. L. Jolly and E. Boyle, Coulometric difficulties of working with the GeHj" ion, Potentiometric Titrations of Weak Acids in the Ge(CHj)3" ion will be used as a reagent Liquid Ammonia, Anal. Chem. 43_, S14 (1971). for preparing some germanium compounds. 4. J. Webster and W. L. Jolly, The Stannonium Ion in Acid Solutions, Inorg. Chem. 10, 877 (1971). c. Studies in Liquid Armenia 5. .'. Webster, M. Millard, and W. L. Jolly, lx Iho synthesis and Charactorization of the The reaction of hydrogen with Cu(NH2)x in ammonia to yield a copper hydride species llalostannnncs, Inorg. Chem. 1£, 879 (1971). corresponds to activation of hydrogen by the strong base NH?". Other strong bases have 6. W. L. Jolly, E. Boyle, and M. Javet, Th° been used by other workers. Efforts will be Use of Indicators in Acid-Base Titrations in made to develop a general base-catalyzed Liqi 1:d Ammonia, J. Inorg. Nucl. Chem. 33, 783 procedure for the preparation of metal (19 V O. hydrides from molecular hydrogen. Preliminary results indicate that at 0° CuH is obtained 7. H. S. Kuo and W. L. Jolly, The Reaction in pyridine by the reaction of Cul with H2 of Potassium Dihydrogen Arsenide with at 2000 psi, using potassium t-butoxide as the Halobenzenes, J. Organometal. Chem. 30, 361 base. (1971).
8. R. M. Dreyfuss and W. L. Jolly, The Reaction of Potassium Germyl with Organic d. Boron Hydride Chemistry Halides, Inorg. Chem. 10, 2S67 (1971).
We shall attempt to quench some of the P. W. L. Joll/, The Acidities of Germane and reactive intermediates formed in the the Phenylgermanes, Inorg. Chem. 10, 2364 hydrolysis of BH4" by treatment of solutions (1971). containing these intermediates with amines, ammonia, and hydroxide ion. Possibly a 10. K. A. Strom and W. L. Jolly, The Cata continuous-flow reactor can be devised for lyzed Hydrogenation of Tolane and Stilbene preparing such adducts on a large scale. in Liquid Ammonia, J. Org. Chem. 36, 3649 (1971).
11. W. L. Jolly, Encyclopaedia Britannlca, e. Sulfur-Nitrogen Chemistry 1971 ed., "Selenium."
Further synthetic applications of lC'SyN" 12. W. L. Jolly, Encyclopaedia Britannica, will be pursued. Possibly the SyN'group can 1971 ed., "Tellurium." be attached to transition metals. We hope to clarify the nature of the interaction between Papers presented S4N4, aluminum chloride, and nitronethane. 1. W. L. Jolly, "Recent Studies of Sulfur- Nitrogen Compounds," Symposium on Theoretical Chemistry and Sulfur Chemistry, New Orleans, La., Feb. 18, 1971 (UCRL-2053S).
7. 1971 PUBLICATIONS AND REPORTS 2. W. L. Jolly, "Electron Pair Repulsion Theory of Structure" and "X-Ray Photoelectron William L. Jolly and Associates Spectroscopy," Stanislaus State College, Turlock, Calif., Jan. 1971 (A.C.S. Visiting Journals and books Scientists Program).
1. S. A. Lipp and W. L. Jolly, The Reaction 3. W. L. Jolly, "Intermediates in the Hydrol of Tetrasulfiir Tetranitride with Sulfuric ysis of Borohydride and Diborane," "X-Ray Acid, Inorg. Chem. 10, 33 (1971). Photoelectron Spectroscopy," and 'Adduct 9
Formation and Displacement Reactions of the 4. P. A. Finn and W. L. Jolly, The Nitrogen Germyl Ion, Ge^"," Distinguished Scientists Is Binding Energies of Transition Metal Lectureship, Michigan State University, East Nitrosyls, L3L-122, July 1971. Lansing, Mich., Nov. 8-12, 1971. 5. W. L. Jolly, The Estimation of Core 4. W. I. Jollv, "X-Ray Photoelectron Electron Binding Energy Shifts Using the Spectroscopy," Departmental Seminar, Purdue Concept of the Equivalence of Equally Charged University, Lafayette, Indiana, May 6, 1971; Cores, LBL-182, Sept. 1971. Orange County, Calif. A.C.S. Section Talk, Feb., 1971; Lockheed Research Laboratory 6. P. A. Finn and W. L. Jjlly, NitTOgen Is Research Colloquium, March 11, 1971; Binding Energies of Some Azide, Dinitrogen and Departmental Seminar, Iowa State University, Nitride Complexes of Transition Metals, Ames, Iowa, Jan. 22, 1971. LBL-186, Sept. 1971.
5. N. L. Jolly. "Chemical Shifts in Core- 7. P. A. Finn and William L. Jolly, The Level Binding Energies," International Nitrogen Is Binding Energies of Transition Conference on Electron Spectroscopy, Pacific Metal Nitrosyls, LBL-122, Sept. 1971. Grove, Calif., Sept. 9, 1971 (LBL-182). 8. N. H. Mendelsohn and W. I. Jolly, The UCRL and LBL reports Reaction of Heptasulfur Imide with Diborane, LBL-446, Dec. 1971. 1. R. M. Dreyfuss, Reaction of Potassium Germyl (Ph.D. thesis), UCRL-20502, Jan. 1971. 9. F. T. Wang and W. L. Jolly, A Kinetic 2. W. L. Jolly, The Structures of Sulfur- Study of the Intermediates in the Hydrolysis Nitrogen Compounds, UCRL-20535, March 1971. of the Hydroborate Ion, LBL-461, Dec. 1971.
3. W. L. Jolly, J. D. Illige and M. K. 10. P. A. Finn, A Study of Diborane Dihydrate Mendelsohn, The Prediction of Gas-Phase Lewis and X-Ray Photoelectron Spectroscopy of Acid-Base Dissociation Heats with an Empirical Inorganic Compounds (Ph.D. thesis), LBL-41S, Four-Parameter Equation, LBL-123, July 1971. Nov. 1971. 10
Robert E. Cormiok, Principal Investigator sphere. " Taube and co-workers have in vestigated the temperature dependence of the 1. WATER EXCHANGE FROM THE FIRST COORDINATION 170 line width in a solution of Cr3* in HiO.6 SPHERE OF SOME DIAMAGNETIC IONS Alei has measured the chemical shift of the bulk water resonance in a scries of solutions James W. Neely of Ci3* in water.7 Since the water molecules in the first coordination sphcTC of CrCV^O)^3* The studies of water exchange from the arc substitution inert on on NMR time scale, first coordination sphere of diamagnetic ions both Taubc's and Alei's data imply a definite was continued during the first half of 1971. interaction between the d electrons of Cr3* By using manganous ion to eliminate the bulk and water molecules beyond those in the first water signal through broadening, the rates of sphere. Taubo and co-workers wore unable to exchange of bound waters could be observed determine the oxact mechanism for second directly in favorable cases. The exchange sphere relaxation nor were they able to deter from aluminum ion was restudicd from 0 to mine the lifetimes of the wator molecules in 140°C because of the large entropy of activa the second sphere. tion found earlier by less precise measure ments. The new values of &!* and t& sre A systematic study of both the transverse 15.6 kcal and 0 e.u. respectively and the and longitudinal relaxation times of the extrapolated first order rate constant at 25° oxygen-17 in the bulk water of an aqueous is - 16 sec"1. solution of CrOljOJfi has been performed as a function of temperature. The transverse The previously reported value for 0 relaxation times were obtained from the half- exchange in solutions of thorium ion was shown width of the NMR adsorption. The longitudinal by its acidity dependence to arise from an relaxation times were obtained by progressive unidentified hydrolyzed species of Th(IV). rf power saturation of the I#R absorption.
Probably the observed oxygen-17 is in the Both Ti anc' T2 when plotted as a function of form of bridging hydroxide or oxide groups. ('!', °Kj"l on a semilog plot give approxinatcly The unhydrolyzed Th apparently exchanges straight lines. The Tj of the paramagnetic too rapidly to be measured by the present solution is very close to the Tj of a solution 3 technique. of pure wator, but the T2 of the Cr(H20)6 *
solution is appreciably shorter than the T2 Solutions of In3* showed the presence of of pure water. A detailed treatment of the bound oxygen-17 having a fast but measurable data using conventional nuclear magnetic exchange rate. The acidity dependence showed relaxation theory leads to the conclusion that a hydrolyzed indium species was responsi that the mechanism for the relaxation of the ble. The water exchange of unhydrolyzed In3* water molecules in the second sphere is appears to be too fast to detect. primarily scalar coupling. Using Alci's data, a scalar coupling constant has been calculated. Comparison of all known rates of water From this coupling constant, from Melton and exchange from the first coordination sphere Pollak's9 determination of the electronic of metal ions has been made in terms of a relaxation times of Cr3*, and from the assump simple electrostatic model involving the tion that there are eight water molecules in charge and radius of the metal ion. The the second sphere [one on each face of the results for ions having no crystal field ef Cr(H20)g3* octahedron], one can calculate fects show a reasonable correlation. 10 2.9 x 10" sec as the lifetime of H20 in the second sphere at 2S°C.
Due to experimental difficulties the data 2. OXYGEN-17 NMR STUDIES OF WATERS IN THE used in this study are loss precise than SECOND COORDINATION SPHERE OF Cri+ desired. At present several experimental modifications arc being nnde so the experi Willium L. Earl ment can be duplicated with higher accuracy and precision, after which it is hoped that In several studies of the transverse re second sphere relaxation can he studied in laxation times of solvent molecules in solu other paramagnetic systems. tions of paramagnetic metal ions, low tempera ture broadening has been ascribed to inter actions between the electronic system of the 1. R. E. Connick and K. n. Stover. J. Phvs. paramagnetic ion and the nuclei of solvent Chem. 65, 2075 (1961). molecules beyond the first coordination £'. Z.Tiiz and S. Meiboom, J. Chem. Phys.
r 11
40, 2686 (1964). Chom. Soc. 74, 5052 (1952). 37 J. C. Sheppard and J. L. Burdett, Inorg. 3. S. Meiboom, J. Chem. Phys. 34, 375 (1961). Chum. S, 921 (1966). 4. J. S. Babiec, Jr., C, II. Lang ford, and T. R. Stcingle, Inorg. Che*. 5, 1362 (1966). 4. RESEARCH PLANS FOR CALENDAR YEAR 1972 5. K. Mithrich and R. E. Connick, Inorg. Chem. 6, 583 (1967). Robert E. Connick 6. M. V. Olson, Y. Kanazawa, and II. Taubc, J. Chem. Phys. SI, 289 (1969). It is hoped that some improvement in exper 7. M. Alci, Jr77 InoTg. Chen. 3, 44 (1964). imental conditions can be made that will lead 8. J. P. Hunt and il. Taubc, J.~Chcm. Phys. to measured values of water exchange rates 18, 7S7 (19S0). (or improved lower limits) for Sc3*, In3*, 57 B. Melton and V. Pollnk, J. Phys. Chew. Th4*, and Zn2+. At the present an additional 73, 3669 (1969). factor of 5 in sensitivity might, be possible. The chromium (III) second coordination sphere water lifotimes will bo measured more 3. OXYGEN-17 NMR STUDY OF THE RATI: OF HATER carefully and it is anticipated that the EXCHANGE FROM PARTIALLY COMPLEXES) NICKEL ION technique can bo extended to several other species. ThoMS Roland The studies on iminodiacetate complexes of The oxygen-17 NMR bulk water resonance sig nickel ion will be continued and, if time nal line width of a solution of nickel ion permits, other nickel complexes will be and iminodiacetic acid (IDA) was measured investigated. earlier as a function of temperature. The data were not sufficiently precise to allow The computer simulation of water exchange one to decide whether the two different types reactions has been interrupted because of the of coordinated water were exchanging at the absence of Dr. Alder, but will be resigned in sane or different rates. The present study 1972. is being undertaken to establish the point. Earlier it was shown that the Niz* - nitrilotriacctate complex does show two distinct water exchange rates. 5. 1971 PUBLICATIONS AND REPORTS Because three possible nickel species (the Robert E. Connick and Associates mono IIM complex with three sites occupied by water, the bis, TDA coaplcx with no sites Journal occupied by water, and the unconplexed hexaquo nickel ion) exist in equilibrium in 1. James Ncely and Robert E. Connick, Oxygen- a solution of Niz" and IDA, both the dis 17 Nuclear Magnetic Resonance Studies of sociation constants of IDA and the complcxing Aqueous Nickel Ion, accepted for publication constants of mono and bis_ IDA coaplexea in J. Am. Diem. Soc. nickel must be *XKMI accurately. The existing data2 were checked spectrophotometrically. Paper presented Equilibria data indicate that in a solution whose Ni/IDA ratio is l.S/1, the maxim*) 1. Robert E. Connick, First Coordination concentration of the mono IDA complcxcd Sphere of Aquo Ions, invited paper presented nickel will be present when the ptl is adjusted at Gordon Research Conference on Inorganic to pi) * 8.5. However, an increased oxygen-17 Chemistry, August 2, 1971. relaxation rate in the neutral region produces a broadening of the resonance signal which UCRL reports occurs over the range pH * 4.S to pU * 9.5 with maximum broadening at pH •» 7.3 A 1. James Katson Neely, Oxygcn-17 Nuclear study of this neutral region broadening is Magnetic Resonance Studies of the First being done as a function of temperature so Hydration Sphere of Diamagnetic Metal Ions that XeF2 with AsFs in , nhydrous HF has yielded a lleil Bartlett, Principal InveBtigator r + new complex cation (F-Xe)7-0,SF] (see 1. NOBLE GAS AND HALOGEN CHEMISTRY Table I): FXeOSOjF «• FXeP • AsF6 - [(FXe)203SF]*AsF6". All efforts to prepare Rudiger J. News, Douglas E. McKoe, Barbara K. the (XeOS02F)+ ion have failed so far. Morrell, Kevin Leary, and Neil Bartlett Xenonfyi) Structure. The crystal structure Xenon(II) Fluoride Derivatives. Although of the 1:1 XoF« adduct with RUF5 lias been cMnounds have been derived from the reactions completed and is reported below. RSOiH • XeF2 - RSOjXeF • HF and 2RSO3H • XeF - (RSO3J2XC • ZHF (R - CFj and CHj) it Xenon (VI) Chemistry. Treatment of has not proved possible to prepare the com PdF3-Pdl:4 mixtures with X0F7, in high pressure pounds pure, and their tendency to detonate at fluorine, produces XoFg derivatives of Pd(IV). tenperatures above - 0° has limited our Rcdi~tion with XoF'2 "ny provide a valuable studies. The interaction of FX0OSO2F and route to MF4. Table I. The Raman spectrum of [(FXe)2OjSF] [AsFfi] (values in cm , relative intensities in parentheses). ,-(a) •(b) [(FXe)2F] [AsF6] '*' [(FXe)203SF' AsFfi [SOjF] 369(7) \>s(AsF6") 372(6) 410(8) 409p(S-F) jSl(sh) S66 588(94)\>(XeF)sym 563(88) S7S(sh)v2(AsF6") S73(sh) «(S02) 600 (100)v (XeF)asym 581 (100) 588(23) .592' 633(12Z.5) f 68309)Vj(AsF5") 685(13.5) 875(2) 786v(S-F) 1035(9) .1082v(S-O)sym 1090(3) 1350(5) 1287u(S-0)asym (a) F. O. Sladky, P. A. Bulliner, and N. Bartlett, J. Chem. Soc. (1969) 2179. (b) K. Nakanoto, Infrared Spectra of Inorganic Coordination Compounds, 2nd ed. Wiley-Interscience, New York and London, 1970), p.114. 13 The interaction of xenon oxide tetra- tion. The fluorine ligands, of the RuFg" fluoride with fluoroacids and bases has been groups nearest the xenon atom, form a dis studied. The acid-base behavior of XeOFj) is torted Archimedian antiprism with the four very like that of IF5. The 1:1 compound equatorial F ligands of the XeFj* ion. The CsF-XeOF^ had been reported previously,1 but packing is exactly what one anticipates for our studies also demonstrate the existence an XeF5+ ion packing with octahedral anions, of the 1:3 addict CsF-3XeOF4. With antimony the distortions being consequences of polar pentafluoride,XeOF4 forms a 1:2 compound ization and packing effects. The axial bond XeOF4-2SbFs, which had been reported pre viously,! and a i:i compound XeC^-SbFc (m.p. 104°C). Raman spectra of the SbFc complexes indicate that they are the XeOFT* + _ + salts XeOF? [Sb2Fii] and XeOF3 [SbF6]-. The Raman spectrum of the cation is like that of the isoelectronic molecule IOF3 and is related + to that of IF4 . Krypton Chemistry. The need for larger quantities of KrF2 for the study of its oxidative chemistry has given rise to a care ful evaluation of available methods. The electric discharge method, although capable of giving yields of the order of a gram-per- day, has proved to be unpredictable. Our first experience of the high energy method^ (10 MeV proton bombardment) indicates that this will provide more krypton difluoride more reliably. The first studies of the interaction of XcOF4 with KrT^SbjFn" are inconclusive but suggest that XeOF4 may not be oxidizable by this reagent. The system KrF2 + XeOFs" may generate Xe(VIII) more effectively. 1. H. Selig, Inor£. Chem. S, 183 (1966). 2. D. R. Mackenzie and I. Fajer, Inorg. Chem. i, 699 (1966). 2. STRUCTURAL INVESTIGATIONS OF HIGH VALENCE CATIONS Donald D. Gibler, Barbara K. Morrell, and Neil Bartlett As part of our comprehensive survey of the factors which control tho fluoride ion donor and acceptor abilities of the non-transition element fluorides we have undertaken a number of crystal structure determinations, two of + + which, XeF5 RuF6" and SF3 BF4*, have been completed this year. rXeFi;1 [RuFfil". This structure was under taken for two reasons, firstly to accurately determine the XeFg geometry and secondly to provide for a comparison with the XeF*RuFg structure determined, in these laboratories, last j'ear. The structural unit of XeFs*RuFg" is shorn in Fig. 1. The cation behaves like Fig. 1 (a,b,c). The crystal structure a pseudo-octahedron with the non-bonding and ion shape of [XeFc] [RuFfi]". xenon valence electron pair in the sixth posi (XBL 716-6868, XBL 716-6869. XBL 716-6872) 14 lo, \M,i) b_ 1.497J' 006A is, however, 0.07A shorter than that n. orted* for PF3. This similarity of shape a. ,d bond shortening, with increase in nuclear charge of the central atom of isoelc. ronic species, is akin to that already observei foT the XeFs , IF5, TeFs" species.^ Fl c Fl In SF4 C^y) the equatorial bonds are of 1.545*0 0P3A and the nearly linear axial bonds a •• 1.646*0.003A. The contraction in forminc the cation is consistent with all bonds i the cation being electron-pair bonds and the axial bonds in the SF4 molecule being single- lectron bonds. Y. Y. N >rino, K. Kuchitsn, and T. Moritani, Inorg. ( icm. 8, 867 (1969). 2. B. I. Morrell, Structural Studies of Xenon I' roridc Complexos (Ph. D. thesis), LBL-156 Sept. 1971. 3. REST RCH PLANS FOR CALENDAR YEAR 1972 Neil Bar-lett Noble Gas and Halogen Chemistry. The attempts to synthesize the Xe(VIII) species XeOF5 a J XeF^O will continue in association with the appraisal of KrF2 and KrF* as oxidizin agents. Structural comparisons will continue co be made between xenon and related iodine s^ ecies. We are planning to studv Il"2 , 1*"2% and IF4" as well as XeF3+ so that comparisons more meaningful to theory can be made between electronically related noble-gas and halogen fluoride systems. Attempts will also be made to prepare and Fig. 2. (a,b). The cr stal structure and ion characterize IOFg" and IFg" as well as XeOFg. shape of [SF ]+ [BF.]\ (XBL 7110-4512, XBL 7110-7597). Transition Metal Fluorides and Oxyfluoridej. Xenon hexafluoride is a sufficiently good fluoride-ion donor that it is able to stabi in XeF5+ is significantly shorter than tne lize unusually high oxidation states in the equatorial bonds, and this is consistent with transition metal fluorides. Thus although it the bond order being h ;;her for the former is difficult to prepare PdF^, the XeFg adducts than for the latter. ".he [RuF6]~ ion is less with PdF4 are obtained readily. It may be deformed in [XeFsHRu.1 ,]" than in [XeF]+ possible, using XeFg as the fluorobase, to [RuFgl". This is attr mtable to the unique prepare Ni(IV) and even Au(TV) complexes by polarizing axis of the with high precision and is shown in Fig. 2. at ~350°: XeF2 + XeFg - 2XeF4. Such a scheme The structure establis! s the salt formulation has promise at least for the preparation of SF3+BF4_ unambiguously. Interestingly, the pure PdF4- SF3+ ion is seen to ha\ the same bond angle (F-S-F) as in the isnelectronic relative ThR crystal structures of PdF/t, PtF4, and PF3(97.8±2°). The S-F bond length of IrF4 appear, on the basis of powder data, to 15 be related. Efforts will be made to solve 4. 1971 PUBLICATIONS AND REPORTS the crystal structure of PdF,j, from powder data if necessary. Magnetic studies will be Neil Bartlett. and Associates made of IrF^ and RhF^ in order to establish the +4 oxidation state, and the thermal Journals stability of these compounds will be better defined. 1. N. Barrett and M. Wechsberg, The Xenon Difluoride Complexes XeFo-XeOF^; • Fluoride bridging in OSO3F2, Re02F3, IrFs XeF2-XeF6-AsF5 and XeF2•ZXeFj•2AsFs and Their and related compounds will be studied by Relevance to Bond Polarity and Fluoride Ion crystallographic and vibrational spectroscopic Donor Ability of XeF2 and XeF^, Z. anorg. u. techniques. allgem. Chem., 385, 5 (1971). The nature of the AggFjg phase obtained by 2. Neil Bartlctt, M. Wcchsberg, G. R. Jones, fluorinating AgF in anhydrous l!F will be and R. D. Burbank, The Crystal Structure of further investigated. Xenon (II) Fluoride Huorosulfatc, FXCOSO2F, Inorg. Chcm., (LBJ.-170, Nov. 1971). Silicon Fluoride Chemistry. We will con tinue to attempt to prepare polymeric (SiF20)x Paper presented but will also explore (SiF2)j, polymer as a reducing agent towards transition metal 1. Noil Bartlett, Noble-Gas Chemistry and fluorides. Its Significance, pnesented as the Dannie- Heineman Prize Address, Gottingen, Germany, The Superfluoride Ion, he will continue Nov. 26, 1971 (LBL-416). to explore the synthesis of F2", principally LBL report by the oxidation of superoxide: 02" • F2 * F " • 0 . 2 2 1. Barbara K. fbrrell, Structural Studies of Xenon Fluoride Complexes (Ph.P thesis), LBL-156, Sept. 1971. 1(» B. SO .10 STATE CHEMISTRY AND PHYSICS 2.0 1. CAUllUMinitlC l-VIIli-XG: HH< A SiNGI.. T GRffNI) Cu-Fe, 640 at ppm H-0 STATI- !N Oil-'c .-. I kOe • JkO» Baylor B. Triplett and Non-an 1.. I'hill ps - lOfcOs • lOkOe •» 2flkO« The ground state of ;i single mngncti • im • SOIiOe purity in a netal continues to he :in uii olved • MkOe problem in spite of the intention (both the oretical :uul experimental > that it has ]••- x 1.0 tiivtil. In particular, it is not clear i.hether i>r not till' (jrouml state is a singlet, fi,riior !-.'.-;>t capacity ini-asurei.i'iils on tul'o ami D] V in :er« magnetic field liave suggested tRi; the u ground state is a singlet.' Additional vi - < deuce supporting this conclusion lias been obtained hy analysis of the field dependence of llie heat capacity of Ciil"e. !__!..,^.j_ . "0.2 0.5 1.0 2.0 5.0 The heat capacities of HI- and 610-at.-p| i TIK) Ud'e samples in magnetic fields are shown in Figs. 1 ami I. The data for the Hl-at.-ppn Pig. r. The heat capacity of 640-at.-ppn tu!:e sample in low fields were indistinguislpble per nolc of l'e, as a function of magnetic fron the reio-ficld data, and have been arait- field. The dashed cuivc represents the single- ted for clarity. Measurements on a 195-at.- impurity ;ero-fieU! behavior. (XBI. 715-3512) ppra sanplc have sliov.i) that /:C/c is the sane as for the Kl-at.-ppn sample for the fields and temperatures i:i I:ig. I. (At is the heat represented by the dashed curve is not propor capacity in excess of that of pure Cu and is tional to concentration and mist therefore be the concentration.) Hie dependence of /.(7c associated with irpurity-ircpurity interactions. on magnetic weld show i-i l"ij>. 1 is therefore Tile application of a mgnctic field broadens characteristic or' single-i,". "rity behavior. the interact vm contribution to X and shifts in rig. 2 the dashed curve represents tt\e it to higher temperatures. Qualitatively, the single-ircpurity icro-fiold linit established effect of magnetic field on it for the 610- by iseasurcmits at lower cuicent rations. The at.-ppn simple can be easily understood. In ;ero-ficld heat 'Vipacity \a excess of tliat ;cro field, uncompensated spins contribute to '.C at tonieraturcs correspondi'.;£ to the effec- 1.3 zivt fields (generated by other uncompensated nins) they experience, lor the oJO-nt.-ppn Cu-Fe, eJ o» pprn s sple the rero-ficld heat capacity shows that H«0 ti ?se fields are S 1.5 kOc. In ;ui applied < 30kCe fu-ld greater than this value ti'.eir contri * 3RkOa bution to ;.C is shifted to higher tempera «*t ^ 1.0 tures. I:i 33 kOe it is alir.ost entirely deter- «o " ~~—"~--^. nim 1 by the applied field and occurs at tea- B i pcraturcs above i K. Thus, the reduction of M iC/c at low temperatures by a 3S-k0e applied field is the sua of two effects--thc shift of •>D* the impurity-ircpurity interaction contribution 03 K > to higher temperatures and the sriill reduction •Nu . (J of the irpurity-ccndiiction-clectron interaction contribution shown by the data for tlie 81-at..- Tianithenr.oilynanic equality Fig. 1. The heat capacity j! Sl-ar.-ppr, ftil-c per mole of l:e, as a function .-. irupgnetic field. (XM. "15-3514I1) T=»B 'T en 17 where x= M/H. At temperatures below 1 K the 2. M. D. Daybell, W. P. Pratt, Jr. and differential susceptibility of dilute CuFe is IV. A. Steyert, Phys. Rev. Lett. 21^ 353 (1968). the sum of a T"*'2 term that saluratesTn 3. J. L. Tholence and R. Tournier, Phys. Rev. fields of the order of 1 kOe and a constant Lett. 25_, 876 (1970). term.2 M is therefore the sum of a T"1'2 term that is independent of field in high field and I temperature-independent term. The large positive values of 3 (C/T)/£H predicted by Eq. 2. LOW TEMPERATURE HEAT CAPACITY OF CERIUM (1) for the T"1'2 term are observed, but only in regions of c and T where interaction effects Mary M. Conway and Norman E. Phillips a;- important. They are apparent in Fig. 2, ft example, for IM kOe and 0.4 tance between CMN and liquid •'lie is abnormally A counting system that uses commercial com low, and decreases with decreasing tempera ponents [Nal(Tl) counters, photomultiplier ture,' probably as a consequence of spin-flip tubes, and a pulse-height analyzer] has been interactions between Ce$* ions and 'He nuclei.' assembled and connected to the LBL computer To achieve useful energy transfer between a facilities. This apparatus has been used for Y-ray-anisotropy temperature measurements in metal and 3||e by the same mechanism, it seems to be necessary to have weakly interacting, the temperature region accessible with the uncompensated magnetic impurities. Rare earth dilution refrigerator and will be important impurities are the most promising candidates, in experiments at lower temperatures that are and, to investigate their properties, we have now being initiated. °"Co in single-crystal ou undertaken heat capacity measuitsnents on a Co thermometers have been used and Co and series of AuGd alloys. Preliminary analysis 59Mn in Fe thermometers have also been pre of the data shows that the system behaves pared . approximately as expected--there is no evidence for a Kondo type of spin compensation even in A number of Go resistance thermometers have a 0.01 at. % alloy, and the impurity-impurity been investigated for possible use ut temper interactions are much weaker tnnn for 3d im atures lower tlian those at which this typo of purities. thermometer had been used previously. Several have been found that are usable to 50 mK and some have properties at that temperature that 1. W. R. Abel, A. C. Anderson, W. C. Black, suggest they will be useful to considerably and J. C. Wheatley, Phys. Rev. Letters 16, lower temperatures. 273 (1966). ~ 2. A. J. Leggett and M. Vuorio, J. Low Temp. Several reasons for questioning our working Phys. 3_, 359 (1970). temperature scale (preserved on Ge resistance thermometers) near 4 K have recently arisen. A number of thermometers were recalibrated against ^He vapor pressures (T5lp an<' an error 4. PRESSURE DEPENDENCE OF THE LOW TEMPERATURE of 3 or 4 mK in the earlier calibration was found. This has lead to a correction in our HEAT CAPACITY OF V203 temperature scale between 3.5 and 15 K which Samuel D. Bader, Baylor B. Triplett, and affects heat capacities in :hat region by up N. E. Phillips to 0.5t. At the same time our scale was com pared with the new NBS acoustic scale. At zero pressure V2Q3 undergoes a tran sition from a nonmagnetic metallic phase to an antiferromagnetic insulating phase on cool ing. The transition is suppressed by pres 6. DILUTION REFRIGERATOR sures of approximately 20 kbar (at liquid He temperatures). This transition has attracted Gary P. Schwartz considerable theoretical interest, and direct measurements of the electronic heat capacity Modifications to the existing dilution as a function of pressure would contribute to refrigerator initially designed and built by an understanding of its nature. We have under B. B. Triplett are nearly completed. The taken experiments of this type in collaboration inclusion of a high speed diffusion ejector with Dr. D. B. McWhan of Bell Telephone Lab pump coupled with a new and larger scaled oratories. Preliminary results have been backing pump will allow faster cool-down rates obtained at approximately S and 20 kbar, and and increased cooling capacity for future ex indicate an increase in the electronic heat periments. More complete vibration isolation capacity and a decrease in the nuclear heat has been achieved and a mixing chamber temper capacity with increasing pressure. ature of 24 mK as measured by a °"Co Y-ray thermometer has been consistently obtained. S. TEMPERATURE MEASUREMENTS 7. LOW TEMPERATURE, HIGH PRESSURE CALORIMETRY Samuel D. Bader, Mary M. Conway, H. Tracy Samuel D. Bader and Mary M. Conway Hall, Jr., Norman E. Phillips, Gary P. Schwartz, and Baylor B. Triplett The high pressure calorimeter originally used in this laboratory had both cylinder and A number of improvements in our working pistons machined from a standard Cu-Be alloy temperature scale and temperature measuring that contained 0.3$ Co. The Co is introduced apparatus have been made. as a grain growth inhibitor which, with the 19 appropriate sequence of heat treatment steps, will be initiated. The first step will be to helps to give the alloy its strength. For our compare the powdered CMN temperature scale purposes, however, the Co has the unwanted with a y-ray anisotropy thermometer by adia side effect of introducing a substantial mag batic demagnetization of a cell containing netic heat capacity. We have therefore at liquid %!e, powdered CMN, and a yray ther tempted to construct a high pressure cell from mometer attached to copper fins. Some pre a Be-Cu binary alloy. However, it has turned liminary information on the Cu-3lle Kapitza out to be impossible to achieve the degree of resistance will be obtained in the same exper hardness in the binary alloy that the supplier iment. In later experiments the field depen ted promised, and several cells failed for dence of the CMN-Jllc Kapitza resistance will this reason. Since the failures were due to be studied to test theories of the magnetic insufficient hardness in the pistons, several energy transfer. The experiments may then be other piston materials (tungsten carbide, ce extended to metallic systems containing mag mented aluminum oxide-titanium oxide, and netic impurities. single crystal, and sintered alumina) were investigated. High-density sintered alumina Heat capacity measurements with the now was found to be the most suitable and we now high pressure calorimeter will be made. The have a tested cell with a Bc-Cu binary body first measurements will be on V2O3, Ce, and II. and end caps and sintered alumina pistons. A tin manometer has been tested with this coll We plan to modify the 3|lo cryostat that was and it will be used for measurements on U and used for measurements on ''lie by the subsiti- Ce. tution of a dilution refrigerator for the 3|!c stage, and to extend our measurements on the heat capacity of liquid The system was tested with a new high homo 2. M. M. Conway, N. E. Phillips, T. H. Gebellc geneity, low remnant field, 16 kOe filamentary and N. A. Kuebler, Low-Temperature Heat Capa superconducting solenoid. In the first run, cities of the Clathrate Salts Ag70sN03 and from an initial temperature of 2 K, the chrome Ag708HF2, J. Phys. Chem. Solids 31, 2673 alum stage cooled to 70 mK. (1970). The calorimeter •••.•orking space allows heat 3. N. E. Phillips, J. C. F. Brock, M. II. capicities and magnetic susceptibilities to Lambert, and M. F. Merriam, Superconductivity be measured simultaneously. This will be and Electronic Structure of a-Phase Indium valuable in experiments to understand the Alloys, in Proc. 12th Intern. Conf. Low Temp. bulk and filamentary superconducting proper Physics. Kyoto, Japan, Sept. 4-10, 1970; ties of a-uranium at zero pressure. Thfc new edited by Eizo Kanda, (1971), pp. 353-54. design will also allow the use of a supercon ducting tin nanometer in high pressure cal- 4. B. B. Triplett and N. E. Phillips, Low orimetry below 1 K. Temperature Heat Capacities of Dilute Solu tions of Fe and Cr In Cu, ibid., pp. 747-48. 5. T. L. Thorp, B. B. Triplett, W. D. Brewer, 9. RESEARCH PLANS FOR CALENDAR YEAR 1972 M. L. Cohen, N. E. Phillips, D. A. Shirley and J. E. Templctori, Search for Superconduc Norman E. Phillips tivity in Lithium and Magnesium, J. Low Temp. Phys. 3, S89 (1970). A series of related experiments on the powdered CMN temperature scale and the Kapitjia 6. B. B. Triplett and N. E. Phillips, Cal resistance between magnetic materials and 3He orimetric Evidence for a Singlet Ground State 20 in CuCr and CuFe, Phys. Rev. Letters 27, 1001 (1971). 7. B. B. Triplett and P. L. Phillips, Low Temperature Heat Capacity of Nin ,, Rhn TO, Phys. Letters 37 A, 443 (1971). •" 21 George Jura, Principal Investigator 1. PRESSURE (UP TO 40 KBAR) AS A PARAMETER IN THE STUDY OF VIBRATIONAL EXCITON LEVELS AND CRYSTALLINE POTENTIALS OF MOLECULAR CRYSTALS: NAPHTHALENE AND N-PARAFFINS* Che-Kuang Wu and George Jura 5 -0.01 The spectra of naphthalene, several long chain hydrocarbons, and polyethylene were determined to pressures of 40 kbars. J I I L J I L Fermi resonance was found in the spectra 100 120 140 ISO ISO 200 220 240 260 280 of naphthalene. Davydov splitting was fol TCK) lowed for all of the molecules. It was pos sible to determine the repulsive forces for Fig. 2. (XBL 7110-741.5) several types of motion from the change in splittings. capacity with pressure at a group of tempera Abstracted from LBL-169 (Ph. D. thesis). tures. On the whole, the results are as good as can be expected. Considering that the derivative is obtained numerically, and the variation in the heat capacity is so small, it is remarkable that at only three 2. HEAT CAPACITY OF IRON AS A FUNCTION OF temperatures was the derivative positive. PRESSURE AND TEMPERATURE* The positive derivative is, of course, impossible. Albert Fan Yee These results show that the Debye tempera The technique for the determination of the ture has increased between t and 10 degrees specific heats of metals as a function of between the pressures studied. This in temperature and pressure was developed to the crease is in accord with earlier estimates. degree that reliable data were obtained from about 25 to 85 kbars. Abstracted from LBL-180 (Ph. D. thesis). Figure 1 illustrates the agreement between the data obtained with the present pulse technique and normally raeasured heat capacity 3. SPECIFIC HEATS AND HEATS OF TRANSITION data for iron. The maximum difference of OF BISMUTH AS A FUNCTION OF TEMPERATURE AND 0.5% is most satisfying. PRESSURE Figure 2 shows the variation of the heat Berardo Juardo and George Jura I ' I I I • I I I I I I I I I i.o Work is being done on the specific heat of o.a bismuth and the direct determination of the 0.6 heats of transition between the polymorphic forms of this semimetal. I a4 Z 0.2 The preliminary results show that the S o a present pulse technique will be able to --0.2 determine the heats of transition between 8 -0.4 crystalline forms. A large number of experi C -0.6 ments have been made on the III-V transition. -0.8 -1.0 This particular transition has been chosen -1.2 ' • ' • ' I I I I I for study because the sample can be recycled, 100 120 140 160 160 200 220 240 260 2S0 T PK) and there is no need to make a new sample for each deterndnation. At the moment, we cannot Fig. 1. (XBL 7110-7453) make a comparison between the best value 22 that we have obtained and that which would In the case of silver chloride it was be obtained from an application of the also noted that there was a large decrease Clasius-Clapeyron equation. The latter gives in the heat leak constant above 90 kbars, a as the heat of transition Ht: pressure at which there is a known poly morphic transition in the salt. 4 Ht = 4.21 10" T , If it is possible to develop the technique where T is the absolute temperature. Our to the degree that good results can be ob best value is 560 cal mol"*, but we are tained consistently, then it is evident that ignorant of the temperature at which this this experiment is worthwhile. measurement has been made. It is necessary to determine the resistance of Bi as a function To date, four different approximations of temperature so that this necessary havo been raade in order to obtain all of the quantity can be determined. However, the desired results. The best is given by the result is in the ball park. equation 2 It appears that heats as low as 25 cal -iE/dt - I R"/C -k/C fE-Fo + at) mol'l can be determined with the present technique. wfceio I; is the voltage drop across the sample, I the current, R* the temperature derivative of the resistance, k the heat leak constant, Ep the voltage that would be observed in the 4. RESEARCH PLANS FOR CALENDAR YEAR 1972 sample at ambient temperature, a a constant, and t the tine. This expression reproduces George Jura, with Tang-hu? Chen the data quite well to times of about 5-7 milliseconds, gives values of G, that in a. Positron Annihilation in Antimony agreement with the limiting value, and reasonably constant values for the heat leak The v. rk on positron annihilation in constant. If the time is stretched, then metallic antimony is continuing. At the there is a systematic change in all of the present time, the angular dependence is being quantities. measured at 60 kbars. The results at the lower pressures are in accord with the Fermi We are attempting to change the instrumen model. The region of interest is now being tation without the expenditure of money. One approached. of the difficulties with an experiment of this type is the fact that about 30 E vs. t b. An Attempt to Determine the Heat readings must be made off of each photograph. Conduction of Non-conducting Material as a A digitalized output of some kind is essential Function of Temperature and Pressure" before this sort of experiment can be per formed efficiently. Mr. Chen is also working on the specific heat problem. He has been able to roughly Hopefully, we will be able to borrow and reproduce Mr. Yee's work. He has not had try out different methods to digitaliie the the experience to develop the necessary results. artistry to obtain as self-consistent results. The present technique of obtaining specific heats from the limiting slope of voltage vs. 5. 1971 PUBLICATIONS AND REPORTS time for a constant current pulse pushes the presently available electronics to the limit, George Jura and Associates and in addition, requires artistry in the measurement of the slope recorded on a LBL reports Polauid film. 1. J. L. Przybylinski, A. F. Yee, and In theory it is possible to use a longer G. Jura, Positron Annihilation in Ytterbium pulse, and in addition to the heat capacity Metal to 80 Kilobars, LBL-160, Aug. 1971 obtain information concerning the heat (submitted to Physical Review). conductivity of the surrounding medium. So far, it has been possible to show that 2. C. K. Wu, Pressure Cup to 40 kbar) as a there is an increase in the heat leak con Parameter in the Study of Vibrational Btciton stant as a function of pressure for both Levels and Crystalline Potentials of silver chloride and pyropholite. The heat Molecular Crystals: Naphthalene and leak constant for the latter is considerably N-Paraffins (Ph. D. thesis), LBL-169, Oct. higher than for the silver chloride. 1971. 25 J. A. F. Yee, Heat Capacity of Iron as a Function of Pressure and Tcnperaturc (Ph. D. thesis), LBL-180, Nov. 1971. 24 Gabon1 A. Somovjai, Principal Investigator LBL-121. 1. LOW ENERGY ELECTRON DIFFRACTION STUDIES On leave of absence from Laboratoire de OF SURFACES Mineralogio, University de Strasbourg, France. a. Low Energy Electron Diffraction Studies Present address: School of Chemistry, of High Index Crystal Surfaces of Platinum" University of Bradford, Yorkshire, England. Bernard Lang, Richard K. Joyner, and Gabor A. Somorjai Several high Miller Index crystal surfaces of platinum have been examined by low energy electron diffraction (LEED) and have been shown to consist of low index (111) or (100) terraces of constant width, linked hy steps of monatomic height. A characteristic diffraction pattern and the schematic diagram representing the surface are shown in Figs. 1 and 2 re spectively. The ordered stepped surfaces were found to be Temarkably stable in ul* igh vacua* up to 1500'K. The presence < ordered atonic steps may be considered as a general structural property of high index surfaces regardless of the chemical bonding in the crystal. A nomenclature is suggested to identify the surface structures of stepped surfaces. Fig. 1. Diffraction pattern of a clean stepped surface of platinum, Pt(S)-[9(m)x(111)1. (XBB 7110-4951) Fig. 2. Schematic representation of the atomic structure of a stepped surface, Pt(S)-|9(in)X(lll)]. (XBL 716-6801) 25 b. Low Energy Electron Diffraction Studies hydrogen appears to decrease die rate of of CHemisorbed Gases on Stepped Surfaces of cracking reactions which tend to deactivate Platinum * the platinum surface. Bernard Lang, Richard Joyner, and The reactivity of various low index and Gabor A. Somorjai stepped crystal faces is being compared. There is evidence for considerable differences The Chemisorption of hydrogen, oxygen, between various faces and, particularly, sur carbon, carbon monoxide, and ethylene was faces with terraces of (111) orientation ap studied by low energy electron diffraction on pear fur more efficient in yielding toluene ordered stepped surfaces of platinum which than surfaces with terraces of (100) orienta were cut at angles less than 10° from the tion. (111) face. The chemisorption characteristics of stepped platinum surfaces are markedly dif d. The Preparation of Atomically Clean ferent from those of low index platinum sur Platinum Surfaces * faces and they are also different from each other. Hydrogen and oxygen which do not chcm- Richard IV. Joyner, John L. Gland, and isorb easily on the (111) and (100) crystal Gabor A. Somorjai faces chemisorb readily and at relatively low temperatures and pressures on the stepped Selected structural and chemical properties plutinum surfaces used in this study. In con of low index and stepped platinum surfaces are trast to the ordered adsorption of carbon reviewed, and the properties of various con monoxide and ethylene on low index faces, the taminants in platinum, especially carbon and adsorption of 00 was disordered on the stepped hydrogen, are discussed. Methods of cleaning surraccs and there is evidence for dissocia of the different platinum surfaces have been tion of the molecule. Carbon formed several developed, and their effectiveness monitored ordered surface structures and caused faceting using Auger electron spectroscopy and low on the stepped surface, which are not ob energy electron diffraction. The cleaning served on low index platinum surfaces. There techniques developed for the different types appears to be a much stronger interaction of of platinun crystal surfaces should also be chemisorbed gases with stepped surfaces than applicable to the cleaning of polycrystalline with low index planes that roust be caused by samples. the differing atomic structures at the steps. There is also evidence for the differing reactivities of the various stepped surfaces, LBL-414. LEL-17S. e. Comparison of Backscattering Intensities for Low Energy Electrons from Various Surface Atoms c. Catalysis by Single Crystal Surfaces: Mario R. Martin and Gabor A. Somorjai The Dehydrocyclization of n-Heptane~to Toluene The atomic scattering intensities |f(aE)| are calculated for two backscattering angles Richard H. Joyner, Bernard Lang, and (158° and 180°) for a number of atoms com Gabor A. Somorjai monly encountered in the investigation of sur face structures using the technique of low Using a modified LEED chamber, a complex energy electron diffraction. The results, chemical reaction could be monitored at plat plotted as a function of the in-ident elec inum single crystal surfaces of less than tron energy, make available the relative back- 1 cm* area, The reaction studied is the de scattering intensities of a number of sub hydrocyclization of n-heptane into toluene. strate and adsorbate atoms, and should be of While LEED and Auger electron spectroscopy assistance in choosing the appropriate energy are used to ascertain the structure and the range for the investigation of surfaces and cleanliness of the substrate, a quadrupole adsorbed layers. mass spectrometer detects the appearance of the peaks characteristic of toluene. f. The Use of Atomic Scattering Factors in Calculations o± Low Energy Electron Pit- The reaction occurs in a static system con fraction 3eam Intensities taining a mixture of n-heptane and hydrogen (total pressure - 2 x 10"4 Torr with a Mario R. Martin and Gabor A. Somorjai ratio Il2/hc > 1) at temperatures of the crys tal between 400 and 700 K. The presence of Arguments are presented to justify the use 26 of isolated atomic scattering factors in low 1 energy electron diffraction calculations. Re I 1 sults of electron scattering from liquid sur faces are compared v Lth gas phase scattering 3.0 data for bismuth am lead. The effect of the liquid structure fa tor is investigated and shown to account for less than IS* of the re • flected electron ir tensity from the liquid V0 bismuth surface. he atomic scattering fac 2 tors for electron; from atoms in the vapor and condensed pha ^s differ most notably in 2.0 the shapes and ma .nitudes of the respective forward scattering peaks. Inclusion of a moderate number ^ ~ S) of phase shifts to *2o3 describe the at< tic potential is expected to yield sufficier accuracy for surface struc ture detormina on by low energy electron diffraction intensity calculations. 1.0 *°0.92 g. Low Energy Electron Diffraction Studies of the Surface Structures of Chemisorbed Aromatic' MSIF.-ules on the Willi and i>UlW) Single Crys' al Surfaces n 1 1 1.0 2.0 3.0 4.0 John L. Gland and Gabor A. Somorjai Oxidation Number The surface structure of many aromatic hydrocarbons were investigated on the (111) Fig. 1. The chemical shift of the vanadium and (100) crystal faces of platinum. Naph L2M2 3^2,3 Auger transition in the vanadium thalene, nitrobenzene, pyridine, and cyclo- oxides as a function of the vanadium oxida hexane form ordered surface structures upon tion number. (XBL 7111-764S) adsorption at 2S°C while other aromatic com pounds, benzene, aniline, azuline, and n- butyl-K-nzene do not. Low temperature ad 1 1 1 sorption studies .and work function change measurements are in progress to further ex plore the course oc ordering in the adsorbed 3.0- monolayer. • V02 2. AUGER ELECTRON SPECTROSCOPY OF SURFACES 2.0 a. Auger Electron Spectroscopy Investigations of the Surface Chemical Composition of *2o3 Vanadium Oxides, and Uxi3ized Vanadium: Chemical Shift and Peak Intensity Analysis * • Fred J. Szalkowski and Gabor A. Somorjai 1.0 - V00.92 - The Auger spectra of vanadium metal and of the oxides VOn^, V2O3, and VO2 have been studied. Chemical shifts on the order of 0.6 eV per oxidation number have been observed for the vanadium inner sh 1 I tion potential as measured by the heat 3.0 b. Reactive Scattering of N2O Molecular Beams""from Clean and Carbon-Covered [1001 • Platinum Single Crystal Surfaces t vo2 Thermal energy molecular beams of N2O have been scattered from a platinum C100) surface 2.0 maintained in the temperature range 900 to ,1 150°K in an ultra-high vacuum chamber. The angular distribution of the scattered beams v2o3 was monitored with a quadrupole mass spec o trometer, the target surface structure by low energy electron diffraction, and the sur 1.0 - face composition by Auger spectroscopy. The 000 •V00.92 effect of surface temperature and surface 6> o structure on the surface decomposition of N2O has boon studied. At 1200°K, using a clean surface, the angular distribution of one of the reaction products, NO, follows the cosins 1 1 law but deviations from this behavior seem to 3.0 occur and there is evidence for "direct" 1.0 2.0 scattering if the surface is covered by a h /hv n layer of carbon. At still higher temperatures N 0 reacts with the carbon layer on the sur Fig. 3. The chemical shift of the vanadium 2 L M face and removes it by the oxidation reaction L2 •2,3^'2," 3 Auger transition as a function of N2O + C CO • N->. The mechanisms of reactive the oxygen KW to vanadium L2M2 5V half-peak •* scattering of N26 are likely to involve the height ratio for the vanadium metal oxidation formation of CN surface intermediates whose runs. The known vanadium oxides arc used as formation is by an exothermic surface reference points. (XBL 7111-7647) reaction. LBL-444. Present address: Sandia, Livermore. +LBL-133. 3. M3LECULAR BEAM SCATTERING FROM SURFACES 4. STUDIES OF THE VAPORIZATION KINETICS OF Lloyd A. West and Gabor A. Somorjai HYDROGEN BONDED LIQUIDS* a. Scatter—, of Polyatomic Molecular Beans Fenton R. McFeely and Gabor A. Somorjai from the Pt(l 0 Crystal Surface The vacuum evaporation rates of glycerol, The non-reactive scattering of thermal ve ethylene glycol, and triethylene glycol have locity molecular beams of CO21.C2H2, NO, CO, been measured in the temperature range 5-50°C N2, NO?, 0 , NHj, D2, H2 have been studied 2 using a microbalance. The activation enthal from the Pt(100) single crystal surface that pies of vaporization of the three liquids were was held at 900°K to prevent the chemi'jorption found to be different from their enthalpies of of 00 from the ambient. It was found that the vaporization. The vacuum evaporation rates spatial distribution of the scattered species for glycerol and triethylene glycol were about were non-cosine under most of the experimental one-'.iird of the maximum rate that can be conditions, thus indirating the lock of rapid calculated from the vapor pressures, and one- energy accomodation between the clean single twentieth of the maximum rate for diethylene crystal surface and the incident molecules. glycol. It appears that breaking one or There is evidence, however, that increased more hydrogen bonds at the surface is the surface roughness on an atomic scalo enhances rate-limiting step in the mechanism of vapor the probability of energy accomodation be ization of these largely hydrogen bonded tween, the surface and the incident polyatomic liquids. molecules. At a constant surface temperature there is a qualitative correlation between the scattered intensity maxima and dispersion out the known strength of the gas-solid interac- I.BL-127. 5. ia.s>.y«3i PIANS Hi' (.•u!:^\;; O..\K IT: '•: the Vap'.ruati .n :-'echan:sr. of '> ^v single Ci.-staj-, '. Cher. rcys. 55 fS), 3624 I19"]i. (labor A. sorvoriai ~. iieni.;rd Ling .crd Gahor \. ->or::or;air Struc surface structure analysis usifii'. low energy tures of Chtcisori t-d Gases on Single Crystal eh tron diffraction and computer L.jJcul.it ion -.urfaces: Classification ar.d fiulcs of 1'orrna- will be carried out usin>: v;*nadion .ind platir. tion Ipublished i:i f" rendu .>. de Chinie urn single crystal -airfaces and surfaci trie Piiysiouc, I9~l, '. . "-«, p. 1223. (i_nglish tures of adsorbed gases. version, IJ("RI. 203S5). •\uger spectroscopy studied will be extended M. 1). L. ilowlett and G. A. Sonorjai , Vacuum to liquid metals, alloys, and polvdispersed Vaporization studies of Lithiun Fluoride particles, and to hydrocarbons adsorbed on iingle Crystals, '. I'hys. Chen. J_3, 41140 U97J). metal surfaces. lJJ.j Citalytic surface reaction studies using 2. I. li. '-Iclccly J»J G. A. Sonor.iai, Studies stepped platinm single crystal surfaces will of the Vaporisation Kinetics of Hydrogen be extended to include the dehydrogonation Bonded li<|uids, LBl.-ir, July 19"!. of cyclohcxane and the cycliration of n- butylbenzcne. 3. 1.. A. J.est and G. A. Soreorjai, Reactive Scattering of ."oO Molecular Beans frcn Clean and Carbon Covered (100) 1'latinun Single Crystal Surfaces, IJiL-133, July 1971. 6. 19"! IlIHLICYiiONS AM; Mill's J. M, Fin*, ':. H. Martin, and G. A. Soffiorjaj, Gabor \. sectorial ajid \ssociates Comparison of Kacl-scattering Intensities for !>3w l'.nergy Llectrorr. frcn Various Surface Journals and book Atoms (I!, l.i, Be, C, 0, Al, Si, S, V, Cr, Ni, Cu, Ag, P'. An;. 1J1.-144, Aug. 1971. 1. II. H. ljarrell m:'. ••'.. A. Sorao.-jai, Low Energy Mectron Diffraction, in Advances in 3. I). I.. Hewlett, Vacuira Vaporisation Studies Chemical Physics i"iw lev 5 Sons, New York, of l.ithiua fluoride Single- Crystals (M.S. WT), p. 215. thesis), UCK1.-2 2. '".. A. Somonai and F. '. ';alkowski. b. M. n. Martin, 1.. 'A. Falicov, and G. A. Simple iu les to Predict the Stiucturo of Soraorjai, Hie Use of Altraic scattering Factors Adsorbed Gaser- on Crystal Surfaces, J. Chian. in Calculations of Low Lnergy Electron i)if I'hys. 54_(1). 589 (1971). fraction Bcaia Intensities, U1L-158, Aug. 1971. 3. I. A. West, L. I. KoraX, .••nd G. A. 7. Ii. Laiig, R. n'. .Jo)Tier, and G. A. Sctnorjai, S'arsorjai, Molecula". Bean Scattering from IJOK linergy Klectror Diffraction Studies of Single Crystal Surfaces Under Ultra-Hi&fr Chenisorbed Gases on Stepped Surfaces of Vacuum Conditions, •). Vac. Sci. Tech. B_ (2i, I'latinum, LBL-17S, Oct. 197!. 450 (1971). 8. B. Lang, K. W. Joyr.i»r, and G. A. Socorjai, 4. L. A. West and '". A. Somorjai, Effects of Low ijiergv Electron Difraction Studies of Surface Disorder, \arious Surface Structures High Index, LB!.-121. Oct. 1971. of Chcnisorbed Gases and Carbon on Heliun Atonic Beam Scattering fro™ the (100) Sur 9. R. IV. Jevner, J. L. Gland, and G. A. face of Plat iinn, J. Chen. fhys. S4_ 17), Somorjai, The Preparation of Atomjcal'y Clean 2864 (1971). I'latinun Surfaces, IBI.-414, Nov. 1971. 5. C. V. Lou and G. A. Sonorja.i, Studies of 10. G. A. Sonorjai, Structure of Surfaces and the Vaporization Mechanism of Gallium Arsenide- Internal Interfaces, LBL-438, 1>.. 1971. Single Crystals, J. Cheia. 1'hvs. SS (9), -1SS4 (1971). 11. T. M. French, Low Energy Electron Dif fraction Studies of Insulating Single Crystal 6. J. Gordon Davy and G. A. Sorccrjai, Studies Surfaces (Ph.I*, thesis), LBL-155, Oct. 1971. 29 12. L. A. Kest, Studies of Gas-Surface Inter faces, 1-BL--JS5, Dec. 1971. actions at the (100) Crystal Face of Platinum by Molecular Bean Scattering, Low Energ}' Elec 14. F. J. Szalkowski and G. A. Somorjai, tron Diffraction, and Auger Electron Spectros Auger Electron Spectroscop)' Investigations of copy, LBL-107, Nov. 1971. the Surface Chemical Composition of Vanadium, the Vanadium Oxides, and Oxidized Vanadium: 13. G. A. Sonorjai, The Surface Structure of Chemical Shift and Peak Intensity Analysis, and Catalysis by Platinum Single Crystal Sur LBL-444, Dec. 1971. 30 hee F. Donaghey, Principal Investigator This program is concerned with physio- chemical methods in the preparation, crystal growth, and processing of solid state- materials, and with the impact of processing on physical properties. The synthesis aspect of this program currently concerns mass transport and interface attachment kinetics of chemical vapor deposition, Czochralski -rystal growth, and spine] synthesis. The characterization aspect concerns the measure ment of process-induced defect structure, magnetic ordering, and electrical and ionic transport properties of inorganic compounds. 1. CZOCIIRALSKI CKYSTA1. CUOU'lH Fig. 1. Spiral subgniln boundaries revealed by heavy etching in a Go crystal grown along Lee F. Oonaghey the (100 > at 5 in./hr with a rotation speed of 13 rpm. X9.5. (XBB 722-770) The industrially important Czochralski method of crystal growth is under study to develop fundamental and practical methods for controlling defect density, compositional homogeneity, and interfacial morphology. The methods of .»tudy involve a correlation of defect and substructure density with crystal growth conditions, and a characterization of the interfacial liquid region through the hydrodynamics of forced convective mixing of the liquid phase. Compositional inhomogencities and disloca tion density were studied in germanium, grown along the <100 > . Growth rate variations corresponding to the rotation speed and to rf heating fluctuations were disccrnable along the lateral surfaces, and (110) facet angles indicated the curvature of the crystallization interface. Spiral subgrain boundary coales cence was observed to develop along the axis of those crystals where dislocation sources were stabilized by a concave crystallization Fig, 2. Mixing of a neutral buoyancy dye in interface. The subgrain boundaries near the ethylene glycol during counter rotation of center of a <100 > germanium crystal are the crucible and simulated crystal. The dye shown in Fig. 1. was introduced at the edge of the simulated crystal. (XBB 722-769) The mass transfer to the crystallization interface is driven by the counter-rotation of the crystal and crucible. As such, the mixing during Czochralski growth were con aass flux to the crystal can be solved by ducted with fluids of different kinematic extending Cochran's solution to the case viscosity, in a glass cell shown in Fig. 2. where the fluiV. far from the interface is Contrary to earlier studies, the Taylor- rotating with a fixed velocity. Without the Proudman isorotational cells were found to influence of edge effects, a stable solution decrease in size with increasing kinematic requires a net flow toward the interface, a viscosity of the fluid. Nonmixing vortices condition readily met by crystallization. were observed at the edge of the simulated crystal where the critical Rayleigh number is Experimental studies of forced convective reached. The rate of mixing between boundary 31 layer cells of the crystal and crucible de catalytic activity. In addition, a number of pends markedly on kinematic viscosity and cation-deficient sulfospinels similar to instabilities in the liquid phase. In8/3?1/3S4 f" 4/3 In2s?) a™* formed by the substitution reaction with chromium, 2 Cr^+ + S - 2 Cr3+ + 3/4D + s2", have recently been reported. Preparative methods for producing 2. CHEMICAL VAPOR DEPOSITION large simple crystals have not been success ful to date because of poor control over Saleem A. Shaikh and Lee I;. Donaghey nucleation, the occurrence of undesired side reactions, and difficulties in maintaining Chemical decomposition and etching in gas sample purity. These compounds thus offer an or vapor streams is frequently used in semi interesting challenge from a synthesis point conductor crystal and device technology. The of view. In addition, the vacancy density conditions usually favored for epitaxial can be varied by the substitution reaction layer formation by chemical decomposition, as above, thereby allowing rate-dependent well as for chemical vapor etching, cor processes such as solid state reaction, responds to diffusion controlled Kinetics. sintering, and ionic transport to bo studied. Yet chemical cost factors and more stringent requirements for ternary and mixed binary In an initial study, the compound CdCr^Tofl compounds favor a trend toward interface- was prepared by the topochomical reaction kinetics controlled deposition or etching. between CdTo and CrClj. The reaction pro In this study, the mass transfer and deposi ceeded in sealed silica ampoules through the tion kinetics of Group III-V compound alloys vapor phase and produced numerouo crystallites are examined, and requirements for reactor nucleated on the walls of the ampculci;. scale-up are deduced. Crystallite growth was inhibited by the con densation of an unidentified second phase. Analytical methods for the calculation of deposition rates have been extended to ac count for boundary-layer induced changes in local rates along the direction of vapor flow, 4. TRANSPORT IN SOLIDS and to correct for these changes by tempera ture programing. In addition, models of Richard Iteadrick and Lee F. Donaghey mixing in large-scale reactors were examined. In practice, servicing and batch loading The application of selective ionic trans requirements can exert an overriding influence port in solids is uider study toward the on reactor design, and lead to growth defects development of electrochemical sensors for related to incomplete mixing in the vapor chemical reaction monitoring. A survey of phase. current electrochemical cell methods for the measurement of heats of formation, diffusion, A research reactor was constructed for and activity was made. Errors arising from the study of the deposition kinetics of vapor phase transport between open electrodes, inorganic compound alloys. The reactor con electronic conductivity of the electrolyte, sists of three independently controlled exchange reactions between the electrolyte temperature zones for saturation, reaction, and an electrode, and thermal transients and deposition. High Reynolds number entries were assessed quantitatively with regard to are incorporated at the entrance to two different cell designs of the type, Pt|Ni:NiO!|Zr0 + 7.St CaO||Me:MeO. The mixing chambers. 2 optimum design was found to require isolated and evacuated electrode compartments, with The kinetics of the Group III-V compound cell potentials measured electrometrically. alloy GaAsxPi.x deposited through the reaction of gallium chloride with arsine and phosphine on germanium substrates is currently under study. 5. MAGNETIC SUSCEPTIBILITY MEASUREMENTS Rajani B. Agnihotri and Lee F. Donaghey 3. CHALCOGENIDE SPINEL A novel coil system was constructed to Rajani B. Agnihotri and Lee F. Donaghey generate the inhomogeneous magnetic field required for magnetic susceptibility measure- The sulfo" and selerospinels involving m...ts by the Faraday method. The system, transition metal cations are currently of consisting of eight D-shaped coils wound from interest because of their ferromagnetism, copper strip, was specifically designed to magnetic semiconducting properties, and their fit over the pole faces of a conventional 4 inch diameter electromagnet ' such a way 7. RESEARCH PIAKS FOR CALENDAR YEAR 1972 as to maximize the image currents supporting a linear field gradient while siiirumizing the Lee F. Donaghiy effect of return loops. a. Chemical Vapor Deposition The magnetometer developed with this coil system allows the continuous adjustment of Measurement of chemical vapor deposition a field-independent gradient and has several kinetics of GaASjjPj^ on Ge substrates are advantages for magnetochemical studies over continuing. Defect generation in composition the shaped-pole-face magnetometer. The field graded layers will be investigated with gradient-induced force is directly proportion topographic x-ray diffraction. Methods of al to the sample magnetization; and because producing graded junctions with very low the gradient can be rapidly switched on and defect density will be examined as an impor off, the field-induced force can be distin tant aspect of all graded junction devices. guished from time-de|iendent changes in sumple weight. Furthermore, by designing the D-coils b. Solidification with an inner diameter larger than the dc magnet pole pieces, the lateral and vortical Recent experimental measurements of high linearities of the field gradient is increased. velocity solidification kinetics indicate a lack of agreoment with any existing theory; A liquid helium dewar with variable tempera possible mechanisms will be assessed, and ture control is currently being fitted to the additional data taken. magnetometer to allow measurement over a range in temperature. The temperature- Hydrodynamic studies of mixing during dependent measurements are essential to Czochralski growth will continue to ascertain deduce the high temperature paramagnetic effects of thermal gradients on mass transfer. moment from the Curie-Weiss law, and to Efficient mixing is essential in this method identify low temperature ordering transitions. of growth to insure microscopic homogeneity. The apparatus will be used to measure the magnetic ordering in transition metal spinels. c. Transport in Solids The data, in conjunction with conductivity measurements, are useful in identifying the The applications of selective ionic trans cation distribution on interstitial sites port to thermodynamic and kinr 'z measure and the temperature dependence of ordering. ment and to chemical process mj.-itoring will be explored further. A method utilizing two reference electrodes to overcome errors in ternary phase activity measurements i' 6. COMPUTER SIMJ1ATION currently under development. The transport properties of spinels will be studied Lee F. Donaghey further by means of electrical conductivity measurei^nts toward evaluating their potential The coarsening of precipitate particles in as selective ion filters. ternary steel alloys was studied by numerical simulation. When the diffusivities of carbon d. Magnetic Properties and alloying elements differed markedly, the coarsening process passed through regimes Studies of defect chalcogenide spineh will kinetically limited by the diffusion of continue toward defining optimum synthesis carbon, by mixed control, and by the dif conditions and magnetic properties. Magnetic fusion of the alloying element. The pre measurements will be combined with electrical cipitate particle size distribution showed a and diffraction measurements to deduce lower cutoff for alloying elements which cationic interstitial ordering and stability segregated preferentially to the alloy carbide, of these compounds. while an upper radius limit appeared if the alloying element segregated to the matrix surrounding the precipitates. Numberical simulation of the transition from interface reaction control to diffusion control showed 8. 1971 PUBLICATIONS AND REPORTS that the transition spans a particle radius interval from approximately 0.2S D/k to Lee F. Donaghey and Associates 40 D/k, where D and k are the chemical diffusivity and interface reaction constants, Journals and reports respectively. Transient coarsening of non- steady-state particle distributions was 1. S. Bjorklund, L. F. Donaghey, and achieved with an N-particle ensemb. approach. M. Hillert, The Effect .f Alloying Elements 33 on the Rate of Ostwald Ripening of Cementite 64th Annual AIChE Meeting, Nov. 29, 1971, in Steel, to appear in Acta. Met. San Francisco. 2. L. F. Donaghey, Hydrodynamic Mechanisms 3. L. F. Donaghey, Compositional Inhomo- for Composition Banding During Crystal Growth, geneities in Malt-Grown Crystals, Dept. of Chem. Engr., Nov. IS, 1971. 34 C. ELECTROCHEMISTRY Charles W. Tobias, Principal Investigator linear potential scan is applied. Apparent limiting currents due to fast current increase 1. FUNDAMENTAL STUDIES OF IONIC MASS or potential scan rates can be satisfactorily TRANSPORT interpreted in terms of a pure-diffusion model. a. Measurement And Interpretation of Limiting Currents* b. Distribution of Ionic Mass Transport Rates in a Flow Channel Jan Robert Selman and Charles W. Tobias Utiol Landau and Charles W. Tobias Applications of the limiting current method for measuring mass transfer rates are re Local maximum mass transport rates were viewed. To date almost 100 publications using measured, using copper deposition from this method have appeared, with recont appli acidified copper sulfate solutions on a rela cations in the area of stirred tanks, parti tively largo (600 cm ) segmented coppor,plate, culate systems, mixed convection and turbu rariieddod in the wall of a flow channel. The lence. pilot plant scale flow circuit and its instru mentation were roportod earlior. The conditions for valid measurement and interpretation of limiting currents are dis Experiments carried out so far (up to Rey cussed. An ideal limiting current curve is nolds number of 10,000), produced well-defined constructed for the copper deposition model local limiting current curvos (Fig. 1). In the reaction. The effect of migration due to the laminar flow range, current distribution along electric field is computed for this reaction,, the plate, at maximum transport rates, follows taking into account that the bisulfate ion is a 1/J-power dependence on the distance from not completely dissociated. The use of inte the leading sdge, in agreement with prediction gral diffusivities, which reflect the effect from theory. of migration as well as that of variable physical properties, is required in the cor Rtf«r»nc« •Ketrodtjxxillon from Itading tdge (en) relation of mass transfer coefficients cal 0.75 6 12 36 culated from limiting currents. 3mJn -J 3mln-] 3min-j(- CuSOii' 106M HjSO,• I.3M Rl* Ionic Mass Transport by Combined Free and Forced Convection, Z. Pnysik. Chcm. 229, 145- 166 (1965). c. MecJ; -Isin of Electropolishing of Copper Koaru Kojima and Charles W. Tobias The anodic dissolution of copper in phos phoric acid was studied by both gnlvanostatic and potentiostatic techniques in the absence of convection in a cell configuration which assured uniform current distribution. The diffusion of copper phosphato [Cu(H,PO.)2] has been identified as the rate limiting process. At the current plateau the concentration of phosphoric acid at the surface is far from zero (pH •> 2). The characteristic current plateau in electropolishing is caused by the formation of a nultiltyered oxide film. On the basis of the understanding of the trans port mechanism involved in the polishing pro cess, a quantitative method has been developed for estimating the current density to be ap plied for optimal polishing conditions in laminar flow.1 Fig. ]. Plates forming the small-gap electrolysis cell. a) Mylar spacers 1. LBL report (Ph.D. thesis) to be issued in b) Centerline pressure taps, 75 u in early 1972. diameter c. Scaling 0-ring d) Sectioned electrode c) 7° flared exit 2. ELECTROLYSIS AT HIGH CURRENT DENSITIES The overall length of each plate is 23 cm. (XBB 7112-6128) a. Mass Transfer at Closely Spaced Planar Electrodes due to entrance to and exit from the channel, Raul E. Acosta, Rolf H. Muller, and Charles W. and minimization of the deflections of the Tobias cell walls when, under pressure. The former wa; achieved by providing a rounded entrance A new cell was designed to conduct limiting and a flared exit (7' diverging walls), while current measurements, using the potassium wall deformations were minimized by building ferri-ferrocyanide redox system on planai the cell out of 3.2 en thick stainless steel nic'el electrodes in a flow channel. The plates. Calibration and preliminary trial sectioned electrodes were designed to allow runs using the 0.5 mm thick spacer have been the study of mass transfer both in the entry completed. and in the fully developed flow regions; they extend 20 or more hydraulic diameters down stream from the point where mass transfer 1. Raul E. Acosta, Dieter Landolt, Rolf H. begins. Since the mass transfer boundary Mullcr, and Charles W. Tobias, "Electrolysis layer is extremely thin (of the order of at High Current Densities: Mass Transfer 10"4 cm under turbulent conditions) the in Considerations," I.M.R.D. Annua] Report 1970, sulating strips separating the electrode sec IJCRL-20500, pp. 39-40, April 1970. tions are only 25u in thickness. The cell de sign also provides for easy variation of both the channel width (1, 1.5, or 2 cm) and height (0.2 to 0.5 mm); the channel proper is formed b. Periodic Phenomena in the Anodic by Mylar spacers held in place by the two Dissolution or' topper plates that constitute the upper and lower walls (Fig. 1). Other important considera .John F. Cooper, Rolf H. Muller, and Charles W. tions in the design of the flow cell included Tobias minimization of the electrolyte pressure drop Electrode potential oscillations are known 5b to occur during dissolution of a variety of Constant Current," paper presented at 138 th metals.1.2 In the anodic dissolution of meeting of the Electrochemical Society, copper, periodic changes in anode potential October 4-9, 1970. occur in a number of electrolytes.3 Of particular interest are the exceptionally uniform, saw-tooth oscillations that ac company the galvanostatic dissolution of cop c. Anodic Behavior of Selected Metals at per in sodium chlorate solutions. The phe High current Densities^ nomenon nay be related to the desirable pro perties displayed by this electrolyte in Daniel J. Jacqiiot, Rolf H. Muller, and Charles electrochemical machining applications. K. Tobias In galvanostatic disolutlon of copper in The anodic dissolution of two heat- sodium chlorate solutions, the anode potential resisting alloys, llastclloy X and Inconol cycle consists of a steady rise in potential X-7SP, and of a low carbon steol, SAE lOlfl, to about 10 wits, followed by an abrupt fall has been studied in a channel cell of 0.5 mn to less than 1 volt. This cyclic rise and gap width under laminar flow conditions fall in potential corresponds to the periodic (Re - 350) in 5 H NaCl, NaClO. and NaNO., growth and rupture of a solid cuprous oxide electrolytes. Overall coll voltage and anode film on the metal substrate.* and cathode potential measurements were ob tained under galvanostatic conditions.in the A tentative explanation for these oscil current density range of 1 to 35 A/cur. lations involves^two reactions: Micrographs and roughness analysis of the (1) Aqueous Cu reduces CIO," ultimately to anode surfaces provided additional informa CI" by the net reaction: tion. 1 H* • CuCCy "" • i/6C103" - Die two alloys demonstrated similar elec trochemical behavior regardless of the nature of the electrolyte. Neither an abrupt jump Cu** (n-l/6)Cl" • l/2H 0 , (1) 2 in potential with increasing current density nor oscillations of potential were observed. (2) Aqueous CI assists in the conical The anodic dissolution of the low carbon dissolution of Cii-0 by complcxing iii*: steel, on the other hand, was strongly in fluenced by the choice of anion. The onset H* • l/2Cu20 • n Cl" - of a high-voltage dissolution mode above a critical current densit) has been observed in 1 NaCl and NaNO, solutions, while oscillations CuCCl,,) "" • 1/2H 0 . (2) 2 of the anode potential (Fig. 1) occurred at different current densities in all solutions. These two reactions may give rise l.o a pro MadeIs which involve the formation of anodic duction of CI" (and a pitting dissolution of films have been explored to explain the ob the cuprous oxide film) which is self- served phenomena. accelerating. The role of self-accelerating processes in periodic chemical phenomena has been mathematically treated by Boniioflffor. * Support for this mechanism may be found in the observed accumulation of chloride in the anolyte and in the pitting of the anode just prior to film rupture and removal. Further evidence will be sought in the simultaneous measurements of Cl" production and potential change at a rotating disk anode. 1. U. F. Franck, Z. Elektrochem. 62, 649 0958). Fig. 1. Oscillations of anode potential 2. K. F. Bonhoeffer et al., 2. Elektrochem. during the dissolution of low carbon steel 52, 29, 60, 67, 149 (1948). (SAE 1018) in S N NaCl at IS A/cm2. TT K. Kinoshita, Studies on the Anodic Dis (XBL 7111-4716) solution of Copper at High Current Densities, (Ph.D. thesis), UCRL-19051, Sept. 1969. 4. J. Cooper, R. H. Muller, C. W. Tobias, "Periodic Phenomena in Copper Dissolution at Smmary of LBL-428. 37 3. ELECTROCHEMISTRY IN NONAQUEOUS SOLVENTS Activity coefficients at each temperature are being determined as a function of the alkali a. Electrochemical Behavior of the Alkali metal chloride molality m,. Metals in Propylene Carbonate The feasibility of the electroreduction of Jacob Jom6 and Charles K. Tobias the alkali metals in propylene carbonate at room temperature is suggested as a new The stability and the reversible behavior eiectrorefining and separation process. t of the alkali metals were established in pro is proposed to modify the well-known mercury- pylene carbonate solutions. Lithium, sodium chlorine cells process, by replacing the potassium, rubidium and cesium were electro- amalgam decomposition stage with a nonaoueous deposited, in pure form, at room temperature propylene carbonate electrolysis cell, in out of solutions of their chlorides. which the alkali metal is dissolved fr.m its amalgam and deposited in a pure metallic form The charge transfer kinetics of the at the cathode. clectrodeposition-dissolution process was investigated, using micropolarization and pulse techniques. Exchange current densities were found to be on the order of 0.01 - 1.0 rnA/crn2. The enthalpy of activation at zero polarization is on the order of 7 - 10 kcal/mol (Fig. 1). 4. RESEARCH PLANS FOR CALENDAR YEAR 1972 Charles K. Tobias a. Fundamental Studies of Ionic Mass Transport Study of tH distribution of limiting currents on planar electrode surfaces in channel flow will be continued. The flow regimes considered include laminar and tur bulent forced convection (to Re-100,000) as well as turbulent free convection combined with laminar bulk flow. This study, employing the copper deposition reaction as a vehicle, is paralleled by precise optical observations of the mass-transfer boundary layer by means of a traveling Mach-Zehnder interferometer (see section on "Optical Studies of Inter- facial Phenomena.") The scope or tMs inves 3.0 J.l 3.2 tigation will be extended to include a study of reduction of boundary layer thickness by Fig. 1. Determination of the enrlialpy at introduction of suspended non-reactive solids zero polarization, AHL, from a plot of i« to the electrolyte, and by periodic wiping of the sin-face ("windshield wiper" model). electrode; This continuing program concerning the quan A - solid Cs electrode. (XBL 721-20141 titative elucidation of effects of hydro- dynamic tlow of electrochemical reactions occurring at non-uniformly accessible sur Accurate thermodynamic properties of the faces, is aimed at increasing the throughput alkali metals and their chlorides in propyl capacity of electrochemical reactors--a ene carbonate-aluminum chloride (lm.) solution severe limiting factor in beneficial are being determined in a cell without trans applications. ference: PtCsJ/MCsVMClOnj), b. Electrolysis at High Current Densities AlCl3(lm)PC/TlCl (s)/Tl(Hg) (1)/Pt (s), The mechanism of detail processes involved in tlie high speed anodic dissolution of where M is Li, Na, K, Rb, Cs. metals and intermetallic phases vill be further analyzed: a) Transport rates to and Standard cell potentials are calculated from electrode surfaces (to 0.2 mm gap) will at 2S, 30, and 35*^, in order to establish a be measured and correlated, using a segmented scale of standard electrode potential of the nickel electrode and the ferrous-ferrocyanide alkali metals in AlClj- PC solutions. reaction, b) The regular oscillation of 3i? potential in the galvanostatic dissolution of 3. D. Landolt, R. H. Muller, and C. W. Tobias, copper in solutions containing chloride ions Transport Process in ECM, in Fundamentals of will be critically examined to establish the Electrochemical Machining, edited by Charles applicability of the model proposed by L. Faust, (The Ulectrocnemical Society, Inc. Bonhoeffer for periodic chemical reactions, 1971), pp. 200-226. c) By using a locally modified commercial ECM machine the development of test shapes 4. J. Cooper, R. II. Mullcr, and C. W. Tobias, and procedures will be continued, with Periodic Phenomena in Copper Dissolution at emphasis on the selection of proper opera Constant Current, ilid., pp. 300-31S. tional ran$.e of practical elcctromachjning parameters. 5. Jan Robert Selman and John Newman, Free- Convection Mass Transfer with a Supporting c. Electrochemistry in Nonaqueous Solvents Electrolyte, J. Electrochem. Soc. 118, No. 7, 1070 (1971). Further characterization of the electro chemical behavior of alkali-motal-halidc Papers presented couples in propylene carbonatc-AlCl. solvent will include determination of activity 1. Lectures presented to local sections of coefficients of the alkali halidcs in the Electrochemical Society, "Electrolysis PC-lmAlCl-. The practical feasibility of at High Current Densities" at: Cleveland, metal refining operations (amalgam to pure Washington, Philadelphia, and Columbus; also metal] will be established. Analogous studier Columbia University, New YorJc, "Electrolvsis (i.e., thermodynamic reversibility, chirge at High Current Densities" (Jan. 1971). transfer kinetics) will be initiated on the lanthanide series. LBL report and UCRL reports 1. Kirk Watson Beach, Optical Methods for the Study of Convective Mass Transfer Boundary S. 1371 PUBLICATIONS AND REPORTS Layers on Extended Electrodes (Ph.D. thesis), UCRL-20324, Sept. 1971. Charles W. Tobias and Associates 2. Jan Robert Selman, Measurement and Inter Journals pretation of Limiting Currents Ph.D. thesis), UCRL-20SS7, Sept. 1971. 1. D. Landolt, R. H. Muller, and C. W. Tobias, Crystallographic Factors in Hiijh-Rate Anodic 3. Kenneth George Hcllyar, Cinematic Studies Dissolution of Copper, J. Electrochem. Soc. of the Anodic Dissolution of Metals at High U_8, No. 1, 36 (1971). Current Densities (M.S. thesis), LBL-1S7, Sept. 1971. 2. D. Landolt, R. H. Muller, and C. W. Tobias, Anode Potentials in High-Rate Dissolution of 4. Jan Robert Selman and John Newman, Copper, J. Electrochem. Soc. 118, No. 1, 40 Migration in Supported Electrolyte Solutions (1971). with Free Convection, UCRL-20322, Jan. 1971. 39 Rolf H. Killer, Principal Investigator Cuffffdt der.sity ImA/cm2) 5 .0 IS 1. OPTICAL STUDIES OF INTERFACIAL PHENOMENA Conecnltoliuft qfadierl (M Cu SO4 cm- ) 5 10 a. Intcrferoaetry of Mass Transfer Boundary Layers Frank R. SfcUrnon, Rolf H. Mullcr, ami Charles K. Tobias When intorfcromctric Measurements of nass transfer boundary layers arc node, light deflection in the refractive index field gives rise to errors in the resulting interferograms. These errors have been calculated, using a previously developed technique1 for a family of concentration pro O.l 0.2 O.J 0.' files shown in Fig. 1, with the inside sur Iftltttoool rtfroclivr inati Qiaditnl tern"') face of the cell wall facing the light source Fig. 2. Error in boundary layer thickness in focus. Figures 2-4 show the calculated due to light deflection as a function of errors in boundary layer thickness, inter- concentration gradient at electrode- facial concentration, and interfacial con- electrolyte interface. (XBL 719-4291) Reiioctive index Cell thickness Glass wall thickness 1.335 1.340 (era) (cm) T o 1.0 1.0 • 1.0 O.S A 1.0 0.2 A 0.2 O.S o 0.2 0.2 centration gradient as a function of inter facial concentration gradient. The cell thickness is seen to bo a crucial factor, while the gloss wall thickness has little effect. In the above calculations, a linear rela tionship between refractive index (n) anJ CuS0.| concentration (noles/litcr) (C) was assumed. The following oore precise relation ship has been derived from scasurr-ants with a standard deviation of 0.000015'> and will be used in the future (valid for 2S°C). njj • 1.3324236 • 0.029808 C - 0.0064948 C2 • 0.0CS1937 C3 - 0.002804S C4. 0.0 as 0.2 0.3 0.4 The choice of the mathematical function Concenirition (MCuSQ,) which relates concentration and distance in a boundary liver is crucial in the analysis Fig. 1. Concentration profiles of boundary of interfcrograms. Presently, a 4-tena layers employed for the computations shown expression approximating an error function in Figs. 2-4. (XBL 719-4293) is used, but its complexity causes difficulty 40 b. Double Beam Intcrfcromctry for Electrochemical Studies Rolf H. Muller Principles of double beam interferometry for the study of refractive index fields and surface topographies have been viewed. Potentials and limitations of interferometric techniques have been analyzed and the capabilities of different interferometer designs compared. An extension of the con ventional interpretation of interferometric results has been indicated. c. Cinematic Studies of the Anodic rifacial rL'ffac• iva indor. gradient (cm" ) Dissolution of Metals at High Current Fig. 3. Iirror in intorfacial concentration Densities^ duo to light deflection. Other designations as in Fig. 2. (XBL 719-4292] Kenneth C. ilellyar, Rolf II. Miller, and Charles W. Tobias Current density (mA/cm*) _— T Previous observations on the formation of Q 5 10 15 20 E If—i 1—I—t—I 1—I—I—I—|—I 1—I—I—1—1—I 1 1 S solid reaction products in the high-rate Inlertaciol concentration gradient (MCuSOo cm'1) « F anodic dissolution of copper have been 0 5 ' ' 1 1 | 1 1 extended to metals and alloys which are 0.2 c technically shaped by electrochemical machin fi -4 ing. "\ 0.05 - 1 err -o.t - s/» •0.15 ~ «N. \0 -0.2 0.0 0.1 0.2 0.3 Interfociol refractive ir.de* gradient (cm"1) w Fig. 4. Error in interfacial concentration gradient due to light deflection. Other designations as in Fig. 2. (XBL 719-4289) in fitting calculated interference fringes to observed fringes. The use of different 2-term expressions is presently being explored to analyze interferograms of a relatively well-known boundary layer resuming from transport by diffusion and migration only. The purpose of this study is to provide a direct experimental check of the optical computations, and result in a suitable expres sion for future use in this work under the more complex conditions of convective trans $mw port. Fig. 1. Geometrically distributed pores in the oxide layer on high-purity copper after dissolution in 2 N KNO3 at 40 A/cm^. Focus 1. IMRD Annual Report for 1970, UCRL-20500, is on the Diane aEbve the surface (200X). p. 46. (XBB 712-769A) 41 Color motion picture photography com tests, the instrument will ^o used for the bined with electrical measurements of the study of transient surface ayers in high- cell voltage (at constant current) have been rate anodic metal dissolution. used in a study of the anodic dissolution of copper, nickel, carbon steel, and three nickel-based alloys in 5 N solutions of NaCl, Abstract of M.S. thesis, LBL-165. NaN03, and NaC103 flowing at 10 cm/sec fRe - 750) at current densities of 1 and 20 A/cm2. Some observations have also been made in free convection and forced convection up to e. Principles of Ellipsometry 150 cm/sec (Re = 12,000) and current densities 2 Rolf H. Muller up to 40 A/cm . Theoretical and experimental principles of The formation of solid reaction products ellipsometry, capabilities and limitations could be observed under many conditions; of the technique for the study of surfaces and however, contrary to the experience with thin films have been reviewed. Factors copper, the evolution of copious amounts of which need to be considered in the operation gaseous products, particularly in the dissolu of ellipsometers of present design have been tion of nickel and carbon steel, often considered and directions of future develop obscured the electrode surface. A new type ments indicated. of geometrically distributed pores has been observed in oxide layers on copper formed during the dissolution in nitrate solutions (Fig. 1). f. Optical Interference in Thin Films Summary of M.S. thesis, LBL-157. Charles R. Brown and Rolf H. Muller Series of generalized interference colors have been computed previously* for the d. Ellipsometry of Surface Layers optical interference in transparent films on any supporting medium. In addition to Hiram Gu, H. Joerg Mathieu, and Rolf H. Muller the usually considered phase change due to the optical path in the film, these color Ellipsometry has been used in two studies series also consider the phase change due to of surface layers on platinum substrates: reflection at both film surfaces. As a chemisorption of carbon monoxide and electro test of the colorimetric computations, the lytic film formation. In the study of thickness of films, derived from interference chemisorption of carbon monoxide, good quali colors, will be compared to thickness tative correlation was found between determinations which do not depend on the ellipsometric and Auger electron spectroscopic optical properties of the interfaces. measurements. The study of electrolytic Transparent solid films on different sub films showed that ellipsometry can be applied strates, to cover the widest possible range to transparent thick film measurements as well of optical phase changes due to reflection, as to thin film measurements. Profiles of will be prepared for use in these tests. 2.62 N and 6.05 N KOH films on a platinum electrode obtained by ellipsometry compared J. H. Turney, M.S. thesis, LBL-171. well with those obtained by the observation of interference colors. Ellipsometric measurements of a 0.94 N H,SO* film could be interpreted in terms of a dielectric film, 2. RESEARCH PLANS FOR CALENDAR YEAR 1972 while data of KOH films suggested absorption characteristics of the film which may be due Rolf H. Muller to the inclusion of foreign particles. Detailed procedures for the alignment of The computed effect of light deflection ellipsometers of fixed and variable angle of errors on the interferometric study of mass incidence have been given.* transfer boundary layers will be checked by conducting measurements on well-defined Construction of a Faraday-cell-operated boundary layers involving only diffusion and automatic ellipsometer with fast response migration. Simplified iteration procedures and high resolution (electronic design by will be developed for use on boundary layers D. E. McClure, mechanical design by W. T. resulting from convertive transport. Mass Giba) is in progress. After the development transport limitations for electrolysis in of suitable readout systems and performance narrow flow channels under high electrolyte 42 velocity and chemical factors responsible for 3. D. Landolt, R. H. Muller, and C. W. oscillatory phenomena in anodic metal Tobias, Transport Processes in ECM, in dissolution will be established (see section Fundamentals of Electrochemical Machining, E). edited by C. I. Faust (Electrochem. Soc., 1971), p.200. Construction of the automatic ellipsometer will be completed. Performance tests ivill 4. J. Cooper, R. H. Muller, and C. W. Tobias. determine the capabilities of the present Periodic Phenomena in Copper Dissolution at design. Transient anodic layers will then Constant Current, in Fundamentals of be investigated with the instrument. Electrochemical Machining, edited by C. L. Faust (Electrochem. Soc.7 1971), p. 300. Independent checks of the film thickness determination by use of generalized inter LBL reports ference color series will be conducted with transparent solid films on different sub - 1. J. H. Turney, Profiles of Supermeniscus strates. Electrolyte Films on Partially Submerged Gas Electrodes (M.S. thesis) LBL-17I, Aug. 1971. 2. K. G. Hellyar, Cinematic Studies of the Anodic Dissolution of Metals at High Current 3. 1971 PUBLICATIONS AND REPORTS Densities (M.S. thesis),LBL-157, Sept. 1971. Rolf H. Muller and Associates 3. D. Jacquot, R. H. Muller, and C. W. Tobias, ECM Studies of Engineering Materials Journals and books in a Flow Channel Cell, LBL-428, Dec. 1971. 1. D. Landolt, R. H. Muller, and C. W. Tobias, 4. H. Gu, Ellipsometry of Surface Layers Crystallographic Factors in High-Rate Anodic (M.S. thesis), LBL-165, Dec. 1971. Dissolution of Copper, J. Electrochem. Soc. 118, 36-40 (1971). 5. R. H. Muller, Double Beam Interferometry for Electrochemical Studies, LBL-115, Dec. 2. D. Landolt, R. H. Muller, and C. W. 1971. Tobias, Anode Potentials in High-Rate Dissolution of Copper, J. Electrochem. Soc. 6. R. H. Muller, Principles of Ellipsometry, 118, 40-46 (1971). LBL-187, Dec. 1971. 43 Sohn Newman, Principal Investigator The results firmly verify the experimental fact' that the Nusselt number varies as 1. DIFFUSION IN AQUEOUS KOH SOLUTIONS (ws/wf)0-45( where ws and Wf represent solids and gas phase mass flow rates respectively. Keinal Nisancioglu and John Newman Satisfactory agreement is obtained at low solids loadings with the Danziger correlation^ Efforts to measure the diffusion coef of heat-transfer data for gas-solids flow ficients of aqueous KOH solutions* have been systems. The theory predicts a decrease in continued. Leaks from the diffusion cell the dependence of Nu on the gas-phase Re with still presented considerable technical dif increasing solids loadings. ficulties. Measurements, which show no signs of leakage, are given in Table I. These data are yet to be tested for reproducibility. Table I. Diffusion coefficient of KOH in water at 25°C. Concentration DxlO' 2 (moles/liter) (cm /sec) 0.266 2.681 0.502 2.603 0.512 2.617 0.984 2.746 1.042 2.805 1.244 2.974 1. Kemal Nisancioglu and John Newman, BW3 Annual Report for 1970,UCRL-20500, p. 51. 1. C. A. Depew, Heat Transfer to Flowing Gas- Solids Mixtures in a Vertical Circular Duct (Ph.D. thesis), UCRL-9280, July 1960. 2. W. J. Danziger, Ind. Eng. Chem. Proc. Do;. 2. HEAT TRANSFER TO TURBULENT GAS-SOLIDS Dev. 2, 269 (1963). SUSPENSION FLOW IN PIPES Kemal Nisancioglu 3. MASS TRANSFER TO A PLANE BELOW A ROTATING 1 2 Experimental work ' has shown that the DISK suspension of fine solids in gas can produce considerable increase in the heat transfer Robert V. Homsy and John Newman coefficient as compared with clean gas at the same gas Reynolds mmier. A theoretical A large disk rotates parallel to a model is proposed, which treats the particu stationary plane in which a circular mass- late phase as a continuum, to calculate the transfer section is embedded. The convective temperature distributions of phases for tur diffusion equation has been solved for high bulent flow of a suspension in a pipe. Schmidt number using a diffusion-layer Solution schemes are developed for the entry approach. There is a region near the center and fully-developed regions. Numerical of the mass-transfer section where the solutions are obtained for the asymptotic diffusion-layer solution breaks down. The case of large axial distance down the pipe, convective diffusion equation in this region for which experimental data available for is the same as that encountered at the rear comparison.1>2 of a sphere in Stokes flow (region 5).* 44 Consequently it is possible to use the results profile was used to predict rates of momentum obtained there to predict mass transfer rates and mass transfer for turbulent flow between near the center of the mass-transfer section. rotating concentric cylinders. Two models Furthermore it should be possible to measure were proposed, and numerical studies mass transfer rates to the plane anc'. thus initiated. Comparison with published exper avoid many of the experimental difficulties imental results indicated that: (1) predicted previously encountered at the rear of a friction factors were low by a factor of sphere.2 one-half compared to observed values, (2) predicted Stanton numbers for mass 1. P. H. Sih and John Newman, Mass Transfer transfer were low by an order of magnitude. to the Rear of a Sphere in Stokes Flow, The behavior of both these factors did, Intern. J. Heat Mass Transfer 1£, 1749-1756 however, depend on the Reynolds and Schmidt (1967). numbers for the systems in the manner sug 2. Ping Sih, Mass Transfer to the Rear of an gested by the experimental results. Object at Low Reynolds Number Flow, UCRL- 20509, May 1971. 6. CORROSION OF ROTATING DISKS 4. DIFFUSION Or A BINARY FLUID ACROSS AN INITIALLY SHARP ISOUNDARY; BOUYANCY-DRIVEN Nader Validat and John Newn-ui HYDRODYNAMIC INSTABILITY DUE TO HEAT OF MIXING Tie experiments conducted by LaQue show that copper and iron disks rotating in sea Robert V. Homsy water are corroded differently. The copper disk is corroded from the periphery, while Recently Heines and Westwater have report in the case of iron the greatest attack occurs on the surfaces towards the center that ed experimental work investigating the effect move more slowly. The purpose of this work of heat of mixing upon Marangoni convection. is to study the corrosion of iron disks in The hydrodynamic instability phenomenon here more detail. is complicated, since it is driven by both surface tension variations (the Marangoni effect) and an adverse density gradient due The surfaces toward the periphery of the to the heat of mixing (the bouyancy effect). disk move with a higher velocity, and the It would seem profitable to isolate these result is that the amount of oxygen that two effects so as to identify their respective reaches them is higher compared to the amount roles. of oxygen transferred to the central part of the disk. Consequently, an oxygen This work treats the instability due to concentration cell is formed, and a current density changes arising from the mixing of a passes through the solution from the center completely miscible, binary mixture across an to the periphery. The anodic and cathodic initially sharp horizontal boundary. Time- reactions are: dependent basic state similarity solutions to the temperature and composition fields are Fe - Fe++ + 2e" , developed. A linear, "frozen time" stability + analysis of the equations of change yields a 02 + 4H + 4e" - 2H20 . complex eigenvalue problem from which a critical Grashof number may be determined. Newman's equations^ were used to find the Thus, the onset time of instability may be iron and oxygen current distribution, as well predicted for given fluid properties and as the potential distribution on the surface initial concentration difference across the of the disk (by using suitable polarization boundary. curves for iron and oxygen). 1. H. Heines and J. W. Westwater, "The The results show that the iron current Effect of Heats of Solution on Marangoni density (which is proportional to the rate Convection" (1971), Department of Chemical of corrosion) increases slightly from the Engineering, University of Illinois, Urbana, center of the disk, goes to a maximum, and Illinois 61801. then drops sharply to a low value at the periphery. The increasing part near the center is almost flat. 5. EMPIRICAL CORRELATION FOR TURBULENT COUETTE FLOW The distance of the point of maximum corrosion rate from the center is not Charles Mohr and John Newman constant. It is a function of different parameters of the system, such as rotational The Wasan-Tien-Wilke eddy diffusivity speed, oxygen concentration of the solution, 45 conductivity of the solution, and radius of mation to within 124 over the range of the disk. frequencies studied. An asymptotic solution for small frequencies was determined, and The potential distribution shown that the work is in progress to develop a singular electrode-solution potential difference perturbation expansion valid for large increases from the center toward the periphery frequencies. and, as a result, this potential difference is more positive near the edge even though this portion of the disk is acting as the cathode (and the central part is the anode). 9. FLUID FLOW IN A PROPAGATING CRACK John Newman and William H. Smyrl 1. F. L. LaQue, Corrosion 13, 303t (1957). 2. John Newman, J. Electrocfrem. Soc. 113, The fluid flow in a sharp, propagating 1235 (1966). crack is shown to produce cavitation in the fluid near the crack tip, at finite crack velocities. For crack shapes which differ by having "rounded" tips, one may have 7. MASS TRANSFER COEFFICIENT IN POROUS FLOW- cavitation only above certain crack velocities. THROUGH ELECTRODES These fluid flow results would be important in the calculation of mass transport charac Harry H. Yip and John Newman teristics in stress corrosion cracking. The mass transfer coefficient for packed beds at low Reynolds number flow (~10"Z) is studied by means of two porous flow-through 10. MASS TRANSFER TO A ROTATING SPHERE AT electrodes made of packed nickel spheres. The HIGH SCHMIDT NUMBERS electrochemical system chosen for this study is an equimolar solution of potassium John Newman ferricyanide and potassium ferrocyanide in excess supporting electrolyte, potassium The Lighthill transformation is used to nitrate. calculate1 limiting rates of mass transfer to a rotating sphere on the basis of shear- The mass transfer to the packed nickel bed stress calculations found in the literature. is followed through the cathodic reduction of the ferricyanide at limiting current. The mass transfer coefficients obtained can then 1. John Newman, Mass Transfer to a Rotating be compared to the theoretical values predicted Sphere at High Schmidt Numbers, J. by the empirical correlations for packed Electrochem. Soc. 119, 69-71 (1972). beds.* At very low Reynolds numbers one sus pects that the correlations will no longer be valid, and the mass-transfer coefficient will, instead, behave according to the pre 11. MASS TRANSFER IN ELECTROCHEMICAL SYSTEMS diction by the Graetz analysis.' John Newman 1. F. Yoshida, D. Ramaswami, and 0. A, Hougen, Three major reveiw works were completed.^-"^ AIChE Journal 8, 5-11 (1962) 2. J. Newman, The Graetz Problem, UCRL-18646, The first is a text or research monograph Jan. 1969 (see also UCRL-20547, June 1971). treating in part D the behavior of electro chemical systems on the basis of thermo dynamics, surface phenomena, and transport processes, developed in the first three 8. WARBURG IMPEDANCE OF A ROTATING DISK parts of the book. ELECTRODE In the second, the principles of current John Newman, Charles Mohr, and Robert V. Homsy and potential distribution and mass transfer in electrochemical systems are developed The impedance of a rotating disk electrode with a view toward their application in subjected to an alternating potential was analyzing and understanding the processes of calculated by numerical solution of the localized corrosion, such as pitting and convective diffusion equation. The resultant corrosion in crevices, under deposits, and complex (Warburg) impedances were found to in regions near gas-liquid-solid contact agree with a Nernst diffusion layer approxi (water-line corrosion). 46 The third, aimed at electroanalytical Newman, Analysis of Porous Electrodes with chemists, places emphasis on the mathematical Sparingly Soluble Reactants, J. Electrochem. techniques used to treat electrochemical Soc. 118_', 1251-1256 (1971).* systems. An original example here is a singular-perturbation treatment of the Debye- 2. L. Hsueh and J. Newman, Concentration Huckel problem at low concentrations, showing Profile at the Limiting Current in a that the formula of Debye and Hiickel for Stagnant Diffusion Cell, Electrochim. Acta activity coefficients should be expected to 16_, 479-485 (1971). fail at lower concentrations for electrolytes of higher charge types. The unifying con 3. L. Hsueh and J. Newman, The Role of cepts are outlined for elliptic regions in Bisulfate Ions in Ionic Migration Effects, boundary-layer problems, and results are Ind. Big. Chem. Fundam. 1£, 615-620 (1971). presented for the mass-transfer rate to short electrodes in uniform shear. 4. A. M. Johnson and J. Newman, Desalting by Means of Porous Carbon Electrodes, J. Electrochem. Soc. 118, 510-517 (1971).* 1. John Newman, Electrochemical Systems (Prentice-Hall, Englewood Cliffs, to be 5. J. Newman, discussion, J. Electrochem. published). Soc. U8_, 1966-1967 (1971). 2. John Newman, Mass Transport and Potential Distribution in the Geometries of Localized 6. J. R. Selman and J. Newman, Free- Corrosion, LBL-408, Nov. 1971. Convection Mass Transfer with a Supporting 3. John Newman, The Fundamental Principles Electrolyte, J. Electrochem. Soc. 118, 1070- of Current Distribution and Mass Transport in 1078 (1971). Electrochemical Cells, UCRL-20547, June 1971. 7. J. R. Selman and J. Newman, High Sc Limit of Free Convection at a Vertical Plate with Uniform Flux Condition, J. Heat Transfer 9_3_, 12. RESEARCH PLANS FOR CALENDAR YEAR 1972 465-466 (1971). John Newman 8. W. H. Smyrl and J. Newman, Limiting Current on a Rotating Disk with Radial Activity coefficients will be included in Diffusion, J. Electrochem, Soc. 118, 1079- a calculation of limiting currents for 1081 (1971). copper deposition from solutions of copper sulfate and sulfuric acid. The effect of 9. C. Tsonopoulos, J. Newman, and J. M. ionic migration on limiting currents should Prausnitz, Rapid Aging and Dynamic Surface be estimated for turbulent-flow situations. Tension of Dilute Aqueous Solutions, Chem. The current distribution on a rotating Eng. Sci. 26, 817-827 (1971).* spherical electrode will be treated below the limiting current. The relaxation of the Paper presented potential when the current to a disk elec trode is interrupted will be predicted. 1. John Newman, "Mass Transport and Potential Distribution in the Geometries of Localized Activity coefficients of mixed electrolyte Corrosion," invited lecture at the U. R. solutions can be predicted on the basis of Evans International Conference on Localized data for the binary systems. The accuracy Corrosion, Williamsburg, Virginia, December of this scheme will be tested. 6-10, lfi-71. Measurements will be continued on the diffusion coefficient of aqueous KOH solutions and on mass-transfer rates in flow-through UCRL and LBL reports electrodes, and measurements will be initiated on the kinetics of dissolution of Cd(0H)2 1. J. Newman, Mass Transfer to a Rotating and mass transfer to tjie plane below a Sphere at High Schmidt Numbers, UCRL-20599, rotating disk. "* June 1971. 2. J. Newman, Mass Transport and Potential 13. 1971 PUBLICATIONS AND REPORTS Distribution in the Geometries of Localized Corrosion, LBL-408, Nov. 1971. John Newman and Associates 3. J. Newman, The Fundamental Principles Journals of Current Distribution and Mass Transport in Electrochemical Cells, UCRL-20547, June 1. J. S. Dunning, D. N. Bennion, and J. 1971. 47 4. J. R. Selman and J. Newman, Migration in 6. W. H. Sinyrl and J. Newman, Detection of Supported Electrolyte Solutions with Free Nonuniform Current Distribution on an Disk Convection, UCRL-20322, Jan. 1971 Electrode, UCRL-20380, February 1971. 5. P. Sih, Mass Transfer to the Rear of an Object at Low Reynolds Number Flow fPh.D. Not supported by IMRD funds. thesis), UCRL-20S09, May 1971. 48 D. PHYSICAL CHEMISTRY Leo Brewer, Principal Investigator has been completed and has been submitted for publication. Additional work is necessary to 1. ENERGIES OF THE ELECTRONIC CONFIGURATIONS clarify the ultraviolet spectra; this study OF THE SINGLY, DOUBLY, AND TRIPLY IONIZED will be continued by Dr. Tellinghuisen at the LANTIIANIDES AND ACTINIDES University of Chicago. Leo Brewer Present address: Department of Physics, University of Chicago, Chicago, Illinois. The method previously developed, 1 for cal culating energies of spectroscopic levels of Present address. Physikalisches Institut the neutral lanthanide and actinide atoms dcr Universitat, Basti, Switzerland. from the thermodynamic properties of the metals has been extended to provide a tabula tion of the energies of the lowest spectro 5. R15I-ARCH PIANS i-OR CALENDAR YI'AR 1972 scopic levels of the ions of these elements. Energies are tabulated for the lowest spectro Leo Brewer scopic level of the configurations involving 4f, 5d, 6p, and 6s electrons for the lanthn- 'flic determination of thermodynamic data for nide ions and Sf, 6d, 7p, and 7s electrons for high temperature molecules is difficult, and the actinide ions. most methods do not yield sufficiently ac curate results. The development of a com bination of radiative lifetime measurements 1. L. Brewer, J. Opt. Soc. Am. 61_, 1101-1111 on a molecular beam with absorption or omis (1971). sion intensity measurements in a high tempera ture King-type furnace will be continued in the hope of obtaining an accurate method utilizing the third law of thermodynamics. 2. RADIATIVE LIFETIMES OF TiO AND VO King furnace measurements are being taken for TiO and VO, and molecular beam measurements Shih-Ger Chang are being attempted with TiO and high tempera ture atomic vapors. Attempts to measure radiative lifetimes by using a molecular beam have not been success Alloys of solutions of transition metals of ful to date, but absorption intensity measure the third to sixth groups with the noble ments in the King furnace have yielded Nf metals are of interest because of the unusual values. Attempts will be made to develop ly strong interactions, which are attributed more intense molecular beams of TiO to check to electron donations by the noble metals to the King furnace measurements by a direct the transition metals with vacant d orbitals. radiative lifetime measurement. The proper representation of the variation of the thermodynamic properties with concen tration is not known and presents a particu- larily important challenge. Vaporization 3. THERMODYNAMIC PROPERTIES OF SOLUTIONS OF measurements as well as other high tempera THIRD AND FOURTH GROUP TRANSITION METALS WITH ture equilibration studies are planned to PLATINUM GROUP METALS characterize the variation of Cibbs energy with concentration. John P. Stiga Manuscripts are in preparation on the fol Vaporization studies are in progress to lowing topics: the thermodynamics of alloys establish the dependence of Gibbs energy upon of Zr and the noble metals, the transition concentration in solutions of strongly probability of the Swan bands of C,, predic interacting metals. tion of saturating phases and soluBilities of intermetallic compounds in metallic iron, the gaseous and matrix spectra of MgO, the low- lying electronic states of CaO, the thermo 4. PHOTOCHEMISTRY OF IODINE AND ULTRAVIOLET dynamic properties of Mo and its compounds, SPECTRUM OF I the thermodynamic properties and phase dia 2 grams of the alloy systems of the lanthanide * + and actinide metals, and the spectroscopic Joel B. Tellinghuisen and Karl Wieland levels of the monohalides of the Group II The work on the photochemistry of iodine metals. 49 6. 1971 PUBLICATIONS AND RETORTS 2. I. Brewer, Spectra of the Actinides and Their Metallurgical Properties, presented at Leo Brewer and Associates Third International Transpluton;um Element Symposium, Argonne National Laboratory, Journals and books Oct. .0-22. 1971 1. Leo Brewer, Bibliography on the High UCRL and LBL reports Temperature Chemistry and Physics of Materials, Part II, A. Spectroscopy of Interest to High 1. [.. Brewer and Chin-An Chang, Electrical TempcraLure Chemistry, B. Reactions Between Conductivities of Pb/Xe and Ph/SF, Films, Gases and Condensed Phases, edited by M. G. UCRL-20S81, July 1971. ° Hocking, Vol. If, Nos. 2, 3. 4 (1971). 2. L. Brewer and J. Ling-Fai Wang, Tnfraral 2. L. Brewer and J. Tellinghui-;on, Detection Absorp'ion Spectro of Tsotopic Ozone Isolated of Iodine Atoms by an Atomic Fluorescence in Rare Gas Matrices, I.BL-1-11, Aug. 1971. Technique: Application to Study of Diffusion and Wall Recombination, J. Chcm. Phys. 54_, 3. L. Brewer, Relevancy of Research, I.HI.-195, 513:5-3 (1971). Oct. 1971. 3. L. Brewer, Energies of the Electronic 4. Cliin-An Chang, Matrix Insolation Studios: Configurations of the Lanthanide and Actinide Model for Static and Dynamic Matrix Shifts, Neutral Atoms, J. Opt. Soc. Am. 6]_, 1101-11 I.BI.-173, Sept. 1971. (1971). 5. L. Brewer and Chin-An Chang, Matr'j 4. L„ Brewer and J. Wang, Absorption Spectrin) Isolation of Pb and Pb , LBL-149, Ai 1971. of the Mg- Molecule in Solid Rare Gases, J. 2 Mol. Spectry. 40, 95-102 (1971). 6. L. Brewer and Joel Tellinghuisen, Quantum Yield for Unimolecular Dissociation of I. in 5. L. Brewer, Energies of the Electronic Visible Absorption, LBL-400, Nov. 1971. z Configurations of the Singly, Doubly, and Triply Ionized Lanthanides and Actinides, 7. P. II. Handel and C. Kittel, Van der Waals J. Opt. Soc. Am. 1. L. Brewer, The Complexity of High Energy Chemistry, presented at SO, Conference, University of California, Berkeley, May 20-21, 1971. new -3 ':•. ::ai*ric, ."r :'>-*£r.i:.* :>V>JS'z^.zt^r the int'.'rmolccular exchange interactions, the l. i.xTRonurnox excitoi bandwidth, the phonons, and tempera ture. In the limits of high and low tempera Ijiarles R. Harris ture, exciton migration can be described by two mechanisms irrespective of the explicit Research in these laboratories has centered form of phonon scattering one adopts. In on zero-field optically detected magnetic the first case, the high temperature inco resonance studies of i.Tolocular triplet states. herent limit, the excitation propagates Experimental and theoretical studies have through the solid via intcrmolecular exchange resulted in the development of new techniques interactions in a random walk fashid. In for tin. determination of the rates and this limit the exciton is essentially local mechanisms of both intr- and intramolecular ized on a single molecule ind interacts with energy transfer processes involving molecular a phonon or localized molecular vibration at excited states. Hiesc techniques have also every lattice site; consequently, phonon- been used to determine the symmetry proper exciton scattering destroys the wave character ties, electron distributions, and rtolecular of the exciton in a time fast compared to geometry of the lowest triplet states in a the propagation of the excitation. In this variety of aromatic and nzaaromatic molecules. limit the delocal'.zed stationary states of In a preliminary study we have succeeded in 'he crystal, the k states, are not good observing zero-field electron spin transitiors eigenfunctions for a description of the ex associated with coherent triplet Frenkcl citon for times exceeding the time between exc.tons. These measurements provide a scattering events, 10~12-10"'3 sec. The fundamentally new tool for studying the inter second limit, low temperature coherent action important in the transport of energy migration, is expected when the density of in solids such as the intermolccular exchange localized molecular vibrations and phonon interactions, exciton-phonon sc ittcring pro states are sufficiently low that phonon- cesses, and exciton-trap intera tions. exciton scattering occurs on the average less frequently than energy migration from mole cule to molecule in the crystal. Under such circumstances the energy should be propagated 2. nil-: DYNAMICS or COHERENT KXI'ITQN over many lattice sites before phonon scatter MI CHUT I OX ing intervenes, causing a change in the k state of the exciton and consequently a change in Anthony 11. Francis and Charles R. Harris the velocity of propagation. In such cases the stationary crystal states are good The d>Tiamic properties of energy transfer eigenfunctions for the problem, and the dynam in molecular crystals is an area that has ics of the migration is described by the generated extensive experimental and theoreti- group velocity of the wave packet of k ca' investigations by many researchers. Both states, V(k), while the derealization of the optical prcperties associated with singlet the wave packet can be thought of as the and triplet excitons and the magnetic average distance between scattering centers properties of paramagnetic excitons have been [i.e., the coherence length). The importance studied at high temperatures. Despite these of this type of migration is that it proidxles facts, little is known experimentally about an extremely effective means of transporting the dyn-Jiiu.es of excitons at low temperatures. energy over long distances in times much Because of the long lifetime of photoexcited shorter than those required in the random triplet states in molecular crystals, triplet walk limit, since on the average the direction Frenkel excitons are particularly interesting, of the migration is unchanged for much longer for they provide a direct mechanism for periods of time and consequently the mean extensive energy migration over long distances free path of the Frenkel exciton is greater. via small intermolecular exchange inter A theory fur microwave band-to-band transi actions. The problem of understanding the tions in triplet Frenkel excitons in the dynamics cf exciton migration in molecular coherent migration limit for a one-dimensional crystals rest? in properly assessing the crystal has been developed in these labora importance of exchange coupling between tories. The principal features of the theory electronic states on adjacent molecules, are: fa) the frequency of zero-field electron vibrational coupling between adjacent spin transitions is related to both the molecules, and phonon-exciton coupling. anisotropy of spin-orbit coupling in the Progress has been made in these laboratories first Brillouin zone and the zero-field spin in understanding the relationships between dipolar llamiltonian, fb) the intensity of the 51 transitions yields the bandwidth and there the phosphorescence, and indeed advantage of fore the effective exchange interactions this fact has been made in these laboratories between molecules in addition to the density in a preliminary experiment which demonstrated of states functions of the band, and finally, that the effective rate constant for trapping (c) the EPR transitions should be homogeneous of triplet excitons is implicitly dependent in nature since, as in the high temperature upon the coherence lifetime of the exciton. limit, nuclear contributions to the line- Both the theoretical and experimental bases width vanish. In addition the experimental for this phenomenon have been developed verification of coherence has been achieved in these laboratories. By varying the at liquid He temperatures. The method used concentration of traps by suitable doping of to study the coherent migration limit of molecular crystals one can prepare a set of triplet excitons is a combination of optical crystals whose trap concentrations span the spectroscopy and magnetic resonance. limits in which the exciton is trapped in a Specifically, the phosphorescence from the time fast compared to phonon scattering, triplet exciton band is isolated from the referred to as the concentrated trap limit, trap emission by a high resolution scanning or--in which the exci'- trapped in a time slow compared to on scattering--the monochromator at temperatures sufficiently dilute trap limit. In .• former case a low to insure that some of the spin alignment microwave perturbation of the sublovel of the triplet magnetic sublevels produced population in an oxciton k state is not by selective intersystem crossing from distributed to other k states via phonor.- singlet states is preserved for times on the oxciton scattering before trapping and order of the triplet lifetime. A microwave consequently the microwave-induced perturbed field is the;-, used to induce transitions population at k travels with a single group between the spin sublevels, and the resulting velocity of the wave packet that is associated change in the phosphorescence intensity is with k. The trapping rate therefore depends monitored as a function of microwave upon a specific velocity V(k). Under these frequency. circumstances, the population of exciton states at a particular value of k is given by Preliminary experiments at low temperatures the density of states, D(k), and D(k) = indicated that in one molecular crystal [VOO]"1; consequently, the effective studied fl.2,4,5-1Jtrachlorobenzene) the population of excitons trapped is the same dynamics of the Frenkel excitons could for for all k. The trapped population is there the most part be described in the coherent fore proportional only to the Boltzmann limit at 4.2°K; however, neither the coherence factor across the exciton band. In the time nor the explicit nature of phonon- dilute trap limit when the exciton scatters exciton scattering could be ascertained with the phonons sufficiently before trap although limits were placed on the coherence ping, a microwave perturbation of the spin time. alignment at a particular value of k will be distributed to other k states via phonon- exciton scattering. The resulting trap spin alignment will be modulated by a set of 3. THE DYNAMICS OF COHERENT 1KIPLET EXCITON k states which have been prepared initially TRAPPING by the microwave perturbation ana then by phonon scattering. In the limit that the Michael D. Fayer, Anthony H. Francis and microwave-prepared k population is scattered to all k states, the effective population of Charles B. Harris excitons trapped is determined by an average group velocity (V > over all the k states, If a crystal is doped with an impurity and thus trapped population would be propor which either serves as a trap or induces a tional to the exciton density of states trap, the electronic energy associated with function and the Boltzmann factOT since the exciton will be distributed to traps at (V >_ 1 * DTK). a rate dependent upon both the trap concen tration and the velocity of propagation of the exciton. If a microwave field induces a change in the spin alignment of the mag.'ietic sublevels of the exciton band, this pertur bation should be carried to the trapping 4. OPTICALLY DETECTED MAGNETIC RESONANCE AND sites via triplet exciton migration and would THE MEASUREMENT OF INTRAMOLECULAR ENERGY result in a change of :.he trap spin alignment. TRANSFER PROCESSES IN TRIPLETS Consequently, any property dependent upon the individual trap spin sublevels will be Charles B. Harris and Robert J. Hoover modulated by the microwave perturbation of the excitons. One propert/ known to be As a consequence of spin-orbit coupling, modulated by the spin sublevel population is phosphorescence from the lowest triplet 52 state of a molecule to the ground state alignment in the magnetic sublevels, and thus vibrational manifold is selective from the the functionality of the phase angle is individual magnetic sublevels and can be restricted only to ratios of radiative rate thought of as a sun of three components, one constants from the magnetic sublevels being from each of the vio-field magnetic sublevels: connected by the microwave field since the total depletive rates from the spin sublevels TX, Ty and tz. The amplitude of each com ponent is determined by the radiative rates can be considered constant at any one and the populations, which are in turn temperature. These ratios are, however, determined by intersystem crossing rates, different for vibrations of different overall total decay rates, and spin lattice relaxa symmetries because of selective spin-orbit tion effects. Advantage has been taken of plus vibronic and/or spin-vibronic coupling. these facts in the method of phosphorescence- microwave double-resonance (PMDR) spectro It has been shown that, within the frame scopy, a technique developed in these and work of the Born-Oppenheimer approximation, other laboratories. PMDR isolates two of the ratio of the radiative rate constants the magnetic sublevels with a microwave fkx/ky) to vibrations of the same total field and monitors the change in phosphore symmetry are equal, since vibronic coupling scence it induces at low temperatures within the triplet manifold affects all three (1.3-4.2°K). The sublevels active in emission spin sublevels equally and any differences in and the degree of spin alignment in the the nuclear potential functions of the spin lowest triplet state have been determined by sublevels caused by spin-orbit interactions these techniques in several molecules. This are extremely small. Thus the Pranck-Condon information is useful in determining triplet factors Cfiv) in the ratio k^/lsy" » V$TQ/K? fy state properties such as the relative rates cancel, and the ratio is determined by the and routes of intersystem crossing, as well ratio of the electronic contributions to the as the specific spin level contributions to transition moment fk|") for a given symmetry the phosphorescence in various idbronic T and polarization. Consequently the ratios transitions. are independent of the particular vibration being monitored in PMDR, provided the overall Specifically, the kinetics of a microwave- symmetry of the transition remains unchanged. induced change in triplet phosphorescence Measurement of the phase angle has allowed resulting from the saturation or inversion of us to separate vibrations according to their the populations of two of the three triplet symmetries in phosphorescence spectroscopy. spin sublevels has been treated in the case Experimental results on a variety of molecules where the microwave pulse is short compared in their excited triplet states have been ob to any radiative or radiationless process tained. and is applied repeatedly in the presence of continuous exciting optical radiation. At the time of the pulse the phosphorescence intensity will acquire a new value because of 5. PROPERTIES OF MOLECULES IN EXCITED TRIPLET the microwave-induced population change and STATES will relax back to some boundary value as a function of time after the pulse. The Michael J. Buckley, Charles B. Harris, and solutions of the differential equation yield Rodney M. Panos all the salient intramolecular energy transfer rates and relative populations of the individual spin sublevels. The determination of the orbital symmetry of excited triplet states of aromatic and sub A second technique has been developed for stituted aromatic molecules is a problem which determining these rates that depends upon has received a considerable amount of attention the quadrature phase relationships of the from theoreticians and experimentalists alike. decaying microwave-induced change in triplet In addition to benzene, the series of halo- phosphorescence. The importance of these substituted benzenes has been extensively phase relationships in PMDR spectroscopy is studied. However, the unambiguous assignment that they are expected to be the same for of orbital triplet symmetries of these mole measurements made monitoring phosphorescence cules has, m part, eluded the investigators. to vibrational states of the same overall symmetry. They are also independent of In the last few years the development of microwave power, and hence to the degree of optically detected magnetic resonance techni saturation of the magnetic sublevels, and to ques (ODMR) has provided experimentalists with the polarization of the microwave transition a powerful new tool, yielding data which, when moment. Since the phase angle is independent incorporated into the wealth of other spectro of the amplitude cf i repetitive signal it scopic data available, may remove the ambigu is also independent of the extent of spin ities in the assignment of the first excited 53 triplet state of many molecules. The value of in molecular crystals. This will be ap ODMR techniques lies in their ability to proached from two types of experiments. The selectively examine the individual magnetic first will be to observe, as a function of sublevels of the triplet state. It is in this temperature, the averaging effect of the context that we have studied the properties exciton optically detected magnetic resonance of the first excited triplet state in sub resulting from exciton-phonon collisions. stituted benzenes. In this regard the theoretical basis for the experiments will have to be formulated. Specifically, phosphorescence microwave Secondly, the observation of coherence times double resonance (PMDR) spectra of in excitons and the scattering processes paradichlorobenzene have been obtained by will be probed by utilizing picosecond laser monitoring low temperature (~1.3°K) trap techniques. Both methods, magnetic resonance phosphorescence and used to make symmetry and laser studies, should allow us to study assignments of various vibronic bands present the dynamics of energy transport over a wide in the spectra. In addition, ratios of range of time scales and temperatures. Final radiative rate constants of individual ly, a systematic investigation of radiation- triplet spin sublevels to several vibrational less relaxation phenomena associated with ex bands, the intrinsic lifetimes of all three cited triplet states will be started. Speci triplet spin sublevels, and selective fically, wo anticipate that the experimental polarized phosphorescence from individual data obtained from the investigation of the spin sublevels to various vibrational bands isotope effect (deuterium vs. hydrogen) on the have been measured. These data, along with radiative and radiationless rates from the in explicit considerations of the para-chlorine dividual triplet spin sublevels in planar perturbation on the aromatic ring, have been analyzed to yield the relationships between aromatic molecules can be used to distinguish the orbital symmetries and the sign and the various theoretical models. magnitude of the zero-field electron spin dipolar interactions in the triplet state for several halobenzenes. On these bases an assignment of the orbital symmetry of the lowest excited triplet state of p- 7. 1971 PUBLICATIONS AND REPORTS dichlorobenzene h»s been made. In addition, other properties of the excited triplet state Charles B. Harris and Associates related directly to the electron distribution have been determined from the optically Journals detected magnetic resonance. The structures uf the three zero-field transitions have been 1. C. B. Harris, Microwave-Induced Time interpreted in terms of a Hamiltonian Dependent Effects in Triplet Phosphorescence, incorporating the electron spin-spin, chlorine J. Chem. Phys. S4_, 972 (1971). nuclear quadrupole, and chlorine hyperfine interactions. It has been concluded from an 2. A. H. Francis and C. B. Harris, The analysis of the spin Hamiltonian parameters Relationship Between Microwave Spectroscopy, that the triplet spin density in the chlorine Coherent Triplet Excitons and Density of out-of-plane orbitals is small, the chlorine States Functions in Molecular Crystals, electric field gradient is significantly Chem. Phys. Letters £, 181 (1971). reduced in the \TT* state as compared to the value for the ground state, and that the 3. A. H. Francis and C. B. Harris, The molecule most likely is distorted to a C%fo configuration. Observation of Coherent Triplet Excitons and Density of States Functions in Molecular Crystals, Chem. Phys. Letters 9, 188 (1971). Finally, the same studies as outlined above 4. A. H. Francis and C. B. Harris, Coherent have been extended to both 1,2,4,5- Triplet Exciton Trapping in Molecular tetrachlorobenzene and 1,2,4,5- Crystals, J. Chem. Phys. S5_, 3595 (1971). tetrabromobenzene. 5. C. B. Harris and R. J. Hoover, Quadrature Phase Relationships in Phosphorescence Microwave Double Resonance Spectroscopy, 6. RESEARCH PLANS FOR CALENDAR YEAR 1972 Chem. Phys. Letters 12, 75 (1971). Charles B. Harris UCRL and LBL reports Our major emphasis will be to explicitly 1. M. J. Buckley and C. B. Harris, Optically determine the nature of processes that destroy Detected Magnetic Resonance in Zero Field, the coherent transport of electronic energy UCRL-20390 Rev, Feb. 1971. 54 2. M. J. Buckley and C. B. Harris, The Intramolecular Energy Transfer Processes, Lowest Triplet State of Substituted Benzenes: LBL-118, July 1971. I. The Optically Detected Magnetic Resonance of Paradichlorobenzene, LBL-128, July 1971. 5. M. J. Buckley, C. B. Harris and R. M. Panos, The Lowest Triplet State of Substituted 3. M. J. Buckley, C. B. Harris and R. M. Benzenes: II. The Phosphorescence Microwave Panos, The Assignment of First Excited Double Resonance Spectra and Orbital Symmetry Triplet States by Optically Detected Magnetic of the -VIT* State in Paradichlorobenzene, Resonance (ODMR) Techniques, LBL-106, July LBL-409, Nov. 1971. 1971. 6. M. J. Buckley, Optically Detected 4. C. B. Harris and R. J. Hoover, Optically Magnetic Resonance of Molecules in Excited Detected Adiabatic Inversion in Phosphorescent Triplet States (Ph.D. thesis), University Triplet States and the Measurement of of California, Berkeley, Lawrence Radiation Laboratory Report UCRL-20390, Feb. 1971. 55 2/3 Ronald R. Herm, Principal Investigator The Kg/g fits shown for the first five quenching gases in Table I are all less than 1. PHOTODISSOCIATION OF Nal VAPOR AND THE but comparable to predictions based on an ENERGY" DEPENDENCE OF THE QUENCHING OF orbiting model in which one assumes that Na*(3p 2p) BY FOREIGN GASES* quenching occurs if and only if the two ap proaching reactants can surmount the barrier Boyd L, Earl, Ronald R. Herm, Shen-Maw Lin, in the effective potential, formed by adding and Charles A. Mims the centrifugal repulsion to the true long- range dispersion attraction. The cross sec Rate constants, lo,, for collisional quench tions exhibited by SO2 and I2, however, are ing of gaseous Na*(3p 2p) have been determined too large to be understood in terms of the by measuring the attenuation of steady-state long range forces between the neutral D-lines fluorescence as a function of the pres roactants. Because lx>th of these gases are sure of quenching gas. The Na*(3p ^P) D-lines known to form stable negative ions, these fluorescence was excited by direct photodis- very large values of Qq(g) probably arise due sociation of Nal vapor by means of monochro to a transfer of the Na* valence electron to matic, 2100 - 2500 A radiation. (Preliminary the quenching gas m large intermolecular results of this work were reported in the 1970 separations. IMRD Annual Report.) Abstracted from a paper submitted to the An expression has been derived for the dis Journal of Chemical Physics [preprinted as tribution in laboratory speeds of an atom pro LBL-102). duced by photodissociation of a diatomic mole cule at thermal equilibrium. This expression has then been employed to calculate the dis tributions in speeds of the photodissociative- ly generated Na* atoms. These computed Na* 2. MOLECULAR BEAM KINETICS: REACTIONS OF speed distributions have subsequently been Mg, Ca, AND Ba WITH Cl2, Br2, AND IC1 employed to deconvolute the dependence of the measured quenching rate constant on irradia Ronald R. Herm, Shen-Maw Lin, and Charles A. tion wavelength, X, so as to arrive at abso Mims lute quenching cross sections, Qq, as a func tion of relative collision speed, g. Over the A general study of the reactions of gaseous range of g studied in this work (-1.0 - group IIA alkaline earth atoms by the crossed 2.5 km/sec), the measured quenching cross sec molecular beam technique is in progress in tions indicate a speed dependence of Qq(g) • our laboratory. The apparatus employs two s Ks/g^' , with s in the range of 4 - 6. Values thermal beams crossed at right angles. The of the K4, K5, and Kg fits for the seven product reactively scattered in the plane of quenching gases studied in this work are given the reactant beams is detected by means of a in Table I. space-charge-limited electron bombardment ionizer and quudrupole mass filter mounted in a rotatable, ultra-high vacuum chamber. Table I. Best experimental values3 of Kg for Product velocity analysis is not possible as Na* quenching cross sections varying as yet with this apparatus so that one expects /s to obtain reliable estimates of the center- Ks/g* . of-mass (cm.) product angular distributions, but only rough ideas of the product recoil 9 11 Quenching 10 K4 10" K5 10 K6 energies. gas Laboratory LAB MX product angular distri 0.88 0.78 co2 1.6 butions have been measured for reactions of C H 0.93 0.82 1." Mg, Ca, and Ba with Cl2, Br2, and IC1. The 2 4 alkaline earth metals are potentially bivalent CH CN 3 1.2 1.1 2.3 and thus might have been expected to react CF3C1 1.2 1.1 2.2 with the halogen molecules by means of long- lived intermediate complexes. However, C H 1.3 1.2 2.4 6 6 analysis of the Ca and Ba+Cl2 and Br2 data 1.8 1.6 3.3 indicate that the alkaline earth monohalide so2 2.6 product is scattered forward (i.e., in the h 2.9 6.0 M atom direction of incidence) in the cm. a ~2" 37^ system, Indicative of a short-lived, ultra- K is given in units of cm - (cm/sec) . direct reaction mechanism. Indeed, the 56 sis of the Mg+Cl2 and Br2 reactions indicate that the MgX cm. angular distributions still peak forward, but are somewhat broader than those of the corresponding Ca or Ba reactions. The MCI products of the M+IC1 reactions appear to scatter forward in the cm. system, with much the same behavior as the corresponding CI2 or Br2 reactions. However, the MI products, of comparable or somewhat lower intensities than the MCI products, are pro duced with very much broader, almost isotropic cm. angular distributions. 1. J. H. Birely, R. R. Herm. K. R. Wilson, and D. R. Herschbach, J. Chem. Phys. 47, 993 (1967). ~ Laboratory Scattering Anglo, 8 3. MOLECULAR BEAM KINETICS: REACTIONS OF Ca AND Ba WITH HI, N0 , S0 , PCI3, SnCl , CCI4, Fig. 1. Mass spectrometrically measured BaBr 2 2 4 product LAB angular distributions from the CH3I, CH2I2, AND 2-NITR0PR0PANE Ba+Br2 reaction are shown by the data points obtained during three different experiments. Ronald R. Herm, Shen-Maw Lin, and Charles A. The solid curve through the data points is Mims the EjBr number density LAB angular distribu tion back-calculated for E' = 2.5 kcal/mole Because we know of no previous kinetics from the CsBr cm. angular distribution re studies of gaseous alkaline earth metal atom ported for the Cs+Brj reaction in Ref. 1. reactions, the molecular beam apparatus de The dashed curve shows the calculated number scribed in the study of Mg, Ca, and Ba reac density distribution in centroid angles for tions with halogen molecules has been employed an energy-independent collision cross section. to study the reactions of Mg, Ca, and Ba Also plotted is a kinematic diagram showing with HI, NO,, SO,, PC13> SnCl4, CC14, CH3I, the source velocities for the most probable CH?l2> and 2-nitropropane. The results collision, the corresponding centjoid vector, indicate large (i.e., comparable to those of C, and relative velocity vector, V; the cir alkali atoms) cross sections for reactions cles indicate the BaBr recoil speed for a of Ba with these gases (except that Ba+CH2l2 few possible E' values (kcal/moie). has not yet been studied), whereas the cross sections for the Mg atom reactions are too (XBL 7112-2261) small to detect in our apparatus; Ca atoms exhibit a behavior intermediate between that of Mg and Ba. Reactions of these gases are striking qualitative feature of these reac discussed in individual set-- below because tions is their very close similarity to the they are found to span a wide variety of reactions of the analogous alkali metals. behaviors. Figure 1 illustrates this striking similarity by the almost perfect fit of the measured BaBr product angular distribution from Ba+Br, M. + HI --By virtue of the light mass of to a LAB angular distribution calculated from the H atom products of these reactions, the the cm. CsBr angular distribution reported1 MI product must be formed in the laboratory for the Cs+Brj reaction. This BaBr LAB moving with practically the centroid velocity. Thus, these experiments can provide no infor angular distribution was calculated by trans mation on the reaction angular or energy forming the reported CsBr cm. angular dis partitionings. We have calculated the ap tribution for masses appropriate to the Ba+Br. propriate distribution in centroid vector reaction, assuming that the BaBr and Br prod-' directions for right-angle collision of ther ucts always recoil with some one value of the mal beams. The calculations were performed recoil energy, E', while averaging over the for a reaction cross section, QR, correspond thermal velocity distributions in both re- ing to an activation energy, E*, in relative actant beair:. kinetic energy, E, i.e., QR » 0 for E < E* and QR = Qo(l-E*/E) for E > E*. Comparisons Results for the other reactions have yet of the measured MI angular distributions with to be analyzed carefully. Preliminary analy these calculations show that: (1) E* * 57 0 kcal/mole for the Ba+HI reaction; this observation is more naturally interpreted (2) E* >» 3 kcal/mole for the Ca+HI reaction; in terms of a reaction producing Cal? + CH2 and (3) no Mg+HI Reaction is observed, sug products. This latter situation would give gesting a large E . Thus, these experiments kinematics somewhat analogous to the M + HI establish bounds on the Bal and Cal bond en reactions in that the heavy Cal2 would be ergies because the HI room temperature beam found in the LAB with a velocity close to carries negligible vibrational and little ro that of the center of mass of the collision tational energies. Thus, D0(BaI) > D0(HI) = partners. 71 kcal/mole and D0(CaI) > 60 kcal/mole. Analogy with the features of the alkali metal reactions with hydrogen halides suggest that D0(CaI) •» 70 kcal/mole and D0(MgI) < ~70 kcal/ 4. MOLECULAR BEAM KINETICS: MAGNETIC mole. DEFLECTION AND PRODUCT VELOCITY ANALYSIS OF THE Li + Br2 REACTION NO?, SO2, andJ2Jj)2CHN02--The Ba+S02 and 2-nitropropane reactions give large Lara A. Gundel, Ronald R. Herm, and Charlotte BaO+ signals, indicating sharp forward product M. Sholeen scattering as found in the Br2 reaction; Ba+N02 has not been studied. The Ca+N02 re Although crossed beam studies employing action shows analogous behavior, whereas no product velocity analysis have been reported reaction of Ca with S02 or 2-nitropropane is for the reactions of K atoms with halogen observed. Although the BaO+ mass spectrometer molecules, the only reported crossed signal from the 2-nitropropane reaction could beam studyl of the analogous li atom reactions + be due to a BaN02 product, no BaNO/ or BaN0 simply measured total integrated (over recoil signal is observed and it seems likely that speed) LiX product laboratory (LAB) angular BaO is the reaction product in view of its distributions. This is unfortunate because high bond strength. The analogous K and Cs Ref. 1 argues that, within the alkali family, reactions with N02 and S02 have not been the distribution in product recoil energy measured in beams. The reaction of K with should be most sensitive to the features of nitromethane is reported to yield an almost the potential surface (at least to the manner isotropic KN02 product cm. angular distribu in which the reaction exoergicity is divided tion. between reactant attraction and product repul sion) for the case of the Li atom reaction by M + SnCl^, PCI?, and CCla —Ba reacts with virtue of its lighter mass. Recently, we &1CI4, PCI3, and CCI4 with reaction cross have measured the distribution in LiBr product sections decreasing in the order listed; Ca LAB speed and angle from the Li+Br2 reaction. has been observed to react only with &1CI4. Only the MCI ion signal has been seen mass- The apparatus was identical to that of spectrometrically. However, this need not Ref. 1 except for the inclusion of a slotted rule out a MC12 product, as any MCI, product disk velocity selector (SDVS) between the would be expected to fragment alirost ex beam collision zone and surface ionization clusively into a MC1+ daughter ion upon detector. Thus, the two thermal reactant electron bombardment ionization. Preliminary beams (with thermal velocity spreads) were analysis of the SnCl4 reactions indicates a crossed at 90°, and product scattered in the sharp forward scattering of the products. Very plane defined by the intersecting reactant rough analysis of the Ba+PClj and CC14 data beams was measured. After collimation by a suggest behavior different from that of slit arrangement, only LiBr product with a &1CI4, but has not yet provided product particular, but variable, velocity was trans angular distributions. mitted through the SDVS. An inhomogeneous magnetic field followed the SDVS and deflected M + CH.I —Both the Ba and Ca+CHjI reac- aside non-reactively scattered Li atoms, tions scatFer the MI product predominately thereby permitting a direct measurement; of backward in the cm. coordinate system, in the scattered LiBr. agreement with the behavior of the analogous K and Cs+CHjI reactions. Ultimately,a center-of-mass (cm.) polar LiBr product flux map, i.e., the flux of LiBr Ca • CH7I7 --A large reaction cross section scattered into cm. «.olid angle ui, centered is observed for the Ca+CH2I2 reaction, with in the reactant plane at polar angle 6, with the Cal+ mass spectrometer signal providing recoil speed u^g-, will be obtained from a product angular distribution which is almost these data by seeking the best back-calculated coincident with the E* - 0 calculated distri fit, averaging over beam velocity distribu bution in centroid angles. If Cal and CH,I tions, to the LAB data. At this point, are the reaction products, this would imply however, time has permitted only a rough, very small product recoil energy. However, nominal inversion of the LAB data to the cm. 58 90'c.m. 5. RESEARCH PLANS FOR CALENDAR YEAR 1972 Li + Br?-»LiBr + Br Ronald R. Herm The main thrust of our 1972 research ef forts will be the final assembly of a "super neutral crossed molecular beam machine," similar in design to that developed by Prof. Yuan Lee of the University of Chicago, which will make possible studies of a wide variety of neutral collision phenomena. Richard Behrens has been involved in the design of this machine during the past year, and we project a tentative start-up data of late 1972. This machine will replace (indeed, will pirate parts from) the universal detector machine currently employed in the study of alkaline earth atom reactions. Thus, Richard Behrons, Shen-Maw Lin, and Charles Mims (the 180' c.m.-«-=- 0*c.m. latter two graduate students should finish during 1972) will continue our studies of non-alkali reactions. We plan to complete our studies of the reactions of alkaline earth SO'c.m atoms (jiost of the remaining studies will be of Sr Mid Ba) and go on to examine reactions of alkaline earth dihalides and, possibly, Fig. 1. "Nominal inversion" cm. map which monohalides. Additionally, the new "super shows contours of constant cm. LiBr flux machine" will be used to study non-metallic intensity for scattering into solid angle us, systems; currently, study of carbon atom centered about polar angle 9 in the plane of reactions appears promising. reactant beams, with LiBr recoil speed ULiBr arid corresponding product recoil energy E' Charlotte Sholeen and Lara Gundel (who (kcal/mole). Included is the velocity vector should probably finish during 1972) will con diagram used for t>« nominal inversion. tinue to measure distributions in product angle and energy for Li atom reactions; Na (XBL 7112-2262) will be studied as well. Additionally, Lara Gundel should finish her study of the chemiluminescence, induced by 2537 A This nominal inversion proceeded by assigning irradiation, of the Hg* + CI, reaction (this both Li and Br2 fixed characteristic veloci work was reported in the 1970 IMRD Annual ties (chosen so as to correspond to the most Report). Charlotte Sholeen will design probable collision event). The cm. polar equipment to make possible measurements of flux map obtained by this procedure, shown differential cross sections for scattering of monoenergetic Li beams off of potentially in Fig. 1, illustrates, in agreement with reactive collision partners and possibly previous studies, that most of the reaction initiate these studies. exoergicity appears as product internal excitation. Figure 1 further illustrates that the distributions in product angle and Boyd Earl will continue the alkali halide recoil energy are approximately separable. photodissociation studies reported on in this Time has not permitted a quantitative com report. He has already modified the equipment parison of the map shown in Fig. 1 with to permit photodissociation in the vacuum UV. similar data published for the K+Br, reaction. These modifications should make possible However, the following qualitative comparisons quenching studies of: (1) first excited state of the Li and K+Br2 reactions are possible: alkali atoms by Br2 and CI2, which require the CI) the LiBr cm. angular distribution is corresponding alkali bromide or chloride as broader by about a factor of 2; (2) the excited alkali atom precursors and (2) the distribution in product recoil energy, E', higher electronically excited states of the is also broader for the Li reaction; and (3) alkali atoms. the nost probable E' value for Li+Br2 is slightly higher than that for K+Br,. Lambert Loh will continue the measurements, reported in the 1970 IMRD Annual Report, of the energy dependence of relative values of 1. D. D. Parrish and R. R. Herm, J. Chem. the total inelastic collision cross sections Phys. S\_, 5467 (1971). of metastiable Hg* atoms. Experiments will include studies of the following collisions: Deflection Analysis of Reactions of Li with (1) Hg- + Tl -i- Hg + Tl*; (2) Hg* + Li -* Hg + N02, CH3N02, SF6, CC1„, and CH3I, J. Chem. Li*; C3) Hg* + Lil - Hgl + Li*; (4) penning Phys. S±, 2518 [1971)" ionization of the heavier alkali atoms; and (5) possible ion pair production in collisions LBL report with alkali or thallium halides. 1. Boyd L. Earl, Ronald R. Herm, Charles A. Mims, and Shen-Maw Lin, Photodissociation of Nal Vapor and the Energy Dependence of the 6. 1971 PUBLICATIONS AND REPORTS Quenching of Na* (3p 2p) by Foreign Gases (LBL-102, Sept. 1971), J. Chem. Phys. (to be Ronald R. Herm and Associates published). Journal 1. David D. Parrish and Ronald R. Herm, Molecular-Beam Kinetics. II. Magnetic Johnston I.D.(lt) Page 1 60 Harold S. Johnston, Principal Investigator 0 and O3. Under stratospheric conditions of temperature, pressure, and solar irradiation, 1. INTRODUCTION some of these reactions are much more impor tant than others. In terms of formation and The purpose of this project has been the destruction of odd oxygen, the dominant "direct observation of unstable intermediates reactions are: in gas-phase reactions" under conditions of mild radiation fields, such as direct sunlight at the earth's surface. fThe method of flash Creation of odd oxygen: photolysis has been very successful in the (1) 02 + hu (below 240 nm) •* 0 + 0 direct observation of intermediates under (2) N0+ hv (below 400 nm) +N0 + 0. conditions of very high light intensity. 2 Some reactions follow a different mechanism b. at low and at high light intensities, and we were interested in certain complex Interconvorsion of related species of odd oxygen: atmospheric reactions.) The emphasis has (3) O + O7+M + O3 + M been on simple inorganic free radicals: CIO, (4) O3 + hv (190-350, 450-650 nm) + 02 + 0. C100, HO, H0O, UNO, NOj. The role of free radicals in converting NO to NO2 and in the formation and destruction of ozone has been Destruction of odd oxygen: of especial interest. I. The 0X system: (5) 0 + O3 * 02 + O2. This line of work was found to have an II. The NOx system: unexpected practical application in March (6) NO + O3 * NO2 + 02 1971. At a meeting of the Department of (7) N0 + 0 NO + O-T } catalytic cycle Commerce Advisory Board for SST Environmental 2 Effects, Boulder, Colorado, March 1971, I net 0 + O3 •* 0 + 0 . discovered that the SST planners had over 2 2 looked (or in one case incorrectly discounted) III. The HOx system: a very important variable in the problem, (8) HO + 03 - HOO * 02 ( catalvtic __.,,. namely, the oxides of nitrogen. The rate (9) HOO + 0 - HO + 0 I cataIytlc c5rcir- constants for the photochemical reactions of 2 the oxides of nitrogen are well known, and net: 0 + O3 •* 02 + 02 . if NOx (NO and N02) is as much as 0.1% of ozone in the stratosphere, it is a strong In all three systems (Ox, NOx, a™1 *Px) there catalyst for ozone destruction. The NOx are additional reactions and catalytic cycles increment from SST exhaust is this order of that destroy odd oxygen, but insofar as rate magnitude, and thus a large effect of NC^ on constants are known they are unimportant the global ozone shield is possible. However, compared to those listed above. the exact effect of NOx from the SST exhaust on stratospheric ozone depends strongly on The oxides of nitrogen (NO, N0 ) act as a (1) the amount and distribution of the natural 2 NOx background and (2) the amount and distri catalytic couple for the destruction of ozone, bution of NOx after artificial sources are and the free radicals based on water (HO, introduced into the stratosphere. HC) also act as a catalytic couple with the same net chemical reaction. The catalysts are not consumed in this process, and one mole cule of N0X in time can destroy a large number of ozone molecules. The steady-state 2. CATALYTIC REDUCTION OF STRATOSPHERIC OZONE assumption is quite good for N02 and for HOO: d[N02]/dt * 0, d[H0O]/dt « 0; and the dif Harold S. Johnston ferential equation for odd oxygen including these steady-state assumptions is: The photochemistry of ozone in the presence of water vapor and the oxides of nitrogen d([03]+[0]) represents a complicated situation involving 2j1[02] - 2k5[0][03] scores of elementary chemical reactions. —at Many of these reactions act to interchange related species with each other: 0 and O3, NO and NO?, HO and HOO, etc. A relatively 2k7[0][N02j - 2kg[O][H0O]. small number act to increase or decrease "odd oxygen," that is, the interconvertible pair, To a first approximation, the rate of 61 formation of ozone and atomic oxygen, Br2, and I2) have been studied up to high 2Ji[02l> *s tne same> with znd without the temperatures, to 18,000°K in one case. The catalysts N02 and HOO. The destruction of chemical reaction is the simple breaking of odd oxygen (essentially ozone) is 2^5103][0] one chemical bond in pure dry oxygen, the destruction rate is X +M-X + X + M. increased by the additional amount 2kg[0][HOO] 2 in the moist stratosphere, and it is further The collision partner M has been argon for increased by 2ky[0][N02] by oxides of most cases, but it can also je other noble nitrogen. The rate constant 1:7 is about the gas atoms or X or X2. A naive view of this same order of magnitude as kg; but, according process leads to the expectation that the to my estimates, the concentration of N02 in activation energy should be at least as large the natural stratosphere greatly exceeds the as the bond dissociation energy. In all concentration of HOO; and the rate of the cases, however, the observed activation energy catalytic cycle III is much slower than the is far less (up to tens of kilocalories per catalytic cycle II. At 220°K, » normal tem mol) than the bond dissociation energy. In perature in the lower stratosphere, the rate a recent discussion of this phenomenon one constant k7 exceeds the rate constant kg by author1 stated: 'This discreponcy between the factor 4600. Thus if N02 is one-tenth En and D0 has been attributed to vibration- of one percent of ozone, the rate of ozone dissociation coupling by some kineticist.s. destruction by N02 catalysis is 4.6 times However, attempts; to demonstrate quantitative faster than the rate of ozone destruction in ly that vibration-dissociation coupling is the natural oxygen system, and the steady- responsible for this discrepancy have not state concentration of ozone would be reduced been successful thus far. This problem by the square root of one plus 4.6, /5.6 • remains unresolved." We have submitted an 2.3. With an increase or decrease of N02 article on this subject that gives a theory about this reference point (0.1* of O3), the of this effect. It employs no adjustable reduction of ozone varies relatively slowly, parameters, but it uses empirical very nearly as the square root of N02. translational-vibrationul energy-transfer constants. This theory involves a non- For the projected fleets of supersonic equilibrium distribution over vibrational transport planes, a large portion of the lower states. It quantitatively accounts for the stratosphere would be loaded with NOv (NO and low activation energy paradox for all cases, N02) in amounts as great as 0.1* of the and it gives fairly good absolute values to present ozone concentration. Large reductions the rate constants. of ozone are indicated. The problem is too complicated to be stated simply in terms of fractional reduction of the ozone shield. 1. M. Warshay, J. Chem. Phys. 54. 4060 (1971> The total problem involves many technical considerations: the rate of radiation absorp tion throughout the stratosphere and how it 4. MOLECULAR MODULATION SPECTROMETRY is changed by a reduction in ozone, the present distribution of NOx in the strato Alan B. Harker, Richard A. Graham, and sphere, the rate of conversion of NOJJ to Harold S. Johnston HNO3, and the role of turbulent diffusion ani. large-scale air transport. These problems A new infrared molecular modulation are considered in a series of reports and spectrometer has been built which is designed articles. The simple argument is that the to give very convenient low-resolution catalytic cycle II is very much faster than scanning from 800 to 3200 cm"1. The dis the natural process I, which it imitates, persing element is a circular, continuously and the amount of NO^. to be added by full- variable, multicoated optical filter. No scale SST operation is sufficient to have a new results are yet available from this strong effect on the ozone layer. system. The ultraviolet molecular-modulation apparatus has been remodelled from a single- 3. DISSOCIATION OF HOM3NUCLEAR DIATOMIC reflection 4 meter optical parti to a M3LECULES AT HIGH TBIPERA.TURES multiple-reflection 40 meter optical path. With this system we can follow reactants, John Birks and Harold S. Johnston such as ozone or oxides of nitrogen, in the range of 10"° atmosphere, and modulated radi During the past 10 years the kinetics of cals can be followed a factor of 100 or more the thermal dissociation of homonuclear lower. This apparatus and the SO meter infrared apparatus can be used in series or diatomic molecules (H2, D2, N2, 02, F2, CI2, 62 in parallel to study the same reactions. NO, + hv(visible) + NO + 02 Recently we have been carrying out extensive calibration and cross-comparison studies in + N02 + 0 the system involving NO, NO2, N2O5, 03, HNO3, and H2O2. We are studying the rates of the elusive reactions: HOO + NO -* ? and Also, we plan to study the photolysis of HO • N0 •+ HNO,. This work is in progress. dinitrogen pentoxide and the kinetics of its 2 reaction with oxygen atoms N 0 + hv - 2N0 + 0 5. FLUORESCENCE OF SULFUR DIOXIDE 2 5 2 N 0 • 0 - 2N0 • 0 Douglas R. Martin and Harold S. Johnston 2 5 2 2 We are studying the fluorescence of sulfur dioxide where the molecule is excited by In each case above there are two alternate narrow-band ultraviolet radiation (10 A band reactions, and the effect of the oxides of width monochromator) and observed with a nitrogen on ozono in thu lowest stratosphere moderate resolution (1 A band width depends on which predominates: monochromator), The process is carried out in a large (22 liter) glass bulb with silica N0 • 0 N0 • 0 windows. The pressure region is down to 1/2 2 3 * 3 2 micron. The effect of total gas pressure, NO, + hv NO + 0 including inert gases, on the rate and wave •* 2 length distribution of the fluorescence is NO + 0 •* N0 + 0 being examined. As was developed in a 3 2 2 similar study of NO2 fluorescence,! such measurements give detailed information about net: 20 + hv(VISIBLE) + 30 the vibrational-translational energy-transfer 3 2 rates for a highly excited triatomic molecule. NO, • 03 - N05 • O, Stephen Swartz and Harold S. Johnston, N0 + hv ->• N0 + 0 Kinetics of Nitrogen Dioxide Fluorescence, 3 2 J. Chem. Phys. 51, 1286 (1969). 0 + 02+M-<-03+M net: no reaction 6. RESEARCH PLANS FOR CALENDAR YEAR 1972 Harold S. Johnston An elusive reaction of the hydroperoxyl free radical with nitric oxide is of con 1. Laboratory studies of reactions important siderable importance and interest in both in the stratosphere. the troposphere and stratosphere Our laboratory apparatus and experience HOO + NO - HO + N02 make it possible for us rapidly to obtain the rate constants needed to fill in certain Work is underway to try to measure the rate gaps of knowledge concerning photochemical of this reaction. A combination of infrared reactions in the stratosphere, especially and ultraviolet molecular modulation spectro the reactions of NO3, N205 and HNO3. We metry is required to separate this reaction plan to study the wavelength and temperature from the many other simultaneous reactions dependence of the quantum yield of nitric in the system. acid vapor The fluorescence of sulfur dioxide gives HN03 + hv - HO + N02 information about singlet-triplet conversion in a simple molecule, about cc'lisional •* H + N03 deactivation of a highly vibrationally excited molecule, and about the superposition of The wave length range of interest is from states in highly excited molecules. Further 320 mm down to 200 ran. We plan to study the experimental work will be done using narrow wavelength dependence of the products of the band excitation and narrow band detection of photolysis of the free radical NO3 fluorescing radiation. 63 7. 1971 PUBLICATIONS AND REPORTS 2. H. S. Johnston, Catalytic Destruction of Stratospheric Ozone. Submitted to National Journal Aeronautical and Space Council, Washington, D. C, April 2, 1971. 1. H. S. Johnston, Reduction of Strato spheric Ozone by Nitrogen Oxide Catalysts from Supersonic Transport Exhaust, Science 173, 517-522 (1971). UCRL reports Papers presented 1. II. S. Johnston, Catalytic Reduction of 1. H. S, Johnston, The Role of Oxides of Stratospheric Ozone by Nitrogen Oxides, Nitrogen on the Ozone Level of the Strato UCRL-20568, June 1971. sphere with Respect to the SST. Presented to the Department of Commerce Advisory Board 2. T. T. Paukert and II. S. Johnston, Spectra for SST Environmental Effects, Boulder, and Kinetics of the Hydroperoxyl Free Radical Colorado, March 19, 1971. in the Gas Phase, UCRL-19J09 Rev. July 1971. 64 Bruce H. Mohan, Principal Investigator He H2 -> He H + H AH = - 8 eV 1. MOLECULAR ORBITAL CORRELATION DIAGRAMS .AM) has never been detected. Use of the correla ION-MOLECULE REACTION DYNAMICS* tion diagram shows that the configuration He + H. will+evolve to H + He H (la 2a). The Bruce H. Mahan molecule He H in this excited configuration is+unstable with respect to dissociation to By examining the symmetry and modal proper He + H. Thus only inelastic and elastic scat ties of the orbitals of reactants, collision tering can result from collisions of these intermediates, and reaction products, we have reactants. developed a number of diagrams which show how the reactant orbitals evolve to product orbit - Correlation diagrams can also explain why als during a collision. An example is given in the CO (H-, H) HCO* and N, (H-, H) N-H reac Fig. 1, which shows+the correlation diagram for tions proceed by direct interaction mechanisms the system (He H H) in linear geometry. Start despite the substantial stability of the inter ing with the electronic configuration corre mediate ions H.CO and N,H. . Several other sponding to He + H, , it is easy to see that applications of the diagrams are discussed in reactants should evolve to products in their the paper. Despite their qualitative nature, ground states. Thus the reaction should pro correlation diagrams provide a useful guide to ceed readily, and indeed, the reaction the understanding of collision processes. He >• H -• He H + H AH = 0.80 eV 2 Abstracted from J. Chem. Phys. 55_, 1436 (1971), and UCRL-20550, March 1971. is well knoim. In contrast, the very exother mic reaction 2. DETERMINATION OF INTERM0LEOJLAR POTENTIALS FROM LARGE ANGLE ELASTIC SCATTERING* ^IHeH*) Eric A. Gislason, Ming-Hwa Chiang, Bruce H. Mahan, Chi-Wing Tsao, and Arthur S. Werner It has become common practice to use the energy dependence of the total scattering cross •10 section and the variation of the small angle differential cross section to deduce inter- molecular potentials. These techniques require that data be obtained with an apparatus of high angular resolution, since the small angle cross sections vary rapidly with angle. In contrast, the differential cross section varies only rather slowly with angle at large (~ 180°) scattering angles, and reliable measurements -20 can be made with an apparatus of low resolu tion. K(He) In these papers, we show how large angle IslHeH*) scattering data can be analyzed to yield relia ble intermolecular potentials in the relatively high energy repulsive regime. An expansion for (He+Hj)* (HeHHT HeH*+H the scattering angle valid at large angles is -30 developed, and used to calculate large angle differential cross sections for repulsive Fig. 1. A correlation diagram |or the lowest power, exponential, and shield Coulomb poten orbitals of the linear (He H H) system. The tials. Numerical formulas are given which al orbital energies are only semiquantitative, low one to choose values for the parameters of except for the atomic species. The lowest- these potentials from the observed angular order correlations are indicated by the dashed dependence and magnitude of the differential lines which cross. (XBL 721-5900) scattering cross section. As an example, the 65 potential for the Ar -He interaction in the Ne + H. NeH + H AH = - 5.4 eV range from 2.2S - 1.25A (0.6 - 10 eV) is deter mined from our measurements of the elastic has been a source of confusion and mystery in scattering in this system. The method is successful, and its wider application should the field of ion-molecule reactions for a lead to an increased understanding of inter- decade. It has been suggested that the product molecular potentials at small internuclear molecule is too excited vibrationally to be separations. stable, or that the charge transfer channels Abstracted from J. Chem. Phys. SS 3929, 3937 Ne Ho Ne + H„ (1971], and UCRL-20712 and UCRL-2u"713. Ne + E, Ne + H + H 3. DECONVOLUTION OF M3LECULAR BEAM INELASTIC compete with and overwhelm the reactive chan SCATTERING DATA* nel. However, it has been found that these charge transfer channels have very small cross Keith T. Gillen and Bruce H. Mahan sections, and so the competition argument is not the correct explanation for the failure to Laboratory measurements of the elastic, form NeH . inelastic, and reactive scattering of molecules and ions are to some degree distorted by ap By using molecular orbital correlation dia paratus resolution effects which arise from grams, it is possible to show that Ne + H- finite slit widths, band passes of velocity must produce NeH in its first electronically analyzers, and thermal motion of target gas. excited state, and not its electronic ground The value of such measurements increases if the state. The excited state of NeH should dis effects of finite resolution can be removed at sociate to Ne + H. Thus one can predict that least partially by deconvolution techniques. at low (< 4.J eV) relative energies, the scat tering of Ne by H- should be largely elastic, In the paper, two common deconvolution tech perhaps somewhat inelastic, and not reactive. niques and their application to large angle At higher energies there should be marked molecular scattering are considered. One meth od, the iterative unfolding technique, is shown to contain an error in its common mode of Ne+(Hs,H6)Ne*(75.0«V) , application. This error is corrected, but 12.5 oV Relative Energy (+90° other limitations intrinsic to the method are shown to exist. The second deconvolution method, the integration-fitting technique, which involves a folding together of the resolution functions and various assumed scattering functions, is shown to be more flexible and generally more satisfactory than the iterative unfolding meth od. The integration-fitting method is applied to some inelastic scattering data from this -iO'A laboratory, and is shown to lead to inelastici Btom ties which are significantly different from Profllt those deduced from the untreated data, or from the iterative unfolding treatment. The refined data are in agreement with the predictions of a simple theory of inelastic collisions over a substantial range of energies. Abstracted from a paper accepted by J. Chem. Phys., and LBL-199. Fig. 1. A+contour map of the specific inten 4. NQNREACTIVE SCATTERING IN THE Ne - H, AND sity of Ne scattered from He. The origin of Ne - He SYSTEMS the coordinate sysjem is the center-of-mass velocity of the Ne - He system, and zero Bruce H. Mahan and John S. Winn scattering angle is defined as the direction of the Ne beam. The large circle marked The failure to observe the exothermic reac Q = 0 represents the locus of elastically tion scattered Ne . PCBL 721-5901) 66 Ne^lHj.HjHJ^daO.ieV) 5. DYNAMICS OF THE REACTIONS OF 0 WITH H,, D2, AND HD i. L 13.66 eV Relative Energy t+ Ming-Hwa Chiang, Keith T. Gillen, Bruce H. Mahan, and Arthur S. Werner We jjave nearly completed our studies of the 0, - H, system and its isotopic variants, which has proved to be the system which is perhaps the richest in variety and amount of dynamical behavior of any yet investigated. Three papers have been published on the reac tions in this system, and the data on isotope effects, product intensity ratios, and detail ed velocity distribution."! is currently being analyzed. At relativj energies less than 5 eV, col lisions of D, and H, proceed through a long- lived collision complex H,0- which decomposes in three channels: -HC2 *H Fig. 2. A+contour map of the specific inten sity of Ne scattered from H,. +Note the +H li 0 -*• OH* + OH relatively high intensify of Ne near 180°, °2 2-*( 2 2' as compared with the Ne - He system in - H 0 + 0 Fig, 1. Note also that the maximum intensity 2 in this region occurs inside the elastic Q = 0 circle, thus indicating that this large angle The product ratios HO, / OH / H20 are ap scattering is inelastic. (XBL 721-5902) proximately 10/1/1 in our molecular beam ex periments. However, the same decomposition can be observed in the mass spectrum of H202, and in this case the product ratios are inelasticity due to the occurrence of the 0.8/1/2.6. The discrepancy between the two dissociative process types of experiments is very interesting, and suggests that the collision complex has sig + Ne + H2 - Ne* + H + H. nificantly different properties in the two cases. Figure 1 shows |he distribution of velocity vectors for the Ne (He, He) Ne process, which We have applied RRKM unimolecular decom provides a convenient reference pattern for the. position theory to the dissociation of excited scattering of Ne by H,. The scattering of Ne H,0, , and are able to predict product ratios by He is completely elastic, although there is tnat agree with the mass spectrometric re some loss of intensity at scattering angles sults, but not with the molecular beam re near 180°, due to a charge transfer+process. sults. However, conventional RRKM theory Figure 2 shows the scattering of Ne by H, at treats the angular momentum properties of the nearly the sane energy. It is clear that the complex in a simplified way, which-can lead small angle scattering is elastic, whereas the to large errors if the reduced masses in the large angle scattering is slightly inelastic. various channels differ substantially. Upon This inelasticity becomes more marked at higher treating the problem with the phase space relative energies. It :> also evident that the theory2 of chemical reactions, which handles intensity near 180° is ljigh, and in excess of angular momentum in a more accurate way, we that observed in the Ne - He system. Thus find qualitative agreement between the pre Ne is not being lost by charge transfer. In dictions of the theory and both the molecular all respects, the behavior of this system is consistent with the predictions based on orbit al correlation diagrams, and thereby 1. M. H. Chiang, E. A. Gislason, B. H. Mahan, strengthens confidence in the usefulness of C. V. Tsao, and A. S. Werner, J. Phys. Chem. these diagrams. 75>, 1426 (1971), and references therein. 2. J. C. Light, Discussions Faraday Soc. 44, 1. B. H. Mahan, J. Chem. Phys. 55^, 1436 (1971). 14 (1967). 67 beam and mass spectrometric data. The signif analysis and for reactant state selection will icant difference between the two experiments be explored and developed, in order to extend is that in the former, the collision complex the range and detail of our experiments. All has significantly more angular momentum than these projects are aimed at increasing our in the latter. The intensity of products in general understanding of chemical reactions, channels where the reduced mass is very small and specifically at clarifying the higher is very sensitive to the total angular momen- energy charged particle collision phenomena fm, thus accounting for the excess HO. in which are important in discharges, atmos the molecular beam experiments. pheric and re-entry phenomena, and radiation chemical systems. The value of phase space theory in ex plaining product intensity ratios has been demonstrated. We are currently exploring the application of the theory to velocity distri 8. 1971 PUBLICATIONS AND REPORTS butions and isotope effects. Bruce H. Mahan and Associates 6. THERMAL ION-MOLECULE REACTIONS IN GASEOUS Journals METHANE-OXYGEN MIXTURES* 1. M. H. Chiang, E. A. Gislason, B. H. Mahan, Robert L. Horton and Bruce H. Mahan C. W. Tsao, and+A. S. Werner, Dynamics of the Reactions of 0. with H, and D,, J. Phys. Ion cyclotron resonance mass spectrometry Chem. 75, 1426^(1971). z i was applied to the methane-oxygen system. The connections between reactant and product ions 2. E. A. Gislason, B. H. Mahan, C. W. Tsao, were established by using ion cyclotron and and A. S. Werner, Dynamics of the Reaction of ion ejection double resonance, and by examin N with H , J. Chem. Phys. 5£, 3897 (1971). ing some of the ion-molecule reactions occur 2 ring in mixtures of oxygen and chloromethanes. 3. B. H. Mahan, Molecular Orbital Correla The ion-molecule reactions which occur in pure tions and Ion-Molecule Reaction Dynamics, methane and pure oxygen were also investigated. J. Chem. Phys. 55, 1436 (1971). The reactions of oxygen ions with methane were found to be very slow individually. 4. E. A. Gislason, Approximate Differential Apparently this is not due to energy barriers Cross Sections at Large Scattering Angles for in the reaction paths, but to competition bet Simple Repulsive Potentials, J. Chem. Phys. ween many channels and because of rather 55, 3929 (1971). extensive geometric changes that are required as the reactions proceed. The fact that 37 5. M. H. Chiang, E. A. Gislason, B. H. Mahan, reactions had to be considered in this system C. W. Tsao, and A. S. Werner, Large Angle. is an indication of the complexity of the ion Elastic Scattering of Ar by He, J. Chem. molecule chemistry of relatively simple gas Phys. 55_, 3937 (1971). mixtures, and of the power of ion cyclotron resonance to sort out reaction mechanisms. 6. C. Maltz, Barrier to Complex Formation in n the Hydrogen Atom-Chlorine Molecule Reaction, Abstracted from LBL-114. Chem. Phys. Letters 9, 251 (1971). UCRL and LBL reports /'. RESEARCH PLANS FOR CALENDAR YEAR 1972 1. Arthur S. Werner, Dynamic Isotope Effects Bruce H. Mahan in Ion-Molecule Reactions (Ph.D. Thesis), UCRL-20363, April 1971. We will continue and extend our studies of vibrational energy transfer, with particular 2. Robert L. Horton, Thermal Ion Molecule emphasis on systems where the conventional Reactions in Gaseous Methane-Oxygen Mixtures theoretical approximations are expected to (Ph.D. Thesis), LBL-114, Aug. 1971. fail. Electronically, inelastic processes will be investigated in N - He and similar sys 3. Keith T. Gillen and Bruce H. Mahan, tems, where detailed calculations of the in Deconvolution of Molecular Beam Inelastic elastic cross sections are possible with as Scattering Data, LBL-199, Oct. 1971 (accepted yet untested theoretical techniques. A new by J. Chem. Phys.). reaction system, slightly more complicated than 0- - H-, will be sought and investigated Paper presented to further test the phase space theory of chemical reactions and the predictions of 1. Bruce H. Mahan, " Gaseous Ion Recombina molecular orbital correlation diagrams. New tion," Gordon Research Conference on Radiation experimental techniques for product velocity Chemistry, July 1971. 68 William H. Miller, Principal Investigator (rotational excitation) were both investi gated regarding this question, namely how 1. THE CLASSICAL S-MATRIX FOR ROTATIONAL much averaging is necessary to quench these EXCITATION; QUENCHING OF QUANTUM EFFECTS IN quantum effects. Results indicate that sum MDLECULAR COLLISIONS* mation over even a few quantum states is often sufficient to make a completely classi William H. Miller cal treatment appropriate. A previously developed theory in which Abstracted from J. Chem. Phys. 54, 5386 exact solutions to the classical equations of (1971). motion for a complex collision system (i.e., numerically confuted trajectories) can he used to generate Che classical limit of the quantum mechanical S matrix (the "classical 2. COLLISIONAL QUENCHING OF METASTABLE S matrix") for the scattering process has HYDROGEN ATOMS* been applied to rigid rotor-atom collisions (i.e., rotational excitation). Figure 1 William H. Miller, Charles A. Slocomb, and shows the comparison with essentially exact Honry F. Schaefer quantum results; when it is realized that this is a highly quantum-like system (only Metastable hydrogen atoms (i.e., the 2s flj « ± 2 transitions are classically allowed), state) are easily quenched by collision with one sees that the "classical S matrix" other species because of the degeneracy with results are quite good. the 2p state which decays radiatively to the ]s ground state. The quenching is thus a To obtain such good agreement for transi two-step process tions between individual quantum states it is absolutely necessary to take account of H(2s) + X + H(2p) + X , quantum interference effects. If one sums (or averages) the transition probabilities H(2p) * H(ls) + hv . over several final (or initial) quantum states, however, the interference terms are If X is a molecule whose lowest non-zero found to average out quite rapidly. The multipole moment is qj_ (e.g., L = 1 is a linear atom-diatom system (vibrational dipole moment, L = 2 a quadrupole moment, excitation) and the rigid rotor-atom system etc.), then the cross section for quenching is 1 1 1 1 I i 1 1 1 1 1 y 2 1 13 a(v) = CL Giqj/hv) ^ "* , 030 a) ,• where u » Tsz^ is the 2s ** 2p transition dipole h of H, qL is the L-pole moment of X, v is the relative collision velocity, and C^ is a 0 20 ry* • I00r 0.10 - ,•, - ~~~-*—T*Z--* V 0 IIII i 1 i i i i - i I I,.I, Fig. 1. A visual comparison of the exact quantum Tesults (points connected by solid line) and uniform semiclassical results (points connected by dotted line). For 100.0 clarity of the display the order of transi tions along the abscissa is chosen so that Fig. 1. The cross section for He + H(2s) •* the quantum values increase monotonically. He + H(2p) as a function of relative collision (XBL 711-6468) energy. (XBL 712-6507) 69 constant; e.g., Cj - 5.924, C2 - 4.206. 4. CLASSICAL PATH APPROXIMATION FOR THE Equation (1) appears to give good agreement BOLTZMANN DENSITY MATRIX* with experimental results. William H. Miller If species X has no multipole moments, then the cross section is more difficult to Diagonal matrix elements of the Boltzmann calculate. Such calculations have been operator exp(-SH) give the particle density carried out, however, for tha case X - He for a canonical ensemble: atom, and Fig. 1 shows tlie quenching cross section as a function of relative collision energy. Recent experimental measurements1 p(x) e A classical-limit approximation for the 3. MOLECULAR RAMSAUER-TCWJSEND EFFECT IN VERY propagator in Eq. (2) gives LOW ENERGY He4-He* SCATTERING* William H. Miller p(x) ~exp [i*(x)/h] , (3) If the collision energy is sufficiently where *(x) is the classical action along the low, then only s-waves contribute to the total trajectory that goes from position x back to elastic cross section: this same x in time t. Surprisingly, the classical path approximation for the propa gator, with t set equal to -ihB, does nor. 2 2 o(E) - 4ir(h /2uE) sin nQ(E) , (1) give the usual classical approximationTor the density: where u is the reduced mass of the two col lision partners, E is the relative collision E 2 1/2 energy, and T|pt ) is the s-wave phase shift. Pr,(x) - (2m/4irBh ) exp[-BV(x)l , (4) If the potential consists of a long range attractive part and a repulsive core, one can show that the s-wave phase shift varies where V(x) is the potential energy; the rea with energy such that it is zero at some son for this is that the relation between t energy Eg. If Eg is so snail that higher and S involves a factor of h, so that the partial waves have not begun to contribute classical result in Eq. (4) automatically appreciably [so that Equation (1) is valid], implies "short imaginary times" (of order h). then the cross section for scattering has The classical-limit approximation for the a zero (actually just a deep minimum) at propagator [Eq. (3)], however, does not imply relative energy Eg. a short time limit, so that the resulting approximation for p(x) is valid for large A WKB analysis for the s-wave phase shift 8 (ie., low temperature) as well as small shows that the system 4He-*He is the only S (high temperature); the only requirement atom-atom system for which these conditions for its validity is that the potential V(x) may possibly be satisfied. (The Ramsauer- be "smooth." Townsend effect is well known from electron- atom scattering.) The energy Eg at which The expressions that result from this the 4He-4He cross section is expected to classical path approximation for p(x) can.be exhibit this minimum is Eg * 2°K, raising the evaluated explicitly under various limiting speculation as to whether such a dynamical conditions and one finds, for example, that effect could in any way contribute to the it gives the exact quantum result for p(x) peculiar properties of liquid helium at in any region of x for which the potential is these low temperatures. a quadratic function of x. Abstracted from Chem. Phys. Lett. 10, 7 Abstracted from J. Chem. Phys. 55, 3146 (1971). _ (1971). ~ 70 S. THE CLASSICAL S-MATRIX FOR LINEAR REACTIVE COLLISIONS OF H + Cl2* William H. Miller and Charles C. Rankin Exact (i.e., numerically computed) classi cal trajectories for the linear H + Cl2 reactive collision have been used to construct the (.iassical limit of the quantum mechanical S matTix (the "classical S matrix") for reactive (and non-reactive) transitions be tween individual vibrational states of the diatomic molecules. This use of classical mechanics to construct transition amplitudes (rather than transition probabilities directly) has the advantage that quantwn superposition is incorporated in this classical-limit theory; previous applications have shown that the resulting quantum effects (interference and tunneling) are quite significant when one considers transitions between specific quantum states. The general approach differs only slightly from the previous study of non-reactive linear atom-diatom collisions. Figure 1 shows a typical result for the classical trajectory function n2(q,)--the final vibrational quantum number (either reactive or non-reactive) as a function of the initial phase of vibration, for fixed values of the initial quantim number and initial translational energy. There is one interval in which the function is quite smooth and well behaved, and another e , • * 7 • • / • • 6 • •• •• Fig. 2. Transition probabilities from a initial states 0 (bottom), 1, and 2 (top) of CI2 to various final vibrational states n. 5 _"_«.«. r o • _ f «0 •• • *: The points connected by the solid line are a reactive transitions (final diatom • HC1), ° ° _ 4 0 and the points connected by the dashed line T • • • are non-reactive transitions (final diatom » 6. PREDISSOCIATION IN THE SCHUMANN-RUNGE 3PO + 3P 0 BANDS OF 0 * 1 1 1 1 2 1.25 1.5 1.75 2.0 William H. Miller and Ifenry F. Schaefer r(A) — Fig. 1. Theoretical potential curves fo" The Schumann-Runge absorption bands in O2 3 1 the nu, n„, and B^Z,/ states of 03. T.ie result from excitation of the ground X*S„" experimental B state curve is that of D. L. 3 state to the B 2u" state. Certain lines are Albritton, A. L. Schmeltekopt, and R. N. Zure significantly broadened, indicating that (unpublished), and the dashed curve is the the excited B^Sy" state is predissociated by repulsive state inferred by J. N. Murrell some other excited potential curve that and J. M. Taylor, Mai. Phys. 16_, 609 (1969). crosses it; the %u state has often been The energy scale is relative to the v = 0 assumed to be this other state. level of the X32n ground state. (XBL 716-6821) _ We have computed potential curves for the 3 3 B 2 " and IIU states (see Figure 1) of sufficient accuracy, however, to conclude that it is not the %u state that is res Abstracted from J. Chem. Phys. 5S_, 4107 ponsible for the observed predissociation. (1971). Other possible states are of a different spin multiplicity, and this is apparently the reason for their never having been considered previously. A semiclassical analysis for the Franck-Condon factor which appears in 7. RESEARCH PLANS FOR CALENDAR YEAR 1972 the predissociation width, however, leads to an estimate of the electronic interaction William H. Miller that stTongly suggests that spin-orbit coupling is the interaction responsible for In the coming year our research efforts the predissociation. The B'Sy" state, will be concentrated on two major problems: therefore, would be coupled by this interac tion to singlets and quintets just as (1) Classically forbidden processes strongly as to other triplets. (tunneling)• The tunneling, or weak transi tion problem is, in my opinion, the principal The ^Iln potential curve has also been unsolved problem in our general semiclassicaf calculated, and although it is a much more theory. Not only is it intrinsically likely candidate as the state responsible interesting from a theoretical point of view, for the predissociation, our calculations but it is potentially the single most indicate that it, too, crosses the B-state important practical outcome of all the too low (see Figure 1) to produce the observed "classical S-matrix" work. Thus the inter predissociation widths. The two quintet ference features, seen when several transi 5 5 states. nu and Eu", both are expected to lie tions are classically allowed, are terribly above lnu, and it may be one of these that interesting, and it is indeed exciting to corresponds to the 'otserved"dotted curve in see that they are accurately described by Fig. 1. classical dynamics + quantum superposition-- 72 but they tend to be averaged out quite easily in quenching collisions. Many of the appli so that a purely classical approach is cations of our work in this general area, usually adequate in these cases. When all therefore, will deal with collisions invol transitions are classically forbidden, ving metastable atoms and molecules. however, a purely classical treatment is completely inadequate, but ope would still (a) Atom-Atom Collisions. The semi- like to know the largest of these small classical description of electronic transi transition probabilities. tions in slow atom-atom collisions is basical ly a solved problem, due primarily to We have already had some success in Stuckelberg's solution of the two state treating tunneling transitions, although avoided crossing problem 40 years ago. We there were some practical difficulties. To plan to apply the Stuckelberg theory (slightly the extent that it was possible to carry out refined) in a fairly rigorous way to a num the procedure, it appears that classical ber of atom-atom collisions. One such dynamics used within this framework describes problem currently in progress is electronic classically forbidden transitions as ac excitation of 0+(4S), 0+(4S) • He - 0*(ZD) + curately as it does classically allowed He, or inversely, collisiona] quenching of ones. the motastable species 0+(2D). Our overall goal in this program is to document the Treating classically forbidden transitons validity of a semiclassical description oi' within the framswork of the classical S matrix such electronic transitions when the semiclas sical theory is pushed to its limit. is essentially a problem of analytic continuation--analytically continuing clas sical mechanics through classically forbidden (b) Atom-Diatom Collisions. We are cur regions. As soon as our techniques for nu rently engaged in efforts to develop a merical analytic continuation are more refined general semiclassical theory of electronic we plan to calculate "classical S-matrix" transitions in complex collisions, i.e., ones elements for rotational-vibrational transi with internal heavy particle J?grees of tions in He + H2 collisions in three dimen freedom (rotation and vibration) as well sions. This is in keeping with our notion of as translation, such as A + BC. Since applying these methods to the most quantum classical dynamics (plus quantum super like systems to see just how far toward the position) is an accurate description of quantum direction the basic semiclassical heavy-particle dynamics (rotation and idea (classical dynamics + quantum super vibration, as well as translation), one position) is meaningful. wishes to retain a classical description of them. Electronic degrees of freedom, however, (2) Electronic transitions: roetastable must certainly be treated quantum mechanical species. The other principal area is that of ly, so that one is forced to mix classical collisions that involve more than one elec and quantum dynamics per se (and not just tronic state of the overall system. Although quantum superposition). TFis complicates collisions between sr^cies in their ground matters considerably, but we feel that we electronic states often take place within will soon have the correct formal solution one electronically adiabatic state, this is to the problem. It is interesting to see almost never the case when at least one of that in this general theory an electronic the collision partners is in an electronically transition, between adiabatic potential excited state. energy surfaces emerges as a special type of classically forbidden process. Our progress A particularly interesting and important with the tunneling problems described above, sub-class of such processes are those that therefore, will also find application in the involve collisions of an atom or molecule in prob]em of electronic transitions. a metastable electronically excited state; i.e., it is only metastable species (those The number of applications which come to which cannot decay radiatively by optically mind, particularly those related to paragraph allowed transitions) that live sufficiently (2), are endless. Collisions which produce long to play a significant role in collision or destroy the ^S or *D metastable states oi" phenomena. In most situations, in fact, it 0, the Is and JA metastable states of 02, anrf is collisions that determine the effective A*S tate of N2, and metastable states of lifetime of metastables. The importance of othe common atmospheric species are extremely metastable species in atomspheric processes impoitant, At higher elevations, in the is well known; they are reservoirs of large ioni:phere, metastable states of 0, 0+, O2, amounts of electronic energy that is released °2+> ®2> N2+ all Play important roles. 73 8. 1971 PUBLICATIONS AND REPORTS Papers presented William H. Millar and Associates \. W. H. Miller, "Classical Correspondence Principle," Cambridge Conference on .Molecular Journals Energy Transfer, July 19-23, 1971. 1. W. H. Miller, The Semiclassical Nature 2. C. C. Rankin and h'. H. Miller, "The of Atomic and Molecular Collisions, Accounts Classical S-Jfatrui: Ljiicar Reactive Colli Chem. Res. 4_, 161 fl971). sions of H + CIj," Vllth Intematioml Conference on th'fr Physics of Electronic and 2. W. H. Miller, Additional WKB Inversion Atomic Collisions," Amsterdam, July ?V30, Relations for Bound-State and Scattering 1971. Problems, J. Chem. Phys. 5£, 417-5 (1971). UCRL and LBL reports 3. W. H. Miller, The Classical S-Mav.rix for Rotational Excitation; Quenching of Quantum 1. David A. Yuen, I. Optically Detected Effects in Molecular Collisions, J. Chem. Nuclear Qutidrupole Resonance anJ Transferred Phys. 54, 5386 (1971). Hyporfinc Coupling in Molecular Crystals; II. Eloctior.-Atomic Hydrogen Scattering: Two 4. W. H. Miller, C. A. Slocorob, and H. F. Approaches (M. S. tliesis), IOV 20702, June Schacfer, Collisional Quenching of Motastablo 1971. Hydrogen Atoms, J. Chom. Phys. 55, 926 (1971). 2. W. II. Miller, C. A. Slccomb, and H. F. 5. W. H. Miller, Molecular Ramsaucr-Townsend Schaefer, Mcilccula!.- Autoionization Lifetimes Effect in Very Low Energy He*-He* Scattering, and Cross SEC?ions for Penning Ionization: Chem. Phys. Lett. 10, 7 (1971); addendum, 11, Numerical Results for He*(ls2s 3s) + H(lszS), 152 (1971). UCRL-20709, Ju»e 1971. 6. W. H. Miller, Classical Path Approximation 3. K. H. Miller, Classical-Limit Green's for the Boltzman Density Matrix, J. Chem. Function (Fixed-Energy Propagator) and Phys. 55, 3146 (1971). Classical Quantization of Non-Separable Systems, LBL-1Z6, Aug. 1971. 7. C. C. Rankin and W. H. Miller, Classical S-Matrix for Linear Reactive Collisions of 4. W. H. Miller, Classical-Limit Quantization H + Cl2, J. Chem. Phys. 55, 3150 (1971). of Non-Separable Systems: Phase Space Quantiza tion and its Relation to the Dynamical Quan 8. H. F. Schaefer and W. H. Miller, Curve tum Condition, 1BL-162, Aug. 1971. 3 Crossing of the B3L„" and It States of 02 and its Relation to Predissociation in the 5. W. H. Miller, Classical Limit of Fredholm Schumann-Runge Bands, J. Chem. Phys. 55, 4107 Theory for Elastic and Inelastic Scattering; (1971). Inability of Phase Space Integrals to Describe Inelastic Transitions, LBL-172, 9. K. H. Miller, Simple Error Bound for Sept. 1971. Coupled-Channel Scattering Calculations, Chan. Phys. Lett. 11, 535 (1971). 6. S. M. Honistein and W. H. Miller, Quantum Corrections (Within the Classical Path Approximation) to the Boltzmann Density Matrix, LBL-455, Nov. 1971. Rollie tl. Myers, Principal Investigator it may be an experimental problem due to in- homogeneous broadening for the Cu^+. 1. SPIN RELAXATION IN a-MS04-6H20 The magnetothermodynamic work of Fisher Michael R. St. John and Rollie J. Myers et al.^ on the water proton spin-lattice relaxation in a-NiS04 • GII2O gives a proton In our previous reports on the magnetic T^ value which nicely follows the activation properties of a-NiSO,j-6H20 at helium temper energy corresponding to Pb. These T, values atures it was pointed out that the magnetic do, however, show a rather complicated pre- substatcs form an exciton band. These excitons exponential factor. The theory for the are thermally populated in a lattice of inter exciton effects upon Tj values are more com acting Ni2+ (S=l) ions possessing a non-degen plicated than on line widths, for Ti depends erate (D > 0) ground state. Since D * S cm"1, entirely upon the fluctuations in the magnetic these excitons are fairly dilute at T < 2GK. field which are transverse to the applied In the molecular field model the thermal field. A corresponding theory for Ty is still steady-state population of those excitons can being developed. be related to an average molecular field which results from the interactions between the T. A. Jindo, Single Crystal Studies of Na^+, but the fluctuations in exciton density Hydrated Transition Metal Ions by Electron should give rise to fluctuations in this Paramagnetic Resonance, UCRL-20386, Jan. 1971. molecular field. 2. T. J. Swift and R. E. Connick, J. Chem. Phys. 37, 307 (1962). Spin relaxation can usually be related to 3. R.~A". Fisher, G. E. Brodale, E. W. Homung, the fluctuation in magnetic field at the site and W. F. Giauque, J. Chem. Phys. 49, 4096 of a spin. Since we have not been able to (1968). determine the spin relaxation parameters for the Ni2+ ions, we are forced to use a spin probe for this study. We have shown* that 2. ELECTRON-ELECTRON DOUBLE RESONANCE IN Cu^* should be a satisfactory probe since its SOLUTIONS EPR g-values are sensitive to the average molecular field in the a-NiS04-6H20 and Joyce T. Yarnell and Rollie J. Myers NiSe04-6H20 lattices. The technique of electron-electron double We have also previously observed that the resonance involves the simultaneous application EPR lines of the Cu2+ broaden rapidly as the of separate pump and analysis EPR frequencies. temperature is raised above 1.2°K in The analysis signal is detected and one looks for relaxation effects which involve the pump NiSeO4-6H20, while they are unaffected by a temperature change in the diamagetic power. It has been previously used1 to study ZnSe04-6H20. intramolecular relaxation among the hyperfine states of radials in solution. The simplest theory for this broadening can be derived by using the methods of Swift and We have assembled an electron-electron Connick.2 If the g-value shift is the dominant double resonance apparatus both to study term (flu mechanism), then the exciton contri intermolecular radical relaxation and as a + bution to 1/T2 for Cu2 is equal to Pb/rb. method of measuring rates of intramolecular In this relation Pb is the fraction of excitons spin-lattice relaxation. in the Ni2+ lattice and rb is the exciton lifetime on a neighboring Ni2+ ion. At 1.5°K The cavity consists of a Varian bimodal we observe an EPR line width at x-band for (V-4536) cavity combined with the coupled 2+ Cu in NiSeO4-6H20 of about 100 G. Since section of the optical transmission (V-4534) the D value of the N'i2+ should be about cavity. It supports two crossed modes. A 4.8 cm"1, this line width gives tb •* 10"11 sec. TE102mo( *e ls use{1 t0 observe the EPR spectrum. Such a lifetime is consistent with exchange This mode is contained completely in the interaction between Ni^* ions with J « 0.2 cm"1. bimodal section and is excited by a coupling hole and metal screw at one end. The second Our initial measurements of the variation mode is a TEIQJ which is used to pump or of the Cu2+ EPR line width with temperature saturate the lines. It covers the entire indicate that it has an activation energy of length of the cavity and is excited by the only 2.6 cm'1 instead of the expected4.8 cm"1. loop coupling of the optical section. The The cause of this discrepancy is unknown, but two modes are set to a variable frequency 75 difference by placing quart: on a side wall 3. MEASUREMENT OF TRANSITION METAL ION of the optical section and using a Teflon RELAXATION IN SOLUTION screw on the end of the cavity. One of the major problems in the experiment is the Thomas M. Hynes and Rollie J. Myers leakage of the pump power into the observing arm. Better isolation was achieved by using Work is continuing on the measurement of a quartz flat cell to hold the sample, and very fast electron spin relaxation times of by using a rejection filter attached to a transition metal ions in aqueous solution. three-port circulator placed between the Ions studied so far are Mn2+, Cu2+, Cr3+, Gd3+, observing bridge and the cavity. The filter Co2+, Ni2+, and Ti3+. Three methods have been itself is a Varian cavity tuned to the pump used in this present study. frequency and critically coupled. The first method is simply observation and The spectrometer used is the Varian V-4502 measurement of the EPR spectrum. Spectra of spectrometer, with the usual klystron and the hexaquo complex of Ti3+ in solution have bridge used for observation. A separate bridge not previously been reported. Several con is used for the pump. Its major feature is the centrated solutions of Ti^+ were run under use of a 20 W Varian VA-6186 traveling wave favorable conditions, and a wide signal of tube (TWTj to amplify the power from a klys some 2000 gauss was observed in each case. 11 tron. Putting this much power into the cavity This indicates a T2e of about 3 x 10" sec, results in a cooling problem, which is cur Measurement of this signal at various temper rently solved by passing nitrogen gas through atures should help establish the mechanism of the cavity and running the pump only for short relaxation of Ti-* in solution. time intervals, A better cooling system is planned. In certain cases, measurements of the bulk water proton NMR line width in these solutions When two radicals are present in solution gives information about electron relaxation they can interact either by means of exchange times. At least three conditions are neces or direct dipole-dipole interaction. The sary: the dominant mechanism of interaction intermolecular spin relaxation effects are between the proton and the paramagnetic ion quite different in these two cases and our must be known, the chemical exchange of pro first attempt at electron-electron double tons from bulk water to water bound to the resonance has been to study these relaxation metal ion must be faster than the rate of effects. proton relaxation while bound, and (in the case where dipole-dipole interaction dominates The two radicals chosen for this particular scalar coupling) the electron relaxation time study are Coppinger's radical and vanadyl must be faster than the rotational correlation acetyl acetonate dissolved in benzene. Our time of the ion. Ni2+ and Co2* may satisfy first experiments involved the simultaneous these conditions for certain temperature resonance of pump power on the center of the ranges. The bulk water proton line widths Coppinger's resonanre (g • 2.0045) and analysis have been measured for these two ions at power on one of the eight vanadyl hyperfine 60 MHz from -3 to 70°C. Results for both ions lines. With S1V (Mi = -3/2) transition this are sufficiently different from similar work* involves a frequency difference of about at 26.5 MHz (6.2 kG) to indicate field de 380 MHz. We have not been able to observe pendencies of the electron relaxation. The a double resonance effect and so we have experiments must be repeated at other field switched to pumping the vanadyl and observing strengths before detailed conclusions can be the Coppinger's radical. Since we need only made. observe the sign of the effect (increase or decrease) to differentiate between exchange The third and most general method involves and dipole-dipole interaction, we see no measurement of the shortening of the nuclear reason why our intermolecular interaction relaxation time (T^) of protons on cations experiment will not work once optimum condi in aqueous solutions of paramagnetic cations. tions are found. Since dipole-dipole interaction is the only important relaxation mechanism, the increase The intramolecular experiments involve in the reciprocal of the transverse relaxation pumping and analysis of the same hyperfine time of cationic protons at distances r from transition. Since most pump power sources paramagnetic cations of spin S is are rather noisy, this looks to be an ideal way to make T^ measurements for broad reso nances in solution. TtrisS^eV^ET 2N r T. M. P. Eastman, G. V. Bruno, and J. H. Freed, J. Chem. Phys. 52, 321 (1970). (7TI * —r~r) • 76 u where s is the Larmor precession frequency of at other field strengths. ar e tne the electron, and ye and YH mag- netogyric ratios of the electron and proton, Measurements on Ti^+ at room temperature respectively. The summation is over all para indicate a relaxation time of roughly 4 x 10"-1-! magnetic ions in solution, and the r"6 factor sec. The interpretation is less certain than can be evaluated by use of electrolyte solution for other ions, due to the large amount of acid theory if the ionic strength is not too high. meeded to prevent hydrolysis. The quantity TI may be either a molecular diffusion time or Tie, whichever is shorter, T. R. Hausser and G. Laukien, Z. Physik 153, whereas T2 is the shorter of diffusion and 394 (1959). T^g. Rotation is not important here, since the 2. A. M. Chmelnick and D. Fiat, J. Chem. Phys. electron spin and the proton are on separate 47., 3986 (1967). species in solution. If Tj and T2 are shorter e e 3. J. W. Neely and R. E. Connick, UCRL-19126, than the diffusion times, the proton resonance Nov. 1969. broadening becomes inversely proportional to the electron relaxation times. Calculations indicate that 10"10 sec is the largest relax ation time that can be measured by this method. Experiments at different magnetic fields can 4. RESEARCH PLANS FOR CALENDAR YEAR 1972 separate the contributions of Tie and T^p, and at different temperatures we can check the Rollie J. Myers importance of diffusion. Our work on the effects of exchange in the a-NiS04'6H20 lattice should be concluded. Broadening of protons on a cation is inde Low temperature EPR spin-lattice relaxation + pendent of the actual cation used. Trijnethyl measurements will be made on Ti(H20)(; and sulfonium has been used mainly, both because some related ions in single-crystal solids. of its chemical stability and because its narrow proton resonance allows a small absolute Measurements will be made on low-spin Fe3+ amount of broadening to be seen after the chelates. A number of these chelates have paramagetic cation is added to the solution. importance in biological systems and they can only be studied by EPR at liquid helium tem Experiments were done on solutions of Mn , peratures . Cuz+, Gd3+, Cr3+, Ni2+, and Coz+ at 60 MHz from -3 to 50°C. Broadening in solutions of Our work oil electron spin relaxation in Mnz+, Cuz+, Gd3+, or Cr3+ is large, and in solution will be continued. The electron- creases with decreasing temperature, having electron double resonance experiments should an activation energy in each case of about be developed as a direct method for the meas 3.5 kcal/mol. The observed broadening is in urement of electron spin-lattice relaxation good agreement with calculations which assume times in solution. a Debye-Huckel type ionic distribution, a distance-dependent diffusional correlation time, and diffusion constants of 1 x 10"5 atrsec"!. Apparently little can be learned about relaxation times of these four ions at 5. 1971 PUBLICATIONS AND REPORTS 60 MHz by this method. Rollie J. Myers and Associates Broadening in solutions of Coz+ or Niz+ also increases with decreasing temperature, but with Journal and presented paper activation energies closer to 1 kcal/mole. Calculations give the relaxation times of Coz+ 1. C. A. Arrington, Jr., A. M. Falick, and and Niz+ at room temperature to be close to R. J. Myers, Electron Paramagnetic Resonance 11 1 7 1 x 10"12 Sec and 3 x 10"- sec, respectively. Spectrom of 02( Ag)--Its * 0 Hyperfine Coupling These times are somewhat longer than those and Electronic ana Rotational g Values, J. deduced from experiments at other fields.2'' Chem. Phys. 55, 909 (1971J. If the time for Ni2+ is correct, it would indicate that an EPR signal might be observable 2. A. Jindo and R. J. Myers, Electron Exchange in a concentrated solution. Since calculation for Hydrated Ions in the a-NiS04-61120 Lattice-- of these times depends on proper evaluation Electron Paramagnetic Resonance Study of of the T~° term, it would be helpful to get an Hydrated Transition-Ions in ZnSeO4'6H?0, experimental check on it. Hopefully, a cal presented as an invited paper at the Fourth ibration ion will be found (perhaps Niz+ or International Symposium on Magnetic Resonance, Fe3+) which has a relaxation time shorter than Weizmann Institute, Rehovat, Israel, Aug. 1971. diffusion, yet long enough to be measured by EPR. Again, these experiments must be repeated 77 UCRL reports 2. J.J. Chang, Relaxation Time Measurements in Electron Paramagnetic Resonance CPh.D. 1. A. Jindo, Single Crystal Studies of Hy- thesis), UCRL-19691, Feb. 1971. drated Transition Metal Ions by Electron Paramagnetic Resonance (Ph. D. Thesis), UCRL-19691, Feb. 1971. K'umeth S. Pitzer, Principal Investigator b. Atomic and Ionic Structure 1. RESEARCH PLANS FOR CALENDAR YEAR 1972 After completing our review of existing quantum mechanical wave functions and associ Kenneth S. Pitzer ated properties of heavy atoms (i.e., transition metals and heavier), we will a. Aqueous Electrolytes proceed to make improved atomic calculations, if feasible and worthwhile, and then proceed Aqueous solutions of several solutes, to the calculations of the energies and usually electrolytes, are of enormous practi related properties of chemically important cal importance in situations all the way from species. Particular attention will be given the separation processes for spent nuclear to ionic species and to oxides, and to ele fuel to the biological processes in the ments which show qualitative differences human body. While very dilute solutions of from the properties of lighter elements in single solutes were investigated intensively the sumo group of the periodic table. and, in most cases, successfully several decades ago, the theoretical basis for con centrated solutions, for nixed solutes, and c. Other Topics for solutions containing multiply charged ions of each sign is far from satisfactory. In addition, one or more of the following Several closely related investigations will topic: will be pursued: (1) nuclear spin attack these problems with the objective of species effects at very low temperatures for making possible predictions of the properties Ho, HO, V>2, and various isotopic methanes; of mixed electrolytes of realistically high (2) carbon vapor species with particular concentration from a minimum of experimentally reference to the triple point of carbon; determined parameters. (3) theories for pure fluids and solutions based upon definite molecular models and Statistical theories are needed which statistical mechanical principles which lead are applicable to mixtures of ions of dif in a simple case to corresponding states ferent sizes as well as different electrical behavior; and (4) unusual types of internal charges. Separately, but with guidance from motion in molecules. theory, various schemes for empirical representation of the properties of mixed electrolytes will be tested. X-ray scattering studies are planned for a few solutes with 2. 1971 PUBLICATION very heavy ions where ion scattering will dominate that from the solvent in order to Kenneth S. Pitzer check the radial distribution predicted theoretically. Journal 1. K. S. Pitzer, Thermodynamic Properties of Aqueous Solutions of Bivalent Sulfates, accepted for publication in the Transactions of the Faraday Society (submitted in June 1971). 79 NUCLEAR SCIENCE Donald R. Olander, Principal Investigator 1. AN APPARATUS FOR INVESTIGATION OF GAS- lowing two-branch, two-site mechanism explains SOLID REACTIONS BY M3DULATED MOLECULAR BEAM both the hysteresis and the surface chemistry: MASS SPECTROMETRY Richard H. Jones, Donald R. Olander, Wigbert CCOads -° C0(g) • S J. Siekhaus, and James A. Schwarz A o (g) + s An apparatus for the study of gas-solid 2 A reactions which produce a gaseous reaction + S C0 product is described. The reactant gas con °ads B * < >ads tacts the solid as a molecular beam travelling a in vacuum. Gaseous reaction products are - C0(g) * SA or SB, monitored by a mass spectrometer as they are emitted from the surface. The technique re where S^ and SB represent the two types of lies heavily upon the modulation of the molec active sites on the surface, n is the sticking ular beam, which improves sensitivity by probability of O2 on A-sites and k the emission rates of CO and C02. The phase Reaction of molecular oxygen as a modulated lags of these products relative to the molecular beam with the basal plane of impinging reactant beam indicated that the pyrolytic graphite was investigated. surface reactions were strongly affected by diffusion of oxygen in the grain boundaries The graphite surface participated in the then into the grains of the pyrolytic graphite reaction in an unusual fashion. Oxidation structure. The proposed reaction model is: tends to create a highly reactive surface while thermal annealing tends to deactivate the surface. The imbalance of these two o2(gJ .20 •COfe) competing processes results in slow changes ads ' in surface reactivity during the course of an experiment, which is manifest as hysteresis in the rate of CO production. The surface 0? chemical reaction occurs on a much smaller time scale than the processes responsible for where n denotes the sticking probability of hysteresis. The data indicate that the fol 02 on the surface and k is the rate constant 80 describing the first-order conversion of the theory of rotating disk mass transfer. adsorbed oxygen to carbon monoxide. The Kinetic expressions for the oxidation of the diffusional processes (represented by the basal and prism planes were obtained. The double arrows) reduce the surface concentra rate expression of Strickland-Constable1 fell tion of adsorbed oxygen. Ogjj and Og denote in between the rates determined here for the oxygen dissolved in the grain boundaries and basal and prism orientations. in the grains, respectively. n '— This double diffusion process so strongly Abstract of paper submitted to Carbon. demodulated the product signals that the ap 1. J. R. Walls and R. F. Strickland-Constable, parent reactivity of the prism plane was less Carbon 1_, 333 (1969). than that of the basal plane. This reactivity inversion is peculiar to the ac modulated beam method and would not occur in dc (steady state) experiments. The reactivity of graph ite which had been annealed to 3000°C was 4. THEORY OF THE GAS CHKTRIFUGE* found to be an order of magnitude larger than that of the unannealed material. This Donald R. Olander increase in reactivity was due to reduction of the demodulation effect which resulted The technical basis of the gas centrifuge, from closing off diffusional paths in the bulk a device for enriching uranium isotopes on during annealing. a large scale, is reviewed. The rotational motion of the centrifuge generates a centrif Hysteresis of the type found in the basal ugal force field some 100,000 times greater plane reaction was observed in the prism than that of gravity. Under the influence plane reaction as well. Approximately one of this force, the heavy isotope concentrates tenth as much CO2 was produced by the reaction at the outer rim of the spinning bowl or as CO. Reaction with a beam enriched in rotor. In addition, small thermal or mechan oxygen-18 did not show complete statistical ical disturbances purposely introduced at mixing of the isotopes of oxygen in the COj the axial ends of the rotor couple with the product. This observation suggested that a centrifugal force fields to establish a slow significant portion of the CO2 product was circulatory motion in the centrifuge. By formed by a direct reaction of surface carbon means of this internal "counter-current," the with molecular oxygen. The oxygen contained gas most highly enriched in uranium-235 in the CO2 product produced by this mechanism appears at the bottom of the device, from are the partners in the impinging 0 molecule. 2 where it may be withdrawn as product. Abstract of paper to be published in J. Chem, Phys. The separative properties and the hydro dynamics of the circulatory flow in the gas centrifuge are analyzed. Conditions for optimal operation and device efficiencies are calculated. 3. THE RATE OF OXIDATION OF THE BASAL AND PRISMATIC SURFACES OF PYROLYTIC GRAPHITE IN THE TRANSITION REGIME BETWEEN CHEMICAL AND The extent of enrichment in each centrifuge DIFFUSIONAL CONTROL* is quite small, although considerably larger than attainable in a single stage of the Rajarama Acharya and Donald R. Olander gaseous diffusion process. In order to pro duce uranium of enrichments required for Reaction rates of oxygen with the basal and light water reactors (~ 3* U-23S), many prism faces of pyrolytic graphite have been centrifuges in series are needed. To supply measured, using a rotating disk contactor. the uranium required for the nuclear economy Temperatures ranged from 1100 to 2000°K, and oxygen concentrations from 0.054 to II in of Western Europe, very many individual helium and argon at atmospheric pressure were centrifuges connected in parallel are re employed. The intrinsic rate of the surface quired. The British-German-Dutch venture in reaction ami the magnitude of the mass trans uranium isotope separation may require upwards fer coefficients attainable from the rotating of a million centrifuges. disk placed most of the experiments in the transition regime between the limits of chemical and diffusional control of the over Abstract of. a review to be published in all process. Surface chemical rate constants Advances in Nuclear Science and Technology, were extracted from the data with the aid of Vol.6. 81 LASER VAPORIZATION OV SOLl'DS mi; VISCOSITY op ^DL^^•N URANIUM DIOXIDE* Robert A. Olstad and Donald R. Olander I Ian-Chung Tsai and Donald U. Olander The quadrupole mass spectrometer in the T.ic viscosity of UO? is needed to analyse laser vaporization apparatus was calibrated gross fuel movement in'various potential for iron by measuring the mass-56 signal from accident situations in an LMFBR.* This trans an iron sample heated at steady state by port property has been measured from the electron bombardment. The temperature melting point to 3000°C, using an oscillating dependence of the mass spectrometer signal cup viscometer. The apparatus and method of was in excellent agreement with the known data analysis have been described previously.' heat of vaporization of iron. Such informa Approximately 80 g of \X>i was contained in tion permits absolute vaporization rates to a 1.6 cm i.d. tungsten crucible bored from be determined from mass spectrometer signals an arc-cast ingot. The lid and support rod obtained during tfie rapid vaporization were electron-beam welded onto the crucible transient produced by a laser pulse on the under vacuum. Prior to scaling, the UOj was surface. baked in a hydrogen atmosphere at 1600*C for ten hours. One of the major problems in the lns.!>i The oscillating crucible method yields tho pulsing experiment is the determination of kinematic viscosity,2 which was converted to the temperature history of the vaporizing absolute viscosity by use of liquid UO2 surface during the millisecond or less density data of Chrlstonscn.* transient generated by the laser pulse. Both theoretical and experimental approaches to Measurements were made on two different this problem have been pursued. A computer samples of UO?. A least-squares fit of the program which calculates the temperature rise data plotted in Tig, 1 yields a viscosity at of a material heated by a pulsed laser was the melting point (2800*C) of 8*1 cP. The written. The program solves the heat con activation energy is IS kcal/molc with an duction equation, with allowances for estimated precision of *8 kcal/mole. This temperature-dependent material properties, large variation is due to the scatter of the time variation of laser power, and ablation data and the rather small temperature range of the material. The laser input power is over which measurements were made. The an important parameter in the calculation. average deviation of the data from the best Although the power in the pulse can be fit straight line on Fig. 1 is U cP. The measured, the fraction of this energy actually viscosity measured here is considerably less absorbed by the surface cannot be directly than the value of 25 cP currently assumed determined. The latter is the difference for accident analyses.1 between the power delivered by the pulse and the portion reflected from the surface. A 1 "I " -m -T" photodiode system was developed to measure ..., OAJtor the reflected energy. The fast-response MUM MMMMUT infrared pyrometer designed to experimentally wo. • 1 oeti.ttn measure the temperature transient was cali O 1 oct.-i. un brated for iron from 900 to 1200°C. )- t A A - _ A A - Initial studies of transient vaporization O will be made on iron, since its vapor pres - O A ~ - sure is well known and the condensation - • - coefficient is unity. Following this aspect, - • - vaporization of zirconium hydride will be - investigated. Since this material is not - commercially available, an apparatus for producing samples of the desired shape and 3 3.2 J.I Jo hydrogen-to-zirconium ratio was constructed. «tn ft-11 Purified hydrogen at a pressure-temperature combination which produces the desired Zr/H ratio is passed over high-purity lx bar Fig. 1. (XBL 721-5913). stock for several hours. The composition of the hydride formed is determined by weight gain. Measured compositions agree very well with those predicted from the phase diagram Brief submitted to the Trans, of the Amer. for the preset conditions. Nucl. Soc. 82 1. Argonne .National Liborator>', Reactor 2. R. P. Ombcrg and I). K. Olander, Effect of Development Progran Progress Report, A.NL-7872, Condensation in the Boundary Layer on Mass p. 8.1 (Oct. 1971). Transfer from a Rotating Disk. II. Experi 2. J. S. Finucane and 1). R. Olander, The mental, Phys. Fluids l±, 1834 (1971). Viscosity of Uranium and Two Uranium- chromiim Alloys, High Temp. Sci. 1^ 466 (1969). 3. D. R. Olander, Description of the Hydrogen- 3. J. A. Christonscn, Thermal Expansion and Metal Interaction by a Morse Potential Func Change in Volume on Melting of Uranium tion, J. Phys. Chem. Solids 32_, 2499 (1971). Dioxide, J. Am. Cer. Soc. *6, 607 (1963). Paper presented Investigation of the High Temperature Reaction 7. RESEARCH PLANS FOR CALENDAR Y1SAR 1972 of Oxygen with Graphite by Molecular Beam and Mass Spcctrometric Techniques, International Donald R. Olondcr Conference on Solid Surfaces, Boston, Oct. 1971. The molecular beam studies of graphite reactions will be continued with an investiga UCRI. and LBL reports tion of the reaction of pyrolytic graphite surfaces with water vapor. 1. !). R. Olandor, Technical Basis of the Gas Centrifuge, UCRL-20S06, Jan. 1971. Construction of a molecular beam system for the study of reactions of ionic solids 2. R. II. Jones, Kinetics of the Reaction with halogen gases will begin. The first of pyrolytic Graphite with Molecular Oxygen system to be examined will be the reaction of studied by Modulated Molecular Beam Mass fluorine with uranium dioxide, which produces Spectrometry, I.BL-104, Nov. 1971. the volatile hoxafluorido of uranium. 3. R. If. Jones, D. R. Olander, W. Siekhaus, The gas phase radiolysis of oxygen induced and J. A. Schwarz, Reactions of Modulated by 1 MeV protons will be continued. Molecular Beams with Pyrolytic Graphite. I. experimental Apparatus and Data Inter The laser pulsing work on iron and pretation, LBL-189, Oct. 1971. zirconium hydride will be continued. 4. D. R. Olander, W. Diekhaus, R. Jones, and Measurement of the viscosity of uraniun J. A. Schwarz, Reactions of Modulated Mole dioxide and ccsiun by the high temperature cular Beams with Pyrolytic Graphite. II. oscillating crucible viscometer should be Oxidation of the Basal Plane, LBL-190, Oct. completed. 1971. 5. D. R. Olandcr, R. Jones, J. A. Schwarz, and W. Sickhaus,' Reactions of Modulated 8. 1971 PUBLICATIONS AN) REPORTS Molecular B»ams with Pyrolytic Graphite. III. Oxidation of the Prism Plane, LBL-191, Donald R. Olandcr and Associates Oct. 1971. Journals 1. R. P. Cwberg and 0. R. Olander, Effect of Condensation in the Boundary layer on Mass Transfer from a Rotating Disk. I. Theoretical, Phys. Fluids 1£, 160S (1971). II. Metallurgy CRYSTAL IMPERFECTIONS i. FIKIJ) IOS FAI-OROX MICHOSCOPF. OR- •RYVVTIONS ON PHUT'fiCTOIl) CkClflll OF Ni ,1V Itua C'hing long aiul Jack Washburn An ir.;;ot of Ni-20 at.? IV was prepnn d l>y melting 99.95?. pure Ni ami (LI:. :J1S tin :stcn wires together in an argon arc furnace. This ingot was cut into small pieces am' reel' cd twice for hotter nixini;. Use ingot was final ly aged for 90 hours at 1300°C and qucnc'icd into oil at room temperature. "Ihe ingot was then swaged into a 0.01 in. wire. 'Ihe wires were annealed at 1309°(.' for 9 hours, foil iwed by an oil quench at room temperature. On'crcd and partially ordered wires were prepared iy annealing the as-quenched wires in evacuat d Fig. 2. FIM micrograph of u partially or quarts tubes at various temperatures below dered alloy aged at 850°C for 30 min. the ordering temperature (970 t 10oC) and !Ucro[jr;.ph' taken at liquid nitrogen tempera quenching into water at room temperature. ture for better definition of antiphase Field ion tips were prepared from the wires boundaries. (XBB 715-1892) by elcctropolishing. The tips were cooled by helium gas punpcd from a liquid helium container during imaging. Iho .''S-quenched alloy is more than 90? a-phase with particles of the tungsten-rich solid solution distributed throughout in a coarse dispersion. Therefore only the a-phase is usually observed in a FIM tip. The a-phase quenched from 1300°C shows numerous dark lines which intersect thciraelvos, forming many small regions of 50-100 A diameter. The only reasonable explanation for these lines is that some long-range order with a domain size of 50-100 A exists in the a-phase I ig. 3. FIM micrograph of an alley aged .it 9. 0CC for 3 hours.' (XBB 7010-4494) after quenching from 1300°C. (Fig. 1). It was found that ordering starts just above 750°C. Mien the alloy was annealed at 850°C for 30 min, the whole specJTnen was filled with small domains with an average size of 59-100 A as shown in Fig. 2. Each of these domains seems to have reached a quite advanced stage of long-range order even Fig. 1. Field ion photograph of as-quenched though the composition of the ra-phase is alloy which was taken while s slow field quite far from the ideal 20 at.* IV. tvhen the evaporation was going or. Nute the numerous alloy was annealed at 920°C the ordering dark lines. (XBB 715-1882) mechanism was different. The ordered domains 84 thin foils outside the electron microscope G> 1000°C) and photographing the same area under identical diffraction conditions, it has been possible to study the self climb of coplanar pairs of edge dislocation loops, b = l/2<101 ) . Quantitative measurements were made for the rate of self climb at several temperatures and a value of 60,300 * 3500 cal/mol for the activation energy for pipe diffusion along edge dislocations and (7.5 ± 4.3) x 10"18 cmVsec for the pre- exponential D0ap were obtained. To study the pipe diffusion along screw dislocations, a configuration where a dislocation loop near ft* '• # i I " Fig. 4. A fully ordered Ni.W field ion micrograph. (XBB 715-1883) grew slowly into the disordered a-matrix, eventually becoming very large (Fig. 3). Finally when the annealing time was suf ficiently long, the entire volume of a FIM tip became a single domain of ordered 8-phase (Fig. 4). i 3 I These observations lead to the following oa conclusions about the ordering mechanisms for Ni4W: at low temperature the mechanism is a homogeneous reaction, with domain size and the sense of each domain being predetermined during quenching, but at high temperature the mechanism is nucleation and slow growth of long-range ordered B-phase domains into the disordered a-matrix involving diffusion of IV to the a-8 interface and solution of tungsten-rich y-phase particles. Because nucleation occurs only at a few centers, probably at the interface between a and y phases, the domains are very large when they I eventually grow into contact. 2. TRANSMISSION ELECTRON MICROSCOPIC STUDIES ON MAGNESIUM OXIDE Jagdish Narayan and Jack Washburn Fig. 1. Coplanar groups of loops at 1, 2, a. Pipe Diffusion in Magnesium Oxide and 3 came closer together and formed single loops of approximately conserved area. By developing the techniques of annealing (XBB 721-20) 85 the center of thin foil is connected to both its surfaces by a screw dislocation of the same b-vector 1/2(101] was found and the shrinkage Tate of dislocation loop measured at different temperatures. The activation energy for pipe diffusion long screw dis locations was found to be 62,700 ± 2000 cal/mol and the corresponding pre-exponential factor 3 D0 was 1.3 x 10 " cmvsec. (See Fig. 1) b. Bulk Diffusion in MgO The annealing kinetics of dislocation loops in plastically deformed and subse quently annealed MgO thin foils have been studied by transmission electron microscopy techniques. The effect of loop-to-sink distance on the rate of climb has been directly confirmed. The rate of shrinkage of dislocation loops, which were near the center of the foil and relatively isolated from other loops, was measured in the tem perature range 1080 to 1427°C. The rate is controlled by oxygen ion mobility mid dif fusion occurs probably by vacancy mechanism. A value of 110,000 ±4200 cal/mol for activa tion energy for bulk diffusion and (1.37 * 0.26) x 10"2 cm2/sec for D0 were obtained in this temperature range. (See Fig. 2 0 Fig. 3. Smaller loop (marked by arrow) slipped along its glide cylinder because of its interaction with another loop of opposite /M t b-vector. (XBB 721-1) coalescence of a large elongated loop with a smaller loop of opposite Burgers vector by prismatic slip of the latter along its glide cylinder. After coalescence the part of the big loop (denoted by the arrow) breaks up by pipe diffusion. This also provides direct evidence of presence of both interstitial and vacancy dislocation loops in plastically de formed and subsequently annealed magnesium oxide. (See Fig. 3J Fig. 2. Dislocation loops which are relative ly isolated and near the center of the foil d. Nucleation of Prismatic Dislocation Loops are shrinking due to bulk diffusion. on Impurity Precipitates in Magnesium Oxide (XBB 721-21) SinRle Crystals Prismatic dislocation loops were observed c. Prismatic Slip in Magnesium Oxide on impurity precipitates between slip bands after annealing of plastically deformed Quantitative measurements of displacements magnesium oxide single crystals. These loops of dislocation loops along their glide nucleated on any of the six (101} slip planes. cylinders due to interaction with relatively The habit plane of the loops showed no cor immobile loops were made after various an relation with the active slip plane during nealing treatments, done outside the electron deformation. By using the results c>f dynami microscope. The Peierls stress was found to cal theory of electron diffraction :.ncluding decrease with the increase in the size of the absorption in the transmission electron loop. The set of pictures clearly shows the microscopy, the majority of the loops were identified as b = 1/2 <101 > interstitial type. Impurity particles which nucleated dislocation loops were found only in MgO which contained a relatively higher percentage of calcium and aluminum impurities. The nature of dislocation loops is discussed on the basis of possible chemical form of these impurities and Crowdian mechanism of plastic deformation around impurity particles. (See Figs. 4 and 5.) 3. ATCM PROBE FIELD ION MICROSCOPE Kaj G. Stolt and Jack Washburn The atom probe field ion microscope mentioned in the last report has been im proved in three respects. A new specimen holder has been constructed that allows for more effective cooling of the tip and easier manipulation of it. A channel plate ion-to- electron converter has been built into the microscope for image intensification. An electrostatic lens is mounted at the entrance to the flight tube to focus the ions onto the detector. The construction of an ion gun capable of producing beams of inert gas ions with a maximum accelerating voltage of 40 kV has been completed. The gun has been connected to the atom probe field ion microscope for in situ ion bombardments of field ion tips. 4. RESEARCH PLANS FOR CALENDAR YEAR 1972 Jack Washburn 1. Transmission electron microscopic studies of ion-implanted silicon are underway. The goal is to correlate the defect structure (from electron microscopy) with the electri cal properties of substrates. 2. The atom probe field ion microscope is being used in connection with the ion gun for studies of ion bombardment of metals. The primary objective is the measurement of Fig. 5. Size of the loop decreases from g - 200 to g =• 200 (Figs. A and B). Figure C is the dark Peld with S > 0. So bottom of the foil is in uctter contrast. Figures D and E are dark field pictures with S < 0, so top is in better contrast. This deter mines the slope of the plane on which loop lies. Therefore the nature of the loop is established (interstitial type). (XBB 711-32) 87 ion ranges by controlled field evaporation of 5. 1971 PUBLICATIONS AND REPORTS a bombarded field ion tip, using the atom probe to identify the bombarding icns. Jack Washburn and Associates 3. Experiments on stainless steel (321), Journals Al, Cu-Al, and Ni-Co are being done to study the effects of flux, fluence, temperature, 1. E. Nes and J. Washburn, Precipitation in stacking fault energy, and dislocation density High Purity Silicon Single Crystals, J. Appl. and arrangements on the radiation-induced Phys. 42_ (9), 3562 (1971). growth of voids. Irradiation will be carried out in situ, using a high voltage microscope 2. Y. Miura and J. Washburn, Mobility of at higher temperatures. A configuration in Prismatic Dislocation I la If-Loops in Copper, which most of the dislocations are connected J. Cryst. Lattice Defects Z_ (4), 197 (1971). to voids and precipitates is being con sidered as the one which will help reduce 3. E. Nes and J. Washburn, Radiation Induced swelling. Precipitation in Silicon During High Voltage Electron Microscope Observations, J. Appl. 4. Switching behavior of amorphous Phys. 42 (9), 3SS9 (1971). semiconducting glasses upon voltage pulsing is being studied. Experiments, to correlate 4. II. C. Tong and J. Washburn, Field Ion the microstructure of the resulting phases Microscope Observations on Peritectoid Growth with the electrical properties (conductivity of Ni4W, J. Cryst. Lattice Defects 2, 221 Hall constant, mobility, etc.) are in pro (1971). gress. 5. E. Nes and J. Washburn, Precipitation in 5. Mechanisms of operation of a homologue Silicon Caused by Electron Radiation Damage of the cuprous sulphide/cadmium sulphide During HVEM, Jernkont. Ann. 155, 527 (1971). photovoltaic cell are being investigated. The goal is to make thin film devices out of Paper presented zinc sulphide similar to those made from cadmium sulphide and develop a low-cost 1. J. Washburn, Annealing of Loops and photovoltaic (solar) cell suitable for large- Tetrahedrf., in Proceedings of Conference on scale manufacture. Radiation Damage, Albany, New York, June 9-11, 6. The effects of neutron and electron irradiation damage on annealing phenomena UCRL and LBL reports in magnesium oxide are being studied. 1. J. Narayan, The Kinetics of Self Diffusion and Dislocation Glide in Magnesium Oxide (Ph.D. thesis), LBL-406, Nov. 1971. 2. J. Narayan and J. Washburn, Nucleation of Prismatic Dislocation Loops on Impurity Precipitates in Magnesium Oxide Single Crystals, UCRL-20347, June 1971. 3. J. Narayan and J. Washburn, Self Climb in Magnesium Oxide, UCRL-20335, June 1971. B. MECHANISMS OF PLASTIC DEFORMATION John E. Dorn (deceased Sept. 24, 2971) and John W. Morris, Jr., Principal Investigators depend on test temperature, solution concen tration and, below room temperature, 1. RESEARCH ON MECHANISMS OF DISLOCATION asymmetrically on crystal orientation. The GLIDE asymmetric yielding behavior was found to be in reasonably good agreement with the pseudo- a. Dislocation Glide Through a Random Array Peierls mechanism proposed by Dorn and of Obstacles Mukherjee. However, the degree of asymmetry in the Mo-Re alloys was found to be less than Dale H. Klahn, John E. Dorn, and John W. that for Mo, in contrast to the theoretical Morris, Jr. predication. This project develops a numerical study of c. The Effect of a Solute Atmosphere on the glide of a dislocation through a random Dislocation Glide"" array of obstacles. The obstacles are ideal ized as square potential barriers, but other Choi K. Syn, John E. Dom, nnd John W. wise may have a wide range of properties, Morris, Jr representing, for example, solute atoms, small precipitates, or repulsive dislocations The solute atmosphere formed around an cutting the glide plane. The dislocation is edge dislocation in a crystal containing a idealized as a continuous line of fixed energy known concentration of substitutional impurity per unit length which interacts with the was computed, using a model which takes into obstacle and with an applied stress according account the effect of the atmosphere on the to the usual relations governing dislocation vibrational entropy of the dislocation, ac motion. A computer code has been written cording to a theory due to Baur. It was which establishes a random array of obstacles found that inclusion of this entropy term of given properties over a plane, and follows causes a significant decrease in the degree the motion of a dislocation through this of solute segregation to a dislocation. Using array under an applied stress. Using this this result, the force required to separate a code, the yield strength in glide has been dislocation from an equilibrium solute studied as a function of the temperature, atmosphere in an anisotropic crystal was the properties of the obstacles, and the geom computed. The results were specialized to etry of the obstacle array, and the mean glide the case of dilute silver additions in Zn, velocity is studied as a function of applied and compared to experimental data obtained stress, temperature, obstacle properties, and through observation of the migration of low- obstacle geometry. angle grain boundaries in Zn crystals con taining known concentrations of silver. The b. Asymmetric Slip in bcc Crystals theoretical models are now being further re fined to more adequately account for the de Chih-an Liu, John E. Dorn, and John W. tails of grain boundary motion. Morris, Jr. The yielding behavior of bcc crystals has been reported to depend asymmetrically on 2. RESEARCH ON PLASTIC DEFORMATION crystal orientation at low temperatures. The asymmetric yielding behavior was generally as a. Mechanisms of Steady-state Creep at sumed to be due to the dissociation of the Elevated Temperatures a/2 <111 ) screw dislocations. The present project was initiated to investigate the al Farghalli A. Mohamed, K. Linga Murty, John E. loying effect on the asymmetric yielding be Dorn, and John W. Morris, Jr. havior of bcc crystals. The Mo-Re alloy system was chosen because it was believed High temperature creep has been studied in that auJing Re to Mo would lower the stacking Al-3Mg solid solution over the stress range fault energy of Mo and increase the tendency ~10"° G to ~10"3 G and over a range of tem of disso ation of the a/2 <111 > screw dis peratures. Double-shear type specimens were locations, hence increase the asymmetry. employed. A composite plot of dimensionless Single crystals of two Mo-Re alloys, 10 at.% parameters ykT/DGb versus T/G (Fig. 1) indi and 20 at.% Re, each in three different cated three distinct regions: region I at orientations, were tested under tension at low stresses in excess of 6 x 10"S G, region II temperatures. The yield stress and slip band from 10"5 - 6 x 10"s G, and region III below formation of these crystals were found to I.O"5 G. Plots of the logarithm of modulus- 89 • i 1 1 'i •• ' r b. High-temperature Transient Creep of Metals o"' . and Alloys Kamal E. Amin, K. Linga Murty, and John £:. K>"* - A general empirical equation was developed AI-3Uf where EQ is the instantaneous strain on Fig. 1. Creep data on Al-3 Mg plotted as loading, e the secondary creep rate, Ke the logarithm of the compensated strain rate rate constant, and B the ratio of initial to a function of logarithm of the stress de secondary creep rate. The experimental creep picting the transitions from glide to climb data on several bec and fee metals and alloys and climb to viscous regions. (XBL 7112-2304) correlate quite well with the proposed mech anism. The constants S and K were found to be independent of temperature and stress. The proposed formulation becomes inapplicable for correlating creep data in polycrystajs at low stresses because of the significant contribu tion of grain-boundary sliding to the total compensated strain rate versus the reciprocal creep at these stress levels. of absolute temperature yielded a value of approximately 33 kcal/ncl for the activation c. Superplasticity in Fine-Grained Alloys energy for creep in each of the three regions. This value is in essential agreement with the The temperature and the stress dependence activation energy for self-diffusion. Values of steady-state strain rates in Zn-22 Al for the stress exponents were found to be 3.2, eutectoid were studied by tensile and creep 4.1, and 0.9 respectively in regions 1, II, testing, using double shear type specimens in and III. The values for regions I and II are a normalized stress (x/G) range of ~10'° to consistent with the theoretically predicted ~10'3. The stress dependence of the strain rate-determining mechanism for these regions, rate revealed three distinct regions: low viscous glide and dislocation climb, stress region (I) with a stress exponent of respectively. -I, moderate stress region (II) with -2, and high stress region (III) with ~4. The tem Region III corresponds to an anomalously perature dependence of the strain rates in high rate of creep under low stresses. This the three regions yielded values of 16.3, anomalous behavior had previously been ob 16.2, and 27.4 kcal/mol for the activation served by Harper and Dorn in studies on pure energy for deformation in the regions I, II, Al. To test whether this anomalous creep is and III respectively, hhoreas strong grain confined to aluminum alloys, creep tests were size dependence of the steady-state strain carried out on Pb at stresses below 10"' G. rates was observed in regions I and II, the Anomalous creep was again observed, indicating data on all specimens coalesced into a single that this unexpectedly high rate of creep at line in region III. From the experimental low stress may be common, at least in fee determination of creep parameters, mechanisms alloys. The rate-determining mechanism of of deformation are identified. Dislocation this anomalous creep is now being investigated climb and Coble creep were found to be the 90 (2) Effect of solute atmospheres on glide. Hie mechanism of migration of pinned low- angle grain boundaries will be investigated to obtain a theoretical model more consistent with experimental data. Our current model assumes rigid translation of t'.ic low-angle boundary. This assumption will be replaced by a model which recognizes the possibility of boundary motion through critical fluctua tions in boundary shape as the boundary bows out under stress. Ilystorisis effects in the experimental data on'boundary motion will bo investigated, using mcxlels involving obstacles to tlie glide of grain boundary dislocations. These obstacles may lie in the bulk of the crystal or may ho located on its lateral surfaces. (3) High temperature croop. The anomalous creep found at low streSiios in Al-Nk; alloys and in Pb will be studied to identity its rate-controlling mechanism. Possible dis location mechanisms will be reviewed tj identify those which may lead to the observed Fig. 2. Mechanical data on superplastic behavior. It is anticipated that the possible Zn-22 Al depictins the three regions: mechanisms will lead to marked differences region I: Coble creep, region 11: super- in the dislocation structure of the alloy plastic creep and region III: dislocation during low-stress creep. Transmission elec climb. (XBL 7113-2505) tron microscopy will then be used to deter mine the dislocation structure. Once the rate-controlling mechanism of this type of controlling rochanisms in regions III and ] creep is known, other materials which should respectively. (See Fig. 2.) exhibit this type of creep will be identified and studied experimentally. Superplasticity. The mechanism of supcrplaststii c flow in Zn-22 Al will be investi- 3. RESEARCH PLAN'S FOR CALENDAR YEAR 1972 gated. This research will involve theoreti cal studies supplemented by transmission elec John h\ Morris, Jr. tron microscopy of strained samples. Research on other supcrplastic alloys will be inaugu Four research objectives have been identi rated. fied for the coming year. CI) Dislocation glide through obstacles. The numerical axle will be used to identify those geometric properties of on obstacle array which must be known to predict the J. 1971 PUBLICATIONS AVI) REPORTS average glide velocity. Previous work has shown that at realistic temperatures the John li. Horn, John K. Mirris, Jr., and mean glide velocity is largely controlled by Associates only a few of the configurations assumed by a dislocation during its passage over a plane. Journals and books Those arc the most efficient pinning configu rations. Our approach will be to identify 1. I). Klahn, 0. Austin, A. Mukhcrico, and the geometric properties of these configura J. fi. Itorn, "The Importance of Geometric tions, to associate configuration geometry Statistics to Dislocation Motion," in with the efficiency of pinning, and to deter Proceedings: Symposium onStatistical and mine the expected density of these pinning Probabilistic Problems in Metallurgy, Seattle, configurations in a random array of obstacles. Aug. 1971 (in press). The results will he used to establish con stitutive relations for strain rate in an 2. K. L. Murty and J. E. Dom, On the Non- idealized crystal deforming through thermally equilibrium Factor for Wuclcation Rates, activated glide of free dislocations. J. I1iys. Chem. Solids (in press). 91 3. K. L. Murty, F. A. Mohamed, and J. E. 2. K. L. Murty, M. L. Vaidya, and J. E. torn, Effect of Vacancy Sinks or Sources on Dorr:, High Temperature Deformation Mechanisms Serrated Yielding due to Solute Locking, in Superplastic Zn-22 Al Eutectoid, LBL-459, Scripta Met. (in press). Dec. 1971. 4. K. E. Anin, K. L. Murty, and J. E. Dom, 3. K. L. Murty, F. A. Mohamed, and J. E. Paramctrization of Low-Temperature Deformation Dom, Viscous Glide, Dislocation Climb, and Characteristics in Single Crystals of Molybde Newtonian Viscous Deformation Mechanisms num, J. Mat. Sci. (in press). Durir.j! Craep in Al-3mg, LBL-442, Dec. 1971. 5. F. A. Mohamed, K. L. Murty, and J. E. 4. K. E. Amin, Analysis of iligh-temperature Dom, On the Kinetics of Ordering of CusAu, Transient Creep of Pure Metals and Alloys, Mat. Sci. and Big. (in press). (Ph.D. thesis), LBL-163, Aug. 1971. LBL reports 5. Chih-an Liu, Effect of Re Additions on Anisotropic Slip in Mo Single Crystals, 1. F. A. Mohamad, K. L. Murty, and J. E. (Ph.D. thesis), LBL-129, Oct. 1971. Dom, Sorratcd Yielding in Ordered and Dis ordered CujAu, LBL-193, Nov. 1971. 92 THERMODYNAMICS OF INHOMOGENEOUS MATERIALS John W. tforris, Jr., Principal Investigator apparent metastability may result from size or boundary effects on the thermodynamics of This is a new program, to be inaugurated small clusters of atoms, or it may result from in 1972. activation barriers opposing structural trans formations in small clusters. Research will •>e inaugurated to obtain a better understand ing of the magnitude of size effects and the 1. RESEARCH PLANS F W CALENDAR YEAR 1972 nature of transformation barriers in isolated small atom clusters. John W. Morris, Jr. f4) Morphological changes during growth Four research t. sks will bo undertaken of precipitates. The distribution and morphol during the coming .'ear. ogy of precipitate phases forrood during the processing of engineering solids have an (1) Theory of -gntinuous interfaces in one- important influence on mechanical behavior. component fluids Recent research [J. W. In particular, laminar precipitates formed Morris, Jr., J. lem. Phys. (in press)] led to along grain boundaries may have a devastating the development >f a general model for a effect on the strength of engineering alloys, continuous two-pnase Interface in a one- and substantial research efforts have been component fluid. An intriguing consequence devoted to eliminating them. A better under of that model wis the result that material standing of the causes and rates of morpholog near an inter*• -e will be subjected to a non- ical changes in precipitates is needed to give classical syst Ti of stresses, including theoretical support to this process research. multipolar strjsses of tensor order greater Research under this program will specifically than 2. During I he coming year this model study the breakup of a planar precipitate will be further used to study the connection during growth, through developing analytic between the theriKOynamic "surface tension" models which take into consideration both (a scalar) and thi:- multipolar stress system. the diffusion effect on precipitate shape and It is hoped that experiments can be identified changes in the nature of the precipitate- through which these multipolar stresses may matrix interface during growth. be measured. (2) Concentration gradient effects in binary so utions. Recent research [J. K. 2. 1971 PUBLICATIONS AND REPORTS Morris, Jr., Phil. Mag. 73, 1041 0971)] led to the conclusion that t.'ie free energy density John W. Morris, Jr., and Associates within a binary solution can depend linearly on an imposed concentration gradient only if Journals the syraaetry of the solution permits a unique, directed axis. The importance of this result 1. J. W. Morris, Jr., The Influence of a lay in the consequence that linear gradient Concentration Gradient on the Free Energy of effects cannot intrude in the early stages of a Binary Solution, Phil. Mag. 23, 1041 (1971). spinodal decomposition in typical solutions. An the other hand, it is possible that linear 2. J. W. Morris, Jr., The Two-phase Fluid gradient effects influence the later stages Interface at Equilibrium: A Continuum Ntodel, of s-pinodal decomposition, and affect other J. Chem. Phys. (in press). important phenomena such as domain boundaries in ferroelectric or ferromagnetic solutions. Papers presented rHs possibility will be studied. We also .tend to initiate research on mass transport i. J. K. Morris, Jr., and P. L. PeBruyn, On .i solutions containing short-range property the Theory of Continuous Two-Phase Fluid ;radients. Interfaces, presented to American Chemical Society, Los Angeles, April, 1971. (3) Thermodynamics of small rystems. The structure of small precipitates formed inder 2. J. W. Morris, Jr., The i>iUcle3tion of highly supersaturated conditions [e.g., by- Metastable Structures, presented to the vapor deposition, splat cooling, or reaction Northern California Section of the ATME, in solution) often differs from that antici Dec. 15, 1971. pated from bulk thermodynamic data. As recently discussed [J. W. Morris, Jr., Proc. Second International Conference on Che- teal Vapor Deposition, Am. Chem. Soc. 0970)1 this 93 RELATION OF STRUCTURE TO PROPERTIES Gareth Thomas, Principal Investigator and how to measure the volume fractions of transformed phases in relation to the tensile STEELS properties (the law of mixtures seems to be well obeyed); and (3) efforts to improve the Introduction. The current programs include properties of martensite, utilizing the (1) continuation of the investigation of the strengthening principles of ausforming but role of composition on the transformation without involving mechanical deformation structure and tensile properties of martensit- during processing (see Fig. 1. in next sec ic steels, particularly to correlate the tion) . apparent relation between embrittlement and transformation twinning (Table t)*; (2) x investigation of the orientation dependence Invited paper presented at AIME Symposium of plastic deformation induced transformations on Martensite, Las Vegas, 1970, by G. Thomas in metastablc austcnitcs using single crystals (ICRL-19648), Met. Trans. 2,2373 (1971). Table I. Structure of body centfrt-d ferrous martens itcs formed on cooling through Ms-Mf? Alloy system Ms Substructure/Composition Fe-C >3S0°C C < 0.3 mainly dislocated laths <2S0°C C > 0.6 mainly twinned plates Fe-Ni >-S0°C Ni < 25 dislocated laths -30°-150',C Ni > 30 mid-rib twinned only and below Extent of twinning increases with I Ni, ana lower M Fe-Ni-C As above Increase in carbon for sane nickel content and vice versa enhances twinning Fe-2SM-0.3C-4.Sto <-77°C All mainly twinned. However, •4.7 Cr -16°C ausforming causes precipitation of +1.8SV -3°C carbides, so that the resulting martensite is less twinned. Fe-Ni-Cb-C -260°C C < 0.3 dislocated laths (l/2Mo/l/2Cr) 1S0-260°C C > 0.4 tidnned cobalt does not decrease v .tinning Fe-Ni-Co As above Fe-Cr-C 3S0-400°C Partially twinned Cobalt raises M --does not decrease t twinning Fe-SCr Dislocated laths fable 1 (continued). !-e-8Cr-U: '. -S!i°C Ivin density decrease? with increasing plate size j : F e-5Ni-0.25C 315- t Mainly dislocated Fe-J!*i-0.2SC :sis°c I Mainly dislocated Fe-5Ni-7Mn-0 2SC 65 °C t Mainly twinned l<--7Mn-0.25C 190°C i Mainlv twinned ler a given carbon Iccl, manganese promotes twinning Fe-Ni-Ii Uislocatiors and twins depends on Ni/Ti and heat treatment I-e-Cr-Ni Mctastable ' low stacking fault energy dislocated austenitcs a from r Fe-Mn-Cr-Ni All data here refer to nominal cooling rates. The Ms temperature and structure can be varied by varying quench rate (see text). Conpositson refers tn weight percent. 1. THE LNJiANU-MENT OF STREMCTnflLMNG DisLoevra; Mym-Nsrrc* !-!in Cheng and Gareth Thomas The basis of this work was the investiga tion of improving the tens'le properties of dislocated imrtertsites by dispersion of precipitate* in the austenite prior to the raartensite transformation. The idea is illustrated in Fig. 1. Tw-v types of precipi AUSTRNIfE tation hardcnable auslenitic alloys were used. •URTCNSITE Out is based on re-22.Ni-4Mo-0.28C where the WIEC in TAXES precipitates are T-b^C and are obtained by DISLOCATIONS ausforning and aging, ant! the other is Ft-2S.\'i-2Ti where the precipitates are the a »e, LATHS) coherent FCCy' fAijTi) ordered phase obtained by ausaging. Fig. 1. Scheme illustrating strengthening After the ausienitic Jispersior. treatment mechanisms for dispersions in martensite. both alloy's Here transforntd to rnitensite Sequence I - 3 - 5 is the conventional by quenching in liquid nitrogen and the pro ausfbrming process whereby plastic deformation perties raessured and compared to nartensites of netastable austenite produces carbides obtained by conventional heat treatrent (i.e., prioi to transformation. The new process of no precipitates in austenite). The results dispersing precioitates is achieved by show that prior dispersions increase the ausaging prior to transformation as illus strength of nartensite (Fig. 2), ."ind this is trated by sequence 1-2-5. Sequence 1 - 2 interpreted as being due to an increase in is the normal y-M sequence, but control of u.'slocation density resulting from dislocation composition is necess-irv in order to avoid nultiplicatiun at the particles during the twins (ree Table !). fXBL 70S-1671) 95 Abstract of UCRL-20562; Met. Trans, (in -*0 -|W press). Presently at Borg-Wamer Corooration, Des Plaines, Illinois 60018. if "\ t >' i IIIII i minim** 2. STRUCTURE AND MECHANICAL PROPERTIES OF Fe-Cr-C-Co STEELS* Mathur R. Ragliavon and Careth Thomas As part of a continuing program concerning the microstructures and mechanical properties of steels in which particular attention is given to transformation substructures, the present work is concerned with martensite and bainite in Fe-Cr-C steels with and without Fig. 2. Mechanical properties vs. prior cobalt. Although cobalt raises the Mj tem ausaging for Fe-28Ni-2Ti alloy; a, b show the perature it docs not affect the extent of properties of the ausaged austenite and c, d twinning for the same carbon level and so Mg show the properties after transformation to temperature alone does not control trans raartensito. (XBL 704-670} formation substructure. Thus, cobalt is not effective in retaining dislocated martensite as carbon is increased and in this regard cobalt is not beneficial to toughness. The Mj. temperatures of the steels were relatively high and hence isothermal trans formation yielded mixtures of bainitcs and tempered martensite depending on the tempera ture of transformation. The mechanical properties of the isothermally transformed steels were inferior to that of the tempered steels due to the interference of upper bainite or (tempered) martensite during the isothermal transformation. Thus, in the steels having high Mc temperatures, the twinned tempered martensitic structure had relatively better mechanical properties compared to the isothermally transformed steels. Atter..£s to produce desirable auto- Fig. 3. Martcnsitic structure after quenching tempered -structures by continuous cooling ausaged Fe-28Ni-2Ti alloy (3 hr S00oC). (single heat treatments) were not successful (a) bright field image--the Ni3Ti precipitates and diri not improve the mechanical properties, are visible in areas not in strong matrix since the structure consisted of a mixture of contrast and can be reversed in contrast by bainite and martensite. dark field, (c). The orientation relationship is N-W (c). (XBB 704-1553] Abstract, Met. Trans, (in press). Y •* M transformation. Figure 3 shows exaaples of the structure of the Fe-28Ni-Ti 3. ORIENTATION DEPENDENCE ON DEFORWWICN all'w. INDUCED e AND a MARTENSITE IN Fe-Ni-Cr SIM3LE CRYSTALS* In addition, the stabilities nf the austenitic alloys are such that upon certain Glen A. Stone and Gareth Thomas aging treatments, the alloys transform partially to martensite (due to precipitation), Fe-lStNi-15tCr austenite stainless steel and "composite" materials are obtained single crystals have been used to study the whose strength depends on the volume fraction formation of strain induced e and a martensite and yield strengths of the phases present. as a function of the critical resolved shear 96 stress on the {111} <112 ) systems. The volume fraction of e martensite appears to be a minimum for < 001 > tensile axis, ar.d in creases for the (123 > , <112 > , and (110 > orientations. Strain induced o martensite formation has also been studied by x-ray and electron diffraction, and the y-a orientation relationships lie within 2° of the Kurdjumov- Sachs relation. It has been possible to quantitatively measure the volume fraction of each o and e variant formed in the single crystals. For a single slip orientation ((2l3]fcc) tested at 188°K, the slip system was (lllJfcctOlilfcc- Th° crystal was pulled 1(a) in tension to 0.51 tensile strain forming 2.It c and 6.8% a by volume. All the c mar<-ensite was associated with the (ill)fcc|| COOOlVj. . It was possible to be more specific"for the a martensito, finding 4.2* was associated with the (Ill)fct!| I (llO)hcei [Olllfcclllilllbcc. Presented at Fall AIME, Detroit, 1971; based on LBL-134. PHASE TRANSFORMATIONS Introduction. The major emphases have been on the problem of characterizing short range order in various systems but especially Ni-14>, and the ceramic system LiFesOg, and spinodal transformations. The alloys under investiga tion are Cu-Ni-Fe and Cu-Mi-Al, with emphasis on the mechanism of loss of coherency and potential strengthening by interface disloca tions and by martensitic reactions after spinodal transformation. Work is also now well under way on transformations in lunar and terrestial minerals, using high voltage electron microscopy. 4. EARLY STAGES OF ORDERING IN NijMo* Fig. 1. Electron diffraction patterns of NijsMo, quenched from 1300T. Santosh K. Das (a) [001] orientation showing diffuse (1 1/2 0) SRO spots. The NijNfc and Ni4Mo The very early stages of ordering in NijMo peaks are not so clear in this orientation. have been investigated by transmission elec (b) [110] orientation showing diffuse tron microscopy and diffraction. The short- Ni2Mo spots (irarked by arrows). No Ni,jMo or range ordered state in this alloy is similar, SRO peaks are present in this reciprocal but not identical, to that of Ni4Mo.J On lattice section. quenching from 1300CC, in addition to the (c) [121] orientation showing (1 1/2 0) diffuse peaks at (1 1/2 0) positions, diffuse SRO spots, N14M0, and NiiMo spots. peaks also appear at positions very close to (XBB 713-1132) where the long-range ordered M4M0 and NijMo peaks should appear (Fig. 1). This suggests different from its long-range ordered that in the short-range ordered state this structure. The word cluster is used here in alloy has clusters of Ni^to and NijMo type. the general sense of the term to mean groups Thus, the structure in the short-range ordered of atoms of two species A and B. state may consist of not one but several types of clusters, the structure of which can be On aging at 650°C the 1 1/2 0 SRO peaks 97 phases coexist without any NijMo phase appearing. In the diffraction patterns the Ni4Mo spots are streaked in (120 > directions, similar to that observed in Ni^Mo,! whereas the NioMo spots are streaked in (110 >direc tions [Fig. 3). The l^Ma phase has also been detected in the stoichiometric Ni4Mo after aging at 650°C. The occurrence of Ni2Mo phase at N13M0 compositions can be rationalized in terms of the recent thermo dynamic calculations of the ground states of fee alloys by Richards.' 3t UCRL-20534. Proceedings EMSA Conference, 1971, p. 124 (Claitors Publishers). 1. P. R. Okamcto and n. Thomas, Mat. Res. Bull. 6, 45 (1971); also Acta Met. 19, 825 (1971). 2. M. J. Richards, D. Sc. thesis, MIT, Feb. 1971. 5. FIELD ION MICROSCOPIC STUDY OF ANTIPARALLEL TWINS IN Ni.Mo* 4 t t M. N. Chandrasekharaiah, S. R. Ranganathan, P. R. Okamoto,* and Gareth Thomas While field ion micrographs from concen trated random alloys have been disappointing, Fig. 2. Electron diffraction pattern obtained the images from ordered alloys reveal excel after aging for 4 hr. at 650°C, showing lent detail. Hence, several ordered alloys presence of only Ni2Mo and NijjMo spots. have been examined in the field ion microscope (a) [001] orientation showing Ni4Mo and '0 understand the mechanism of ordering and NijMo spots streaked in (210 > and <110 ) defect configuration.^ Out of many ordered directions respectively. alloys that have been examined, Ni^Mo^ has (b) [121] orientation. OCBB 713-1132) been studied in considerable detail. LeFevre et al. have studied the image characteristics of Ni4Mo and also analyzed the translational antiphase boundary contrast, while a cor relation of the mechanical properties with • structure using x-ray, electron, and field ion 6 o o o microscopic techniques has been presented by Chakravarti et al. Recently both electron and field ior. microscopy have been uced by Okamoto and Thomas5 to study the nature of 6 o o 6 .'.hort-range order in as-quenched N14M0. Super lattice formation in Ni4Mb occurs with a change in lattice symmetry from dis 000 ' • ordered face-centered cubic lattice to the • FimOAMNTlL >CC ' 'Of ordered body-centered tetragonal lattice of A M4W« t*OT the Dla type. The tetragonal axis is parallel • soo yrv to one of the cube axes, and the other two Fig. 3. Sketch of [121] reciprocal lattice axes are rotated either positively or section showing positions of SR0, Ni^Mo, and negatively by an angle 6 = 18.4°. There are Ni Mo spots. (XBB 713-1132) 30 possible ways in which the g phase may 4 nucleate from the disordered a phase, giving rise to a variety of domain boundaries. The gradually disappear and the Dla and Ni2Mo various types of interfaces obtained as a peaks become stronger (Fig. 2). For aging result of the ordering process can be clas- sif" J into three types:" times up to SO hr both the VA.$a and Ni2Mb 98 (a) Translational antiphase boundaries are obtained when the tetragonal e ••"-; of the two domains are parallel but the on^in 01 one domain is on a different sublet1'ce. (b) Antiparallel twin boundaries are the interfaces obtained when the tetragonal axes are antiparallel to each other ami the two a-axes enclose an angle 20 - 36.8°C. The twin planes are (200), (020), (220), and (220) as referred to the face-centered cubic lattice. (c) Perpendicular twins are obtained by the impingement of two domains with their c-axes mutually perpendicular to each other. In an exceedingly elegant demonstration Ruedl et al.6 derived that the three dif ferent types of interfaces give rise to a-fringes, wedge fringes, and 6-fringes respectively and presented evidence for their presence from electron micrographs. Fig. 1. Computer-simulated field ion pattern showing antiparallel twin boundaries in Ni4Mo. The contrast from translational antiphase Fundamental pianos are marked F. boundaries in the FM images can be understood (XBB 7111-SS59) on the basis of the g.ft criterion, and 3 LeFevre et al. identified them in Ni4Mo. computer simulation has provided ample IfiFevre and Newman8 made tentative identifi validity for this theory, no experimental cation of an antiparallel twin by the orienta evidence has been available to show the tion relationship between the two domains. matching of planes of dissimilar indices The presence of perpendicular twins has been across an interface. Tt~may be noted that inferred by Chakravarti et al.* on the basis 1:1 and 1:2 matching of planes of similar of their high density of occurrence and their indices across twin and domain boundaries plate-like morphology. It is clear that have been reported earlier. T is pleasing, unambiguous identification from FIM images therefore, to report 1:1 matchi,.u of (110) has not been achieved. plane of one domain with (020) plane of a second domain across an antiparallel twin 9,10 Chandrasekharaiah and Ranganathan have boundary in ordered Ni4Mo. used computer simulation as an aid ir. the interpretation of field ion images from twin Computer simulation of the six different interfaces. One of the major contrast orientations in Ni4Mo has been performed,11 features is the matching of important planes and boundaries were introduced across domains across the boundary. Usually it occurs that of different orientations to correspond to m planes of (hik^i) of one domain match with the antiparallel and perpendicular twins. n planes of £h2k2£2) of a second domain, Figure 1 shows the contrast to be expected where ra and n are small integers. While when two domains meet across antiparallel Table I. Antiparallel twin interactions in Ni.Mo. Domain I Domain II Twin Planes (220) (020) bet fee bet fee fcc fcc m n m n (020) (260) (110) (240) 4 3 1 1 (130) (680) (110) (240) 7 3 3 1 (110) (420) (310) (860) 3 7 1 2 (110) (420) (200) (620) 3 4 2 3 99 2. R. W. Newman and J. J. Hren, Phil. Mag. 16, 211 (1967). 3. B. G. LeFevre, H. Grenga, and B. Ralph, Phil. Mag. 18, 1127 (1968). 4. B. Chakravarti, E. A. Starke, Jr., and B. G. LeFevre, J. Mat. Sci. 5_, 394 (1970). 5. P. R. Okamoto and G. Thomas, Acta Met. 19, S25 (1971). (57 E. Ruedl, P. Delavignette, and S. Anelinckx, Phys. Stat. Sol. 28, 305 (1968); Mat. Res. Bull. 2_, 1045 (1967). 7. S. Ranganathan, H. B. Lyon, and G. Thomas. J. Appl. Phys. 38_, 4957 (1967). 8. B. G. LeFevre and R. W. Newman, in Applications of Field-Ion Microscopy in Physical Metallurgy and Corrosion, edited by R. F. Hochman, E. W. Muller, and B. Ralph (Georgia Institute of Technology, 1969). 9. M. N. Chandrasekharaiah and S. Ranganathan, J. Appl. Phys. 40, 4835 (1969). 10. M. N. ChandTasekharaiah and S. Ranganathan, Trans. Ind. Inst. Metals (in Pig. 2. Field ion micrograph in Ni4Mo press). showing a thin slice of a second domain 11. M. N. Chandrasekharaiah and matching across antiparallel twin boundaries. S. Ranganathan, Phys. Stat. Sol. (to be The (110)TT - (020)T matching region is communicated). circled. (XBB 7110-5560) twin boundaries. The m:n interactions are 6. ELECTRON METALLOGRAPHY OF SPINODALLY listed in Table I for two different DECOMPOSED (Cu-Mn)j Al ALLOYS orientations of the boundary planes. From Fig. 1 it is seen that the fundamental planes Marcel Bouchard* and Gareth Thomas (marked F) remain unaffected while the promi nent superlattice planes give rise to definite We have studied ternary alloys having ii.teractions. compositions along the tie-line CujAl-Q^MnAl. At high temperatures the alloys have a bec Figure 2 shows a field ion image of Ni^Kb taken at 7°K. A very thin slice of a first disordered structure. Upon quenching, the domain is seen passing through (110) of alloys order first to the B2 structure and at domain II. The ledges of the rings within lower teirrperature further order to the L2j_ the slice match 1:1 with the (110) rings of structure. domain II. This can be compared with the boundary between (020)T and (110)TT in Fig. 1. During aging at temperatures below 340°C, hence, it is concluded that the ledges of the quenched structure decomposes and the rings within the slice belong to (020). This resulting microstructures show all the is also in agreement with the expected loca metallographic characteristics of a spinodal tion of (020) for the antiparallel twin decomposition. The center of the irascibility orientatiL ;. The twin is extremely narrow gap is near the atomic composition (~25A thick) and brings out the usefulness ^2.5Mno.5AI, and experimental results sug of FIM in studying the substructure due to gest that the spinodal temperature is 325 ± ordering. Trace analysis showed that the 10°C. Micrographs in Figs. 1, 2, 3, and 4 plane of the boundary was (020)f . were obtained from an alloy aged inside the cc spinodal. Figure 1 shows the microstructure obtained after short aging at 240t'C. It Materials Research Bulletin (in press); consists of a three-dimensional array of LBL-412. plates parallel to the cube planes. The interparticle spacing measured from photo Department of Metallurgical Engineering, micrographs (^90A) is in agreement with that Institute of Technology, Banaras Hindu calculated from the spacing of the sidebands University, Varanasi-S, India. along 100 in reciprocal space. Figure 2 shows that the sidebands are associated with Argonne National Laboratories, Argonne, both the fundamental and the superlattice Illinois 60439. reflections. In addition, Fig. 2 shows dif fuse sidebands along 110 near the fundamental 1. B. G. leFevrc, Surface Sci. 23, 144 (1970). 100 Fig. 2. (001) reciprocal lattice piano showing the sidebands along 100 associated with tno oarly stngo of the splnodal decomposi IMM tion. (XBB 721-392) Fig. 1. fa) Bark field micrograph using (111) superlattice reflection and showing composi Fig. 3. Periodic microstructure obtained tion fluctuations within ordered domains in after aging the alloy Cu? cMin sAl at 290"C Cu, Mn rAl aged at 240°C for 30 min. for 1400 min. (XBB 721-391) 2 c cwnnn ™ 20) Brij - • • - - tb) [220) Bright field obtaine(XBd Bfro 721-39m a 5 5 different grain. XBB 721-396) 101 * -#?esjE Fig. 4. Bright fieid micrographs of the same area and the corresponding diffraction pattern of the alloy Cuii cMnn M aged at 300°C for 10,000 min. Tnis figure shows the a/2 <100 > type interfacial dislocations. The diffraction pattern reveals that the fig is always parallel to g. (XBB 718-3544) 102 After longer aging the micrestructure consists of rod-shaped particles parallel to the cube axis. Figure 3 shows this stage of the coarsening process just before the coherency is lost. After prolonged aging, intcrfacial dislocations are generated. Figure 4 shows that the network of disloca tions consists of two mutually perpendicular sets of dislocations smoothly curved around the particles. Contrast experiments have shown that the dislocations arc edge in character with b • a/2 (100 ). Corresponding ly, tho diffraction pattern shows the presence of two phases possessing similar cubic struc tures but with slightly different lattice parameters. Tho diffraction patterns can be. indexed in terms of the ordered unit colls of the two phases Cuy\l and C112M1AI. The microstructurcs shown in Figs. 5 and 6 result from aging tho non-symmetrical alloy Cuj oMno.jAl inside the miscibility gap but outside the spinodal. Figure S shows heterogeneous precipitation associated with a/4 (111 ) APB's present in the ordered as- Fig. S. (Ill) dark field micrograph showing quenched structure. The microstructnre heterogeneous nuclontion at a/4 (111 > APB's obtained after prolonged aging is shown in in Cuj.aMnn ?A1 aged at 240°C for 100 min. Fig. 6. The white regions in the micrograph (XBB 721-393) are CujAl. These regions were found to contain a high density of striations after deformation of the spinodally decomposed alloy. This is thought to arise as a result of the formation of martensite in thy CivjAl regions. M. Bouchard acknowledges receipt of a research fellowship from llydro-Quebec, Montreal, Canada. r.VraiPHASF. INTERFACES IN SPINOIWL ALLOYS'" Marcel Bouchard, Ronald J. Livak and Gareth Thomas Detailed electron metallographic studies are being made of interfneial dislocation? which are formed to relieve the clastic co herency strains developed upon long aging of spinodal alloys in the systems Cu-Ni-Fc and Cu-Mn-Al (Fig. i and Fig. 4 of p.jper 6). The stresses generated by the coherency strains ate sufficiently large to exceed the yield strcngtn of the alloys so that the interface disloca Fig. 6. Bright field micrograph showing thfl tions can be gsnerated by plastic deformation. growth of heterogeneously nucleated particles These slip dislocations accommodate the lat in Cu-. fiMnn ?A1 aged at 240°C for 10,000 min. tice mismatch at the interphase interfaces. i-B "•' (XBB 721-394) The geometry of tho arrangements is determined by the operative slip systems, by the degree reflections. As the aging continues, the of long-range order (if any), and by the interparticle spacing increases and cor requirement that the Burgers vector or a respondingly, the sideband spacing decreases. large component thereof also lies in the 103 interface plane. The model proposed for the formation of the interface dislocations suggests that during coarsening, the tut) coherent spinodal phases experience tensile and compressive stresses producing disloca tions which glide on slip planes from op posite interfaces and are then incorporated into the interface. As a consequence, the interface planes change their orientation and (100) habit is lost. In Cu-Ni-Fc {110} faces develop and this shape change is ex pected from the slip model, since in this alloy the dislocations are generated on {11 l Fig. 1. Micrographs taken of the sane area slip planes and their a/2 <110 > Burgers under different diffraction conditions for vectors lie in o (COi) plane adjacent to the Cu-Ni-Fo (alloy 2) aged for 200 hours at plane of Cite interface (Fig. 1). Rcarran.r.e- 77S*C. The dislocations visible in (a) have mont of the dislocation structures at thr 6 • a/2[110] or a a/2(I01], whereas those interfaces may be accomplished by climb. In visible in (b) and (c) have 6 - a/2[0U]. Cu-Mn-Al the dislocations lie in the {001) Note that the dislocations visible in (b) interface and have (100 ) Burgers vectors and (c) appear to be rotating so that they (Fig. 4 of paper 6). Those can only develop lie along 1100]. Notice a tendency for from slip dislocations after combination re pairing of the second set of dislocations; actions and climb. these pairs aro no* dipoles, as shown by the reversal of g from (b) to (c). (XBB 718-3729) Presented by invitation at the International Conference on Grain Boundaries and Interfaces, IBM, New York, Aug. 1971; Surface Science (in press). M4 E. ELECTRON MICROSCOPY Oareth Thomas, Principal Investigator [001] foils of GaAs semiconducting devices in which dislocations are associated with surface A comprehensive review paper of our work and interfacial defects. In each case the in electron microscopy and its applications crystal is sr.. at the exact Bragg condition was given at the 5th International Materials for each reflection along the systematic row Synposium, Berkeley, September 1971. The being used. The work was done on the Berkeley conference proceedings will be published Hitachi microscope operated at 650 kV. early in 1972 by the University of California Press. Figure 1 shows a single dislocation in gallium arsenide and its projection under various operating reflections. In A and B, •he operating reflections 040 and 080, because 1. DISLOCATION OOTTtAST AT HIC3I VOLTAGES* they have relatively large Bragg angles even at high voltages (650 kV), approximate the Richard A. Osiecki, Lutgard C. De Jonghe, low voltage two-beam conditions; i.e., no Walter L. Bell, and Garetli Thomas strong systematic influences aro present. It can be seen that there is little change in Some results of up to 12-bcam contrast image width by using the higher order reflec calculations for systematic orientations1 have tion. The smaller Bragg angle of the [220] indicated that high order reflections at high reflections allows many beams of the systema voltages my provide improved dislocation tic set to operate. Figures 1C-5 show that resolution. For example the contrast profiles the use of these systematic reflections leads for dislocations show a decrease in image to a narked decrease in dislocation image width for increasing order of reflection (i.e., width, especially in C-E, which correspond increasing g.b), thus enabling improved to imaging of primarily the edge components resolution of narrow dislocation dissociations of the dislocations. In F and G the contrast and closely spaced dipoles or superdisloca- is primarily due to the screw conponents of tions. the dislocation and, although not as marked as in C-E, there is again some decrease it This japer provides experimental confirma image width with increasing g. Figure 2 tion that high order systematic reflections shows inclined arrays of dislocations imaged do indeed result in an improvement in bright using various systematic < 220 > conditions. field image resolution, e.g., image widths of The confusion due to oscillatory contrast dislocations can be reduced by a factor of which is present in A, especially hhere the -3. The results have teen obtained from dislocations intersect the surface or where .i.*^ • tin) «-U4H Mg. 1. (XBB 713-1047). 105 f-ig. 2. (XBB 713-1046, »•*» F»S. 3. (XBB 707-33:8). 106 Fig. 4. (XBB 7011-5224 they lie above one another, is sharply reduced image profile and multiple-beam bright-field as the order of the reflection at the Bragg profiles obtained by computer methods for condition increases. For example, the double the possibilities gj.b = 1 and 2. Comparing dislocation images which are readily apparent the trace of the image with the theoretical in the lower portion of B and C are extremely profiles, it is clear that there is excellent difficult to resolve in A. Also, the image correspondence for the width of the central width decreases with increasing order of minima and the positions of all extrema for reflection, as expected. Not only is resolu the gj.b = 2 case. Variations in relative tion improved, but the "noise" in the image intensities in the experimental image are due to contrast phenomena (e.g., oscillations most likely due to thickness variations and in intensity) is largely reduced, i.e., image discrepancies in dislocation depth. An confusion is reduced. illustration that the g2-b = 1 dislocation e:chibits only an observable central minimum is given in Figures Id and g. Thus, there These results seem to have broad applica exists confirmation of the applicability of tion to studies of lattice defects and small images obtained using higher-order reflec coherent particles at high voltages and can tions to determine Burgers vectors in a be applied more generally than the "weak beam" system where the Burgers vector is well method of IVhelan et al. 3 characterized. The method can then be applied to a system where the Burgers vector has not Dislocation image characteristics under been characterized, and there is even exten systematic diffracting conditions have also sive debate and confusion about the possible been found useful for determining ,he allowable Burgers vectors. In the cobalt- magnitude of the product gib where gj is the ferrite spinels it is possible to determine first-order reflection of the systematic set which is the correct possibility. Figure 4 and thus can be of tremendous help when it is illustrates how the possibility of a g,.b = 1 necessary to choose between a number of dif dislocation is tilled out by comparing higher- ferent possibilities. For example, in the order bright field images to theoretical fee system, for an undissociated dislocation profiles and comparing the number and spacing with Burgers vector of the type a/2 (110 > , of primary extrema. when a 220 reflection is used the only three possibilities are gj.b equal to 0, 1, or 2. Figure 3 shows the image of a dislocation Presented at the Stockholm Conference on 107 High Voltage Electron Microscopy, April 1971. Now at the Gordon McKay Lab., Harvard University, Cambridge, Mass. 02138 1. W. L. Bell and G. Thomas, Proceedings of 29th EMSA Conf., 1969, p. 156 (Claitors Publishers). 2. L. C. De Jonghe and W. L. Bell, 7th Intl. Congress, Grenoble, 2_, 357 (1970). 3. D. J. H. Cockayne, I. L. F. Ray, and M, J. Whelan, Phil. Mag. 20, 1265 (1969). 2. DISAPPEARANCE VOLTAGE DETERMINATIONS USING A CONVERGENT BEAM Walter L. Bell Disappearance voltages for second-order reflections can be determined experimentally in a variety of ways.l The more subjective methods, such as Kikuchi line disappearance^ Fig. 1. Selected-area convergent-beam dif and bend contour imaging,3 involve comparing fraction patterns for aluminum (111) systema a series of diffraction patterns or micro tic reflections at 222 Bragg conditions. At graphs taken at intervals throughout the most voltages a strong central maximum exists disappearance range and selecting that voltage in the 222 intensity profile, such as seen at which gives the strongest disappearance effect. 500 kV. At the disappearance voltage of The estimated accuracies of these methods are 430 kV the central maximum disappears. both to within 10 kV, or about 2-4*, of the (XBB 723-1361) true disappearance voltage, which is quite sufficient for using these voltages in further Fig. 2. Enlargements of 222 reflections at Bragg conditions for Al-14 Ag alloy at voltages near the disappearance voltage. Central maxima have similar appearances at equal voltages above and below the disappearance voltage, thus interpolations can be made with reasonable accuracy. PCBB 722-820) 108 calculations. However, it is the necessity .ixes at 90°, has been designed and is now in of determining this information by comparisons use. The development was prompted by the of exposed plates rather than while operating wide range of users' requirements for both the microscope that detracts from the immedi greater range and more precise control of ate usefulness of these methods if there is specimen tilting than available with the reason to perform experiments at an unknown standard Hitachi tilt-rotation stage. The disappearance voltage. following design criteria were established: 1) The stage should be directly inter The convergent beam technique for determin changeable with standard HI! 650 stages, ing the disappearance voltage has been found including hookup to existing drives entering to be a highly objective method when it is the column wall. applicable, i.e., when reasonable crystal 2) There should be two separately con perfection exists and an area of uniform trolled axes of tilt at 90°. These axes are thickness can be found.4 The critericr; for to be oriented so as to parallel the transla determining this voltage is that the central tion axes of the microscope's built-in X-Y maximum disappear from the rocking curve for stage. the second-order spot. Figure 1 shows the 3) A tilt range of * 14° along each axis, systematic set of reflections in aluminum if feasible within the standard immersion when the 222 reflection is placed on the polepioce configuration, should be provided. reflecting sphere, both at 500 kV where a 4) The intersection of the tilt axes strong central maximum exists and at 430 kV should be fixed and centered at the specimen where the central maximum has been eliminated plane, so that a correspondingly centered and hence the disappearance voltage attained. specimen detail undergoes neither lateral nor Figure 2 shows enlargements of the 222 vertical displacement dming tilting. reflection from an Al-14 Ag alloy at voltages 5) The tilting actions are to respond in a range about the disappearance voltage, smoothly and positively to manual fine- which is easily determined by inspection to adjustment controls. Settings made with be 410 - 415 kV. The central maximum reap reference to tilt-angle indicators (graduated pears at voltages only 10-15 kV above and in either actual degrees or arbitrary units) below the disappearance voltage and is similar must be accurately reproducible. in appearance at equal voltages either side as seen in Figure 2. Hence further objecti vity is possible with the convergent beam 4. HIGH VOLTAGE ELECTRON MICROSCOPY OF technique since interpolations can be made INTERFACIAL DEFECTS IN GaAs* while operating the microscope after passing through the disappearance range in steps of, Richard A. Osiecki and Gaieth Thomas say, 5 kV and observing at which voltages the central maximum reappears or where similar- High voltage transmission electron micro looking central maxima are obtained. scopy has been utilized in order to study Furthermore, the accuracy with which a voltage possible causes of electrical breakdown at is determined by this method should easily be interfacial regions in a n+/n"/n+ gallium to within 5 kV of the true disappearance arsenide pulse-generating device (Fig. 1). voltage. Electron optical examination revealed defects [OOI] 1. W. L. Bell, Handbook of Conference on the Applications of High Voltage Electron i Microscopy, Harwell, 1970, p. 35. 2. D. Watanabe e* al., Acta Cryst. A25, 138 (1969). 3. R. M. Fisher, private communication. 4. W. L. Bell and G. Thomas, Proceedings of 27th Annual EMSA Meeting, St. Paul, 1969, p. 158. M* GoAs+Te <25^M 3. AN ORTHOGONAL DRIVE TILTING STAGE FOR THE HITACHI 650 kV MICROSCOPE GaAs+Te James C. Hodges and Gareth Thomas A specimen stage for the HU 650 electron Fig. 1. Schematic of a gallium arsenide microscope, giving up to ± 14° tilt along two pulse-generating device. (XBL 712-6498) 109 Fig. 2. Dislocation "butterfly" observed emanating from the free surface of a vapor- deposited n+ layer. Arrows in c and d point out residual contrast from edge components of dislocations. Foil normal is [001]. (XBB 716-2380) Fig. 3. "Butterfly" observed in the epitaxial/ epitaxial interfacial region in which the n+ layer was solution grown. Foil normal is [001]. PCBB 716-2386) which consist of inclusions rich in silicon of electrical breakdown, and consequently and glide dislocation arrays. These defects silicon or other deposits should not be were observed at the free surface (Fig. 2) allowed to form during device manufacture. as well as at the epi-epi (Fig. 3) and epi- substrate interfaces. The silicon impurities probably arise from quartz tubing used in the Submitted to J. Materials Sci. and Eng. diffusional processing and are responsible for (LBL-440); presented at the Stockholm the production of the dislocation arrays. In Conference on High Voltage Electron Micro turn these defects undoubtedly are the cause scopy, April 1971. 110 5. HIGH VOLTAGE ELECTRON MICROSCOPIC STUDIES Orthoenstatite• MgSiOj belongs to the OF MINERALS pyroxenes which play a major role in terrestrial, meteoritic, and lunar rocks. It Wolfgang F. Muller, Gareth Thomas and H. R. crystallizes in the space group "l-ca, Of Weak special interest is its transformation into clinoenstatite, P 2i/c, which is believed to High voltage electron microscopic studies be a martensitic transformatior (Turner et al., have been started on minerals, specifically 1960). Electron microscopy of orthoenstatite on orthoenstatite and anorthite. The speci as-grown from Bamle, Norway, revealed stacking mens for the electron microscope have been faults parallel to {100} which could be thinned by ion-bombardment, using a modified expected due to structural considerations and Commonwealth Scientific Instruments machine. parallel to {210} which is the main cleavage ANORTHITES • »> * ,• . :<*;** . APOLLO 14 VftL SCHIESOHE, ALPS METEORITE Fig. 1. Examples of the long-Tange ordered domain structures from (a) Apollo 14, (b) the Alps and (c) meteorites. (XBB 7112-6099) Ill plane. Dislocations parallel to [001] have or twinned martensites per se are beneficial been observed. Two orthoenstatite samples to toughness is still r.ot answered. Fe-Ni-C.o have been deformed in the Griggs apparatus alloys have been designed so as to try to of the Geology Department. The conditions resolve this issue, paying attention to were 12 kbar confining pressure, 600°C, composition effects and deformation modes. I0"4sec"l strain rate, and 10 kbar confining Work is continuing on Mo and Cr bearing steels pressure, 600°C, 10"^sec"l. Electron micro to characterize their effect on substructure graphs showed lamellae parallel to {100} as and properties. Work on plastic deformation thin as a few hundred Angstroms. Selected induced martensite in single crystals is area electron diffraction patterns revealed continuing so as to understand more completely the presence of untwinned clinoenstatite t the orientation dependence on martensitc which has its a and b axis parallel to the a formation and to relate the volume fractions and 6 axis of the orthorhombic matrix. transformed to the tensile properties. More emphasis will be placed on mechanical pro Anorthite. CaAl2Si20g, triclinic, is the perties of coarsened spinodals and spinodals endmember of the solid solution NaAlSijOg - undergoing martensitic transformations. CaAlSi208 called plagioclases which belong to the petrologically most important terrestrial I). Non-metals minerals. Anorthite also occurs in meteorites and is a major constituent of the lunar rocks Mechanisms of order and substructure at the Apollo 11, 12, 14, and 15 landing characterization of ceramic oxides and sites. A combined x-ray and electron micro minerals will bo continued. Thinning techni scopic study has been started in order to ques have been mastered and systems under learn about the nature of order - disorder investigation are lithium spinels (of transformations of anorthite which is importance in electronics) and feldspars controversial. The results so far obtained (important in mineralogy). The latter are: Dark field images with reflections of involves interdisciplinary activities with the type h + k even, 1 odd ("c" reflections) geology. operating displayed antiphase domains of decreasing size in a metamorphic specimen from c. Electron Microscopy Val Schiesone, in the Alps; in an eucritic meteorite, and in a volcanic anorthite from The special effect of disappearance voltage Japan. That means: the larger the cooling at HVEM will be utilized for two effects: rate the smaller the domain size. The 1) compositional characterization and 2) anorthite from Japan showed domains on the improved defect resolution, e.g., to study order of 70 to 100A in diameter; the cor in-situ electron irradiation damage. Many responding c reflections were diffuse (indica Eeam contrast effects will also be investi ting some short-range order). Anorthite from gated at HVEM. the lunar basalt Apollo 14 14310 displayed antiphase domains using reflections of the type The velocity analyzing microscope will h + k odd, 1 odd (type "b" reflections). also be applied to compositional characteriza Attempts to image domains using the "c" tion, e.g., carbon segregation in iron based reflections failed. The "c" reflections of alloys. It would be helpful to have an image the lunar anorthite were markedly more diffuse intensifier system added to the analyzing than those of the volcanic anorthite from microscope. Japan. Funds will be needed for building a Examples of the long-range ordered domain goniometric hot stage for the HVEM. It will structures from Apollo 14, Switzerland, and be helpful to update the old Hitachi HU 125 meteorites are shown in Figs, la, b, c with the newer HU12 instrument. Trade-in is respectively. possible. W. F. Miller acknowledges a stipend from Deutsche Forschungsgemeinschaft. 7. 1971 PUBLICATIONS AND REPORTS 6. RESEARCH PLANS FOR CALENDAR YEAR 1972 Gareth Thomas and Associates Gareth Thomas Journals and books a. Alloy Strengthening 1. P. R. Okamoto and G. Thomas, Short Range Order in Ni-Mo Alloys: Structural Considera The critical question of whether dislocated tions, Mat. Res. Bull. 6, 45 (1971). n: 2. Der-hung Iluang and G. Thomas, Structure the workshop conference on Microscopy of Clus and Mechanical Properties of Tempered ter Nuclei in Defected Cyrstals, Chalk River, Martensite and Lower Bainite in Fe/Ni/Mn/C Canada, Sept. 8, 1971. Steels, Met. Trans. 2, 1587 fl971). 17. W. L. Bell and G. Thomas, Applications 3. R. J. Livak and G. Thomas, Spinodally and Recent Developments in Transmission Decomposed Cu/Ni/Fe Alloys of Asymmetrical Electron Microscopy, in Electron Microscopy Compositions, Acta Met. J9_, 497 '[19715. and Engineering Materials; Proceedings of the 5th Annual Symposium on the Structure and 4. T. Y. Tan, IV. L. Bell, and G. Thomas, Properties of Materials, Berkeley, Calif• ', Crystal Thickness of Kikuchi Line Spacing, Sept. 13-17, 1971, edited by G. Thomas Phil. Mag. 24_, 417 (1971). (University of California Press), p. 23 (in press). 5. P. R. Okamoto and G. Thomas, On Short Range Order and Microdomains in the N14M0 18. [.. J. Livak and IV. W. Gerberich, System, Acta Met. 1£, 825 (1971). Intorgranular Fracture of a Spinodally Decomposed Cu-Ni-Fc Alloy, ibid. , p. 64. 6. L. C. DeJonghe and G. Thomas, High Voltage Electron Microscopy Studies of Phase presented papers and UC.RL reports Transformations in Cobalt Ferrites, Sci. and Engr. 8_ (5), 259 (1971). 1. S. K. Las, P. M. J. Fisher, P. R. Okamoto, and G. Thomas, Short Range Order in Au-Cr, 7. G. Thomas, Electron Microscopy Investiga Au-Mn, Au-V and Ni-Mo Alloys, presented at tions of Ferrous Martensites Met. Trans. The Metallurgical Society Division of the 2, 2373 (1971). American Institute of Metallurgical Engineers, at Statler Hilton Hotel, Detroit, Michigan, 8. S. K. Das and G. Thomas, Multiple Oct. 18-21, 1971. Transformation Twinning in Ferrous Martensites, J. Iron Steel Inst. 209, 801 2. G. A Stone and G. Thomas, Orientation (1971). Dependence on Deformation Induced £ and a Martensite in Fe-Ni-Cr Single Crystals, ibid. 9. M. Raghavan and G. Thomas, Structural and Mechanical Properties of Fe-Cr-C-Co Steels, 3. R. Osiecki and G. Thomas, Heterogeneous Met. Trans. I, 3433 (1971). Dislocations in GaAs, High Voltage Electron Microscopy Conference at Stockholm, Sweden, 10. T. Y. Tan, W. L. Bell and G. Thomas, April 14-16, 1971. (UCRL-20525, June 1971). Crystal Thickness Dependence of Kikuchi Line Spacing, Phil. Mag. 24_, 417 (1971). 4. R. Osiecki, L. C. De Jonghe, W. L. Bell, and G. Thomas, Dislocation Contrast at High 11. M. Bouchard and G. Thomas, Electron Voltages, ibid. (UCRL-20700, June 1971). Metallography of Spinodally Decomposed (Cu-Mn)3 Al Alloys, in Proceedings of the 5. J. C. Hodges and G. Thomas, An orthogonal 29th Annual Electron Microscopy Society o"f Drive Tilting Stage for the Hitachi 650 kV America Meeting, Boston, Mass.f Aug. 9, 1971, Microscope, Conference on High Voltage edited by C. J. Arceneaux (Claitor's Electron Microscopy at Teddington, England, Publishers, 1971), p. 126. Spet. 22-23, 1971 (LBL-145, Sept. 1971). 12. R. Osiecki and G. Thomas, Enhanced 6. M. Bouchard, R. J. Livak, and G. Thomas, Resolution of Dislocation Imaging at High Interphase Interfaces in Spinodal Alloys, Voltages, ibid., p. 178. International Conference on the Structure and Properties of Green Boundaries and Interfaces, 13. S. K. Das, Early Stages of Ordering in IBM Watson Research Center, Yorktown Heights, NijMo, ibid., p, 124. N. Y., Aug. 23-25, 1971 (LBL-152, Aug. 1971). 14. T. Y. Tan and W. L. Bell, A Computational '/. D. Huang and G. Thomas, Author's Reply to Method for Image Analysis of a Crystal the Discussion of: "Structure and Mechanical Containing Defects, ibid., p. 84. Properties of Tempered Martensite and Lower Bainite in Fe-Ni-Mn-C Steels," LBL-184, Aug. 15. W. L. Bell, Disappearance Voltage 1971 (submitted to Met. Trans.) Determinations Using u Convergent Beam, ibid., p. 184. 8. M. N. Chandrasekharaiah, S. Ranganathan, P. R. Okamot.o, and G. Thomas, Field-ion 16. W. L. Bell, High Voltage Electron Micros Microscopic Study of Antiparallel Twins in copy of Point Defect Clusters, presented at Ni4Mo, LBL-412, Nov. 1971 (submitted to 113 Materials Research Bulletin). 11. R. A. Osiecki, Interfacial Defects in GaAs Device (M. S. thesis), University of 9. I-lin Cheng and G. Thomas, The Enhancement California, Berkeley, UCRL-20591, July 1971. of Strengthening Dislocated Martensite, UCRL-20562, April 1971 (submitted to Met. 12. G. A. Stone, Deformation Induced Trans.) Martensite in Fe-Ni and Fe-Ni-Cr Single Crystals (M. S. thesis), University of 10. S. K. Das, Short Range Order (Ph. D. California, Berkeley, LBL-134, Sept. 1971. thesis), University of California, Berkeley, LBL-176, Nov. 1971. 114 F. HIGH STRENGTH MATERIALS Victor F, Zackay and Eai'l F. Palmer, Principal Investigators important factor in present day alloy design is the existence of high resolution instru INTRODUCTION' ments capable of discerning elements of struc ture at the submicron and, in some cases, at The newly emerging science of alloy design the near-atomic level. The application of has as its goal the development of electrical, these analytic and physical tools promises to magnetic and mechanical properties superior to accelerate greatly the progress of alloy those of cinrcntly available materials. To design. make progress in achieving this goal for struc tural materials a synthesis of both the sci The examples described below of our re ence and the engineering of materials is pre search efforts in alloy design include the requisite. Knowledge from such disparate following: (P Some concluding studies of fields such as dcTect and alloy theory, phase TRIP steels including the proposed develop transformations, continuum mechanics, process ment of steels resistant to hydrogen cmbrit- ing and joining must he gathered and applied tlement (2) the design of carbon-free low to complex systems. alloy iron base materials intended for elevat ed temperature service (to 700°C) and (3) the The immergencc of modern fracture mechanics development uf low alloy untempcred steels in the past decade permits the quantitative with ultrahigh strength and fracture toughness measurement of the intrinsic fracture tough equivalent to the maraging steels. Concurrent ness and fatigue strength of high strength with these studies are several which are dir steels. Much current work in this and other ected toward development of new methods of laboratories of the world is concerned with characterizing structure; particularly those the relationship of defects and microstructure elements of structure whose size and transient to this intrinsic fracture toughness and nature have made their detection difficult to fatigue strength of steels. From such date. The use of NMR techniques to follow the studies, almost surely, will come new concepts initial stages of nucleation in solids is an for the design of superior materials. Another example of such an effort. 1. PRECIPITATION- STUDIES USING NUCLEAR The Al NMR signal in an Al-4 w/o Cu MAGNETIC RESONANCE alloy has been recorded at 150°C and 230°C as a function of time, using a Varian wide- Elias N. Plaza, Mark W. Perra, and Benjamin line spectrometer. Figure 1 shows the way Francis the intensity of resonance changes as a function of time. From this curve the rate The decomposition of a supersaturated of solute depletion from the matrix will be solid solution during aging will cause a re derived. distribution or clustering of the solute atoms and eventually the precipitation of a Additional studies on NMR on the Al-4 w/o second phase. These changes in the micro- structure of the supersaturated solid solution will cause changes within the local electronic 1.0 • • * * • — and magnetic fields, which will in turn s" * cause changes in the nuclear magnetic • resonance (NMR) parameters of a given nuclear OS • species. a • Al - 4 w/o Cu . \ 0.6 • "AI signal In the Al-4 w/o Cu alloy, clustering T ^30'c occurs very rapidly at room temperatures and O ) I 0.4 - • - below. In order to study this phenomenon by NMR, a gas quenching furnace has been built. This furnace can quench the powder to - 50°C 0.2 at a rate of approximately 40,000 °C/sec and n - O in less than S seconds store the sample at O 100 200 300 400 500 600 700 800 liquid nitrogen temperatures until it is Time ( min) needed. Thus the very early stages of the 27 decomposition of solid solutions at room Fig. 1. Intensity of NMR signal of Al in temperatures and below can be identified Al-4 Cu alloy as a function of aging time. and studied. fXBL 721-2159J 115 Cu and the Al-Ag systems are planned. Also determined with the aid of the electron beam the order-disorder reaction will be studied microprobc analyzer. The matrix Ftructiirc in the Cu-Al system was studied wi^I1 optical and electron transmission microscopy and the grain Finally, a frequency-swept NMR spectrometer orientation randomness WPS examined with the is being built and will allow NMR studies x-ray i.auc back-reflection technique. to be extended to ferromagnetic materials. The second phase in the alloys investigated was formed by a nuclention and growth process which resulted in a complex grain boundary 2. PREDICTION OP THE STRESS-STRAIN BEHAVIOR structure as mentioned previously. This OF POLYCRYSTALLTNE a- IRON CONTAINING HARD problem was overcome by the use of nn allo- SPHERICAL PARTICLES tropic phase change, after the aging treat ment, which refined the grain structure and Russell H. Jones spheroidized the grain boundary network. The final structure was a random dispersion of Ihere has been much effort to derive particles in n sort polycrystnllinc matrix relationships which describe the yielding with the grain boundary network sphcroidized behavior and work hardening behavior of and no longer positioned at a grain boundary. single-phase alloys. The yield strength has Figure 1 gives the thorrao moch.inicnl been related to dislocation morphology, alloy treatment. content, grain size, internal stresses, particle morphology and many other variables. In this study, it has been found that up Also, the work hardening of single-phase to some limiting strain the stre1,.- strain alloys has been related to grain size, behavioT of a two-phase alloy h c.en dislocation density, and stacking fault predicted from a relationship based on energy. particle-dislocation interactions. In two-phase alloys containing hard The yield strength values depended on the particles the largest amount of effort has interparticle spacing in a manner similar to been in the analysis of dislocation-particle that proposed by Orowan, but the best fit interaction mechanisms for yielding. Less was obtained with an Orowan-type relationslup emphasis has been placed on the analysis as modified by Ashby.1 This relationship cf work hardening mechanisms in alloys ^counts for the dependence of the line containing hard particles, and consequently tension with dislocation character and the the mathematical relationships describing critical configuration for dislocation this behavior are less accurate than relation by pass. ships describing the yield strength. The work hardening characteristics of The objective of this study was to utilize the two-phase iron alloys depended on the existing dislocation-particle interaction particle diameter, volume fraction of second theories, which were developed to predict phase, and the strain in a manner proposed the flow properties of single crystals by Ashby.2 Ashby proposed that the flow containing hard particles, to analyze the stress had a parabolic relationship to the flow properties of polycrystalline a-iron strain and up to a limiting strain this with a dispersion of hard spherical particles. relationship was obeyed by the iron alloys Conclusions about the correctness of a tested. At this limiting strain the increase particular model can only be made from tests of the flow stress of the two-phase structure of single crystals. The emphasis in this over that of the single-phase structure study was to use relationships based on reached a maximum and it was concluded that measureable material variables, such as this saturation of the particle hardening particle diameter and spacing, volume fraction was responsible for the deviation of the flow of second phase, matrix Burgers vector and stress from the parabolic relationship. The slopes obtained from plots of the flow stress shear modulus and grain size, to calculate versus the square root of the strain compared the flow properties of two-phase polycrystal very well with the slopes predicted by line alloys. Ashby's work hardening model. Iron alloys with second-phase FejTa volume fractions of 0.73 to 6.14% and particle The flow curves of the alloys were cal diameters of 1250A to 2300A were examined. culated using Ashby's work up to 3$ plastic The distribution of the particle diameters strain (Figs.l and 2). Both the magnitude and randomness of the particle dispersion and the shape of the calculated curves were examined with carbon extraction replicas. matched the experimentally determined data. The volume fraction of second phase was Above 3% strain, the calculated data were 116 greater than th. experimental data with the 50LUTIOI TREAT WENT deviation increasing with increasing volume N S PHASE fraction. The difference between the calcu FIELD GRAIN REFINEMENT AND lated and experimentally determined data PARTICLE 5PMERCX0i:ATlON was caused by the saturation of particle- T TREATMENT IN r*F«2 ° IASE FIELD contributed work hardening and by the conver l^-» GRAIN SIZE CONTROL sion of the flow relationship from that of DURING r"« TRANSFORMATION single crystal deformation to polycrystalline deformation. The difference between the , eooL- experimentally determined and the calculated i I ultimate stress was 15? to 35?, depending on the volume fraction of second phase. It was concluded that the flow stress of a two-phase polycrystalline alloy, with hard particles as the second phase, can be success fully predicted from a relationship which depends on the morphology of the second phase. An improved match betwee: experimental and calculated results could be obtained with D 50 IDO 150 200 250 500 T30 TIME IMIN.IJTEG) the knowledge of the mechanism controlling the work hardening saturation. Fig. 1. Schematic of heat treatment given to I-'e-'I'a alloys. XBI. 7,9-7243 1, M. F. Ashby, Physics of Strength and Plasticity, A. S. Argon, Ed., p. 113, The M. I. T. Press, Massachusetts, 1970. 2. M. F. Ashby, Phil. Mag. 14, 1157 (1966). 30 ALLOY 2 EXPERIMENTAL CALCULATED 0.0 2 0.03 TRUE PLASTIC STRAIN Fig. 2. Iron rich portion of the Fe-Ta phase diagram. (After Sinha and Hume-Rothery2). XBL 719-7229 117 3. THE DESIGN OF LOW ALLOY CARBON-FREE IRON- 100 BASE ALLOYS O ; RUSSEL H. JONES WORK jl FOR Fi - I T, BINARY ALLOY 90 < Sungho Jin uj or Recently several binary systems of Fe-Ta, 60 < Fe-Nb, Fe-Ti, Fe-Zr, and Fe-Hf were studied Z with the objective of developing carbon-free 70 2 age hardenable iron-base alloys. Such alloys o contain a fine dispersion of Laves phase 60 u precipitates fFe;Ta, Fe2Nb, FepTi, Fe2Zr, (J) FRACTURE STRENGTH a and Fe2Hf) but typically are brittle due to UJ the presence of a grain brundary film of the 50 a second phase. In this laboratory the highest strength Laves phase alloys were found to be those containing Fe^Nb and Fe2Ta. In recent work of Russell H. Jones the brittle 30 I- fracture of a Fe-Ta alloy was eliminated by < the use of an a-^y transformation which (9 spheroidized the continuous grain boundary 20 ^-- + """" 20 q network.2 However, the relatively high solution treating temperature required caused 10 excessive oxidation as well as coarsening ELONG. AND R.A. of precipitate particles during the spheroidi- / 0 zation treatment. Previous studies on Laves 3 5 7 phase alloys have been confined to binary —• ATOMIC PERCENT CHROMIUM alloys. Fig. 2. Mechanical properties of the ternary alloys (It Ta) as a function of chromium The objective of this investigation was content. Aged at 700°C for 40 minutes fbefore 16001 1 1 1 1 1 1 r i—r spheroidization). CXBL 7112-2201) i i r 100 _ O; RUSSEL H. JONES' WORK FOR Fi - I I, BINARY ALLOY 90 „ UJ 80 « z 70 z o 60 g a ui 50 * o z 40 *» u 1100 r * F«„ TO *? -+-6,•3* 0 3 5 7 ' 670 °C -ATOMIC PERCENT CHROMIUM Fig. 3. Mechanical properties of the ternary alloys (li Ta) as a function of chromium content. Aged at 700°C for 40 minutes and spheroidized at 1100°C for 10 minutes. J 1 1 I I L (XBL 7112-2202) 2 3 4 5 6 7 6 9 10 II 12 "TANTALUM (at.-%) to determine the effect of chromium addition Fig. 1. Effect of chromium on the Fe-Ta phase on the properties of Fe-Ta alloy with the diagram. (XBL 7112-7695) aim of providing a basis for developing 118 improved high temperature creep resisting 1 1 1 1 1 alloys. This involved the determination of LONGITUDINAL ROOM TEMPERATURE TESTS the approximate phase diagram by dilatometer analysis, study of the aging characteristics, measurement of mechanical properties, and the determination of the heat treatment . 230 required for grain boundary spheroidization. Figure 1 shows the effect of chromium additions on the Fe-Ta pliase diagram. The ct-v transformation temperature was lowered • 0.2 % YIELD STRENGTH about 100°C by adding 7 at.fc chromium, and the y-6 transformation temperature was lowered about 60°C relative to the binary _l L_ phase diagram. Figures 2 and 3 show how the 01 IOO 200 300 400 500 600 700 mechanical properties of Fe-Ta alloy changes AS-QUENCHED with chromium addition both before and after TEMPERING TEMPERATURE, X (I HR ) spheroidization. Finally, by multi cycling (a) through the a-y transformation temperature, a fine grain size could be achieved which v™ 150 resulted in increased ductility. LONGITUDINAL ROOM TEMPERATURE TESTS o K VALUES 1. G. Sasaki, Mechanical Properties of Q • K, VALUES _ Laves Phases, University of California, r Berkeley, Lawrence Radiation Laboratory, UCRL-20301, Dec. 1970. 2. R. H. Jones, Ph.D. thesis, University of California, Berkeley, Lawrence Berkeley Laboratory, LBL-179, Oct. 1971. , UT—~~^ z OT 100 200 300 400 500 600 700 < AS-QUENCHED 4. RELATIONSHIPS BETWEEN MICROSTRUCTURE AND a- TEMPERING TEMPERATURE, °C (IHR) FRACTURE TOUGHNESS IN A SECONDARY HARDENING (b) STEEL Fig. 1. (a) Ultimate tensile strength and Roger D. Goolsby yield strength vs. tempering temperature, (b) Plane strain fracture toughness vs. The relationships between microstructure tempering temperature. [Note: The toughness and fracture toughness in a 0.3% C, 5% Mo values at 150°C and 225°C tempering are steel are being studied. The techniques of apparent ones (KQ), since the fractures transmission electron microscopy, scanning were not completely plane strain.] electron microscopy, fracture toughness (XBL 7111-7625)) testing, and stress wave emission detection are being used to characterize the structure of quenched and tempered structures of this example of this type of precipitation is alloy consist of a ferritic matrix containing shown in Fig. 2(a) and (b), which are bright a dispersed precipitate of FejC or M02C, and dark field transmission electron micro depending on the tempering temperature. It graphs of the 500°C tempered specimen. These has been found that the fracture toughnesses micrographs show FejC precipitates both with of specimens tempered at various temperatures martensite laths and also at lath boundaries. are markedly different, even though the Preferential precipitation of carbides has strength levels of these specimens are been observed for all specimens tempered at similar. This is shown in Fig. 1(a) and (b), 300°C and higher. which are curves of yield and ultii.iate strength vs. tempering temperature and Of particular interest is the high tough fracture toughness vs. tempering temperature, respectively. The sharp decreases in ness observed for specimens tempered at low toughness which occur in the tempering temperatures (< 225°C). It has been found ranges 225-300'C and 500-600°C are due to that the toughness of these specimens is preferential vrecipitation of carbides on sensitive to changes in the austenitizing martensite lath boundaries. A typical temperature. In Fig. 3 is shown the fracture toughness vs. austenitizing temperature 119 data for as-quenched specimens. Of signifi cance is the very sharp increase in toughness -c 150 that exists on changing the austenitizing 1 ! 1 1 temperature from 1060°C to 1115°C. There are several factors which could be responsible for this effect, and research is continuing 120 - on isolating the cause of this enhancement in o o toughness. / • 90 ~ 60 - I t "* o K. values — c 30 • K. values - IC 1 g 800 900 1000 1100 1200 1300 •jj Austenitizing temperature [°C ( I hr ) ] Fig. 3. Plane strain fracture toughness vs. austenitizing temperature for as-quenched specimens quenched into iced brine. [XB1, 721-2163) 5. UNTEMPERED MARTENSITE STEELS OF HIGH STRENGTH AND TOUGHNESS William E. Wood The use of commercial alloys such as the low alloy carbon steel (A.I.S.I, types 41XX and 43XX) at high strength levels is limited by their poor toughness. Recent workl sug gests that the toughness of these steels may be increased by altering the heat treatment. •prelimina- work has been done on a 4130 and a 4340 low alloy steel. A typical commercial heat treatments for these steels consists of austenitizing at about 1600°F and quenching into an oil bath. For the 4130 alloy three heat treatments have been studied. The first consisted of austenitizing for 1 hr at 2200°F in an argon atmosphere followed by quenching first into an ice brine bath and then into liquid nitrogen. The second was Fig. 2. Transmission electron micrographs the same as the first except that the of a specimen tempered at 500°C. These show material was quenched into an oil bath. extensive cementite precipitation on Finally, the third heat treatment was that martensite lath boundaries. Ca) Bright of austenitizing at 1600°F for 1 hr followed by field, (b) Dark field reversal of contrast an oil quench, as is done commercially. The from cero;ntite reflection. (XBB 7111-5621) results are given in Table I. By increasing both the austenitizing temperatures and quench rates, fracture toughness values were increased by 604 over those of the commercial treatment. 120 Table I. Mechanical properties vs specimens were tested in the as-quenched con heat treatment for 4130 dition, which resulted in a yield strength of 250,000 psi for the 4340 and 200,000 psi for Austenitizing Quench K. the 4130. Thus the toughness of the 4340 temperature medium lc , ., approaches that of the maraging steels at sim (°F) fksi-in.1^) ilar strength levels, but without the large and expensive alloying additions required for 2200 IROLN 98.5 the maraging steels. 2200 oil 83.5 Additional work is planned to determine 1600 oil 58.0 the optimum heat treatment. Further studies of the fracture toughness and strength of other commercial steels will be made. Table II. Mechanical properties vs heat treatment for 4340 T. Roger Goolsby, Ph.D. thesis, Dec. 1971, LBL-405, Nov. 1971. Austenitizing Quench K. temperature medium ,/9 (°F) _ fksi-in.1'^ 6. STABILITY AND MECHANICAL PROPERTIES OF IBQLN METASTABLE AUSTENITE STEELS 2200 water Dilip Bhandarkar, Victor F. Zackay, and oil 67.3 Earl R. Parker IBQLN 62.3 The unusual combinations of strength, ductility, and toughness that can be achieved 2200 + 1600* water 61.1 in certain grades cf commercial metastable oil 63.8 austenitic stainless steels are known to depend strongly upon phase transformations IBQLN occurring during processing or testing. Largely by controlling a single parameter, 1600 water austenite stability, it is now possible to oil 40.0 produce metastable austenitic steels having a side variety of useful combinations of mechanical properties. In this study, 1 hr at 2200°F in argon atmosphere followed attempts were made to isolate and characterize by cooling to 1600 °F in the furnace and then changes in stability by systematic variations quenched. in composition, processing history, and test environment. The effects of these changes Preliminary studies have also been done in stability on the mechanical properties on the 4340 steel, but over a wider range of were studied; initially in carbonless Fe-Ni heat treatments. A total of 9 heat treatments alloys, and, finally, in more complex carbon- was employed, using three austenitizing containing alloys (Fe-Ni-C and Fe-Cr-Ni-Mn-C temperatures and three quench rates. The systems). The chemical compositions of the first austenitizing treatment was 2200°F at alloys are given in Table I. 1 hr in argon atmosphere, the second was 2200°F for 1 hr in argon followed by cooling The changes in stability and mechanical in the furnace to 1600°F before quenching; properties of the Fe-Ni alloys, induced by and the third was 1600°F for 1 hr. Three processing and testing, are briefly summarized different quench rates were used: (1) ice below. brine quench followed by a liquid nitrogen quench (IBQLN), (2) water quench, and (3) oil quench. The fracture toughness values for 1) The mechanical properties of a severely each of these heat treatments are given in cold worked alloy, 38N, stable with respect Table II. The formation of quench rocks in to either athermal or isothermal phase their higher carbon steel prevented accurate transformations under all conditions of toughness values from being determined. Again processing, were measured to provide a basis by increasing the austenitizing temperature, for comparison with the more complex and less toughness values were increased by 65% over stable systems that were subsequently studied. the values resulting from a commercial heat treatment. 2) A less stable alloy, 34N, exhibited significant stability and mechanical property changes when either processed (rolled) or It should be pointed out that all these tested below the H4 or the Nfe. When tested 121 Table I. Chemical impositions of alloys. (either mode) with strain. Some of these observations are summarized below. 1) A maximum in the yieli strength of Compositions (wt i) all carbon-containing fully austenitic alloys was observed for specimens processed at PDA Designations C Ni Cr Mn temperatures between 200 and 300°C. The combined electron microscopy, magnetic and mechanical property evidence indicated that 34N < 0.010 33.7 -- - the maximum was due to dynamic strain aging, i.e., to the formation of carbon atmospheres 38N < 0.010 37.8 ... -- around dislocations in the highly deformed matrix. CN28 0.294 28.0 ... 0.5 CN8Cr 0.325 8.0 9.0 2.0 2) Two equations, proposed as quantitative descriptions of the relation between the volume of martensite transformed, V , and the CN12Cr 0.290 12.0 9.0 2.0 a strain F., were compared by using transforma CN16Cr 0.292 16.0 9.0 2.0 tion data obtained on the carbon-containing alloys. A parabolic relationship,1 viz., V - me1'2, was found to be superior, parti CN21Cr 0.287 21.4 9.0 2.0 a cularly at large strains, to the autocatalytic form suggested by Angel.2 3) The relationship between stability at -196°C, below both the Ms and the %, this and mechanical properties, using the stability cold-worked alloy exhibited an unusual coefficient m, was discussed for steels combination of mechanical properties (relative whose stability was varied by systematic to those at room temperature), i.e., a lower changes in test temperature (22 to -196°C), yield strength, a much higher rate of work composition (8, 12, and 164 of Ni) and PDA hardening, and twice the elongation at temperature (250 and 450"C). fracture. The unique deformation mechanism of austenite associated with these properties 4) Studies of the shape of the Va (volume was that of a stress-induced transformation of martens.ite transformed) vs. e (conventional to martensite. strain) curves, coupled with a detailed examination of the features of the stress- 3) At a test temperature (-78°C) between strain curves of several steels, permitted the M^ and the Mg the mechanism of austenite unambiguous characterization of the stress deformation (of a cold worked alloy, 34N) induced mode of transformation of austenite. included both the stress and the strain induced modes. A stabilizing effect was 5) At low strains, the Va vs. e curve for produced by deforming at higher PDA tempera steel, CN8Cr, deformed 70* to 450°C, was tures, which led to an accentuation of the observed to follow closely an autocatalytic strain induced mode. Progressive changes in behavior. the stress-strain curves of specimens pro cessed with increasing PDA temperatures reflected the change in the mode of 1. W. W. Gerberich, G. Thomas, E. R. Parker, transformation. These results strongly and V. F. Zackay, 2nd International Conf. suggested that the stability, the mode of on the Strength of Metals and Alloys, transformation, and the mechanical properties Asilomar, Calif., Aug. 1970. of a carbonless metastable austenitic alloy can be changed by thermomechanical processing. 2. T. Angel, J. Iron Steel Inst., p. 165 (May, 1954). The behavior of the carbon-containing al loys was, in many ways, analogous to that of the carbonless Fe-Ni alloys. However, in addition to the structural changes, chemical 7. THE EFFECTS OF HIGH PRIOR-DEFORMATION effects (and associated changes in stability) TEMPERATURES ON TRIP STEELS were observed. Although the behavior of these steels was more complex, this dis Spencer J. Friedrich advantage was somewhat offset by the fact that it was possible to follow, continuously The intention of this study is to investi and quantitatively by means of a magnetic gate the effects of high prior-deformation technique, the course of the transformation temperatures upon the mechanical properties 122 of TRIP steels. The mechanical properties of rolling parameters, and these parameters are a 13Cr, 7Ni, 4Mo, 0.3C TRIP steel have been difficult to hold constant as the rolling studied after 80% prior deformation at tem temperature is increased. In order to obtain peratures ranging from 450 to 650°C--the more meaningful data, the rolling mill has normal processing temperature for this steel been instrumented to obtain power input data, has been 450°C. It has been found that as which should be less sensitive to small the processing temperature increases, the changes in the rolling parameters. austenite becomes more unstable because of alloy depletion of the matrix by precipitating An summary then, if these steels can be carbides. This increasing instability rolled at elevated temperatures, it will decreases the yield strength and elongation reduce the forces on the processing equipment, by IS 4 at 650°C, but does not appreciably which at the lower processing temperatures affect the ultimate strength. is at the capacity of present rolling mills. Thus commercial production of TRIP steels It is believed that this ability to be will be made easier. worked at elevated temperatures without a sharp drop in mechanical properties is due to the molybdenum content of the steel. The 8. HYDROGEN EMBRITTLEMENT OF TRIP STEELS molybdenum strengthens the material either by solid solution strengthening of the matrix Robert A. McCoy at the elevated temperatures, or by precipi tating as a finely dispersed, stable carbide In the investigation on the resistance of at the rolling temperature. In order to TRIP steels to hydrogen embrittlement (H.E,), determine the contribution of molybdenum to two overall objectives were sought: the ability of this alloy to be worked at high temperatures, a moly-free alloy has 1) The determination and understanding of been designed which has similar alloy contents cathodic charging, H2 environment, and load arJ mechanical properties to the moly- conditions for which alpha martensite TRIP containing steel. The mechanical properties steel can be used safe from H.E. as well as (and rolling forces at temperatures from 450 determining the limiting conditions beyond to 900°C) of the two alloys will be compared. which they become susceptible to H.E. If molybdenum is the key to promoting hot 2) The development of a high-strength strength, then the mechanical properties of epsilon martensite TRIP steel which is the moly-free steel should begin falling -ff relatively immune to H.E. even under the most at a lower temperature than the molybdenum- severe conditions. containing steel. In order to differentiate between a solid solution mechanism and a pre Appropriate experiments were designed and cipitation mechanism, specimens at each roll performed to test several hypotheses regarding ing temperature will be aged after rolling. the mechanical behavior of TRIP steel Also, specimens will be held at temperature specimens (tensile and single edge notch) between rolling passes for twice the normal which had been cathodically charged with hold time of 5 minutes. If solid solution hydrogen or mechanically tested in pure H2. strengthening is the mechanism for allowing In general it was found that TRIP steels are high rolling temperatures, then the specimens much more resistant to H.E. from cathodic with shorter times at temperature will give charging than are martensitic steels of higher strengths, with strength decreasing comparable strength. However, after a severe as a function of rolling temperature and cathodic charge or when tested in pure H2, aging time. If carbide precipitation is the elongation of a TRIP steel tensile specimen could be reduced by as much as 95% the mechanism, then mechanical properties of its standard elongation. However, as should increase with aging time or tempera opposed to martensitic steels, TRIP steels ture, reach a maximum, and then decrease. never embrittled prior to yielding. A TRIP steel notched specimen in an atmosphere of Coupons have been rolled for both alloys at hydrogen gas exhibited slow crack growth due 450, 500, 550, and 600°C, and tensile and to H.E. when loaded in tension such that the aging specimens are being machined. During stress intensity factor exceeded a threshold rolling, roll separation forces were measured. level of about 20 to 30 ksi-in.1/2. From It appears that the roll separation forces Arrhenius plots of the data as shown in drop off rapidly at about 500°C for the moly- Fig. 1, the activation energy of this slow free alloy. The alloy containing molybdenum crack growth was found to be about 10,000 has shown no corresponding drop in roll cal/g-atom, approximately the same as the separation forces. This rolling data is activation energy for hydrogen diffusion in largely qualitative, since the roll separation martensite. force is very sensitive to changes in the 123 Fig. 3. SEM fractographs showing the four corner fracture surfaces of a TRIP steel tensile specimen tested in 30 psig Hj> at room temperature at a strain rate 0.002 min"1. (D) shows slow crack growth area. (XBB 719-4364) SEM fractographic studies of the embrittled specimens showed similarities of fracture surfaces between cathodicallj' charged speci mens and specimens tested in H2. Figure 2 shows the fracture surface of an embrittled TRIP steel tensile specimen which had been given a severe cathodic charge. The crack INVERSE TEMPERATURE, y. growth region originating at the surface abruptly changes to a dimpled surface, Fig. 1. Arrhenius plot for slow crack growth indicating ductile failure. The four corners of TRIP steel SEN specimens tested in 15 psig of the fracture surface of a TRIP steel H - (XBL 719-7266J 2 tensile specimen tested in H2 are shown in Fig. 3. Crack growth originating from the surface is in the lower right corner. Apparently for both cathodically charged specimens and specimens tested in H2, slow crack growth occurs until the material in the reduced cross section can no longer support the load and fails in a ductile manner. The fracture surface of a 154 Mn TRIP steel SEN specimen which contained a small amount of a' and exhibited some slow crack growth is shown in Fig. 4A and B. The slow crack growth area of this specimen resembles that of the conventional TRIP steel specimens whose transformation product is martensite rather than e'. The fracture surface re sulting from catastrophic failure at K^ of a 20* Mn SEN specimen which contained some c' but no a' and exhibited no slow crack growth is shown in Fig. 4C and D. As opposed to Fig. 2. SEM fractographs of a cathodically the rough, delaminated fracture surfaces of charged TRIP steel tensile specimen showing SEN specimens containing a', this fracture slow crack frowth from the surface which surface is characteristic of the ductile occurred during a room temperature tensile failure consisting of elongated dimples as test at a strain rate 0.002 min"1. a result of tensile tear. Such a fracture (XBB 719-4365) 124 surface results in a purely austenitic 10. 1971 PUBLICATIONS AND REPORTS structure. Apparently the e' in this material does not change the mode of fracture of the Victor F. Zackay, Earl R. Parker, and Associates austenite. Journals This investigation demonstrated that a high strength epsilon martensite TRIP steel 1. W. W. Gerberich, P. L. Hemmings, and V. F. highly alloyed with Mn can be developed and Zackay, Fracture and Fractography of Metastable that if no alpha martensite is formed prior Austenites, Met. Trans. 2, 2243 (1971). to or during its use, it will remain relative ly immune to H.E. 2. G. R. Chanani, V. F. Zackay, and Earl R. Parker, Tensile Properties of 0.05 to 0.20 Pet C TRIP Steels, Met. Trans. 2_, 13.1 (1971). 3. W. IV. Gerberich, Fracture Mechanics of a Composite with Ductile Fibers, J. Mech. Phys. Solids 19, 71 (1971). 4. J-P. Dirat and IV. W. Gerberich, A Metasta- blc Austcnite with Plane Stress Fracture Tough ness Near 500,000 psi-in.V2iinternational Journal of Fracture Mechanics 1_, 109 (1971). UCRL and LBL reports 1. V. F. Zackay, W. W. Gerberich, and S. F. Ravitz, Mechanical Properties and Corrosion Re sistance of TRIP Steels, UCRL-20523, Feb. 1971. 2. IV. W. Gerberich, Crack Growth Mechanisms in Fibrous Composites, (Ph.D. thesis) UCRL- Fig. 4. SEM fractographs of SEN specimens: 20524, Feb. 1971. (A) and (B), 15% Mn steel showing slow crack growth; (C) and (D), 204 Mn steel showing a 3. W. W. Gerberich, J. P.. Birat, and V. F. dimpled surface characteristic of a ductile Zackay, A Superposition Model for Environmen failure. (XBB 719-4361) tally Assisted Fatigue-Crack Propagation, UCRL-20538, March 1971. 4. C. S. Atherton, The Relation of Toughness to Microstructure in Spheroidized Steel, (M.S. thesis) UCRL-205 9. RESEARCH PLANS FOR CALENDAR YEAR 1972 5. R. A. McCoy, The Resistance of TRIP Steel to Hydrogen Embrittlement (D.Eng. thesisl, Victor F. Zackay and Earl R. Parker LBL-135, Sept. 1971. The major effort in the coming year will 6. R. D. Goolsby, Relationships Between Mi consist of the further development of the crostructure and Fracture Toughness in a alloys described in the foregoing section. Secondary Hardening Steel (Ph.D. thesis), In particular, the relation between micro- LBL-40S, Nov. 1971. structure and elevated properties of the ternary Fe-Ta-Cr alloy will be explored and 7. S. Jin, Spheroidization of the Grain an attempt will be made to optimize composi Boundary Precipitates in Fe-Ta-Cr Alloys tions and microstructures. The studies of (M.S. thesis), LBL-443, Dec. 1971. the untempered steels will be given the highest priority. The program will include 8. C. F. St. John, Theoretical and Experi basic studies of mechanisms of embrittlement mental Studies of Flow and Fracture in Mono- of untempered commercial steels as well as crystalline Silicon (Ph.D. thesis), LBL-113, continued work on optimization of the heat Aug. 1971. treatment of untempered (or low tempered) special steels. A new project will be 9. R. H. Jones, Prediction of the Stress- initiated having as its objective the Strain Behavior of Polycrystalline a-Iron development of a carbon-free alloy having Containing Hard Spherical Particles (Ph.D. high toughness at cryogenic temperatures. thesis), LBL-179, Oct. 1971. 125 HIGH FIELD SUPERCONDUCTIVITY Leo Brewer, Earl R. Parker, Viator F. Zaokay, a. Criteria for Stability and High Tr of A15 and Milton R. Fichus, Principal Investigators Type Phases In this program three different experiment Larry D. Hartsough and Robert H. Hanmond al investigations into superconductivity have been carried out. These approaches range from Although many new A1S phases have been a basic understanding of superconductivity in found in recent years most of the exploration transition metals to a practical technique for has been guided by crystal chemical concepts preparing superconducting tape. such as position in the periodic table and relative atomic size. This study incorporates the recent recognition that the A15 phases are electron compounds. Thus the 3:1 atomic ratio 1. SYNTHESIS OF METASTABLE SUPERCONDUCTING indicated in the usual chemical notation, A3B, A-15 COMPOUNDS for these phases is not rigorously adhered to, especially in systems where A and B are tran Many potential A15 compounds, some pre sition metal elements. There are still a number of binary systems in which the A15 dicted to be high Tc superconductors have not been found in binary systems in which empirical phase lias not been found. The purposes of correlations suggest they should exist. Often this study were to assemble a set of guide this is the result of the relatively low tem lines for the regions of composition and perature stability of this phase, its formation temperature where these "missing" phases might being prevented by the correspondingly low dif be found and to utilize the guidelines to fusion rates and by the nature of the forma synthesize new A1S phases which might also be tion reaction, e.g., peritectic or peritec- good superconductors. toidal. Thus the usual methods of forming a phase, such as cooling from the liquid state, The empirical criterion for A15 occurrence are not successful. is summarized below in the form of a pro cedure to be used in testing a particular binary system for likelihood of finding the The formation of phases by evaporation A15 phase. Much use has been made of the and co-condensation of the elements is a Engel-Brewer Correlation, including bonding aspects of the A15 phas1 and the temperature useful tool when a number of problems might stability of competing phases in arriving at impede the use of other techniques. Most this criterion. important is the fact that low temperature phases can be formed by this technique, through a variation of substrate tejiiperature First choose A elements from Groups 4 to 6 and subsequent annealing conditions. Elec transition metals and B elements from Groups tron beam evaporation makes high evaporation to the right, up to Group 5A. This criterion rates possible, minimizing contamination of reflects the operation of the chemical for the deposited film from the ambient atmos electrochemical) factor dealing with the phere. tendency of A elements to approach a d^ configuration. It also expresses the neces In this program a study was made of the sity that the B element supply more than one criteria for the occurrence of the A15 phase s,p electron. and was used to predict those systems in which it is most likely to be found. Second, the tendency of A elements to form Coupling this with empirical criteria for the A1S phase decreases approximately in the high Tp in A15 compounds enables us to pre order V, Nb, Cr, Ti, to, Zr, Ta, W, Hf. Here, dict those systems most worthy of explora the extra stability achieved by d-bonding in tion. V-Al is one such system and was A15 for 3d metals, electron transfer and d described in the annual report for 1970 band filling, increased stability of bec in (UCRL-20500) and additional research on the the Sd metals, and failure of Group 4 metals compound V3(Al,Ge) is reported herein. to achieve a full d^ configuration are all taken into account. Co-condensation of the elements by electron Third, choose those systems for which the beam evaporation has been successfully ap size ratio RA/RB (CN12) is near 1. plied to the synthesis of NbrSi for the first time. Also, this technique has allowed a Fourth, the composition at which the A15 study of the NbjGa and NbjGe A15 compounds. phase is most likely to occur is determined by 126 electron concentration. Figure 1 demonstrates volume and e/a may be combined, and Tc the probability of finding the A15 phase as versus electrons per unit volume, n, shows a function of the number of electrons per unit definite peaks within a given class of cell (except perhaps for systems with Ti and materials. The envelopes of maximum Tc for Zr). Often the probable composition can be various A15 phases are presented in Fig. 2. determined by extrapolation from neighboring These envelopes account for all binary and systems for which data exist. ternary A15 compounds formed by the A ele ments listed with each envelope. A means of predicting lattice parameter for the hypothetical A15 phases is necessary in order to use electron concentration as a criterion for determining candidates for WE) high T . This has been done by analyzing /A c the data for binary A15 phases with a multiple -q linear regression model. i i i i i i i 4 0 3,0 6.0 70 Applying the preceding criteria, those (6 ELECTRONS/ATOM systems which appear to combine a high likeli hood of occurrence and be good superconduc (VJ _ tors include V3AI, NbRe, Mo gnRe 45, " 14 M Ru Tc _l °.67 .25 .08. Ta.84Al.16; Ta,84Ga.16> -l _ UJ W.43TC.54M0.03, and W Ru. P 3 ^ 12 O — ~ The V-Al system was chosen for the experi z mental part of this study a\nd its synthesis, w '0 composition r.'i/ige, and temperature stability UJ • V) were reported in the 1970 annual report <3 J^ (UCRL-20500). Subsequent data revealed that I 8 V3AI was no', formed aL substrate temperatures in - 5 greater than 480°C but once formed, remained 6 stable to approximately 660°C. The onset of s superconductivity was increased from 9.6 to tt W 11.1°K as a result of annealing for 8 hours 1« _ —, at 600-625°C. Z3 By analogy with Nb (Al, Ge), the highest 2 n 3 Tc superconductor, the addition of Ge to i i i i i i V3AI was tried but the results were incon n T clusive. V3(A1, Ge) films were prepared by 3r2 36 40 44 46 52 56 ELECTRONS/UNIT CELL co-deposition from two sources, V and an Alx Gei-X alloy with x from 0.4 to 0.9. Pig. 1. In the lower portion the number of Most of the specimens had superconducting existing A1S phases is indicated as a func transitions, before and after annealing, tion of the number of electrons per unit which were lower and broader than those of cell. A similar correspondence between the the binary V3AI. electronic density of states and the number of electrons per atom it' demonstrated in the upper portion of the figure. (XBL 717-7047} v,s, h Finally, a procedure for determining the maximum temperature at which the A15 phase is l likely to be found, based on phase diagram - data, is presented in the thesis of Larry h Hartsough (LBL-119). Also the thirty most [ probable binary systems are listed and dis cussed there. 020 030 040 050 020 030 040 OSO 020 030 040 030 w3 EL£CTH0N5/fl High Tc A1S compounds have been empirically found to occur at certain values of the Fig. 2. Envelope of maximum Tc versus num electron-to-atom ratio, e/a, near 4.7 or 6.4. ber of electrons per unit volume for various The mean atomic volume has also been corre A1S phases. The A elements associated with lated with Tc for the elements and a number each envelope a.e given in the upper right of classes of intermetallic compounds. The corner of each box. (XBL 718-7127) 127 One exception with a source composition Through this work we have been able to x = 0.9, exhibited an increase in the onset confirm the dependence of Tc on the increased Tc from 9.0 to 11.2°K after a 3 day anneal order derived from stoichiometry. Although at ~650°C. Although more work on V3(A1, x) we have been successful in extending these ternary alloys; should be done it does not A15 phases to the 3:1 atomic ratio, the appear that the T„ of V3AI can be increased fact that we did not reach the higher critical to over 20°K by alloying, since the maximum temperatures that otiitr investigators did Tc of pure, well-ordered VjAl is probably no can be attributed to the "noise" in the Nb greater than 12-13°K. evaporation rate. Th» fluctuations in the Nb rate were beyond the range of modulation of the feedback control loop. This resulted in a Nb inhomogeniety as i function of depth, b. Synthesis of M5 Compounds in the Niobium and the annealing temperatures were not high System, Nb^X enough to allow bulk diffusion to homogenize the film. Subhas Hazra and Robert H. Hamror.d 5 The fact that niobium has the highest 2) NbjSjL Hopfield has suggested that superconducting temperature among all the an A.15 phase with a small atomic volume, elements and its presence among those com loading to small Nb-Nb interatomic spacings, could give a high superconducting transition pounds with the highest Tc, encouraged us to investigate the A15 phases in the different temperature. 'Ihe Nb-Si system was chosen systems of the element. We have investigated because of the small atomic radius of Si and the Nb-Ge, Nb-Ga, and Nb-Si systems and have the favorable electron density criterion dis cussed above. While a high T is expected, prepared A1S itt^Si for the first time. c the size ratio is unfavorable for it to form. Past attempts at forming the A15 NbjSi 1) Nb_3Ge_and_Nb_7Ga. In the Nb-Ge and Nb-Ga systems the solubility of Ge and Ga conpound resulted in the discovery of has been reported as less than the stoichiometric tetragonal NbjSi below 1945°C. stoichiometric 3:1 composition within the A15 phase. Matthias et al.1 extended the We have synthesized the A15 phase of solubility of Ge in the Nb3Ge phase by the Nb3Si by electron beam evaporation and co- splat cooling technique, resulting in an deposition of the elements. This phase is stable below 680°C and has a lattice param increase in Tc from 5.S°K (at 18 at A Ge) to ~17°K (at 25 at.* Ge). This result indicated eter, ag, of 5.170 ± .001A at 21 at.* Si. the dependence of Tg on the increased order The A15 phase was identified by the presence derived from stoichiometry. of all expected x-ray diffrr.ction lines. The relative intensities of all x-ray lines in Making use of the rapid quenching a Debye-Scherrer powder diffraction pattern characteristic of vapor deposition and the were in good agreement with calculated concumitant extension of solubility limits values. we proceeded to prepare these two compounds with the hope of extending the Ge and Ga The A15 phase was found in all films deposited on alumina -••'-'*-]--*„ = cu tempera content in the A1S phase. NbjGe was jv— ^ ' up to 25 (±1) at. 4 Ge by vapor deposition tures less than 530°C, with increased order onto low temperature alunina substrates and Si solubility at the higher substrate (300°C). An increase in Ge content in the temperatures. An amorphous, or highly dis A15 phase was observed with decreasing sub ordered structure, was observed at lower strate temp.rature (from 600 to 300°C). For substrate temperatures (100-200°C) and at higher Si concentrations with increasing each substrate temperature the maximum Tc occurred at the Ge rich phase boundary in deposition temperature. Consequently the both the as-deposited and annealed films. A15 phase was not found at the 3:1 ratio The maximum T- onset attained in annealed until a substrate temperature of 450°C was films (700°C for 24 hours) occurred at the used. For films prepared at 530°C, the 3:1 composition and was 15.2°K. atomic ordering was greatest at 21 at.% Si, while the 25 at. 4 region was disordered, as evidenced by the broadened diffraction lines. During our investigation of the Nb-Ga A maximum transition temperature of 9.3°K system Webb et al.^ claimed a stoichiometric was found at 21 at.% Si with a decrease at Nb3Ga with Tc > 20°K, obtained by rapid quenching from the liquid, followed by a low higher and lower concentrations due to che temperature anneal. We have also been able increased disorder of the 25 at. 4 region and to extend the A15 phase to 25(±1) at.* Ga the highly disordered A15 structure in the by deposition at a substrate temperature of two-phase region (bec + A15) at lower Si content. Films deposited at 1000"C resulted 160°C. A Tc onset of 16.5"K was obtained after a 24 horn anneal at 700°C. in the formation of tetragonal NbjSi. 12S Annealing increased Tc cf the as-deposited disordered A1S phase and of the amorphous films at 25 at.i Si, but never exceeded 9.3°K. The results of the structural transformation of amorphous films is summarized by A 610°C. A1c + . 640°C.. ,„, u T + A 680°C. A 24 hr A1S A 170 hr '" 24 hr M5 + T '*" » A1S (improved order) + T(more), where A s amorphous and T = tetragonal. Thus the amorphous phase could not be transformed directly into a pure A]5 phase. This could only be formed by vapor deposition at the proper substrate temperature. In summing up our observations we suggest that in order to maximize Tc for NbjSi, one should deposit the films at the maximum substrate temperature capable of synthesizing A15, probably on the order of 600°C. Even at this temperature the stoichiometric com position may not be satisfactorily ordered ar-.d subsequent annealing should be carried out at temperatures less than ~680°C. Fig. 1. Powder-rolling apparatus. 1. B. T. Matthias et al., Phys. Rev. 139A, (XBB 7011-5225) 1501 (1965). 2. G. W. Webb et al. (to be published). 3. J. J. Hopfield (to be published). approach to such a composite was through powder metallurgy. The procedure was to (1) make a porous framework of the matrix metal (Nb), (2) infiltrate t'le pores with the 2. HIGH FIELD SUPERCONDUCTING COMPOSITES BY alloying element (Sn), and (3) diffusion POWDER METALLURGY heat treat the couple to form the super conducting phase. Mechanical deformation The superconducting materials with the could be introduced between steps (2) and (3). highest known values for the critical magnetic field and the critical temperature are brittle A system was developed for the direct intermetallic compounds. Our objective is to rolling of matrix metal powder into continuous develop procedures for making practical high tape with a controlled porosity and inter field superconductors from such materials connected pores. Figure I is a photograph of with a specific set of characteristics. The the powder-rolling apparatus. The relevant first of these is that the morphology of the parameters in the preparation of the porous material should be amenable to control by tape include the average particle size and thermo-mechanical treatments, since it is distribution of the matrix metal powder, known that the current-carrying capacity of the roll gap and speed, sintering and a superconductor is structure sensitive. infiltration conditions, degree of mechanical Secondly, the material should behave as a deformation, and the time and temperature for 'stable' superconductor--under an impressed diffusion. electro magnetic load it must not revert to the normal state catastrophically. Finally, a. The Niobium-Tin System the material should have adequate mechanical flexibility to permit the winding of solenoids. Prakash Babu and Milton R. Pickus It was thought that these characteristics The powder metallurgy technique was applied could best be achieved in a composite material to an existing and useful A15 compound--NbjSn. consisting of thin interconnected filaments Roll-compacted tape (0.016 in. thick) with of the brittle superconducting phase dispersed good green strength was obtained from niobium in a ductile matrix, and that the best powder using a roll gap and speed of 0.012 in. 129 < 10 to ~ SOu. However, to investigate the influence of particle size a small quantity of powder was classified into graded size fractions. A bize fraction in the range -20 + lOu was found to be optimum with respect to size, uniformity of distribution, and volume fraction (~284) of pores. Figure 1 shows a scanning electron micrographic com parison of the commercial (-270 mesh) and the -20 + lDp powders _, and Fig. 2 shows the distribution of tin in the as-infiltrated tapes (directly related to the respective pore systems). Vacuum sintering the grcon niobium tape for 2-3 minutes at 2250°C imparted excellent ductility with very little loss of porosity. Comparable ductility and porosity could also bo obtained by sintering for only 1/2 minute at 2350°C. The porous sintered tape was infiltrated by immersion in liquid tin for 1 minute at 650°C. At higher immersion temperatures the ductility was decreased by Fig. 1. Scanning electron micrographs of niobium powder. x25 (a) -270 mesh (-53 microns), as received from supplier. (XBB 721-609) (b) -10 + 20 microns fraction extracted from above. (XBB 721-610) Fig. 2, Porous niobium tapes infiltrated with tin (white area). x440 (a) From -270 mesh (-53 microns) niobium and 5 rpm, respectively. Mast of the following powder. (CBB 7110-5259) work was done with the commerical powder (b) From -10 + 20 microns niobium powder having a broad distribution of sizes, from (CBB 7110-5263) 130 Fig. 5. Current density as a function of the magnetic field for selected resistivities in the Nb,Sn filaments. fXBL 7110-7518) in Fig. 4 was obtained by heating a 754 Fig. 3. Tape shown in Pig. 2(b), cold rolled reduced tape for 2 minutes at 950°C. Under to 754 reduction in thickness. x440 these conditions some unreacted tin remains, fCBB 7110-5275) which appears as white areas in the photomicrograph. With the same diffusion heat treatment, but a prior cold reduction of 85-90?, there was observed a marked reduction in the amount of unreacted tin and a cor responding increase in the amount of Nb3&i-- the latter accounting for about 30% of the volume. The average diameter of the NbjSn filaments was about 2u. The superconducting properties of these tapes, despite the very short reaction times employed, compare favorably with the best reported values for niobium-tin superconduc tors. Critical temperatures were in the range 18.1 - 18.4°K. In a steady magnetic field of 80 kG (the maximum available) the critical current density for the NbjSn filaments was ~3 x 10s A/cm2 at 4.2°K. The critical cur rent density in a pulsed magnetic field (dis Fig. 4. Tape shown in Fig. 3, diffusion heat cussed in the report by Terr)' Garrett that treated 2 minutes at 4S0°C. x680 follows) of 150 kG was -6 x 104 A/cm2 at White: Tin; light gray: niobium matrix; 4.2°K. The current density as a function of dark gray: NbjSn. (CBB 711Q-52S3) applied field is shown in Fig. 5. The composite tapes have good stability the formation of intermediate phases, and at and flexibility. As an indication of the lower temperatures the infiltration was incom latter it was found that the tapes could plete. sustain coiling, uncoiling, and even reverse coiling with nc deterioration of supercon In order to elongate the tin into thin ducting properties. filaments, \'.he infiltrated tapes were reduced in thickness by cold rolling. Due to the difference in flow stress between niobium and b. Critical Current Measurements of Nb^Sn in tin, it was found necessary to limit the a Pulsed Magnetic Field amount of reduction to 85-904 to avoid excessive loss of tin. The microstructure Terry N. Garrett of an infiltrated tape cold rolled to a 75% reduction in thickness is shown in Fig. 3. A pulse magnet, capable of producing The large interfacial area between the niobium magnetic fields up to 200 kG, has been made and tin, when the tin is present in fila operable and initial critical current density mentary form, is very effective in diminishing measurements have been made on Nb^Sn tape the diffusion time to attain a given volume prepared by the powder metallurgy technique fraction of N'bjSn. The microstructure shram discussed above. 131 The field is produced through the dis WEI8HT PERCENT O'.UMIMUM charge of a 110,000 uF capacitor bank, with 1 2 3 4 5 6 7 8 9 10 12 14 16 18 a maximum operating voltage of 1.1 kV, 1 1 1 SI | 1 1~L^I II 1 1 1 1 1 1 1 through a helical magnetic coil. The bank ^S. ""*"-,, LIOUIO consists of ten smaller capacitors in ?.oao - ^~-4?° I960" \32.0 parallel. The difficulty of synchronizing the discharge of all ten capacitors via ten 1600 ignitrons was removed by requiring only one ignitron to discharge the entire bank. 1600 / j / 4p-V|H£ Additional problems were encountered in pulsing the magnet for high field strengths 1400 Nb AI / (3200 kG) due to the induced mechanical t forces. 1200 - M^Sn specimens were prepared as indicated lOOO . 1 ll 1 III 1 1 0 5 10 15 20 25 30 315 I 410 45 above, except that they were cold worked ATOMIC PERCENT ALUMINUM over a range from 75 to 954, prior to reacting in the tin bath. The microstructure of Fig. 1, The niobium-rich region of the these samples consisted of a network of con niobium-aluminum phase diagram.-' tinuous filaments of NbjSn and unreacted tin (XBL 7110-7430J in a matrix of niobium. Larger volume fractions of NbjSn were found in the 954 The heat treatments used for the formation cold worked specimens due to the reduced of NbjAl were based on the phase diagram of diffusion distance. Lundin and Yamamoto,* the niobium-rich portion of which is shown in Fig. 1, and on the experimental observation that at temperatures in fields up to 180 kG. The~95S cold worked above 1550°C (the eutectic temperature for a specimen was observed to have the highest mixture of Nb2Al and Nb AI3) the reaction 4 critical current density, a Jc « 5 x 10 between niobium and the aluminum-rich phases A/cm2 was obtained at 120 kG. The other was too rapid to control. Therefore, in the specimens had smaller values but were in ex first series of heat treatments the teirpera- cess of 10* A/cm2. These critical currents ture was held at 1530°C for periods of 1, 5, compare favorably with the results of others 10, 15 and 30 minutes. After 1 minute at 4 4 which range from < 10 to 6.5 x 10 at 120 kG. 1530°C, the critical temperature, Tc was 16.95°K. The Tc values gradually increased with longer times at 153D"C, reaching a c. Critical Temperature of Niobium-Aluminum maximum after 15 minutes of 17.47"K. Since Composites it is known that the Tc of Nb3Al is enhanced by annealing in the temperature range Thomas Tom and Milton R. Pickus 650-800°C, all of the samples were subse quently vacuum annealed for 18 hours at TSO^C. The purpose of this study was to adapt the After this anneal the Tc values were all powder-rolling technique to the niobium- increased by approximately 1/2°K, with a aluminum system, develop a heat treatment for maximum of 17.89°K obtained for the sample the formation of the A15 phase with a high that had been heated for 15 minutes at 1530°C. critical temperature, and determine any correlations between critical temperature and From the phase diagram of Fig. 1, assuming process variables. there is still some MJ2AI present, it appeared that a closer approach to stoichiometry, and Infiltration of the porous niobium tape, therefore a still higher Tc, might be achieved prepared as described by P. Babu in the by a two-step heat treatment prior to the low foregoing section, was accomplished by temperature anneal: heating for 15 minutes irmersion in liquid aluminum for 15 seconds at at 1530°C followed by a shorter heat treat 800°C. Tapes infiltrated under these con ment at temperatures from 1700 to 1800°C. ditions were very auctile and reductions in The effect of the additional short time-higher thickness up to 90S were possible. The temperature treatment was to diminish Tc for infiltrant was primarily aluminum with a thin temperatures greater than 1700°C and times layer of Nb AI3 between it and the niobium longer than ~2 minutes. The highest critical matrix. At lower temperatures infiltration temperature, 18.7°K, was achieved using a was only partial. Immersion at higher second annealing step of 1-1/2 minutes at temperatures or for longer periods at 800°C 1700°C. embrittled the tape as a result of the rapid increase in the amount of Nb AI3 that was It is recognized that the lack of metal- formed. lographic and microprobe information makes 132 difficult a proper interpretation of these data. All attempts to prepare metallographic sections of heat treated niobium-aluminum composites resulted in pull-out of the intermediate phases. No problem was en countered with either bulk samples of the individual phases, from which anodized color standards were prepared, or with as- infiltrated composites. A metallographic examination technique and evaluation with respect to current-carrying capacity is planned, 1. C. E. Lundin and A. S. Yamamoto, Trans. AIME 236, 863 (1966)• Fig. 1. Tlie critical temperature, TC) for "amorphous" transition metal alloys and ele 3. SUPERCONDUCTIVITY IN AMORPHOUS ments in the Sd series. The alloys are from TRANSITION METAL ALLOYS neighboring elements except for Re-Ir and Ta-Re alloys. The dashed curve indicates Tc The absence of long range order revealed a for the corresponding crystalline alloys. correlation between the superconducting tran (XBL 7110-7513) sition temperature, Tc, and atomic-like properties in amorphous 4d transition metal films, as reported in the 1970 annual report.1 The position of the Tc maximum for the There it was shown that Tc exhibited a single amorphous 5d alloys closely corresponds with maximum near the middle of the transition met the maximum number of unpaired d-electrons in al series in contrast to the corresponding the atom, i.e., a half filled d-shell in the crystalline alloys. In order to examine the atom. This occurs at rhenium which has an generality of the T behavior for amorphous 2 5 c s d atomic configuration, with seven electrons transition metals and in the hope of increas per atom. The half-filled d-shell of the atom ing our understanding of these materials we should correspond roughly to a half-filled have examined superconductivity in vapor- d-band in the solid and a maximum in the co deposited Sd transition metals and in a 4d-5d hesive energy. combination. Structural studies are also summarized. Thus an interpretation of the T_ results is provided by the Barisic, Labbe, and Friedel^ (BLF) relation for the electron-phonon coupling a. Amorphous Sd Transition Metals parameter. A, in terms of the cohesive energy due to d-electrons, Ec(d), as Michael M. Collver and Robert H. Hammond The highly disordered or "amorphous" metal films were made by condensing the metal onto a M2> sapphire substrate maintained at 4.2"K. The results for the superconducting critical tem where q0 represents the exponential decrease, perature of vapor-quenched 5d alloys from Hf e"1° , of the d-wave function in the region to Ir are shown ,'n Fig. 1. The dashed curve of overlap, (w^ > is the mean square lattice represents the Tc for crystalline solid solu frequency, and M is the ion mass. tion alloys. We observe a relatively smooth behavior with Tc rising to a single peak in The heat of sublimation has a maximum at W, the 5d films as found in the amorphous 4d but it must be recalled that the cohesive en alloys. However, there are distinct differ ergy entering the BLF expression for A is the ences relative to the 4d results. The peak the contribution due to the d-electrons only. occurs near Re, at an e/a "6.8, while in the The possibility of a cohesive energy peak at 4d series it occurs close to Mo, at e/a * 6.4. Re receives some experimental support from the Also there is a definite change of slope in chemisorption data of Plummer and Rhodin. Tc as a function of e/a near W, contrasted They measured the binding energy of 5d atoms, with the near-linear behavior in the 4d series. on a W surface and found a maximum at Re. Overall, we ' elieve these results corroborate our earlier views regarding the dominance of an atomic-like property. For the cohesive energy to be the dominant factor explaining the behavior of Tc, < w^) 133 must be nearly constant for the Sd alloys and exhibited a decrease in resistivity generally the other parameters affecting Tc, such as the within a temperature < 30°K and were found to electron-electron Coulomb repulsion parameter, be crystalline at 300°K. The Tc of films H , must vary only weakly with e/a. The large which crystallized reverted to or toward the variation in < u)2) found in the crystalline bulk crystalline values while the films which metals and which primarily accounts for their remained amorphous (to 300°K) changed by less variation in Tc, is apparently washed out in than 0.1°K. the amorphous state. This view is supported from a calculation of Tc data and heats of sublimation.* found for amorphous alloys composed of ele ments with a large difference in the number Recent theoretical calculations taking into of electrons per atom. In the 4d series an account electron correlations have obtained amorphous Zr.so RU50 (A e/a = 4) alloy at a triangular shaped cohesive energy curves with composition corresponding to the amorphous a sharp maximum as a function of the number of Tc peak had a Tc * 3.2°K, and in the Sd r electrons per atom, rather than the commonly series the Tc for a Ta_g0 I .40 alloy occurred assumed parabolic maximum. In the amorphous at ~2.6°K; in both cases Tc was reduced by 4d alloys the Tc peak [and presumably K,-(d)] more than 6° from the peak for nearest- is well defined and triangular, and a similar neighbor alloys. In such alloys there is an appearance can be observed i.i Fig. 1 for the increased likelihood of charge transfer from 5d alloys. the d-electron-rich atom to the atom deficient in d-electrons. This may result in a special An extrapolation of the linear Tc behavior sort of short range order capable of altering near the W-Re peak leads to zero Tc near the Tc. beginning and end of the 5d series. However, the Tc data fall below this extrapolation on In conclusion, we find that the super both sides of the Tc peak. The dip to the conductivity in amorphous transition metals left of the peak occurs for IV-Re and Ta-Re exhibits a systematic behavior in close cor alloys and suggests that the filling of the respondence with both an atomic-like property d-band occurs as a function of e/a in the man and its manifestation in the solid state, ner indicated by Tc. The apparent cutoff of i.e., the filling of the d-band and the superconductivity near the high e/a ends of cohesive energy. This correspondence was the amorphoi's TM's may be the result of in found for alloys having small differences in cipient magnetic effects; such effects are e/a. believed to occur in the crystalline transi tion metals. 1. IMRU Annual Report for ]970, UCRL-20500, The transition temperatue after heating to p.137. 300°K is shown in Fig. 2. Specimens which 2. S. Barisid, J. Labbfi, and J. Friedel, remained in the amorphous state up tj 300 °K Phys. Rev. Lett. 25, 919 (1970). are indicated by the arrows. The other films 3. D. M. Newns, FRys. Rev. Lett. 25, 1575 (1970). ~~ b. Amorphous 4d-5d Transition Metal Films Michael M. Collver and Robert H. Hammond The Tc behavior for amorphous transition metals has been extended to an alloy svstem composed of elements in different rows of the periodic table. The Mo-Re system was chosen because it is in the region of e/a, where the Tc peak occurs in both the 4d and 5d series and because it was not possible to prepare Mo-Tc alloys, the latter being radio active. These alloys were prepared by elec tron beam evaporation from a single source and Fig. 2. The critical temperatures after the composition was determined by electron warming to 300°K. The arrows indicate alloys microprobe analysis to within ± 4 at.i. that showed no change in resistivity upon warming and had amorphcus-like diffraction The superconducting transition temperature patterns. (XBL 7110-7514) is shown in Fig. 1 for films in the as-depos- 13 I"' 10- I t 5 500- 0 20 40 60 80 100 Mo ATOMIC PERCENT Rt Re Re Mo Re Fig. 2. The transformation temperature, l'tr. from the amorphous to crystalline state for Fig. 1. The critical temperature, Tc, of Mo- Re alloys in (a) the as-deposited "amorphous" Mo-Re alloy films. The equilibrium phase state and (b) after warming to 300°K. The diagram is superimposed (note different temperature scales). (XBI. 719-7269) dashed curve indicates the Tc for the cor responding crystalline alloys. fXBL 721-2077) c. Electron Microscopy of Vapor Deposited ited state (a) and after warming to 300°K (b). Transition Metal Films As might be expected fro,;i the previous behavior for the 4 and 5d films, a peak in T. is present. Richard loop, Michael M. Col Ivor, and Robert However, it occurs at an e/a * 6.25^6.35, low II. Hammond er than that for either the 4d or 5d alloys. The position of the peak between e/a of 6 and The structures of the v^por-deposited 7 is understood in terms of the half-filled d- transition metal films is expected to play a band, but its position relative to the 4 and considerable role in understanding the super 5d series is not understood at present. As in conducting properties which have been pre the other series an extrapolation of Tc leads sented. An initial effort in this direction to zero Tc at the ends of the transition metal has been pruvided by the electron microscopy series. Thus the atomic-like nature of Tc is observations on the allov systems Y-2r, also observed upon mixing TM rows. Xb-Zr, Xh-Mo, Mo-Ru, Mo-Re, IV-Rc, and IV-Ta. All observations were made after removal of Annealing studies were also performed for specimens from the vacuum system at room this system. Figure 2 shows the temperature, temperature. Ttr, at which the transformation from an amorphous to crystalline structure occurred, Fortunately, many films remained in the with the equilibrium phase diagram superim as-deposited state to temperatures greater posed. At temperatures less than 300° K the than 300°K with no change in their critical transformation temperature was assumed to be temperature and only slight changes, if any, that at which a sharp decrease in resistivity in their resistivity, Flectron micro^jpc occurred. Transformation temperatures observations of these films indicate that any greater than 300°K were directly determined grains, if present, are less than 20A in by either heating on the hot stage of an diameter. This is based on observations of electron microscope or heating in a separate grains < 3QA in pure Ru films under optimum vacuum system (10"'' Torr) and subsequently resolution con-Jit ions. All films exhibiting observing the structure in an electron micro diffuse, amorphous-appearing diffraction scope. In genera] the grain sizes after patterns, after warming to 300°K, lie in two transformation were proportional to Ttr. phase regions of their respective equilibrium 135 crystalline values after heating to 300°K. This behavior fits into the notion that a close-packed structure is representative of the amorphous state; i.e., these elements and alloys are already close tc an amorphous structure. Radial distribution analysis of electron or x-ray diffraction data would be useful in determining whether the short range order is close packed as suggested by the results presented here. d. Stability of Amorphous Transition Metal Alloys Michael M. ColIvor A variety of criteria have boon given for the formation of amorphous metallic structures. The most outstanding of these include limited solubilities of the elements, atomic size Fig. 1. Electron diffraction pattern of differences greater than 10%, a composition "amorphous" Mo_ Ru . (XBB 719-4353) range corresponding to the center portion 68 >32 of the miscibility gap, and an atomically rough or amorphous surface. Based on the phase diagrams. The diffraction patterns for data acquired in our work, these criteria these alloys exhibit three broad haloes, have been found to be too restrictive, at typical of an amorphous structure, but which least for transition metals. Perhaps this is can be ascribed to the first, third, and due to past studies which have concentrated fourth rings of an fee diffraction pattern. on alloy systems composed of hep and fee A typical diffraction pattern is shown in elements. The above factors will aid in preparing and stabilizing an amorphous Fig. 1 for Mo 68 Ru j2- While many of the elements in these alloys are bec, the bec structure to higher temperatures but are not structure is definitely excluded. Heating necessary for its preparation. of several samples to higher temperatures (~800°C in some cases) resulted in the rapid Rather than suggest alternative criteria transformation to an unidentified crystal for the formation of amorphous structures, structure with large grains (~ lu). emphasis is placed on their temperature stability. Amorphous metals can be prepared Those specimens whose Tc reverted to or under such a great variety of conditions-- toward the crystalline value uid whose e.g., gaseous impurities, deposition rate resistivity decreased rather sharply over a dependence, substrate material and tempera narrow temperature range, as a result of ture, mode of deposition, etc.--that any warming to 300°K, were found to have crystal criteria for their formation must specify a line diffraction patterns characteristic of large number of deposition conditions. their cquilibium structures, with grain sizes Consequently, for a given alloy system pre varying between 40 and several thousand pared under identical deposition conditions angstroms. Such patterns were found for we have concentrated upon the variation of films of pure elements and alloys whose tlie transformation temperature (amorphous •• composition corresponded to the solubility crystalline) as a function of composition. A range of these elements. Thus the systems schematic irreversible phase diagram has exhibiting complete solubility, Nb-Mo and been constructed, shown in Fig. 1, which K-Ta, were found to be bec at all com indicates the temperature stability of the positions. amorphous structure (solid curve) super imposed upon the equilibrium phase diagram There were some exceptions to tlic above (dashed figure). One is allowed to cross a behavior; some hep elements (Ir, Ru) and phase boundary in the direction of increasing their dilute alloys failed to exhibit any temperature only, due to the irreversibility sharp decrease in resistivity and were found of the transformation. This figure is based to have crystalline diffraction patterns. on observations on seven allov systems (one However, some structural change had occurred shown in Fig. 2 of Paper 3b, this report) and since there were changes in T toward is capable of qualitatively describing past re- 136 i 1 1 1 1 1 1 1 r single-phase structure prior to separating into two phases. In the two-phase region near the solubility limit the metastable phase is usually that of the neighboring solid solution, as observed ir. Ma-rich Mo-Re and W-rich W-Re alloys. The solubility can be extended across the entire composition range in simple eutectic systems of elements with the same crystal structure and similar atomic sizes. Mare complex behavior is found in the two- phase regions of systems composed of elements of different atomic size and/or different crystal structure, and in systems which form compounds. The transformation often goes through intermediate metastable crystalline phases at temperatures on the order of 1/3 of the average melting temperature. e. The Superconducting Phase Transition in Disordered Transition Metal Films Charles W. Krause, Michael M. Collver, and Robert 11. Hammond In the last three years an intense interest has developed in the superconducting phase transition. The superconducting phase is well o IO «o to ao IOO described by the microscopic BCS theory, * ATOMIC PERCENT • introduced in 19S7. Since then, mort work in Fig. 1. "Irreversible" phase diagram of an superconductivity has centered on unravelling amorphous alloy system. A typical crystalline the details of the microscopic theory and its phase diagram is superimposed. (XBL 719-V293) phenomenological counterpart, the Ginzburg- Landau-Abrikosov-Gorkov theory. A good under standing of the phenomenological driving suits on amorphous and metastable alloy films. mechanisms of phase transitions in general has been attained by workers in thermodynamics In general the transformation temperature and quantum statistical mechanics. The current remains relatively low in the elements effort surrounding the superconducting phase CS 300*K) and increases slowly with increasing transition is an attempt to formulate a fully solute concentration in the single-phase microscopic theory for a transition, that solid-solution region. Since the atoms are between a superconductor and a normal metal. already randomly distributed as in .-. solid solution only small atomic displacements (low In particular, precursors to the phase temperatures) are necessary to trigger the change in superconductors can be observed at transformation to the crystalline structure. temperatures twice the critical temperature and Wiere a pair of elements form a complete above. These precursors were first identified solid solution, e.g., Nb-Mo and W-Ta, the with the enhancement of the conductivity of transformation temperature remains near to or thin amorphous non-transition metal films scales between those Tr_ for the individual above Tc due to short-lived fluctuation Cooper elements [40 to TCK and 130 to ~360°K, pairs. Much work on "clean" and disordered respectively). non-transition metal films is proceeding else where; our equipment and technology allowed a The increase of Ttr is most rapid as the first look at the transition metals. alloy composition moves into a two-phase region. For the uniformly mixed amorphous A Curie-Weiss behavior for the fluctuation- material to transform to the equilibrium induced conductivity, o', was found: two-phase structure would generally require temperatures sufficiently gTeat for diffusion to occur (in an amorphous, structure). The heat released upon transforming would aid in accelerating the transformation. However, where R is the normal state sheet resistance, the material often transforms to a metastable on is tlie normal state conductivity, T = 137 be investigated in the hope of providing an alternate low-resistivity current path, viz. TUNGSTEN Rj-IOTA a Cu interface between NbjSn and the Nb ma THICKNESS >2SOA trix, for short-term instabilities due to flux motion. Use of the pulse magnet to investigate the current-carrying capacity of niobium- aluminum tapes is planned, followed by an ex tension of this work to the Nb-Al-Ge system. 5. 1971 PUBLICATIONS AND REPORTS Leo Brewer, Earl R. Parker, Victor F. Zackay, Milton R. Pickus, and Associates Journals and books Fig. 1. The effect of superconducting fluctuations on the conductivity of amorphous 1. L. D. Hartsough and R. H. Hammond, Solid W is shown here, a' is the conductivity con State Comm. 9, 885 (1971). tribution from the fluctuations and cjn is the conductivity in the non-superconducting state. 2. R. L, Loop, M. M. Collver, and R. H. (XBL 719-7253) Hammond, Fifth International Materials Symposium, University of California, Berkeley, Sept. 1971. T-Tf/Tc, and g is the width parameter. Figure 1 shows typical results. A value of 3.0(±1.0)xl0"s ST1 for g was obtained for nine UCRL and LBL reports elemental and alloy films, measured from (T=0.01+0.01) and varying in Rj| from SO to 1. L. D. Hartsough, Analysis of Prior Claim 200 n. This is in good accord with the values to Synthesis of V3AI, UCRL-20556, April 1971. obtained from the theory and comparable in reproducibility obtained in experiments on 2. L. D. Hartsough, Stability of A15 Type non-transition metal films. Phases as Affected by Electron Concentration (Ph.D. thesis), LBL-119, Sept. 1971. In conclusion, it should be pointed out that 3. Subhas Hazra, Synthesis of Low Temperature the transition metals are peculiar in that the Stable A15 Compounds in the Niobium System d-electrons carry the super-current. Our ex (M.S. thesis), LBL-433, Dec. 1971. periments indicate that no additional modifica tions of present theories are needed to ex 4. N. P. Babu, Correlation of Thermo-mechani- plain the results we obtained for these cally Varied Microstructures with Supercon materials. ducting Properties of NbjSn Processed by a New Powder Metallurgy Technique, (M.S. thesis), LBL-437, Dec. 1971. 4. RESEARCH PLANS FOR CALENDAR YEAR 1972 5. T. Tom, Application of Powder Metallurgy Process in the Formation of Hb^Al Super Leo Brewer, Earl R. Parker, Victor F. Zackay, conductor (M.S. thesis), LBL-188, Dec. 1971. and Milton R. Pickus 6. T. N. Garrett, Pulsed Magnetic Fields and It is planned to design and construct an the Critical Current Densities of Supercon integrated system to prepare long lengths of superconducting tape by powder metallurgy ducting M^Sn Strips in These Fields (M.S. processing techniques. An investigation of thesis), LBL-448, Dec. 1971. the superconducting properties of these tapes in the form of test solenoids and, hopefully, 7. M. M. Collver, Superconductivity in a superconducting magnet will be made. Sim- Amorphous Transition Metals (Ph.D. thesis), plications in the processing techniques, LBL-167, Nov. 1971. namely the use of an inert gas environment for sintering and infiltration and possible 8. R. L. Loop, The Structure of Vapor replacement of niobium by niobium hydride Deposited 4d Transition Metals (M.S. thesis), powders will be investigated. Sane emphasis LBL-177, Sept. 1971. will be placed on obtaining the most desirable particle size distribution. 9. C. W. Krause, M. M. Collver, and R. H. Hammond, ••luctuation Pair Conductivity in Disordered Superconducting Transition The niobium-copper-tin ternary system will Metal Films, LBL-181, Oct. 1971. 138 H. THERMODYNAMICS OF METAL SYSTEMS Ralph R. Hultgren, Principal Investigator 4. HEAT CONTENTS OF SODIUM NITRATE \ND SODIUM NITRITE 1. EVALUATION OF THERMODYNAMIC DATA Jaime Cases and Ralph Hultgren Ralph R. Hultgren, Donald T. Hawkins, and Pramod D. Desai The heat contents are being measured. The project for the evaluation of Thermodynamic Data for Metallic Systems, the 5. RESEARCH PLANS FOR CALENDAR YEAR 1972 first edition of which was published in book form,* has been continued. Current evalua Ralph R. Hultgren tions as they are completed are distributed to all who request them. During 1971, Plans are to complete the work on heat selected values were distributed for the contents of sodium nitrate and sodium following 85 systems: nitrite. We will continue evaluations of the thermodynamic data. Two volumes of the In, IT, OS, Re, Rh, Ru, Sn , Sr, Ta, Zn, Ag-Al, evaluations will be published in the summer Ag-Au, Ag-Bi, Ag-Hg, Ag-In , Ag-Mg, Ag-Mn, of 1972 with the American Society for Metals. Ag-Pb, Ag-Sn, Ag-Tl, Ag-Zn , Al-Ce, Al-Co, Al-Cr, Al-Fe, AJ-Mg, Al-Mo , Al-Pd, Al-Pr, Al-Pt, Al-V, Al-Zn, Au-Cd, Au-Hg, Au-tn, 6. 1971 PUBLICATIONS AND REPORTS Au-Pb, Au-Pd, Au-Pr, Au-Sb , Au-Sn, Au-Zn, Be-Ni, Ca-Ni, Cd-Te, Co-Fe , Co-Ni, Cr-Fe, Ralph R. Hultgren and Associates Cs-K, Cs-Na, Cs-Rb, Cu-Ti, Fe-Ir, Fe-Nb, Fe-Pd, Fe-Pt, Fe-Sb, Fe-Te , Fe-U, Fe-W, Journals and book Fe-Zn, Ga-Ni, In-Ni, K-Na, K-Rb, Li-Sb, Li-Sn, Mg-Ni, Na-Sb, Ni-Pd Ni-Pt, Ni-Sb, 1. R. Hultgren and P. Desai, Selected Ni-Sn, Ni-Ts, Ni-W, Ni-Zn,Pb-Te , Pd-Sn, Thermodynamic Values and Phase Diagrams for Sb-Sn, Sb-Sr, Sb-Zn, Sn-Te , Sn-Tl, Sn-Zn, Copper and Some of Its Binary Alloys. Te-Zn, Tl-Zn. Monograph I. INCRA Series (International Copper Research Association, New York, Thermodynamic data were critically 1971), 204 pp. evaluated during 1971 to obtain tables of selected values. 2. D. T. Hawkins and R. R. Hultyren, The Effect of Ordering on Lattice Heat Capacities: Ordered and Disordered AuCu, J. Chem. Thermo R. R. Hultg-en, R. L. Orr, P. D. Anderson, dynamics 3, 175 (1971). and K. K. Kelley, Selected Values of Thermo dynamic Properties of Metals and Alloys 3. H. Bell and R. R. Hultgren, Heat Capacity (John Wiley, New Yo.k, 1963). of Liquid Bismuth, Met. Trans. Z, 3230 (1971). UCRL and LBL reports 2. LOW TEMPERATURE HEAT CAPACITY OF BINARY 1. J. L. Dickson, Vapor Pressure of Lead ALLOYS and Activity Measurements of Lead-Platinum Alluvs by the Torsion Effusion Method Hong-il Yoon and Ralph R. Hultgrer. (M. S. thesis), LBL-116, July 1971. This work was completed during 1971. Two 2. H. Yoon, I. Low Temperature Lattice Heat articles have been accepted for publication Capacities of Alloys: InPb, AuCd, Mj^Sbj. and a Ph. D. thesis filed by H. Yoon. II. Effect of Ordering on Lattice Heat Capacity: AuOr; (Ph. D. thesis), LBL-101, Sug 1971. 3. VAPOR PRESSURES OF Pb-Pt Alloys 3. H. Yoon and R. R. Hultgren, Heats of Fori at ion of Solid Indium-Lead Alloys, Jerry L. Dickson and Ralph R. Hultgren UCR -18402 Rev., July 1971 (J. Chem. and Eng-. Data, in press). The work on vapor pressures has been completed and Mr. Dickson received his 4. J. Chipman, Thermodynamics and Phase Master of Science degree based upon this Diajram of the Iron-Carbon System, UCRL- work. 205 5 Rev., Aug. 1971 (Met. Trans., in press). 139 I. ADVANCED MATERIALS Robert H. Bragg, Principal Investigator 1. STRUCTURE AMI PROPERTIES OF SOME ADVANCED MATERIALS Robert H, Bragg We have been studying the properties oi Glassy Carbon, GC, which has mechanical aid electrical properties totally different from conventional carbons and graphites. We hbvt also begun studying unidirectional so'ldiftco- tion of eutectics in order to produce unique microstructures with unique properties. Finally we have studied the flaw and fracture of multistrand boron reinforced composite tape. Some delays have been encountered in getting the experimental program underway, but all of the problems are coming under control almost simultaneously. Significant comments are included in the short reports given below. DEPOSITION SURFACE 2. EFFECT OF POLARIZED X-RAY BEAM IN SMALL- ANCLE SCATTERING BY PYROLYT1C GRAPHITE Madan M. Bisral and Robert H. Bragg As discussed in UCRL-20517, unusual resolution and partially unexpected specimen volume effects with newly available Bonse- Hart small-angle scattering apparatus was observed. The principal results were that CI) highly polarized x-ray beams are obtained, (2) these beams make possible the study of materials which exhibit double Bragg dif fraction and make difficult or in some cases impossible the observation of snail angle scattering from, say, cold worked metals, and (3) the specimen volume effect is yet to be explained on theoretical grounds. Figures 1 and 2 illustrate the experimental situation which demonstrates the effect of: the polarized beam orientation relative to the orientation of a textured material. Figure 3 illustrates the specimen vol'jne effect with MgO single crystals. Somewhere between specimens Ca) and (d) there should have been a reversal in the ordinates of the Fig. 1. 'Jhree pyrolytic graphite specimens curves if true small-angle scattering was having different orientations with x-ray beam. operative. The bafal plane in i"a) and (c) is parallel to tfe: electric vector whereas in (b) it is perpetmiicular. fXBL 711-6454) 140 Considerable additional work has been done on the physics of the apparatus itself which has heretofore been unexplored and is worthy of publication, particularly in view of the fact that several theories regarding the in terpretation of data must be revised. Perti nent experimental data are shown in Figs. 1-3. 3. STRUCTURE OF GLASSY CARBON Madan M. Biswal, Peter P. Bihuniak, and Robert II. Bragg Becasuc of the difficulty encountered in locating a suitable source of glassy carbon, a comparative study of available commercial products was made. 'Ihe results of this comparative study were presented at the 10th Biennial Conference on Carbons at Lehigh Universtiy (UCRL-20566). An acceptable source of glassy carbon was selected from the comparison just described and material has been ordered and received. A furnace for the heat treatment has been rebuilt and tested to 2850°c. The low temperature transport properties apparatus has been designed and as a result of competitive bid ding an order has been placed for the con struction of the dewar, and a design of the Fig. 2. Log-log plot of small-angle scatter electronic read-out circuitry has been com ing curves for three different orientations. pleted. Thickness of the specimen is 0.125 cm. (XBL 711-6456) An interesting feature of the comparative study was that some glassy carbons were far more porous than others, in some cases two different sources might give permeabilities varying by a factor of 10^1 This was beauti fully borne out by scanning electron micros copy, where large cavities that we expected to find in the porous material were seen. Data presented at the last Carbon Conference, since verified in our laboratory by Biswal, using transmission microscopy and diffraction, is that the structure of glassy carbon, far from being a system of closed pores made by juxtaposed small crystallites, may be more like a mass of interwound spaghetti. It would be as though the original long chain polymers, ever, when carbonized, still remember their basic f-amework. 4. BRITTLE FRACTURE 1W FILAMENT-REINFORCED MATERIALS K) JO BO 90 100 IIO 8. SECONDS Guillaume R. Kerhuel and Robert H. Bragg Fig. 3. Scattering in the small-angle region in graphite for five different thicknesses. The fracture behavior of boron-aluminum In this case the scattering increases with was investigated. Monolayer tapes of boron decrease in thickness in the ultra-small fibers imbedded in an aluminum matrix by a angular region. (XBL 711-6457) liquid infiltration process were tested in 141 400 vl 1,1. • V-' a 300- • • $$ 200 (a) tik-. 100- ^ Fig. 2. Stress-strain curve of fiber and of k composite. (XBL 717-7079) « Fig. 1. Polished cross sections of the material studied. PCBB 718-3572) -* Eiparlmaniol rnulli ~ Wtlbull'i wlropotoflon simple tension at room temperature. It was •• Ealropolotiofl o' repeatedly observed that the entire gage «ipirlm*nlal ratullt length of the specimen shatters at fracture, and a series of parallel fractures was obtained. The number of broken fibers within the composite was measured at different stress levels and in the broken parts of the speci men. In all cases this number was found to 100 ISO 200 be very small. The fracture distribution along the gage length did not show any symmetry but was not random. In all cases Fig. 3. Cumulative frequency distribution long and short broken parts were obtained. of strength for the fibers and for the matrix. Examination of the fracture surfaces showed (XBL 717-7080) 142 Fig. 1). Thus far, no heat treatments have been attempted. Considering the fact that checkouts began only last November the pro gress has been remarkable. 6. RESEARCH PLANS FOR CALENDAR YEAR 1972 Robert H. Bragg Fig. 4. Typical fracture distributions along the gage length. fXBL 717-7082) Redesign of the high temperature heat treat ment furnace will be done, and it is expected that heat treatments to 3000°C can be ac typical features related to the length of complished. This will make possible the the broken part examined, 'ihough few experi systematic investigation of the kinetics of mental results are available at the present the "graphitization of glassy carbon." It time, it is believed that "secondary" will furthermore provide material for the fractures result from local compression clue systematic investigation of the low tempera to dispersion of stress-waves emitted during ture transport properties which will be the first or "initial" fracture. The performed by a new graduate student. The phenomenon observed depends not only on the work begun in 1971 on unidirectional solid material investigated but also on the geometry ification of eutectics has progressed of the tensile specimen. It is similar to rapidly to the point where it is relatively Hopkinson multiple fractures phenomenon and easy to grow duplex single crystals worthy should be observable with other brittle of investigation. The structure of these elastic and dispersive material. The eutectic materials will be investigated by pertinent experimental evidence is shown in the student who initiated the work and a new Figs. 1 - 4. student whose background is in plasma physics (theoretical and experimental). For the moment the work on fracture of composite materials is being suspended, lfork will 5. EUTECTIC SOLIDFICATION begin on electronic properties in the Spring Quarter. Kwame Ankra and Robert H. Bragg This work is accelerating apace. Excellent microstructures have been obtained and the apparatus works virtually unattended (see 7. 1971 PUBLICATIONS AM) REPORTS Robert H. Bragg and Associates Papers presented 1. M. M. Biswal, P. P. Bihuniak and R. H. Bragg, Structure of Glassy Carbon, presented at the 10th Biennial Conference on Carbon, Lehigh University, Pennsylvania, June 28- July 2, 1971. 2. G. R. Kerhuel, Flow and Fracture of Continuous Filament Monolayer Composites, presented at the Fifth International Materials Symposium on The Structure and Properties of Materials --Techniques and Applications of Flectron Microscopy, University of California, Berkeley, Calif., Spet. 13-17, 1971. Fig. 1. Photomicrograph showing lamellar structure of Al-CuAl2 eutectic unidirection- ally solidified from bottom to top. (X400) (XBB 722-841) 143 UCRL and LBL reports 2. G. R. Kerhuel, Brittle Fracture in Filament-Reinforced Materials (M.S. thesis), 1. M. M. Biswal, Effect of Polarized X-ray LBL-138, Sept. 1971. Beam in Small Angle Scattering by Pyrolytic Graphite (M.S. thesis), UCRL-20517, March 1971. #$*'• }. #% j*J.o •* jf J " V IK *'"*0' 1 ^M^ J&P*I I jjaTj* 1 *te|f * * ^ j^R'^ffife"' tf|% WaM^^'iL l( OJ iw;*!?* ^ "•£*'<••» '• Kf %.,'}' ,?£ A - |»&P»J€-. r^jfji^' H^S S^*r^ * « IHw *MI '-'*'° ' HKLJL If V*l \ * • i. MSP^W5 ...» <®5 Si «,? *3H III. Ceramic Science 14S A. HIGH TEMPERATURE REACTIONS Alan U. Searay, Principal Investigator J_ = J.v cSTFc3H~ (3) 1. THE KINETICS OF VAPORIZATION AND OF ENDOTHERMIC DECOMPOSITION REACTIONS Here C is the clausing correction for the Alan W. Searcy and Dario Beruto* length-to-radius ratio of the pores and ac is the condensation coefficient for the vapor A model for analysis of the kinetics of phase decomposition product. This equation vaporization in vacuuiir>' lias been extended to should apply regardless of the rate-limiting describe congruent vaporization processes in process at the AB surface. which dissociation of the eventual vapor mole cules from active surface sites or separation Equations (1) and (3) and the models that of these molecules from collision with other lead to them suggest experiments that can particles of the self-adsorption layer is rate throw additional light' on the nature of the limiting. Tne equation derived for the flux rate-limiting process in vaporization, decom J when a single vapor species is produced is position reactions, and in gas-solid reactions, P°exp[-(AGa • dGs)/RT] Jv = (1) NATO Research Fellow 1970-71. (2nMRT)m * where P is the equilibrium pressure, AC is 1. A. W. Searcy, in Chemical and Mechanical the free energy of formation of the escaping Behavior of Inorganic Materials, edited by molecule in a two-dimensional gas transition A. W. Searcy, D. V. Ragone, and U. Colombo state, AG? is the free energy of formation of (Wiley-Interscience, New York, 1970), Chap. 6. the reaction site ov surface particle, R is the gas constant, and M is the molecular 2. R. J. Galluzzo and A. W. Scare)', High weight of the vapor molecule. The equation Temperature Science J_ (in press). shows that for dissociative processes of the kinds assumed in the related models that lead to Eq. (1) to be rate limiting, requires that the apparent enthalpy of activation for 2. THE KINETICS OF DECOMPOSITION OF CAJXITE vaporization exceed the enthalpy of the SINGLE CRYSTALS equilibrium reaction and requires the apparent Dario Beruto* and Alan W. Searcy entropy to be within a restricted range of values. Calcite was selected as a prototype materi al (a) for a study of the applicability to The Langmuir equation gives an upper limit decomposition reactions of the experimental to the rates of endothermic decomposition techniques that have proved useful for studies reaction rates just as it does for congruent of congruent vaporization reactions, and (b) vaporization. Furthermore, the vaporization for evaluation of theoretical models developed models that were developed for congruent in thi? laboratory for decomposition reaction vaporization can be used for endotherntic de kinetics (see previous abstract). composition reactions, which can be written in the general form The same apparatus and tccliniques that have been described by Leonard and Searcy* were AB(s) = A(s) + BCg). W used except that the alumina parts that fit about the calcite crystals to define a The predictions of the model can be compared limited area for vaporization were more care to experimental data which are collected at fully machined. As a consequence, vaporiza short enough times so that the reverse reac tion was completely negligible from all of a tion can be neglected. calcite crystal slice except for the defined area of surface. Unless the vapor produced in the decomposi tion reaction can escape by some condensed Scanning electron micrograph photographs phase patii, the growth of a porous solid made in the calcitc study showed the reaction product of decomposition should yield an product to consist of a lu thick layer next to effusion barrier to reaction. Accordingly, an the calcite and an outer layer that grew as expression was derived for Jn, the net flux the reaction proceeded. Both layers had that should escape through a porous product occasional lOu pores, but most of the pores effusion barrier: were smaller than lp in diameter. X-ray dif- 146 fraction studies showed the outer layer to be expected conclusion is consistent with the polycrystalline calcium oxide in its normal view that adsorption layers are more readily (NaCl-typc) structure, with the crystallo- maintained near equilibrium with bulk phases graphic direction oriented preferentially than has been generally believed. normal to the exposed plane of calcitc. the inner layer was also polycrystallinc and oriented) with a structure that is a rhombo- I hcdral distortion of the normal calcium oxide NATO Research Pellow, 1970-71. structure, Earlier evidence from differential thermal analyses indicated that a metastable 1. H. B. Leonard and A. W. Searcy, .1, Appl. modification of calcium oxide exists, but ours Phys. 42, 4047 (1971). is apparently the first visual and x-ray dif 2. K.TT. Stern and L. L. Weise, High Tempera fraction evidence. A surprising result of the ture Properties and Decomposition of Inorganic stud}' is that, while "more kinetic studies of Salts, I't. 2, Carbonates, Nat. Std. Rof. Data the thermal decomposition of CaCOj have been Sor., National Bureau of Standards (U.S.J 30, reported than for any other carbonate, and 32 pp. (1969). perhaps for any other salt",2 and while those studies arc in general agreement that the apparent enthalpy of activation for decomposi tion is about equal to the enthalpy of the * ANISOTROPIC VAPORIZATION RATTS OP OPPOSITE equilibrium reaction, the study in our labora FACES OF ZINC OXIDE SINGLE CRYSTALS: EXPERI tory yielded a value 9 kcal (about 18*) per MENTS AND THEORY* mole higher than the equilibrium enthalpy This difference is e consequence of the higher Robert J. Calluzzo and Alan IV. Searcy vacuum maintained in the study here. The rates of vaporization of opposite basal faces of zinc oxide have been measured as The entropy of activation derived from the functions of temperature. The rates differed experiments on the basis of our model is 37.2 by about a factor of 4. The average apparent eu compared to 36.21 eu for the entropy of enthalpies of vaporization for the reaction reaction for CaC03(g)->CaO(s) • OJ2(g) • The ^.iU(s) - Zn(g) • 1/2 02(g) for the two faces agreement in these values to within the ex were measured to be: zinc-rich (0001) face, pected •indom error of measurement means that 134.6 t 3 kcal/mol; oxygen-rich (0001) face, desorption probably constitutes part of the 140.8 ± 5 kcal/mol, compared to 110.5 ± 0.S rate-controlling process. kcal/mol calculated from thermochemical data for the equilibrium reaction. For calc.itc the decrease in flux with in creasing product layer thickness which is pre The apparent entropies for the reaction dicted by Eq. (3) of tlic previous abstract was were for the (0001) face, 49.9 1 2 eu and for not observed. This result means cither that the (0001) face, 50.2 • 3 eu. These entropies ac, the condensation coefficient, for CO? on to within expected limits of error, arc equal calcite is much smaller than we assumed in to each ether and to the entropy of the using Eq. (3) or that carbon dioxide can equilibrium reaction at the same temperature, escape by some process more effective than 48.0 ± 0.3 eu. The results are consistent effusion. with expectations of a new model1 if desorp tion is rate limiting. The theoretical argu Experiments on the effect of introduced CO2 ments that justify this conclusion are re pressures on the decomposition rates arc viewed. planned to confirm or disprove the assumption made about o^, namely that it cannot be smaller than the vaporization coefficient. Unpublished theoretical analyses in our To appear in High Temperature Science 3, laboratory indicate that the condensation co No. 6 (1971). efficient is equal to tlic vaporization co 1. A. K. Searcy in Chemical and Mechanical efficient, which is known from the short-time Behavior of Inorganic Materials, edited by weight loss measurements. We expect the ex A. K, Searcy, D. V. Kagonc, and U. Colombo periments to confirm the equality. If it is (h'ilcy-Interscience, New York, 1970), Chap. 6 confined, the failure of the porous calcium oxide layer to reduce the reaction rate probably indicates that surface diffusion along the high-surface area channels main 4. THE STABILITIES OF te^gj AND U F (g) tained the activity of carbon dioxide at the 2 6 outer surface of the calcium oxide layer at substantially the activity of carbon dioxide James A. Roberts, Jr., and Alan K. Searcy at the calcite surface. This somewhat un The existence of La2l:(, in the equilibrium 147 vapor of lanthanum fluoride has recently been demonstrated by Skinner and Searcy.1 The measured entropy of dimerization by the reac tion 2 LaF3(g) = La2Ffc(s) calculated by con ventional analysis of the mass spectrometer data appeared anomalous when compared with entropias of similar reactions. Skinner and Searcy reasoned that their anomalous measured entropy was a result of extensive fragmenta tion of the La2F6 parent to LaFj + LaF3 + La + e" under electron impact in the ion source of the mass spectrometer. The LaFt formed in the reaction would be masked by the LaFj which was produced as the major ion species from LaF3(g). Calculation of pressures from+intensities of the minor fragment peak, La2FS» which was not masked, would then lead to an apparent pres sure of the dimer that was below its actual pressure. Skinner and Searcy suggested that their in terpretation of the lanthanum fluoride data might be tested by studying the dimer to mono mer vapor pressure ratios in cerium fluoride vapor. The additional electron on each cerium atoir might reduce the tendency of the dimer to fragment. Using the same experimental teclinique as 66 70 75 Skinner and Searcy, we have exaiiiined the l/T (MIO4) CK-'I equilibrium vapor over cerium fluoride (Fig. 1) Fig. 1. Vaporization data for cerium and made new measurements for the I^Ft/LaFj trifluoride. (XBL713-6568) ratios. The Ce2F$ intensity ratios and new La^FJ/LaFi ratios and absolute intensities for each ion are close to the l^Fc/LaF-S ratios ed value than were the results of Skinner and of Skinner and Searcy, but botn ratios vary Searcy. more steeply with temperature. The results for the lanthanum fluorides are given in We noticed that the data taken at the Table I. The entropies of dimerization we higher pressures fell below what would be ex measured are closer to the value (-35 eu/mol) pected from extrapolation of the low tempera Skinner and Searcy had selected as the expect- ture data. The deviation of the log P vs. l/T curves was greater than could be caused merely by heat capacity changes. The effect Table I. Thermodynamic value at 1577°K. also seemed to be more pronounced for the diner than for the monomer. Thermodynamic LaF3 LaF3 CeF, quantity (Ref. l)(tii is work)(this Cork) This bending over as molecular flow con ditions are exceeded constitutes a behavior AH monomer (kcal) 98.2 98.2 that has apparently not been analyzed pre V 95.8 viously, presumably because in most mass AH dimer (kcal) 123.7 131.5 131.5 spectrometer studies, the measurements are made at lower pressures to avoid any problems 4,* dimerization3 -72.7 -64.9 that might arise from exceeding the limits of -60.1 Knudsen flow. The observed behavior is the AS nonomer (eu) 43.2 43.2 43.7 opposite of the effect observed in torsion effusion^ and weight loss^ studies where an AS 6liner (eu) 40.3 44.6 46.8 increase in pressure was observed. After re examining the basic nature of the mass spec- AS dimej-izationa (eu; -46.1 -41.8 -40.6 trometric and other effusion techniques, we think that this difference in behavior can be at least qualitatively understood. The mass spectrometric technique differs from the other ^alwes fc> the reaction 2MF,, -, M F 3(g) " 2 6 techniques in that it samples a snail fraction (g) of the effusing species at some distance from 148 the orifice and thus is sensitive to scatter insi e bubbles to the vapor pressure of the ing due to background gases or collisions in liqi 1 within vjhich the bubbles form and to the beam itself even after the gas has left the mbi'-nt pressure are quantities of great the cell. The most frequent scattering col impo tance to the theory of nucleate boiling. lisions would perhaps be monomer-monomer col Curr nt analyses of vapor phase nucleationl'3 lisions, because monomers are the predominant are ased on a theory developed by Young, species. But the fraction of dimers scattered Laplr.ce, and Kelvin and made rigorous by from the beam should exceed the fraction of Gibb;. These analyses involve the assumption monomers scattered because the roost probable of a pressure barrier to the formation of events, when relative concentrations are vapc nuclei. They involve the use of the normalized, are collisions in which the dimers, Lapl ce equation because of their slower average velocity, are 2o overtaken by the faster moving monomers, Thus, AP even though the h:,$\ monomer pressure is r ' responsible for bieakdown of Knudsen flow con when P is the pressure gradient across the ditions, the deviations of dimer intensities from the low temperature extrapolated values surface of the bubble, a is the surface ten is more pronounced than deviations of monomer sion and r the bubble radius. intensities. r bbs' analysis involves an error in one of its mderlying assumptions, and a new The possibility raised by Skinner and equa ion for the effect of surfaces on vapor Searcy, that the main frequent dimer peak is pressure was therefore derived.'' In lino masked by the monomer peak of the same mass, with this new analysis we present a theory of remains. For these two fluoride systems we vapo • phase nucleation based on a free energy cannot eliminate ' le relatively high uncer ban er rather than a pressure barrier. tainty in measurer, jnts of dimer temperature dependence. To dr so requires carrying %• consider the formation of a bubble as measurements to lower intensities than are now proc-eding in two stages: first, the trans feasible with our mass spectrometer. Uncer foration of the liquid into vapor at the tainty in dimer data must accordingly be ambient pressure (which also includes any recognized as high. hydi 'Static head), with a free energy change of f /SJirr^AGy where AGy is the free energy We plan to inve ^tigate the fragmentation char e for the liquid-vapor transition per pattern for Al2Clt , which is known to be unii volume of vapor formed at ambient pres present in aluminr n chloride vapor at concen- sure secondly, the free energy of formation trations more tha an order of magnitude high- of the interface, 4Tir2o. The total tree er than is the mor raier. While quantitative energy change when a bubble is formed is predictions of vai lation in fragmentation 4 •? patterns from one netal halide system to J AG j iir AGv 4Tir^a another cannot be nade from our present level r of understanding, lualitative similarities are Since AGy is negative for a superheated strong. If A1C1J lppears as the principal liquid, this function will have a maximum as r mass peak of the c imer, the kind of fragmenta- increases at the critical radius r , tion behavior hype diesized by Skinner and c Searcy will have I ->en experimentally demon- r = -12_ strated, and the t lalysis made by Skinner and c AG. Searcy can be app] Led to the data for both lanthanum and cer. jm fluoride vapors with Further analysis gives reasonable confide nee. 2a- r 1. H. B. Skinner and A. W. Searcy, J. Phys. c F^ Oiem. 75, 108 (19 1). f In/ 2. D.~A\ Schulz nd A. W. Searcy, J. Phys. where f is the ambient pressure and f„ the Chem. 67, 103 (19> 3). fugacity of the superheated liquid. Beyond 3. K.T). Carlson P. W. Gilles, and R. J. rc, dGy/dr is negative and the bubble will Thorn, J. Chem. P ys. 38_, 2725 (19633. grow spontaneously. Values of the critical radius for various amounts of superheating of water at 1 a'tm of ambient pressure are given in Table I. 5. THE FORMATION If VAPOR PHASE NUCLEI IN NUCLEATE BOILING* The fugacity of the vapor inside the bubble fj, can be calculated from the equation Alfred Biichler an Alan W. Searcy fa f ln^i f ln The relationships of the vapor pressure 1 t 0 r a 149 Table I Superheating, T, °K: 10 170 3.23 x 10-3 1.61 x 10"3 3.20 x 10"4 4.15 x 10"5 where 170° is the highest reported superheating.5 Table II. Bubble pressure as a function of radius. Radius Bubble pressure fi, atm. Bubble pressure fi, atm. r, cm. (Tn - 374°K) (T„ - 383°K) 10 8.28 8.40 -4 10 2.20 2.40 10 1.16 10"Z * fl smaller than fo Values of fi as a function of r are given in tensity data were calibrated to yield absolute Table II. Beyond a certain value of r, fj as pressures by using the torsion-effusion tech given by the equation becomes negative. nique to measure the absolute vapor pressure Physically, this means that beyond a certain of the monomer. These data are illustrated radius, fi remains constant and assumes a in Fig. 1. value close to f. o- Linear least squares fits of the absolute pressure data yield the equations. Accepted for Monatshifte fur aiemie. 1. R. Defay, I. Prigogine, and A. Bellemans, (714-1015°K)logloPmonomt.r = Surface Tension and Adsorption (translated by D. H. Everett) (Wiley, New York, 1966). Chap. ,(1.210tO.Q 5)X 10* . IS. 0 6<36t Q 06> 2. M. S. Plesset, J. Appl. Mech. 16, 277 (1949). — (824-1003°K)log P - 3. S. J. D. van Stralen, Int. J. Heat Mass 10 din6r Transfer 11, 1467, 1491 (1968). 4 4. A. W.~Searcy, IMRD Annual Report for 1970, - q-5siiQ.oi9)x io + gzst 0 21f UCRL-20S00, p. 147. 5. F. B. Kenrick, C. S. Gilbert, and K. L. Wismer, J. Phys. Chem. 28, 1297 (1924). where the reported errors are the standard deviations from the means. The enthalpy of sublimation at 298°K for 6. THE VAPOR PRESSURES OF GALLIUM TRIFLUORIDE GaFjf -j was calculated by both the second and MONOMER AND DIMBR thircFlsw methods. Since values for many of the required thermodynamic functions and IWid H. Feather, Alfred Bucliler, and Alan W. molecular constants are unknown, it was neces Searcy sary to use estimated quantities. The result ing second and third law values for the Apparently no measurements of the vapor enthalpy of sublimation at 298°K with standard pressure of gallium trifluoride as a function deviations are 58.85* 0.?6 kcal/mol and 58.27+ of temperature have been reported. Using a 0.17 kcal/mol, respectively. mass spectrometer, we neasuied the temperature dependences of the ion intensities attribut Since calibration runs with the mass spec able to the monomer and dimer in equilibrium trometer using lead as a standard showed with solid GaF3- The mass spectrometer in- systematic errors in temperatures and tempera- 130 higher concentrations relative to the monomers than arc diners r.ith high dissociation energies. T~! L. 1.. Quill, "ditor, The Gicmistry and Metallurgy of Miscellaneous Materials: Thermodynamics (McGraw-Hill, New York 1950J pp. 258-260. 7. KbSLAKCH HANS TOR CAJ.LNDAR VLAR 1972 Alan IV. Searcy Iheoretical work on the thermodynamics of solutions and of surfaces and interlaces will he* continued. A theoretical analysis of the net magnetic moment to be expected for gaseous molecules with coupling intermediate between llund's cases B and C has apparently never been com pleted. This analysis is required for inter pretation of the experimental data that we lig. 1. Vapor pressure of gallium triiluoridc. have obtained for SC2 and for related mole (.\:U.719-7241) cules. Progress has been made on a line of analysis that we hope to complete during the coming year. turc dependence to be low, greater confiJence can be placed in the second law calculat ons From examination of the fragmentation tlian in the third law values, which are nore patterns of vapor molecules foi-med from two strongly dependent on the estimated data. different metals and a halogen we have con These considerations lead to values for tie cluded that ion peaks of monomer fragments enthalpy and entropy of sublimation at 298 K may mark the principal peaks of dimers of of 58.8 * 3 Kcal/mol and 44.(• t 3 eu/mol, metal halide vapors when the dimers are pre respectively. sent at low concentrations. As a conse quence, partial pressures of the dimers which A linear least souares fit of the dimer are calculated by conventional analysis of ciata yields second law values for the enthalpy mass spectrometer intensity data may be too and er.tropy of sublimation at 914°K of 70.98 • low. We plan to investigate the fragmenta 0.88 -ical/mol and 42.76 • 0.98 eu/mol, respec tion pattern for A^Clg, the main species in lively. Die estimated error limit on the aluminum chloride vapor. Since fragmentation enthalpy is = 4 kcal/mol, which yields final patterns for chemically similar molecules are values for the enthalpy and entropy of sub qualitatively similar, the results should in limation per mole of dimer at 914°C of 71.0 * dicate whether or not fragmentation to ions 4 kcal and 47.8 * 4 eu. Combining these re masked by the monomer peals is also likely to sults with the temperature-corrected enthalpy be ertensive for metal halide dimer molecules and entropy of sublimation of the pionomer present at low concentrations. gives values for Che enthalpy and entropy of dimerization at 914°K of -38.5 kcal and -32.2 eu. This value for the entropy is in The extent of deflection by an inhonio- good agreement with the reported entropies of geneous magnetic field of zinc atoms and oxygen molecules that vaporize from a surface dimerization at 900°K for the AlFj, AICI3, z nc AlBrj, AII3, and FeCl3 svstems which are of a single crystal °f i oxide will be com -36.3, -32.4, -34.2, -29.4, and -31.6 eu/mol pared with the deflection of zinc atoms and of dimer, respectively. oxygen molecules of the equilibrium vapor from zinc oxide. Analysis of the kinetics of vaporization of zinc oxide suggests that the The data from Tliis and other similar oxygen iM.lecu.les that leave the surface may be studies were examined in terms of correlations predominantly in a --A excited state rather of Brewer.1 The analysis explains the than in the •>£ ground state and that the zinc apparent paradox that dimcrs which have low atoms may be in the first excited state. In dissociation energies to monomers are found at either case a difference in deflection from that shown Dy the equilibrium vapor species 8. 1971 PUBLICATIONS should be detectable. Alan W. Searcy and Associates We will investigate the effect on the rates of calcite decomposition of pressures of car Journals bon dioxide gas intermediate between the equilibrium pressure and the lower effective 1. II. B, Skinner and A. W. Searcy, Demonstra pressure produced at the surface by decomposi tion of the Existence of La2F6 Gas and Deter tion of calcite crystals in vacuum. The car mination of Its Stability, J. Phys. Chem. 75, bon dioxide may catalyze formation of the 108 (1971). ~ stable form of calcium oxide instead of the metastable form that we have found to be 2. i). R. McKenzie, A. W. Searcy, J. B. Holt, produced in vacuum, but we hope that it may be and R. H. Condit, Oxygen Grain-Boundary Diffu possible to measure the equilibrium dissocia sion in MgO, J. Am. Coram. Soc. 54 (4), 188- tion pressure of calcite relative to the 190 (1971). ~ metastable oxide. If so, we can determine the stability of the mctastable oxide relative to 3. D. II. Feather and A. W. Searcy, The the stable form and also determine the activa Stabilities of Gaseous Sodium Chloride Trimers tion entropy and enthalpy for the metastable and Tetramers, High Temp. Sci. 3_ 155-162 (1971) reaction. •I. R. B. Leonard and A. W. Searcy, The Varia We will undertake exploratory experiments tion of Vaporization Rates with Orientation with barite (BaSCh) and perhaps other sulfates for Basal Planes of Zinc Oxide and Cadmium available as single crystals in an effort to Sulfide, J. Appl. Phys. 42, 4047 (1971). f.iid a suitable subject for study of both the total vaporization rate as a function of tem 5. R. J. Salluzo and A. W. Searcy, Anisotropic perature and of the extent to which equilib Vaporization Rates of Opposite Faces of Zinc rium is maintained for the reaction 2S03(g) • Oxide Single Crystals, High Temp. Sci. 3_ (6), 2S02(g) + 02(g) during the non-equilibrium 491 (1971). process of vaporization in vacuum. At the anticipated decomposition temperatures, UCRL and LBL reports appreciable concentrations of oxygen and of both oxides of sulfur should be measurable 1. A. Biichler and A. W. Searcy, The Formation with our quadrupole mass spectrometer. of Vapor Phase Nuclei in Nucleate Boiling, UCRL-20S52, March 1971. (English version of A new apparatus will be used in investigat a paper to appear in German in Monatshifte ing the kinetic eneigy distributions of fur Chemie.) initially cold gas molecules after they have encountered a hot metal surface. Initial 2. R. Galluzzo, Vaporization Kinetics of Zinc studies will be with inert gases in order to Oxide Single Crystals (M.S. thesis), UCRL- establish the effectiveness of the apparatus, 20528, April 1971. but we intend as soon as possible to apply the apparatus to the study of the kinetic energy distribution among the unreacted portion of reactive gases and of gaseous reaction products. 152 B. MICROSTRUCTURE AND BEHAVIOR OF CERAMIC MATERIALS: GLASS AND CERAMIC-METAL SYSTEMS Joseph A. Paek, Principal Investigator Si02) specimens decomposed with the evolution of SiO and O2 at elevated temperatures and An objective of the research program is to low partial pressures of oxygen. Reactions contribute to the development of a funda also occur in mullite specimens between Si, mental understanding of the factors involved C, or SiC impurities and residual silica-rich in obtaining ceramic materials with controlled glass. The latter reactions were the cause character (which includes microstructure), for the observed formation of bubbles at the and of the relationship of character to interfaces of fused silica-mullite diffusion mechanical behavior at room and high tempera couples at about one atmosphere ambient tures. This objective involves studies on pressure and 1650 to 1800°C. the kinetics and mechanisms of solid state reactions which in general play a part in Abstracted from UCRL-20536, Aug. 1971. the development of microstructure. It also f involves studies on the mechanisms responsible Present address: Corning Glass Works, for the mechanical behavior of single crystals, Corning, New York. and on the application of such knowledge to the understanding of the behavior of poly- crystalline ceramic materials. 3. CAPILLARY PENETRATION OF SILICATE LIQUIDS INTO MgO COMPACTS A second objective of this program is concerned with structural, thermodynamic, and Boon Wong, Ilhan A. Aksay, and Joseph A. Pask electrochemical studies of glass, glass- metal, and ceramic-metal systems. It involves The objective of this investigation is to studies related to wetting, bonding, and the understand the fundamentals of capillary pene nature of the interfaces between dissimilar tration, into porous ceramic compacts, both in phases; to the thermodynamics and kinetics the presence and absence of reactions, from of chemical reactions at such interfaces; and theoretical and experimental viewpoints. to the kinetics and mechanisms of dissolution Equations for capillary penetration were and diffusion in glasses. developed based on Poiseuille's law for capillary flow and the concept of wetting of a solid by a liquid. Experimental studies 1. DIFFUSION KINETICS IN MULLITE were performed on the penetration of a number (3Al203-2Si02) AND ALUMINA-SILICA LIQUIDS of silicate liquids into sintered magnesia (MgO) compacts of about 951 theoretical densi Ilhan A. Aksay and Joseph A, Pask ty. Studies have continued on the diffusion (A) Calcium Magnesium Silicate (CMS) Liquid kinetics in mullite and alumina-silica liquids. The kinetics of Al3+ diffusion in mullite in the temperature range of 1650 to According to the phase diagram determined 1840°C are being studied through the growth by Osborn and Muan,1 monticellite (CaO-MgO- of the mullite phase at the sapphire single Si02> (CMS) liquid is at chemical equilibrium crystal-liquid (SiC^-Al^Oj) interface in with MgO at 1550oC after the precipitation semi-infinite diffusion couples. The glasses of a small amount of MgO. No chemical used are saturated with respect to mullite reactions thus occur between the CMS liquid in order to eliminate dissolution of mullite and the MgO solid during the penetration into the liquid glass. In addition, dif experiments. fusion couples of sapphire single crystal and fused silica are being used above the In such a system, the driving force for Melting temperature of mullite to study the wetting is indicated by the difference in the diffusion kinetics of the aluminum ion in solid/vapor and solid/liquid interfacial fused silica. The diffusion profiles are energies. The contact angle in a sessile being analyzed with the electron beam drop experiment then becomes a measure of the microprobe. driving force for wetting. The contact angle for this system was the largest and the rate of penetration was the slowest. A 2. DECOMPOSITION OF MULLITE* logarithmic plot of penetration distance with heating time at 1550 C indicates an initial Robert F. Davis, Ilhan A. Aksay, and parabolic relationship (slope » 1/2). The Joseph A. Pask negative deviations from a slope of 1/2 at later stages is considered to be caused by Surfaces of 2:1 type mullite (2A1203- a back pressure due to the entrapment of 153 gases inside the capillaries in the compact. 5. MICROSTRUCTURE STUDY OF POLYCRYSTALLINE SPINEL DEVELOPED FROM POWDERS PREPARED BY THE FREEZE-DRYING TECHNIQUE (B_) CMS Liquid with Additions of Sesi- quioxide (R2O3) Chi-Shine C. Tao and Joseph A. Pask Kith the addition of AI2O3, CrjC^, and Freeze drying is the removal of water from Fe203 to the CMS liquid, the chemical the specimen while the water is in the frozen equilibrium of the liquid with MgO was dis state. Processing of oxide materials by the turbed and reactions took place between freeze-drying technique requires four them during the penetration experiments. The operations: mixing, quick freezing, sublima rates and the nature of the reaction affected tion, and reaction on heating. Baker reagent the contact angles and the penetration grade MgS04-71120 and A^CSO^-UHjO were used behavior. The behavior of CMS liquid with as starting materials for the preparation of II AI7O3 indicated a similar behavior to CMS spinel, MgAljO^, powder. Freeze drying liquid. The penetration rates of CMS liquids offers the (Wantages of attaining intimate with 101 AI2O3 and 151 Al20r followed a mixing of the two starting materials, and parabolic relationship, indicating that the submicron crystalline powders after calcina penetration enhancing factors were about tion. Ultrafine powders normally have high identical to the penetration retarding sintcrability and thus produce fine grain, factors. The penetration rate of liquids dense compacts at lower firing temperatures. with both St Fe2P3 and II CT2O3 showed a positive deviation from the parabolic rela The effects of different freeze-drying tionship which indicated a strong contribution temperatures, different aqueous solution to the driving force for wetting by the reac concentrations, and different calcining tion. Sessile drop experiments with liquids temperatures were studied on the basis of the containing 101 and ISt O2O3 showed decomposition reactions of the freeze-dried spreading. For these systems strong positive powders. The powders were characterized by deviation and complete penetration were x-ray diffraction, thermal differential, observed in 5 minutes at 1S50°C. No penetra thermogravimetric, size distribution, and tion was observed for liquids containing 451 scanning electron-microscopic analyses. and 551 F«2°3 due t0 tne raPid formation of Measurement of the half width and integrated spinel (MgFe^) at the solid/liquid interface. intensity of the x-ray diffraction peaks provided information on the crystallite growth and completion of the decomposition reactions. The calcined spinel powders will be sintered. 1. E. F. Osborn and A. Muan. Figure 598 The microstructure will be correlated with in Phase Diagrams for Ceramists, E. M. Levin, the crystallinity and crystallite size of C. R. kobbins, and H. F. McMurdie, Am. Ceram. the powders. Soc., Columbus ,0hio, 1964. 6. LIQUID PHASE SINTERING STUDIES 4. PERMEATION 0? SILICATES IN MAGNESIA AND FORSTERITE COMPACTS* Carl E. Hoge and Joseph A. Pask Abbaraju P. Raju,+ Ilhan A. Aksay, and Liquid phase sintering is an important Joseph A. Pask industrial process for achieving highly dense materials in relatively short times. The CaOCK20)-MgO-Si02 liquids permeated into presence of liquid slags in contact with magnesia (MgO) and forsterite (Mg2Si04) refractory linings in metal furnaces is the compacts when open channels were available primary cause of corrosion of the lining and at the liquid-solid interface. Solid- its ultimate failure. Understanding of the solution formation at the liquid-forsterite interaction of high temperature liquids and interface resulted in the closing off of the refractory oxides is vital in the development open channels. The sessile drop and permea of sintered bodies and the prevention of tion experiments provided information that corrosion. could be used in the interpretation of microstructure development. Basic studies have been started to deter mine the effects of various "sintering parameters" on densification rates and Abstracted from LBL-159, Nov. 1971. resultant microstructures. Particle size and Present address: Zenith Radio Corp., shape, amount of liquid phase, unfired Microcircuit Facility, Elk Grove, Illinois. density, solid-liquid interfacial energy, and 154 pressure are among the parameters being mechanical behavior was exhibited by those investigated. The system chosen for initial specimens having the most uniform distribu study is monticellite-MgO. Others will be tion of boundary phase. selected during the course of the investiga tion. Results of creep tests for specimens containing 5$ of the silicate phase uniformly distributed indicate that at 1200°C deformation is controlled by dislocation 7. EFFECT OF UF ON CREEP OF MgO* processes. The stress exponent n was found to be 3.6 at this temperature, and an Patrick E. Hart"1, and Joseph A. Pask apparent activation energy for creep of 104 kcal/mol was obtained, using the Compressive creep behavior of pore-free incremental temperature technique. translucent polycrystalline MgO specimens1 containing * 0.25 wt.l Li, whose density on At 1300°C creep was founJ to occur as a the basis of single-crystal MgO was * 99.44 result of two deformation mechanisms. At of theoretical, was studied. The Li was higher stresses creep was controlled by dis believed to occur in a fluoride-rich phase location processes, and a stress exponent along the grain boundaries. The strain, at of 3.8 was noted. At lower stresses creep 750 psi and a constant: heating rate of was controlled by viscous flow of the 250°C/hour, first deviated from the thermal boundary phase, and a value for n of 1.0 expansion curve at * 630°C, and a large was determined, llie apparent activation increase in the strain rate occurred at energy for creep at 1300°C was 130 kcal/mol. * 830°C. The increase in activation energy with temperature was consistent with expectations that such an increase should be observed Abstracted from J. Amer. Ceram. Soc. 54_ (6), where two mechanisms affecting deformation 315-316 (1971). operate simultaneously. Present address: Battelle-Northwest, Richland, Washington 99352. At 1400 and 1450°C creep was controlled only by a viscous-flow mechanism. The stress 1. P. E. Hart, R. B. Atkin, and J. A. Pask, exponents were found to be 1.3 and 1.1 at Densification Mechanisms in Hot-Pressing of these temperatures, respectively. The Magnesia with a Fugitive Liquid, J. Amer. apparent activation energy for creep was Ceram. Soc. 53 [2], 83-86 (1970). j i—i— i 1 1—i—i i i i 17 8. HIGH TEMPERATURE MECHANICAL BEHAVIOR OF MgO-CaO-MgO-Si02 MATERIALS William E. Snowden and Joseph A. Pask An investigation of the high temperature mechanical behavior of materials containing S to 154 of a silicate phase in MgO was completed. Specimens were fabricated by hot pressing or hot pressing and annealing techniques. The silicate phase was distributed along MgO grain boundaries and consisted of crystalline CaO-MgO-Si02 (Monticellite) and small amounts of glass. The behavior of these specimens in stress- strain tests was studied over the temperature range 1Z00-1400°C. The degree of uniformity of distribution of the boundary phase was i i i i i ' 5 - found to be most important in affecting the 2xlO" « 6 » IO * , * < « 8 |0"5 high temperature mechanical behavior, particularly at 1200°C where specimen Fig. 1. Strain rate vs. normalized stress deformation was controlled primarily by for creep specimens of MgO containing Si plastic deformation within the MgO framework. Ca0-MgO-Si02 tested in the temperature range At all temperatures the most favorable 1200-1450°C. (XBL 714-6692) 155 greater than 200 kcal/mol. The relation between strain-rate and stress for the creep specimens is illustrated in Fig. 1. The slope of each line corre sponds to the stress exponent n. From the figure the decrease in n with temperature is evident, and the two distinct values obtained at 1300°C for two deformation mecha nisms are indicated. The effects of these two mechanisms on final microstructural appearance are shown in Fig. 2. At 1200°C, where dislocation processes were found to contribute to creep, there is no apparent increase of the thickness of the boundary phase. At 1300 and 1400°C, however, the effects of boundary separation by viscous flow are quite pronounced. 9. MECHANICAL BEHAVIOR OF MgO SOLID SOLUTIONS C. Dokko and Joseph A. Pask In general, ceramic materials are known to be brittle, particularly at low tempera tures. For polycrystalline MgO this is due to the failure to activate the secondary slip systems {001} Previously1 a nearly fourfold increase in the room-temperature compressive yield stress on {110} slip planes has been reported to be possible by alloying NiO with MgO. However, yield stresses on {100} slip planes have not been determined. Thus, in this program yield stresses on both slip systems in single crystals will be determined, as a function of NiO content and temperature. The use of an inert atmosphere will be necessary for high-temperature deformation since a softening effect due to reduction of NiO has been reported^ to occur at 1200°C. In conjunction with the single-crystal study, Fig. 2. Photomicrographs of sections normal alloyed polyci/stalline specimens will also to the loading axis from specimens of MgO be investigated. Specimens will be prepared containing 5* Ca0-Mg0-Si02 after creep tests: by use of hot-pressing and sintering techni [top) ~ 10* strain at 1200°C, (middle) ~ 12* ques from MgO and NiO powders. Similar strain at 1300°C, (bottom) ~ 1U* strain at procedures are expected to be followed with 1400°C. (XBB 715-2117) FeC and MnO additions. 156 In addition to stress-strain tests, creep 090 tests will be performed in order to determine 0 43 the stress component from the steady-state creep rate and to attempt to determine its 040 relationship to the grain boundary strength. r 039 It is expected that higher grain boundary o strengths will be reflected in a stronger 5 030 £ 0Z9 stress dependency because of the activation o of the secondary slip systems resulting in i 020 increasing dislocation interaction. Micro- g • TS:H »l ti*tX structural observations will be made with < 0*3 emphasis on the grain boundary morphologies. on 009 1. T. S. Liu, R. J. Stokes, and C. H. Li, J. Amer. Cer. Soc. 47, 276-279 (1964). Fig. 1. Diffusion profiles for NIO-CaO 2. R. B. Day and R. J. Stokes, J. Amer. Cer. couple with (100) faces in contact after Soc. SO, 44S-448 (1967). 173 hours at 1346°C. (XBL 686-921) [Ni] £ 0.41. The higher diffusivity and 10. FRACTURE OF BRITTLE TRANSPARENT MATERIALS correspondingly greateT flux in the NiO-rich phase resulted in reaction at the interface William E. Snowden and Joseph A. Pask and movement of the crystallographic boundary with the growth of the NiO at the expense of An investigation of crack propagation in the other crystal. The crystallographic brittle transparent materials, primarily orientation of the volume swept out trans glass, subjected to iipact has been started. formed to that of the growing crystal. The To permit study of the effects of stress NiO-CaO diffusion couples also provided wave interaction with stationary and moving information for the corresponding phase cracks, a lens Schlieren system was set up. diagram. Following impact using a magnetic hammer capable of producing impact velocities Abstracted from J. Amer. Ceram. Soc. 54 (3), approaching 1 mm/sec and sharp pulses of 1S2-1S8 (1971). very short duration, the fracture sequence will be photographed using a high speed fram Present address: Systems Group of TRW Inc. ing camera. Redondo Beach, California 90278. To date the work has largely been directed towards determining the limitations and capabilities of the system. A number of problems have been encountered, primarily 12. THE EFFECT OF ATMOSPHERE ON INTERFACIAL related to synchronization. For single- REACTIONS AND WETTING BEHAVIOR OF GLASS-METAL frame tests, however, the effect of a stress SYSTEMS wave intersecting a long crack has been observed. A principal objective of the Carl E. Hoge and Joseph A. Pask study will be to determine the conditions under which branching occurs when a material Interfacial reactions and wetting behavior is subjected to transient stresses. of Na2-2xpexSi20s glasses on substrates of high purity iron (Marz) and impure iron (Armco) were studied using the sessile drop configuration and the electron microprobe. 11. INTERDIFFUSION AND MOVING BOUNDARIES IN Experiments were performed at 1000°C in high NiO-CaO AND NiO-MgO SINGLE-CRYSTAL COUPLES* vacuum and at 1 atm pressure of argon and a 10 to 1 mixture of C0/C02. All ambient Marvin Appel'1" and Joseph A. Pask atmospheres had low partial pressures of O2 and Nil. For single-crystal NiO-CaO diffusion anneals in air at 1346 to 1S27°C, the chemi Decomposition of FeO and Na20 in the cal interdiffusion coefficient was found to glasses was observed at the liquid-vapor increase exponentially with increasing Ni interface. The oxidation of the metallic iron concentration (Fig. 1 shows an example of substrate by Na+ in the glass occurred at the diffusion profiles); for NiO-MgO at 1342 to solid-liquid interface under high vacuum 1346°C, a similar relation was found for conditions, but not in argon or CO/CO2. 157 An adsorbed oxide film is normally present will provide additional information on the on the surface of the iron substrates. This liquidus. Activity data will also be of film, as well as any impurities in the bulk value in quantitative evaluations of glass of the iron, affects the specific surface free reactions with metallic iron. energy slv of the substrate. The wetting behavior of the glasses and the reactions at the interface are mainly determined by 1. P. T. Carter and M. Ibrahim, J. Soc. changes in sfv. Mien the oxide film is Glass Tech. 36 [170], 142-63T (1952). desorbed, sfv increases, spreading by the li 2. A. M. Lacy and J. A. Pask, J. Amcr. Ceram. quid on the solid and an associated redox Soc. S3 (12), 676-679 (1970). reaction at the interface occur. This occurs readily with high-purity iron (Mar?.), but with impure iron (Armco) sfy is Maintained at a 15. RESEARCH PLANS FOR CALENDAR YEAR 1972 low value for an extended period of time due to its impurities Joseph A. Pask a. Solid State Reactions 13. ELECTROCHEMICAL STUDY OF THE CoO-Na2Si20s Diffusion experiments are continuing with GLASS SYSTEM* semi-infinite couples of fused silica and sapphire in the temperature range of - 1800 to Alton M. Lacy+ and Joseph A. Pask 2GC0*C to determine the dissolution and diffusion kinetics of aluminum ion in fused A solid electrolyte electrochemical cell of silica. The kinetics of diffusion of Al3* the typo Pt|Ni:Ni0a-l||ZrO2 • 7.S» CaO|| and the solid solution range in mullitc arc Co:CoOa Na20-Si02 CLASS SYSTEM high velocity impacts has been initiated. Of particular interest will be a determination Robert B. Langston and Joseph A. Pask of the conditions under which branching occurs when a material is subjected to transient The available data on the FeO-Na^-SK^ stresses. Materials under study arc glasses system has been reviewed.* Determination of and sapphire. activities of FeO in the Fen.9sO-Na2Si20s system have been made.2 Additional measure Stress-strain and creep data at elevated ments of FeO activities in the liquid regions temperatures will be obtained en single at 700 to 1100'C will be made. These data crystals and polycrystalline specimens of 158 MgO doped with varying amounts of NiO. 7. J. A. Pask, Ceramic Materials, in Mechanisms and kinetics will be determined Technology Forecast for 1980. edited by and correlated with the microstructure of E. Weber, G. K. Teal, and A. G. Schillinger the specimens. (Van Nostrand Reinhold Co., New York, 1971), pp. 67-78. d. Glass Studies 8. R. F. Davis and J. A. Pask, Mullite, Determinations of activities of oxides in Chap. 3 in Hi^h Temperature Oxides—Part IV sodium silicate liquid glasses by electro Refractory Classes, Glass-Cermics, and chemical methods and calculations of thermo Ceramics, edited by Allen M. Alper (Academic dynamic properties arc being continued. Press, New York, 1971), pp. 37-76. Studies on mechanisms and kinetics of dissolu tion studies of oxides in silicate glasses arc 9. T. G. Langdon, 0. R. Cropper, and J. A. also being continued. Pask, Creep Mechanisms in Ceramic Materials at Elovatcd Temporaturos, in Ceramics in c. Ceramic-Metal Systems Severe Environments Materials Science Research, Vol._5, cTSitcd by W. Wurth Kriegol and llnyne Studies will bo continued on the effect of Polmour III (Plenum Publishing Corp., New York, atmosphere and ceramic and metal compositions 1971), pp. 297-313. on electrochemical reactions in ceramic-metal systems. These studies will be related to 10. T. G. Lan&don and J. A. Pask, Factors the conditions under viHch a chemical bond Influencing the Stress-Strain Behavior of occurs at the interface. Ceramic Materials, ibid., pp. 283-296. 16. 1971 PUBLICATIONS AND REPORTS UCRL and LBL reports Joseph A. Pask and Associates 1. D. R. Cropper, Creep of Lithium Fluoride Journals and books Single Crystals at Elevated Temperatures (Ph.D thesis), UCRL-203S0, Jan. 1971. 1. M. Appel and J. A. Pask, Interdiffusion ant*. Moving Boundaries in N'iO-CoO and NiO-MgO 2. R. F. Davis and J. A. Pask, Diffusion and Single Crystal Couples, J. Am. Coram. Soc. Reaction Studies in the Alumina-Silica S4 (3), 1S2-1S8 (1971). System, UCRL-19644 Rev., June 1971. Z. A. M. Lacy and J. A. Pask, Electrochemical 3. William E. Snowdcn, High Temperature Studies in Glass: III, The System CoO- Mechanical Behavior of HBO-CaO-MgO-SiO-, Materials (M.S. thesis), UCRL-20S67, July 1971. Na2SiiOi;, J- An. Ccram. Soc. 54 (S), 256-259 (1571 J.s - 4. C. E. Itoge, The Effect of Atmosphere on 5. T. G. Langdon and J. A. Pask, Effect of Interfacial Reactions and Wetting Behavior Microstructure on Deformation of Polycrystal- of Glass-Metal Systoms (M. S. thesis), UCRL- line Magnesium Oxide, J. Am. Ceran. Soc. 20334, Jan. 1971. 54 (5), 240-246 (1971). 5. R. F. Davis, I. A. Aksav, and J. A. Pask, 4. J. A. Pask, Nature, Status, and Selection Decomposition of Mullite, UCRL-20536, April of Ceramic Materials in The Characterization 1971. (J. Am. Ceram. Soc, in press.) and Application of Materials, Lecture Scries !*io. 51, AGARD(Aavisory 'Jroup fo:- Aerospace 6. J. A. Pask, Chemical Reactions and Research and Development), North Atlantic Adherence at Glass-Metal Interfaces, A. I. Treaty Organization, 7 p., May, 1971. (UCRL- Andrews Memorial Lecture given at the 20S20, Jan. 1971.) Porcelain Enamel Institute Technical Forum, Oct. 6-8, 1971 at Ohio State University, 5. P. E. Hart and J. A. Pask, Effect of LiF Columbus, Ohio; I.H1.-166, Aug. 1971. on Creep of MgO, J. Am. Ccrran. Soc. 54 (6), 515-316 (1971). (UCRL-19699, Jan. 1571.) 7. A. P. Raju, I. A. Aksay, and J. A. Pask, Permeation of Silicates in Magnesia and 6. J. A. Pask, Structural Ceramics and Forstcrito Compacts, LBL-IS9, Nov. 1971. Design, edited by S. J. Acquiviva and S. A7 Bom (Gordon and Breach, Sew York, 1969); reviewed for Materials Science and Engineering 7 (6), 562-363 (1971). 159 C. RELATION OF MICROSTRUCTURE TO PROPERTIES IN CERAMICS Richard M. Fulrath, Principal Investigator density measurements of the powdered samples, using a helium gas pyenometer. The presence 1. RELATION OF MICROSTRUCTURE TO PROPERTIES of random point defects in the crystallites IN CERAMICS would be expected to lower the bulk density. However, the density was essentially constant Richard M. Fulrath for defect concentrations up to It. Greater defect concentrations produced by doping The primary objective of this research with both Nb and Bi did show the expected program is to quantitatively relate the decrease in density. These results indicated mechanical and physical properties of ceramic that defect concentrations of It or below fin materials to the processing procedures and the micron size particles used) can bo accom microstructure developed in processing. The modated by surface irregularities. However, program encompasses four principal research greater defect concentrations appear as volurre efforts: (1) the processing and properties defects. of ferroelectric, piezoelectric, and ferro magnetic ceramics, (2) the permeation and Dense disks were fabricated of the doped diffusion of gases through and the solubility series for measurement of ferroelectric of gases in ceramics, (3) the mechanical hystoresis. The results, in agreement with properties of brittle matrix composite provious workers in this laboratory, indi systems, and C4) basic studies in sintering cated greatly enhanced remnant polarization ce. amic materials using high temperature for the materials doped with Nb or Bi. scanning electron microscopy. However, the presence of point defects at crystallite surfaces indicates that the This program places emphasis on the defects themselves played no significant role science of processing and the microstructure in polarization reversal. Instead, the developed in the processing of ceramics. effect of the dopants on ferroelectric hysteresis must be assigned to their known effect on the microstructure, viz., increased densification and decreased grain growth. 2. NOX-STOICHICMETRY OF LEAD ZIRCONATE TITANATE Dr. Shackelford was associated with this * program and supported by an IBM Post doctoral James F. Shackelford Research Fellowship. Present address: McMaster University, Hamilton, Ontario, Variations in the stoichiometry of both Canada doped and undoped lead zirconate titanate (PZT) were studied. The basic composition was Pb(ZrxTii-x)03, where x was 0.55, for the series of doped materials and x was 0.60 3. INTRINSIC NON-STOICHIOMCTRY IN SINGLE- for the undoped series. The dopants were Nb PHASE LEAD ZIRCONATE-TITANATE* and Bi. PZT compositions in this range have the rhombohedral structure below the Curie Robert L. Holman and Richard H. Fulrath temperature. The attainment of a single-phase lead Variations in stoichiometry were assumed zirconate-titanate with consistent dielectric to be accomodated by point defects. For properties requires a reproducible control the undoped material, the vapor pressure of of stoichioroetry. Webster et .1.1 have the volatile vapor species, PbO, was varied, observed that the degree of PbO deficiency giving rise to a variation in the PbO: effect the piezoelectric properties. And (Zr, Ti)Oj ratio in the solid. For example, more recently, Atkin2 observed a stoichiometry decreasing the ppjjQ below the value for the effect on the sinterability of PbTi.sZr^ stoichiometric material produced volatiliza which suggested the existence of a rather tion of PbO from the solid, leaving Pb and 0 wide single-phase region relative to the PbO vacancies. For the materials doped with Nb content. To date, phase equilibria investi and 3i, Pb vacancies were assumed to be gations in the lead zirconate-titanate system formed to accomodate charge imbalances. have not determined the width of the Separate weight change experiments have PbCTixZri-x)Oj single-phase regions. verified these defect models. It has been shown3 that one can control Further information about the distribution the extrinsic stoichiometry of lead zirconate- of these point defects was obtained from titanate by controlling the lead oxide vapor 160 (o) CRUCIBLE A SAMPLE AT C PftOK SIMPLE EQUIUMATEO 0 owes o -» o • o \i *ITH c D A SAME COMPOSITION °no AND SAMPLE AT 0 AS SIMPLE \ i 1 *P» »PK> ° Pt COIL' SAMPLE HOLDER SAMPLE AT 0 GIVES 0D-»>Oc*O» "no v: AHO SAMPLE AT E \ XP.0 ° ' Xp.0 WHERE WEIGHT CHANGE O-E (AW) • SINGLE PHASE WIDTH Fig. 3. Activity (apbo)-composition (Xpbo) diagrams illustrating the two-stop experiment TO* WEIGHT - CHANGE CELL to determine tho width of any Pb(TixZrl-x)°3 Fig. 1. PGA weight change crucible and sam single phase region. (XBL 7010-6822) ple. (XBL 7010-6822) SAMPLE ATC ' SAMPLE AT D [ .LIQUO LIOUIO+PZT PZT PZT+ MOLE FRACTION PbO (Xn0) Fig. 2. Activity of lead oxide (apbo) vs. mole fraction of lead oxide (XW)) at constant temperature (1100°C) for any PbO-(Zr,Ti)02 binary. Points A and B refer to the constant lead oxide atmospheres 200- provided by crucibles of these compositions. Point C locates the composition and activity of stoichiometric PZT, while D and E indicate Fig. 4. Weight change vs. time at constant the single phase extrema. (XBL 718-7089) temperature (1100°C). These three experi mental curves are placed together to show the determination of the single phase width (AW) pressure. This technique may be extended and the reproducibility of the experiment. (XBL 718-7091) to determine the width or any Pb(TixZri-x)03 single-phase region, utilizing thermo- gravimetric analysis and atmosphere control. Thermodynamically, when a sivie-phase PZT PbO, until an equilibrium activity balance sample has a different lead oxide activity it attained. The santple's composition will than that provided by the atmosphere, a then remain fixed for as long as the reaction will occur until the sample and the atmosphere provides the constant lead oxide atmosphere have the same PbO activity. If activity. the sample is small, it will alter its stoichionetry, by either gaining OT losing Specifically, a single-phase, porous PZT 161 sample is alternately "equilibrated" with I. IW.VlSSi.v; ,i- SonrUM-POTASSIUM N10BATE the PbO atmosphere established by PZT LlKAMICS* crucibles (Fig. 1) fabricated of the two- phase compositions that provide different Bob R. Powell" constant lead oxide activities. Thn general form of the PbO activity-composition diagram The study of the processing of sodium- at a constant temperature fllOO°C) is potassium niobate (hence forth NKN) cer;i»iics illustrated in Fig. 2. The experiment is started by Stannek in 1969 was continued and diagramed in rig. 3. A sample, initially at included a study of the reactions occurring C, changes its composition to D at the during calcination to form pure sodium- phase boundary, when reacted with a crucible potassium niobate, as well as the sintering of composition A. This sample of composition phenomenon of NKN. D will then alter its composition to F. when reacting with a second crucible B. The com The calcining process is based on the position changes of the sample are allowed reaction: by the vapor-phase "equilibration" of the sample with the atmosphere. The associated 1/2 Na N0 * 1/2 KNO, + 1/2 M^Oj + weight-changes are monitored thcrmogravimct- 3 ? rically. For example, the weight change Closs) from I) to E exactly corresponds to W m the mol * PbO width of the PZT single-phase tNa0.SK0.5) NM3 + 2 * °2 • region. Representative data are shown in Fig. 4. Reversibility is demonstrated by The actual reaction process as determined replacing the sample at E in crucible A and using thermogravimetric analysis, differential returning to composition D, with the same thermal analysis, a gradient firing experi sample weight change. ment and x-ray analysis of the mixed oxides that were calcined at different tempt-ratures These results indicate that two-phase lead is as follows: zirconate-titanate crucibles may be success fully synthesized to maintain constant 1) The nitrates melt and form a large "equilibrium" PbO atmospheres. The crucibles volume of liquid. allow for the continuous weighing of speci 2) At approximately 550°C a low alkali mens during their reaction with the pre compound, (NalQNb^j forms with the selected constant crucible atmosphere at excess (NaK)O intermingled with ;t. temperature. 3) The alkali oxide begins to vaporize an the final reaction temperature is A width of the single-phase region of reached so that NKN forms from the 2.48 JKJI I PbO at 1100°C was determined for outside of the calcining compact and the PbTi.5Zr.5O3 composition. Only minor progresses inward. If enough ;^aK)0 temperature dependence on the overall width is lost some (NaK)Nbg02i remains at of the region was detected. The width the center of the compact. increased by .08 mol i upon raising the tem This second phase, (NaKJNbgOjj was observed perature 50°C to 1150°C. A similar reduced consistently at the compact interior by x-ray width was noted upon lowering the tempera diffraction. ture. The sintering study which included doping Abstracted from R. L. Holman and R. M. pure NKN with 0, 1/2, 1, 2, and 4 ate*ic Fulrath, Intrinsic Non-stoichiometry in percent BaO or 2 mol percent B2O3 led to the Single-Phase PbTi.5Zr.5O3, LBL-147, Aug. 1971. conclusion that any densification of the pellets during firing is due not to a 1. A. H. Webster, T. B. Weston, and N. F. H. sintering mechanism but the particle rear Brict, Effect of PbO Deficiency on the rangement stage of sintering in the presence Piezoelectric Properties of Lead Zirconate- of a liquid phase. Titanate Ceramics, J. Am. Coram. Soc. 50 [9], 490-1 (1967]. In each sintering experiment the pellets 2. R. B. Atkin and R. M. Fulrath, Point densified for about 1 hour and then stopped Defects and Sintering of Lead Zirconate- or de-desified. This rapid initial densifi Titanate, J. Am. Ceram. Soc. 54 [5], 265-70 cation fits the model of sintering in the 0971). ~ presence of a liquid as proposed by Kingery. 3. R. B. Atkin, R. L. holman, and R. M. The 2 mol % B2O3 was added to create a Fulrath, Substitution of Bi and Nb Ions in larger volume fraction liquid and the sin Lead Zirconate-titanate, J. Am. Ceram. Soc. tering curves for B7O3 doped NKN are of the 54 [2], 113-115 (1971). same form as undoped NKN, initial densifica tion to about 80* of theoretical density, at which point dens iiH-i' i- ; : :•'-.''.• : :• - .-.. ,••• .•;, t ht nie- •• f I01' sec"1 reversed -1 i ghf 1 .-. ,-.,:: •' !.'• . 0 1 -;•• ;• having the lower 1 reqi.eiu ••':•• aae < •: .•..-• 1'S"P.cd for Phe purpose f I'.a id.;:': hv Ir ,'vn 1. !':- ' a sill. 1, asing ricuoib grain growth, ilns was '''• '-! 1 iter.-.tni'- !;;' :. i!:e fuse 1 sjjK.a of this the sintering temperature t> -tudv !oi ii''- !ispl.e.- citeir-.a] solution of large additions f.~! jid ! moo a.'drogci'., po ;! 1 s doe t r. he presence of found to limi! the volume i R t ion f I itiui. -..at or-produccd hveroyv'-.. 'mp-irison of the to such an extent that m -,s : fk it i n •:i del with t is iitciaterc .lata showed a p'/2 occurred. pp'ssiee dependcnc1' (••:• the solubility findkat!\< al ii^.i, lationj and a binding The lack of sintering o: 'A'.', is attribute cut itjv of a1".HI ',"• k.al/wol c mdicat ivc of to the crystallography <>*' MN. \K\ has .1 hepji .; b: ndi.-n; . fee structure Kith M.'"' in .1 bod', centered position in the unit •• * il and ' - in the ft faces and Na* and f* at ' (.• -nurs I'his > \i.^! raet'-. <:•:• ' I.'TC• '. Shackelford, IT.e similar to Nai I whul is nored i'v. bavins' a <. lab: lily f ''e.es in 'lass fph. a. thesis), low sm face energ-. . hen. e th'ie 1-. 1 small IH ».'. .'ii'.-JQ, i in. lii" 1 . driving fi.rec for •• i ntt r lug prcs'-nt in \K\. '•resent addr>"-s: '\ 'last-: lln1\ 1 rsitv. It appears that it is toe small to cause Hani I ton, >ut ail", ' :. e!.. extensive sintering. Abstracted from Bob P. Powell, processing of Sodium-Potassium Nioba'e < oramics fV.S. <;I.AKS thesis), 1.BI.-41 s, Nov. IV" I . losep). Present address: M. s. \rmv. '* prot^rar to understand tlv. mechanisms of I he diffusion of ,;.ise- in ulass and (he effects 1 i v.are us parameters on this process 5. 'iifl: SOUIBM.ITY y> f.'isi - <\ ;i.\^* has betn undervav lor the past several years. Pie primare a;-a of interest has been lames V. Shack"! for.1 directed toward imderstaniline these processes in the rr)st basic clis-- composition, fused The statisti-ai thermodynamics of the gns- ••ill -a pure i';. il.'.s Miteria! is also in-glass system provided » basic model of the the most r.ract i ..' I rem the standpoint of solubility. 1-quations were derived whi.h overall materia! transported because- ^ascs gave the solubility (in atoms or molecules diffuse more easily through fused silica per cubic centimeter of glassj as a function than any other -omjvisition. of gas pressure, temperature, f'indamcptal constants, and material parameters, rhrec ihc d.'ffusivi'v of cases i'l glass has been nain categories of solubility were considered: studjed extensile. o\er the past 71) years (a) physical solubility of monatonuc gases, hut none of the experimentors has attempted (b) physical solubility of tvMvitoniic cases, to eharactcr;"e the materia! or explain the and (c) chemical solubility ->f polyatomic mechanisms. !i i.s a well-known fact that fused gascs. The results .'or fa) and .TO -ire silica mav contain from 10 to 1000 ppm essentially equivalent. The results for lei "water" iti the form of hydroxy! ions fOH), are dependent upon the spci'i.. system depend ini; on the methr ' of manufacture--that involved. is. those ::tade by meltinc quartr crystals electrically have low water contents whil" The model was compared with the experi- those iaade by -he oxidation of SiCIj iLsually ta?i3tal results f-r a variety -f systenis. possess unusually larce amounts of water. Case (a) was observed for he! urn and neon j.n Hie presence of o.l wt % •'•' hydroxyl in >;lass fused silica. \'ar:ations of the thcrrai has been shown to affect proiK-rties such as histor>' c*" the fused silica did net -•ppcar t viscositv bv as much a? tenfold. Also it have 3 riiifca5ura!)le effect on physical solubil >.e h'-en siiowii that the thermal history, ity as evidenced in the helium data. 1'ase licti.e temperature, of the glass has an effect ->n its properties. In the fabrication (b) has obser\-ed for hedjoccn in fused 0 silica. Piiysical solubility was measured -I' Jiff us ion membrane the thermal history by a codified Seiverts' technique :uid was and water content are simultaneously altered, characterised bv a linear dependence on dejx?in.!ins upon conditions of manufacture. pressure. Hindis energies for the physical With this in mind, it was decided to pursue ly dissolved, species were of the order ex a studv on the effect of these two param pected for va.; der IVaais r -;ndine,. Vif^ratiotwl eters on the diffusion of lie and \e in fused 163 silica with known water contents and thermal treatments. Previous work in this labora tory has isolated the effects of these variables on the density of fused silica; therefore, by subjecting a diffusion mem brane to a set of known conditions the speci men can be characterized. The diffusion membrane, Fig. 1, consists of fused silica tubing, 0.020 in. diameter with 0.003 in. wall thickness, wound in a 1 in. diameter helix with 38 turns/inch. One end of the helix is sealed leak tight and the other end is sealed to a support tube which eventually enters a mass spoctia- meter. The interior of the coil is kept at a vacuum of approximately 10'" Torr while the outside is subjected to one atmos phere of either He or No. The flow ot gas through tlie wall is recorded on a mass spectrometer as a function of time. By applying the proper boundary conditions to Pick's second law, 3C/3t - D 3zC/3x2 , a solution is obtained in the form of a Fourier series froffi which D, the diffusion coef ficient, can be described by applying the results obtained in the early stages of the gas flow. The diffusion coefficients are subjected to an Arrhenius analysis to obtain the jcrivation energies for this process, the c.tfusion membrane was subjected to various thermal treatments as prescribed by Shackelford et al.1 to obtain various densities. Initial work on the study of the permeability of gases in glass showed anomalous behavior in relation to permeability of gases in glasses other than fused silica. In these other glass compositions it has been shown that as the density of the glass increased the permeability decreased. Fig. 1. Fused silica specime.i geometry. Opposite behavior was reported for fused (BBC 6910-6449) silica. Since the diffusivity is proportional to the permeability, it was anticipated that the results would show similar behavior. The activation energies for diffusion as a func tion of the treatment are given below: Gas ri(diffusion).caI Sample treatment Density Temp. Range,°C He 5114 + 27.S 1000"C, 1 atm H20 2.201 -73 to 164 Ne 8897 + 292 1000°C, 1 atm H2O 2,201 196 to 367 He 4967 +27.0 1200°C, vacuum 2.204 -78 to 165 Ne 8360 + 276 1200°C, vacuum 2.204 194 to 343 The diffusion of He and Ne in fused silica this data is complete, fictive temperature show anomalous behavior; that is, as the effects will be able to be distinguished density increases the activation energy for from water concentration effects. diffusion decreases. Since thermal treat ment and water content have simultaneous Efforts during the past year were also effects on the density, experiments are concentrated on developing a nodel to describe currently underway to determine the activa the diffusion process. The need for a dif tion energies of low-water, 1000°C and a fusivity model is twofold. The solubility high-water-content, 1200°C samples. When of gases in glass has been described by 164 previous workers2 in this laboratory. The systems. A wide range of diamond grir sizes permeability • solubility x diffusivity. were used in preparation of the specimens, Thus by combining the two models the overall and the specimens were sawn under carefully permeation process can be described thermo- controlled conditions. It was found that dynanrically. With knowledge of several the strength of both glass and composites fundamental parameters one can predict the was idenpendent of the diamond grit size permeability of various gases in glasses. used in sawing. There is some controversy in the literature as to whether there is a temperature depend Abrasion of the diamond-sawn surface by ence of the pre-exponentiil term of the SiC grit, however, significantly reduced the Arrhenius equation for diffusion based on uniaxial bend strength. This strength empirical results. The model is developed reduction could be recovered by annealing the in terms of the activated complex theory specimens following the abrasion treatment. giving \3 -hv//2kT j -E AT _ kTd2 (ehv/ZfcT 6h (ehv t™ - e"hv /2kT Knowledge of the vibration frequencies in Examination of the specimens indicated the solution and activated states can lead the strength reduction by abrasion was the to the identification of the temperature result of edge chipping and not the introduc dependence of the pre-exponential term. The tion of more severe surface damage in the vibration frequency in the solution state has tensile surface. been previously2 calculated. Work is cur rently in progress in applying the proper Abstracted from S. Risbud, Influence of potential to the vibration in the excited Surface Damage on the Strength of a Brittle- state and determining the vibration frequency, Matrix Composite (M.S. thesis), LBL-137, assuming that the potential is parabolic. July 1971. Wien completed the diffusion mechanism will be defined explicitly. 1. J. F. Holzgraf, The Characterization of the Matrix of Brittle Matrix Composites, UCRL-20388, Dec. 1970. 1. J. F. Shackelford, J. S. Masaryk, and 2. D. P. H. Hasselman and R. M. Fulrath, R. M. Fulrath, J. Am. Ceram. Soc. 53, 417 Proposed Fracture Theory of a Dispersion (1970). Strengthened Glass Matrix, J. Am. Ceram. 2. P. L. Studt, J. F. Shackelford, and Soc. 49 [2], 68-72 (1966). R. M. Fulrath, J. Appl. Phys, 41_, 2777 (1970). 8. CHARACTERIZATION AND SINTERING BEHAVIOR 7. INFLUENCE OF SURFACE DAMAGE 0?.' THE OF BARIUM AND STRONTIUM FERRITES STRENGTH OF A BRITTLE MATRIX COMPOSITE11' James S. Reed* Subhash Risbud BaO is essentially insoluble in BaO-6Fe203 In a recent investigation of the strength and the commercial magnetic ceramic materials of brittle matrix composites Holzgraf1 found with nominal compositions of Ba0-5.5Fe203 that abrasion of a diamond-sawn surface would and SrO'5.5Fe203 consist of the 1:6 phase reduce the bend strength significantly. This and one or more minor phases. investigation was conducted to determine the cause of this strength reduction because Compacts formed by mechanical pressure previously developed theories for the filtration in a magnetic field consisted of strength of dispersion strengthened glass aggregates of platy particles with planer matrix composites2 postulated that such a faces and both planer and curved edges. strength reduction should not occur. During sintering, planer faces exhibited preferential growth and led to a reduction Vacuum hot-pressed glass and aluminum of edge-face-edge contacts and an increase oxide dispersion strengthened glass matrix in the degree of preferred orientation. composites were prepared. Bend strength Shrinkage anisotropy during sintering is specimens were diamond-sawn from these two preposed to be the result of vacancies 165 diffusing preferentially to face boundaries. 10. RESFARCH PLANS FOR CALENDAR YEAR 1972 After sinterinjr at 1150°C in one atmos Richard M. Fulrath phere O2 the nominal composition BaO-5.5Fe20j consists of BaO-6I-'e203 and BaO-Fe203. The a. Ferroelectric and Ferromagneti: Ceramics presence of the 1:1 phase markedly restricts grain growth and gives more uniform densifi- Studies in the lead titanate-lead zirconate cation characteristics when compared to systems will concentrate on determining the BaO-6Fe203. limits of the single phase composition width by both thernogravimetric and Knndsen SrO-5.5Fe203 sintered at 11S0°C in one effusion techniques. The parameters con atmosphere O2 consisted of SrO-6Fe2C>3 and trolling domain boundary mobility will also two secondary phases with approximate be studied. compositions of Fe203 and Sr0-Fe20j. Compared to BaOS.5Fe203 the nominal composition The processing and properties of lead SrO-S.SFe203 sinters more rapidly, exhibits titanate ceramics will continue to be a slower rate of grain growth, and shows a investigated with emphasis on the processing higher degree of differential firing of sub-micron powders. shrinkage. Lithium ferrite studies will continue After sintering in one atmosphere of 02 with emphasis on the defect structure contri at 12S0°C all systems exhibited exaggerated bution to processing and properties. grain growth. b. Permeation, Solubility, and Diffusion of Gas in Ceramics Professor Reed was associated with this research group during the 1970-71 academic He and Ne diffusion studies in fused Si02 year while on sabbatical leave from the will be completed. The permeation of He State University of New York College of through noncrystalline dense AI2O3 as a Ceramics at Alfred University, Alfred, function of grain size will be studied in New York. order to develop more detailed understanding of grain boundary structure. c. Sintering During the first half of 1972, studies on 9. HIGH TEMPERATURE SCANNING ELECTRON the initial and intermediate stages of MICROSCOPY sintering by direct observation in the hot stage of the scanning electron microscope Richard M. Fulrath will concentrate on spherical metal powders. On completion of the new powder feed During 1971 a major effort was made to mechanism for the plasma spheroidization further develop the capabilities of the facility, ceramic spherical powders will be scanning electron microscope hot stage. A available. Studies on these materials will reliable heating system was designed which start during the latter half of 1972. has been operated over 300 hours above 1000°C, The maximum temperature achieved to date has been 1600°C, which was held for 1 hour. The hot stage has also been operated at 1500°C for 4 hours while continously recording the 11. 1971 PUBLICATIONS ANu> REPORTS shrinkage characteristics of an AI2O3 powder conpact by time-lapse photography. Richard M. Fulrath and Associates The hot stage modifications have also Journals included changes in the secondary electron detection system and thermal electron sup 1. R. B. Atkin, R. L. Holman. and R. M. pression system to increase the useful Fulrath, Substitution of Bi and Nb Ions in magnification available to approximately Lead Zirconate-Titanate, J. Am. Ceram. Soc. SOOOx. S4_ [2], 113 (1971). The hot stage has demonstrated its useful 2. R. B. Atkin and R. M. Fulrath, Point ness in high temperature studies of sintering Defects and Sintering of Lead Zirconate- metallic and ceramic powders and also in Titanate, J. Am. Ceram. Soc. 54_ [5], 265 observing phase changes in metals. (1971). 166 Papers Presented 5. R. L. Holman and R. M. Fulrath, Intrinsic Non-Stoichiometry in Single-Phase Lead 1. R. M. Fulrath, Hi; 'amperature Scanning Zirconate-Titanate: I, PbfTi 5Zr C)OT, LBL- Electron Microscopy, presanted at the 5th 147, Aug. 1971. International Materials Jymposium on The Structure and Properties of Materials--Tech 6. J. W. Sherohman, Sintering of High-Purity niques and Applications of Electronic Submicron Lead Titanate Containing Niobium Microscopy, University of California, Berkeley, (M.S. thesis), UCRL-20596, June 1971. September 13-17, 1971. 7. J. F. Shackelford, P. L. Studt, and R. M. Fulrath, Solubility of Gases in Glass. II. A Polyatomic Model, UCRL-20716, June 1971. 8. J. F. Shackelford and R. M. Fulrath, UCRL and I.BL reports Solubility of Gases in Glass. III. Experi mental Studies of He, Ne, and Hj ^ p ] 1. G. A. Young and R. M. Fulrath, Strength n usec of Internally Strained Brittle Matrix Silica, UCRL-70717, Juno 1971. Composites, UCRL-2039S Rev., Jan. 1971. 9. S. Risbud, Influence of Surface Damage on 2. J. F. Holzgraf and R. M. Fulrath, The the Strength of a Brittle-Matrix Composite Characterization of the Matrix of Brittle (M.S. thesis), LBL-137, July 1971. Matrix Compsoites, UCRL-20388 Rev., Jan. 1971. 10. J. F. Shackelford, P. L. Studt, and 3. J. F. Shackelford, The Solubility of Gases R. M. Fulrath, Solubility of Gases in Glass. in Glass (Ph.D. thesis), UCRL-20399, Jan. 11. He, Ne and H2 in Fused Silica, LBL-401, 1971. Oct. 1971. 4. J. S. Masaryk and R. M. Fulrath, Permea 11. B. R. Powell, Processing of Sodium- tion in Fused Silica, UCRL-18393 Rev., Feb. Potassium Niobate Ceramics (M.S. thr-sis), LBL- 1971. 413, Nov. 1971. r m-m *>* i •*4 Mi«~ ' *-. ISf '# ' • * * 5 *! »' « * Q i '\ .0 * .^. 167 A. THEOHE1ICAL STATE PHYSICS Haroin I. Cohan, Principal Invcttigator of the modulated reflectivity (l/R)(dR/dw). The calculations were compared with the ex 1. ELECTRONIC STRUCTURE OF SEMICONDUCTORS perimental results of Professor Shen's group and good agreement was obtained between ex Marvin L. Cohen periment and theory. Our studies of the electronic structure of The application to niobium was the first semiconductors and insulators were extended pseudopotential calculation for a transition further. Hand structures, optical constants, metal. The results were in good agreement frequency and wave-vector dependent dielectric with APW calculations. The simplicity of the functions and electronic charge densities pseudopotential mothod allows n determination were computed for a number of Materials. of tho enorgy bonds at many points in the Brillouin zone and these calculations make Electronic charge density calculations were detailed density of states and joint density node for Si, Ge, Sn, GaAs, InAs, InSb, ZnSc, of states calculations possible. In addition CdTe, and PbTe. These calculations were done the flexibility of tho pseudopotcntial for each valence band individually and then method allows calculations of the electronic added to give the SUM of the valence charge structure of transition metal compounds. This density. The results were displayed on a is a particularly interesting application plane in the real space cell, giving an since recently these materials have become the excellent picture of the role of the valence objects of intensive theoretical and experi electrons in the bonding. The magnitude of mental research. the bond charge and the degree of covalcnt and ionic bonding was determined for each crystal. 3. SUPERCONDUCTIVITY Reflectivity and modulated reflectivity spectra were calculated based en our Marvin L. Cohen pseudopotentlal band structures. For several semiconductors the results were compared Using current superconductivity theory directly with the experimental measure i^.ts and some assumptions about the normal state of Professor Shen's group. These compwisons properties of solids, estimates of the gave detailed information which could vien maximum superconducting transition tempera be used to fix the pseudopotcntial for ture have been made. The optimum resonant various materials. frequency for the attractive interaction was determined and the role of Umklapp The first calculation of the frcquercy scattering was discussed. Using these results and wave-vector dielectric function for a it was possible to evaluate theories which semiconductor was done for silicon. The have been developed to predict high tempera results showed how semiconductor dielectric ture superconductivity. functions differed from those for metals. In particular, the role of the gap becane An experimental project was also carrind quite evident. The dispersion of the plasma out based on the theoretical suggestions of made was given explicitly; the real and imag our group. Both Li and Hg have been examined inary parts of the dielectric function and theoretically and appear to be reasonable the imaginary part of the inverse dielectric candidates for superconductivity. In both function was given for many wave vectors and coses, however, down to the temperatures frequencies in perspective plots. tested no superconductivity was found. It is hoped that proximity effect measurements using a dilution refrigerator w:tll shed same light on the possibility of superconductivity 2. ELECTRONIC STRUCTURE OF NOBLE. AND at lower temperatures in these materials. TRANSITION METALS AM) THEIR COMPOUNDS Marvin L. Cohen 4. NARROW GAP SEMICONDUCTORS An empirical pseudopotential with a non local d-potential which had been used pre Marvin L. Cohen viously by our group to study Cu was applied to a calculation of Ag, Au, and Mb. The cal We have continued our work on IV-VI culation for Ag and Au included a calculation semiconductors by calculating optical pro- 168 perties, charge densities, basvl structures, 6. 1971 PUBLICATIONS AND REPORTS band gap temperature coefficients, band grp pressure coefficients, superconducting Marvin L. Cohen and Associates parameters and Ferni surfaces for these solids. Because of the importance of these Journals and book materials for infrared devices we explored the band gap properties in detail. These 1. T. L. Thorp, B. B. Triplctt, W. D. Brewer, studies have been helpful to experimentalists M. L. Coi.en, N. E. Phillips, D. A. Shirley, and we hope to obtain new ir.forsetion from J. E. Templeton, R. W. Stark, and P. H. the data about the properties of these Schmidt, Search for Superconductivity in materials. Lithiun and Magnesium, J. Low Temp. Phys. 3 (6), 589 (1970). HgTe-CdTe alloys have been added to our list of narrow gap materials. These alloys 2. J. P. Walter and M. L. Cohen, Electronic are semiconductors at one end of the alloy Charge Densities in Semiconductors, Phys. curve and semiroetals at the other with a zoro Rev. Letters 26, 17 (1971). gap structure in between. Wo have computed the optical constants at both ends of the S. Y. W. Tsang and M. L. Cohen, Calculation alloy curve for visible and UV light. The of the Temperature Dependence of the Energy results are in good agreement with experiment. Gaps in PbTo and SnTc, Phvs. Rev. 83, 12S4 Studies of alloy properties are now in (1971).* ~~ progress. 4. Y. W. Tsang and M. I.. Cchen, Band Ordering in PbTe, Solid State Comm. 9. 261 (1971). S. RESEARCH PLANS FOR CALENDAR YEAR 1972 5. J. P. Walter and M. L. Cohen, Pseudo- potential Calculations of Electronic Charge Marvin L. Cohen Densities in Seven Semiconductors, Phys. Rev. M, 1877 (.1971). We plan to expand our work on the CdTe-HgTe alloys tf explore the dependence of the band 6. M. L. Cohen ind Y. W. Tsang, Theory of gap on alloying, particularly in the zero Electronic Structure of Some IV-VI Semi band gap region. conductors, J. Chem. Phys. Solids 32 (Suppl. 1), SOS (1971).* ~ We have begun studies of the electronic charge density in hexagonal and cubic ZnS to 7. M. L. Cohen and P. W. Andersen, Comtsr.ts explore the changes in going from hexagonal on the Maximal Superconducting Transition to cubic structures. In particular we ex Tciperature, in Proceedings ot "Superconducti pect that the equivalence of the four vity in d and f Band Metals" (in ptcss).* nearest-neighbor bonds in tbe cubic structure is destroyed in the wurtzite case. This work will be extended to amorphous crystals. Not supported by IMRD funds. Calculations arc being planned to use LBL reports the nor.-local pseudopotential scheme for energy band studies of Nb conpounds starting 1. C. Y. Pong and M. L. Cohen, Electronic with NbN which is a high temperature super Energy Band Structure of SnS? and SnSei, conductor. Depending on our successes here Phys. Rev. (in press); LBL-402, Oct. 1971. we will explore other transition aetal compounds. 2. D. J. Chadi, J. P, Walter, M. L. Cohen, Y. Petroff, and M. oalkanski, Reflectivity New frequency and wave vector dependent and Band Structure of CdTe and HgTe, Phys. dielectric function calculations are being Rev. (inptvss); LBL-48J, Dec. 1971. planned for Ge, GaAs, and ZnSe. The effects of IMclapp scattering (or local field 3. J. Stokes, Y. R. Shen, Y. W. Tsang, corrections) will be explored. M. L. Cohen, and C. V. Tuag, Wavelength Modulation Spectra cf Single Crystals of The relation between the band gaps and Silver and Cold ,LBL-451, Dec. 1971. the bonding charge will also be studied. It is hoped that the charge density calculations 4. C. Y. Fong and M. L. Cohep, Pseudo- will aid in helping us to understand the potential Calculation of the Electronic dependence of bonding on the band gap. Structure of a Transition Metal—Niobium, LBL-48S, Dec. 1971. 169 5. Y. W. Tsang and M. L. Cohen, Electronic 6. J. P. Walter and M. I.. Cohen, Frequency Charge Density in PbTe, Solid State Comnunica- and Kave-Vector-Dependent Dielectric Function tions (in press); LBL-486, Dec. 1971. for Silicon, Phys. Rev. (in press); LBL-487, Dec. 1971. 170 B. MAGNETIC PROPERTIES OF SOLIDS Alan M. Portis, Principal Investigator 2. RESEARCH PLANS FOR CALENDAR YEAR 1972 1. ELECTRON RESONANCE IN METALLIC PARTICLES Alan M. Portis Daniej. A. Gordon and Alan M. Portis In collaboration with Professor Walter D. Knight it is planned to develop an extensive Electron resonance has been observed in program for the study of the electronic small particles of lithium prepared by gamma properties of small metallic particles, iall irradiation of lithium hydride at elevated support of this work has been requested from temperature. The resonance line is unusually the National Science Foundation through its narrow, being of the order of 200 milligauss Division for Materials Research. It is hoped wide, as measured in a field of about 3000 that the expanded activities can be funded by gauss. This is probably the narrowest line NSF from midyear. ever observed from electrons in a metal. The line can be saturated quite easily at room temperature and shows a strong Overhauser shift at low temperature. 3. 1971 PUBLICATION Alan M. Portis and Associates Journal 1. Boris Chornik, Nuclear Relaxation in a Dilute Ni_-Pd Alloy, Phys. Rev. £, 681 (1971). 171 C. FAR INFRARED PROPERTIES OF SOLIDS Paul L. Richards, Principal Investigator by an rf current source. The important param eter in the fit is the rf frequency normalized 1. THE JOSEPHSON JUNCTION AS A PHASE-LOCKED by the junction resistance and critical HETERODYNE DETECTOR current ?=hio/2eRIc=0.15. Paul L. Richards and Yuan Taur Measurement of the first-order step yields a linear heterodyne detector for the rf sig In the well-known ac Josephson experiment, nal. We have investigated the properties of if the influence of the rf signal exceeds this detector with a voltage-biased point that of the low frequency noise, then the rf contact junction and a 37 GHz rf source. The signal synchronizes the Josephson currents height of the first step is measured by and dc beats or steps appear. In our experi modulating the voltage bias and measuring ments, we have used a niobium-niobium point the junction current with a superconducting contact Josephson junction which is sealed quantum interference galvanometer (SLUG). in He gas and adjusted at room temperature. The operating frequency and bandwidth can bo This junction is mechanically stable and can selected by adjusting the dc and ac bias be temperature-cycled without significant voltages. Figure 2(a) shows the detector changes in its characteristics. Figure 1 response as a function of ac bias voltage shows the height of the zeroth and first- with an ac modulation of 0.6uV. This narrow order steps measured as a function of applied peak shows that th? source is monochromatic rf voltage at 37 GHz. These curves deviate and strongly synchronizes the junction, ihe significantly from the Bessel's functions detector sensitivity is limited by step J(j and Jj expected from simple Josephson rounding (due to ambient 4K noise) to a theory, but are in excellent agreement with coupled power of ~10"12vf. Figure 2(b) theoretical calculations based on a model shows the detector response from a rounded which assumes that the junction is driven step when the junction is just abo "t to lose its synchronization. Further experiments are in progress to obtain better coupling by operating the junction in a high-Q resonant cavity. 36 37 38 40 Fig. 1. Height of the zeroth order Josephson Frequency (GHz) step [triangles) and first order step (circles) Fig. 2(a). Radiation synchronizes the as a functic' of incident rf power at 37 GHz. junction when the power level is 36.5 dB Both curves -e normalized to the junction below the first minimum of the first induced critical cur. nt. (XBL 7112-2263) step. (XBL 7112-2264) 172 is shown in Fig. 1. The first, third, and fifth peaks in the figure arise from the forward propagating peak at 13.5 cm"!, while the second and fourth peaks have contributions from both the forward peak and the backward peak at 6.7 cm"*. 0.5 1.0 36 37 38 PLATE SEPARATION, mm Frequency (GHz) Fig. 1. Fabry-Perot fringes of far infrared Fig. 2(b). Response of the junction to radiation generated in LiNbOj by picosecond monochromatic radiation at 37 GHz. Here the pulses simultaneously phase-matched at dc bias voltage has been transformed into 13.5 cm-1 and 6.7 The circles are frequencies through Josephson's relation experimental points and the curve is a fiu = 2eV0. The synchronization is destroyed theoretical calculation assuming Gaussian by low frequency noise when the power level laser pulses with a 1.8 psec pulse width. is 39 dE below the first minimum of the first (XBL 714-6714) induced step. (XBL 7112-2265) 2. GENERATION OF FAR INFRARED PICOSECOND 3. THEORETICAL STUDY OF COHERENT PHONON LIGHT PULSES GENERATION BY OPTICAL MIXING IN A ONE-DIMEN SIONAL SUPERLATriCE Kei-Hsiun Yang, Paul L. Richards, and Y. Ron Shen Kei-Hsiun Yang Paul L. Richards, and Y. Ron Shen Because of its broad spectral content, a picosecond laser pulse can generate far in Optical and acoustic Umklapp processes can frared in a nonlinear crystal as a result of occur in a superlattice when its period is beating between its various frequency com comparable with the respective wavelength. ponents. In our exp< -iments we have used a With the help of the lattice momentum, phase- Ndiglass laser and a iNbOj crystal. The matched generation of acoustic phonons with response time of far infrared detectors is frequencies higher than ID GHz appears fea not fast enough to follow the pulses. The sible. We have investigated phonon generation generation of such short pulses can only be by optical mixing in three types of superlat verified by comparing the far infrared spec tice: 1) the medium is periodic for light, trum with theoretical calculations. but not for sound; 2) periodic for sound, but not for light; 3) periodic for both light and For certain crystal orientations of LiNbOj, sound. The calculated acoustic power gener the far infrared spectrum generated contains ated by two ruby lasers beating in a super- two narrow bands of radiation, one of which lattice about 12u thick, composed of 100 KCl-CdS layers is shown in Fig. 1 for 10 dif propagates forward and the other backward ferent Uhiklapp processes. Calculations in inside the crystal. The backward propagating dicate that the detrimental effect of imper far infrared is reflected by the surface of fections in the periodic structure of the the crystal and so can be detected along superlattice increases with the order of the with the forward pulse. One of our measure umklapp process. We also investigated the ments of these two narrow bands of radiation 173 is conventionally done using a Michelson in terferometer with a thin (1-5 mil) Mylar beam splitter. The performance of such an inter ferometer is limited by beamsplitter effi ciency, which is a periodic function of fre quency with an average of typically 304. Techniques have been developed for making free standing wire-grid far-infrared polarizers which have very high efficiencies. It is pos sible to modify a Michelson interferometer in order to use such a polarizer as a beamsplit ter with nearly 1004 efficiency over a wide range of frequencies. Such an instrument is under construction. It is hoped that the greatly improved efficiency will significantly improve our investigative capabilities. 5. GROUND BASED OBSERVATIONS OF THE COSMIC BACKGROUND RADIATION John C. Mather,* Michael W. Werner,''' and Paul L. Richards Big-bang cosmology predicts that the uni •^WW\/WWV.WW1 verse is filled with blackbody radiation at 0 CO 120 IK 2«0 300 360 420 (j (QHil a temperature of ~ 3°K. This radiation has been observed by several groups at wavelengths Fig. 1. Calculations of the sound power P? generated at. frequency ID by beating two ruby greater than 8 mm. However, recent rocket and lasers with powers Pi and Pj; and differene balloon experiments!>^ revealed large discrep frequency cu in a 100-layer KCl-CdS super- ancies at wavelengths shorter than 1 mm, with an excess flux F 1.3 10"9 W/cm2sr; and lattice (781 A KC1 and 400 A CdS). The cal Q culation is done for frequencies at which the indications that this flux is concentrated in difference frequency photon dispersion curve a narrow frequency band. An experimental in the medium crosses the phonon dispersion search for this concentrated flux was carried curve shown in the insert. The stop bands for out from the University of California Barcroft phonon propagation are indicated by solid bars. Station on White Mountain -n Eastern Califor nia. At this 12,500 ft site, which was acces (XBL 721-5950) sible only by helicopter at the time of our observations, the atmospheric transparency generation of high frequency acoustic phonons was greater than 704 in the region of interest by direct absorption of millimeter or sub- around 12 an"*. millimeter wave photons in a piezoelectric superlattice. Since the wavelength of far The instrument use>. was a two etalon rapid infrared photons is much longer than the wave scanning Fabry-Perot interferometer with metal length of acoustic phonons or the period of mesh reflectors. The incoming radiation was the superlattice, optical Umklapp becomes chopped against an ambient temperature black- unimportant. Calculation on an epitaxially body. The input horn could be pointed to var grown GaAs GaP superlattice about 4u thick, ious zenith angles and into a liquid nitrogen composed of 100 layers, shows maximum energy blackbody for gain calibrations. The detector conversion of about 0.01%. If the super- was a liquid-helium cooled InSb crystal, feed lattice were located in a resonant cavity with ing a cooled transformer, room temperature FET a photon Q of 10^, energy conversion could preamp, lock-in detector, signal averaging reach 104. computer, and magnetic tape recorder. By using the averager, confusion from the rapidly chang ing atmosphere was avoided. Resolving powers as high as 150 and peak-to-peak noise of 5°K 4. CONSTRUCTION OF A POLARIZING INTERFEROM were obtained. ETER FOR FOURIER SPECTROSCOPY Observations made on two trips to the moun Richard Aurbach end Paul L. Richards tain covered the range from 0.7 to 1,7 mm wavelength. (Figs. 1 and 2). Good agreement Far infrared Fjurier transform spectroscopy was found with the expected atmospheric emis- 174 1.0 10 ^/^n*^^^ 200 0.6 100 Resolulion 02 12 7 »— V (cm'1) -2/ (cm'1) Fig. 1. 10-14 cm sky emission spectrum Fig. 2. Sky emission in the 6-9 cm region observed in third order (solid), compared with (solid)obtained at lower resolution than theoretical spectrum for 1 precipitable mm H2O Fig. 2, compared with a theoretical spectrum (dashed). Note the O2 lines at 12.3 and 14. for 1 precipitable mm H2O (dashed). The main 14.2 cm"l. This spectrum contains a substan feature appearing is the 6.11 cm"l H2O line. tial contribution from unwanted orders from Note the absence of any strong features in the interferometer and so underestimates the the 7-9 cm"-'- region which could be attributed transmission in the windows. The arrow at to the water vapor dimer, (^0)2- 11.7 an"* shows the position of the emission (XBL 719-7308) feature reported by Beery et al.* No sharp features which can be related to Fo appear on the spectrum. (XBL 719-7307) 6. BALLOON-BORNE FAR INFRARED SPECTROMETER sion from water vapor and oxygen. No bumps John C. Mather, Paul L. Richards, and David weie found on the spectra which could explain P. Woody the supposed excess flux. This experiment is not sensitive to broadband radiation because A far infrared spectrometer has been devel of the poorly known atmospheric emission and oped that can be flown with a high altitude because of overlapping orders in the spectrom balloon for use in astrophysical studies. The eter. Thus, it c-ji be stated that if the ex principle features of the apparatus are a cess flux were concentrated in a line narrower liquid-helium cooled polarizing interferometer than about 0.3 cni'l (this number being a func and an InSb detector. The spectrometer is tion of the assumed £requp;icy) our experiment calibrated for absolute radiometry with an would have been it. anticipated accuracy of better than 0.1°K for a Rayleigh-Jeans source. Fannie and John Hertz Foundation Fellow. This spectrometer is being interfaced to a balloon gondola developed by Professor Kinsey Work partially supported by NASA grants Anderson of the Berkeley Space Sciences Lab NGL 05-003-272 and NGR-OS-003-432. oratory. The system is designed to make high 1. K. Shivanandan, J. R. Houck, and M. 0. resolution measurements of the cosmic back Harwit, Phys. Rev. Letters 21_, 1460 (1968). ground radiation in the 6-20 cm" • spectral ?.. D. Muehlner and R. Weiss, Phys. Letters region. Low resolution measurements in this 24, 742 (1970). spectral region made from rockets and balloons 3. J. G. Beery, T. Z. Martin, 1. G. Nolt, and have yielded contradictory results. C. W. Wood, Nature Physical Science 230, 36 (1971). 17K 7. VORTICITY IK SUPF.RFLUID HELIUM 9. 1971 PUBLICATIONS AND REPORTS William E. Tennant and Paul L. Richards Paul L. Richards and Associates Superfluid helium is analogous to the elec Journals and books tron condensate in a superconductor. If two superfluid helium baths are joined by a small 1. P. L. Richards, The Impact of New Infrared opening, liquid may flow through the opening Sources and Detectors on the Future of Fourier without resistance and thus without a chemical Spectroscopy, 1970 Aspen International Con potential difference until a critical flow ference on Fourier Spectroscopy, AFCRL-71-0019, velocity is reached. If a chemical potential Special Report No. 114, Jan 5, 1971. difference is then applied to the two baths, the flow rate increases only slightly, and 2. G. C, Brackett, P. L. Richards, and W. S. vortex lines begin to cross the orifice at a Caughey, Far Infrared Magnetic Resonance in rate proportional to the chemical potential Fe(III) and Mn(III) Porphyrins, Myoglobin, difference. When the chemical potential is Hemoglobin, Ferrichrome A, and Fe(III) modulated by ultrasonic sound waves at a har Dithiocarbamates, J. Chem. Phys. 54, 4383 monic or subharmonic of the vortex crossing (1971). frequency, the flow between the baths in inter rupted* in a way which is closely analogous 3. J. C. Mather, M. W. Werner, and P. L. to the ac Josephson experiment in a super Richards, A Search for Spectral Features in conducting weak link. We are presently at tempting to observe this effect by means of the Submillimeter Background Radiation, in a temperature modulation of the chemical po Proceedings of the Seventeenth International tential (2nd sound) rather than by the previ Astroplvysical Symposium, Liege, Belgium, June ously used ultrasonic modulation (1st sound). 28, 1971 QJCRL-20704)• 4. K. H. Yang, P. L. Richards, and Y. R. The effects of turbulence on the chemical Shen, Generation of Far Infrared Radiation potential of superfluid helium have also been by Picosecond Light Pulses in LiNbOj, Appl. studied. We have found that the balance be Phys. Letters 19, 320 (1971). tween the gravitational and vorticity contri butions to the chemical potential at equili 5. R. R. Joyce and P. L. Richards, Phonon brium may be used to measure the amount of Contribution to the Far Infrared Absorptivity turbulence in one of the two weakly linked of Superconducting and Normal Lead, in Pro baths mentioned above. Preliminary results ceedings of the Twelfth International Con indicate the absence of the anomalous density ference on Low Temperature Physics, Kyoto," increase in turbulent helium reported by Ruben 1970 (LT-12, published in 1971, pp. 291-292). and Pobell.2 6. D. W. Faries, P. L. Richards, Y. R. Shen, 1. P. L. Richards and P. W. Anderson, Phys. and K. H. Yang, Tunable Far Infrared Radiation Rev. Letters 14_, 5261-4 (1965). Generated from the Difference Frequency Be 2. K. D. Ruben and F. Pobell, Physik 215, tween Two Ruby Lasers, Phys. Rev. A3, 2148 103 (1968). (1971). 7. J. C. Mather, M. W. Werner, and P. L. Richards, A Search for Spectral Features in 8 RESEARCH PLANS FOR CALENDAR YEAR 1972 the Sub-Millimeter Background Radiation, Astrophys. J. Letters 170, L59 (1971). Paul L. l'dchards Projects 1, 3, 4, 6, and 7 will be contin ued and a project will be initiated to study the phonon contribution to the far infrared absorptivity of metals. 176 SUPERCONDUCTIVITY Gene I. Roohlin, Principal Investigator varied between these limits to reproduce the I-V characteristics of still other weak 1. EXTERNALLY SHUNTED OXIDE BARRIER JOSEPHSON links with intermediate amounts of hysteresis JUNCTIONS: GENERALIZED WEAK LINKS (e.g., point contacts and some microbridges). Paul K. Hansma and Gene I. Rochlin Figure 1 shows one of our sample geome tries. There are three S-I-S junctions formed We have shown experimentally that external at the intersections of the ten strips; the ly shunted superconductor-insulator- end ones are externally shunted by evaporated superconductor (S-I-S) thin film Josephson silver strips. Figure 2 shows a family of junctions can reproduce the electrical I-V characteristics with varying amounts of characteristics of many different types of hysteresis obtained by changing the tempera superconducting weak links. By varying an ture of a single shunted junction. The externally applied magnetic field between curves are labeled with the dimensionless a 2 0 and 1 gauss, or by varying the temperature circuit parameter Be (2e/li) ic CR , where by a few degrees Kelvin, the current-voltage ic is the critical current, C the capacitance, (I-V) characteristics of shunted junctions and R the shunt resistance. This parameter can be changed from the no hysteresis limit ca.n be varied with temperature or magnetic characteristic of some weak links [e.g., field, since the critical current, ic, is a superconductor-normal metal-superconductor function of both. (S-N-S) junctions and some microbridges] to the complete hysteresis limit characteristic Due to their versatility, externally of others (e.g., unshunted S-I-S junctions). shunted Josephson junctions may '..ell find The amount of hysteresis can be smoothly applications in areas where other, less controllable, less reliable, and nonadjustable weak links have been used, or in cases where the performance of a device must be altered in mid-experiment to provide optimum instru mental sensitivity. In addition, these shunted junctions have 2.0 1 1 1 Sample 131 > 1.5 f ID E 0.5 "^-o.s o 1 -0.4 ' /\ \ 1 - i 1 1 Tin Silver 0.5 1.0 1.5 2.0 Normalized current i/i. Fig. 1. One type of externally shunted Josephson junction used in our experiments. Fig. 2. Experimental results for the voltage Two of the three S-I-S Josephson junctions vs. current of an externally shunted formed at the intersections of the tin Josephson junction as a function of the films are shunted with silver strips. dimensionless circuit parameter Be- (XBL 713-3043) (XBL 715-3609 177 lOOOp circles by varying the magnetic field, and the solid circles by varying the temperature. The solid curve is the computer-generated theoretical result. In short, what we have constructed and analyzed is a simple, reproducible, reliable, and theoretically well-understood Josephson junction device. At present they serve as models of, and simulators for, the behavior of weak-link devices; they have the potential of replacing less versatile devices in applications of the Josephson effect for sensitive detection of voltage, magnetic field, and infrared radiation. 2. OBSERVATION OF PHONON STRUCTURE IN HIE 03. 10 r TUNNELING DENSITY OF STATES OF SUPERCONDUCTING VANADIUM Paul K. Hansma ar.d Gene I. Rochlin In 1957 Bardeen, Cooper, and Schrieffer (BCS) proposed that the electron-phonon interaction was the cause of superconductivi ty. Since that time there has been an over whelming amount of experimental evidence that this is the case for simple metals. Neverthe less, it has been suggested that for transi tion metals there may be other interactions involved. The most sensitive probe of the super conducting state is electron tunneling. If the electron-phonon interaction is responsible for the superconductivity of a given metal, then there will be phonon-induced structure in the electrical characteristics of a superconducting tunnel junction with that metal as one of the electrodes. In particular Fig. 3. Experimental results for the there will be structure in the second hysteresis parameter a as a function of the derivative, d^I/dV^, at voltages such that circuit parameter B . The circuit parameter c eV + A equals the energies of peaks in the was varied both with temperature (1.12 - phonon density of states, where A is the 3.7°K) and with magnetic field (0 - 1G). superconducting energy gap. (XBL 712-3022) Our measurements on tunnel junctions with provided the first experimental verification superconducting vanadium as one of the elec of the Stewart-McCumber theory of weak links. trodes reveal small dips in d^I/dV^ at This theory predicts the amount of hysteresis voltages of 7.5, 20.S, 25.5, and 28.5 mV in the I-V characteristic of a weak link as above the gap. The positions and relative a function of the dimensionless circuit magnitude of these dips are in excellent parameter gc. The hysteresis is quantified agreement with the phonon density of states with the parameter a, defined as the Tatio of vanadium as measured by incoherent, of the minimum current flowing at non-zero inelastic, neutron scattering, including the voltage to the critical current, ic. For small peak at 25.5 mV found only by the most example, the curve labeled by 6c = 2 in Fig. 1 recent neutron measurements. has a » 0.8. Figure 3 shows both the experi mental and theoretical results for this While our measurements are not yet of hysteresis parameter a, as a function of gC' great enough accuracy to prove that the All of these experimental points were obtained electxon-phonon interaction is the only from a single shunted junction, the open mechanism responsible for superconductivity 178 in vanadium, they conclusively show that it is at least important to the superconductivity of this transition metal. 3. THIN FILM VANADIUM OXIDES Paul K. Hansma and Gene I. Rochlin We have successfully produced thin films of V20j, VO2, and V7O5 from evaporated vana dium films deposited on both mica and glass substrates. The oxide films were formed by the oxidation and subsequent controlled reduction of 3000A thick vanadium films. First V2O5 was formed by oxidation of the Fig. 1. Temperature dependence of the dc clean vanadium surfaces in nir at 500°C for critical current for bridges of various film 24 hours. VO2 could then be formed by thickness. The upper broken curve is the reduction of the V2O5 film in a 1:1 mixture Ambcgaokar-Barntoff calculation for a tunnel of CO and C0 : alternatively, V7O3 could 2 junction. The middle solid curve is the be formed by reduction in pure H . 2 calculation for a bulk superconductor. The lower broken curve is the BCS model without We are especially interested in the VO2 extended phase coherence. (XBL 7111-7533) films because they show a 2-1/2 order of magnitude change in resistivity over a very small range of temperatures near 70°C. There is a possibility that this effect can be exploited for the production of high- sensitivity room-temperature bolometers for the detection of infrared radiation. 4. TRANSITION FROM BULK TO JOSEPHSON- JUNCTION BEHAVIOR IN WEAK LINKS Yeong-du Song and Gene I. Rochlin We have measured the behavior of thin film superconducting microbridges (on the order of 1 to lOu long, 1 to 3u wide, and O 02 04 OS 08 10 REDUCED TEMPERATURE T/Tc 100 - 2000A thick) as a function of tempera ture, size, and the impurity-dependent co Fig. 2. The impurity dependence of the herence length 5. In regions of temperature temperature-induced deviations of the critical and impurity where £. » I, where L is the current. The theoretical curves are the same bridge length, the bridges exhibit a as in Fig. 1. (XBL 7111-7534) Josephson-junction-like behavior, both dc ajid in their response to applied microwave signals. At lower temperatures or higher bulk-like behavior, so 'subharmonic' steps at impurity concentrations, where £ is shortened voltages nV = nt>w/2e (m and n integers) so that 5 « L, the bridges behave like appear and the rf sensitivity diminishes small pieces of bulk superconductor as appreciably. A lumped circuit parameter shown in Figs. 1 and 2. model of the type studied extensively by Hansma and Rochlin adequately describes the We have also investigated microwave- ac and dc response of the bridges to applied induced steps in the microbridge I-V fields in the junction like regime. This characteristics as a function of these same ac behavior indicates that all superconducting parameters. When the junction is in the weak-link rf detection devices can be £. » L regime, (i.e., junction-like) the ciiaracterized by a uniform parametrization induced rf steps are at exactly V = hw/2e, in terms of £ (T, impurity)/d, where d is as one finds for an ordinary Josephson the characteristic size of the weak link. junction. However, as the conerence length If £/d is large, the device will behave like is diminished and the bridge shifts into a Josephson junction; if £/d is small, the 179 device will exhibit bulk weak-link behavior. very small area OS .0005 x .0005 in.) The unusual temperature dependences of many r jperconducting-normal metal sandwiches. such devices can thus be explained as being Equipment has been built to allow precise due exclusively to the temperature dependence optical alignment of these S-N sandwiches of the coherence length and the resulting directly on top of small (.003 x .003 in.) variation in the strength of the coupling. superconducting tunnel junctions exhibiting excellent Josephson diffraction patterns and ve.7 small excess current at low dc biases. 5. SCRATCHED FILM QUANTUM INTERFEROMETERS By this means we are examining the change in the density of states of a superconducting Yeong-du Song and Gene I. Rochlin film induced by a normal metal overlay separatfd by a thin oxide layer from the Figure 1 illustrates an extremely simple, superconductor. 'Die coupMng of the two elec reliable, durable, and reproducible two- trodes of the sandwich can be directly junction quantum interferometer we have built. determined by tunneling from one side of the This device consists of a thin Sn film vapor sandwich to the other, whilf the Joscphson deposited at room temperature in a moderate junction (one electrode of which is the vacuum, and then 'scratched' with a diamond superconducting part of the sandwich) nets stylus to obtain a pair of weak-link bridges as a sensitive detector of any change in the between the two halves of the film. Using superconducting density of states attributable an elastic deformation scratching jig we have to the normal metal proximity. designed, we are able to guarantee the alignment and matched widths of the bridges For this work it is necessary to fp!::icate even for bridges as small as lu x 3u, superconducting junctions with very • • excess current which are stable at ruom 1 1 1 1 1 temperature for the time necessary to produre SAMPLE 29A $%$M 03 a normal metal overlay (2 hours). We have Ic • 2.6 mA T • 164*K VA made such junctions which not only exhibit LOOP AREA<3>C!cm2 the above characteristics but also exhibit 0 excellent Josephson diffraction patterns, recycle up to four times from helium to room rr temperature, store for a week at room 0.2 - tempo-lture, and can be abused with high currents at room temperature without signifi i i i i i 0.12 0.13 0.11 0.15 0.16 cant change in either the low voltage MAGNETIC FIELD CAUSS characteristics, the magnitude of the Fig. 1. Modulation of the dc critical Josephson current, or the magnetic diffraction supercurrent of a double-bridge interferometer pattern. made by scratching a thin tin film. The magnetic field is applied perpendicularly to We are currently investigating properties the device, which is shown in the inset in of these small S-N sandwiches while varying the upper ri^ht-hand corner. (XBL 7111-7528) the contact area over two orders of magnitude from .005 x .005 in. to .0005 x .0005 in. In the absence of an oxide layer the smaller- area sandwiches exhibit properties analogous By operating this device near Tc, the bridges exhibit Josephson-junction-l;ke to thin film microbridges at temperatures near behavior, and this device is effectively a Tc of the superconductor. They are also two-Josephson-junction interferometer. Ihe remarkably stable at high biases (approx. 100 mV) despite high current densities. The lOt modulation of the critical current, Ic, when the magnetic field is applied, is an resistance ratio R^nn mV/R0V = 1000, with order of magnitude greater than is usually the bulk of the increase of resistance obtained in interferometers of this type, and occurring at a few mV. These sandwiches is due to the extremely low inductance of have also been recycled with little change in the film bridges and the enclosed loop. characteristics. 6. TUNNELING INTO VERY SMALL SUPERCONDUCTING 7. TUNNELING IN NORMAL METAL JUNCTIONS NORMAL SANDWICHES Duncan E. McBride and Gene I. Rochlin Michael D. Jack and Gene I. Rochlin We continue to study the properties of thin We have developed a method of producing film tunnel junctions in which neither metal 180 is superconducting, in order to investigate 0.1 to 100 Hz range in semiconductors in thp the effects of elementary excitations on the process of testing the apparatus. These tunneling current. Preliminary results have measurements will be extended to investigate been obtained, 10 ng A1-A1;03 - pure copper the widespread existence of 1/f noise, the junctions in which we have observed changes mechanisms for its generation, and the nature in the tunneling conductance at energies of the spectrum itself. Eventually this corresponding roughly to the phonon excita technique will be applied to noise in systems tions in copper, but experimental resolution such as membranes and nerve axons as well as and sensitivity are not yet sufficient to solids. allow us to make a more definite identifica tion. We are developing additional techniques fo: data collection and analysis, primarily by using a small digital computer, in order 9. RESEARCH PLANS FOR CALENDAR YEAR 1972 to improve the quality of the data and to extract the maximum amount of information from Gene I. Rochlin it. Our tunneling program will be continued, This technique is also being used to with investigations still to be made on CuCr investigate the effect of adding small amounts and CuFe alloys, and the phon"n work on V~to of magnetic impurities (chromium, iron, etc.) be completed. Tho properties of Vx Cy bar to copper and similar metals (gold and silver). riers and of Cr halides will be thoroughly Since electrons interact strongly with the investigated and reported. magnetic atoms, we expect this interaction i.0 affect the tunneling current, which is Ifo shall continue to explore the properties a sensitive probe of the density of states of of the shunted Josephson junctions in both the metals which make up the junction. the ac and dc regimes, and work on a mechan ical analog model of these junctions to ex plore their nonlinear behavior. 8. NOISE Experimental and theoretical work on 1/f noise in a variety of systems will be ex plored by using the computer for long-peiiod Duncan E. McBride and Gene I. Rochlin autocorrelation to derive the low frequency noise power spectrum. We have recently begun an investigation of fluctuations and noise, particularly current- dependent 1/f noise, in a number of diverse We also shall complete the woik on tunnel systems including thin metal films, semicon ing into very small SN sandwiches, by using ductor devices, and tunnel junctions of stable Josephson junctions as probes, and almost any type. We calculate the voltage shall investigate the properties of three autocorrelation function electrode double junction devices produced by growing a thin oxide layer between the superconducting and normal electrodes. CM = | V(t) V(t + T)dt , 10. 1971 PUBLICATIONS AND REPORTS where t is time and T is a time intei-val and Gene I. Rochlin bad Associates from this compute the power spectrum with a Fourier transform Journals 1. P. K. Hansma_, J. N. Sweet, and G. I. p(u) = i j C(T) cos(vr)dT . Rochlin, Properties of Sn-I-Sn Josephson Junctions with External Resistive Shunts, •c. Bull. Am. Phys. Soc. W, 399 (1971). Using our POP-11 digital computer both to take data and to compute the correlation 2. P. K. Hansma, G. I. Rochlin, and J. N. function and spectrum during the experiment, Sweet, Externally Shunted Josephson Junctions: we can easily observe at one time a power Generalized Weak '.inks, Phys. Rev. B£, 3003 spectrum in about a 4-octave range including (1971). any frequency up to 100 Hz and extending indefinitely down in frequency. All this is 3. Y.-d. Song, Transition from Bulk Behavior done without using any filters. We have to Josephson-Junction-Like Behavior in already observed spectra with excellent Superconducting Microbridges, LBL-198 resolution and signal-to-noise ratio in the (Ph.D. thesis and to be submitted to Phys.Rev.). 181 John Clarke, Principal Investigator evaporating small disks of superconductor on both sides. The thicknesses of the foils were 1 A NEW METHOD FOR DETERMINING VERY LOW in the range 25-250 urn, and were chosen to be SUPERCONDUCTING TRANSITION TEMPERATURES sufficiently thick that Josephson supercurrents would not interfere with the resistance meas John Clarke, Stuart M. Freake, and urements. The specimens were mounted in a Michael L. Rappaport dilution refrigerator which could attain tem peratures as low as 18 mK. The resistances We have devised a new technique for deter (~ 10"E 9.) were measured to a relative ac mining the BCS interaction parameter (NV)JJ of curacy of 0.03* with a .•='• -^conducting volt a normal metal and the corresponding super meter (SLUG). conducting transition temperature. The method involves measurement of the temperature de In Fig. 1 we show th variation of the re pendence of the resistance of a superconductor- sistance of two Pb-Ir-Pb specimens with tem normal metal-superconductor (SNS) junction. perature. Below 100 mK the resistance de Because of the superconducting proximity ef creased rapidly, and became zero just above fect, the resistance of such a junction is 57 mK, the transition temperature of our Ir. reduced below that for the normal metal alone At temperatures above 6K the resistance rises by an amount which depends on the interaction rapidly due to the propagation of normal elec parameter and on temperature. Wi'ch the aid of trons into the superconductor. Using a a quantitative theory, the interaction param simple mtdel, we have calculated the tempera eter can be determined from the resistance ture dependence for T < 2K. The results are measurements. shown in Fig. 1, and are in quite good agree ment with the experimental results. At low We have performed a series of experiments temperatures the resistance also depends on on Pb-Ir-Pb junctions. The normal metal was measuring current, I, and magnetic field (see in the form of a high-purity foil, and its sur Figs. 2 and 3). The linear dependence of face was carefully sputtered clean before resistance on Jul at the lowest temperatures is predicted by our simple theory, but a more refined theory is required to quantitatively account for the field dependence. A more detailed theory will be needed be fore accurate estimates of the interaction parameters of normal metals with very low _i i i i i i ' 0.05 0.1 O.Z 0.5 1.0 2.0 5.0 10 Temperature (K) Fig. 1. Variation of resistance with tem perature for two Pb-Ir-Pb samples. Thickness of Ir in 8C was 53±lu, and in 8D, 54±ly. Continuous lines for current density of -2 1.26 A cm"2, dashed line for 0.126 A cm"2, Current density (Acm ) and dotted line is theoretical curve for 8C. Fig. 2. Variation of resistance with current (XBL 7112-4973) at different temperatures. (XBL 7112-49755 182 in order to measure the magnetization of small specimens of various semiconductors to deter mine whether they are superconductors. Ex periments are performed at temperatures down to 18 mK in residual fields less than 10"9 T. Nine different materials have been tested to date. A specimen of tin telluride became superconducting at 44 mK, which is consistent with earlier experimental results on this ma terial obtained by other workers. A gold telluride specimen showed evidence of super conductivity at temperatures below 48 mK, but even at 18 mK the magnetization curve in dicated that only 54 of the specimen was superconducting. This implies that the sample was inhomogeneous, and the superconductivity was possibly due to small pockets of strained tellurium. The remaining seven specimens tested: CU2l'e, Ag?Te, Ag2Se, SbSI, Hg0.79Cd0.2lTe, Pb0 a3Snn.i7Te. Pb0,8Sn0.?Te, 0.5 I 2 5 10 did not show any evidence of superconductivity Flild (gnuii) at 18 mK. Fig. 3. Effect of magnetic field upon re sistance for various current densities at We are currently testing other materials, different temperatures. (XBL 7112-4974) but are seriously hampered by the difficulty of obtaining good, homogeneous samples. transition temperatures can be made. However, our simple theory indicates that at 20 mJC we should be able to detect interaction parameters 3. THE SUPERCONDUCTING FOUNTAIN EFFECT as small as 10"3, corresponding to transition 4nn temperatures of 10" K. John Clarke and Stuart M. Freake We have performed similar experiments with A temperature difference across a superleak Pb-Cu-Pb sandwiches, which also showed a de in liquid helium gives rise to a flow of the creasing resistance at low temperatures. We superfluid through the leak from the colder conclude that copper has a non-zero interaction side to the hotter. This phenomenon is well parameter. 2. A SEARCH FOR SUPERCONDUCTIVITY IN DEGENERATE SEMICONDUCTORS John Clarke, Marvin L. Cohen, Stuart M. Freake, and Michael L. Rappaport The study of superconductivity in semi conductors is of great interest for two main reasons. First, semiconductors are better understood than metals. Second, it is pos sible to make large changes in the carrier density by alloying, and in the band structure by applying pressure, in order to study the dependence of the superconducting properties 5 7 9 on the normal state properties. A number of 6 degenerate semiconductors are known to be Peak-to-peak critical current (I0"" amps) superconducting (GeTe, SrTi03, SnTe), and Fig. 1. The difference in the critical currents have transition temperatures of the order of measured in opposite directions as a function 0.1 K. of the sum of the critical currents, (a) Foil temperature Tf=7.1K.. wire tempera We have constructed a superconducting mag ture TW=5.7K; and 3 netometer in the He-*He dilution refrigerator (c) Tf =4.3K, TW=7.1K. (XBL 7112-4972) 183 known and is called the fountain effect. In an analogous way, if one creates a temperature gradient in a superconductor, one expects a net flow of superfluid pairs, to balance the flow of normal electrons carrying the thermal energy. No direct observation of this foun tain effect in superconductors has yet been reported. We have detected the thermally produced ipPb a f^m supercurrent by measuring the asymmetry of the Glue spots critical current of a Pb-Pb point contact Fig. 1. Bolometer element with Pb-Cu/Al-Pb Josephson junction. The observed critical junction on the loft and Cu/Al heater on the current is smaller if the applied current and the thermally induced supercurrent are in the right. (XBL 7112-4927) same direction than if they are in opposite directions (see Fig. 1). The asymmetry is only of order 10"7A for a 1 K temperature dif ference across a IS junction, which is in good agreement with the value we have calcu lated theoretically. In order to measure this small effect, we have designed and built an electronic system which automatically meas ures the asymmetry and the average critical current. With the aid of this system, we can rapidly take data before the point contact junction, a type which is notoriously unstable, spontaneously changes its properties. 4. A JOSEPHSON JUNCTION BOLOMETER John Clarke and Gary I. Hoffer Fig. 2. I-V characteristics of an SNS junction at two different temperatures The superconducting-normal metal-super (T, using glass substrates. The best of these junctions are presently under investigation. had the following characteristics measured at 5°K: Two circuits which demonstrate the im 7 portance of the current gain are shown in Thermal conductance G 6X10" W/°K Fig. 1. When used as a voltage amplifier, T.iermal time constant T=C/G 1 sec Fig. 1(a), the asymmetric interferometer is current-biased just above its critical cur dc: responsivity S 1.7X10'2V/W dc rent . The voltage gain is 8(R/Ri), where Sis Noise equivalent power NEP 4X10"11 W/4Tz the current gain, R is the parallel resistance of the two junctions, and R^ is the sum of These data were obtained using the Cu/Al the resistances of the input voltage source heater for a power input. The measured re and the input resistor. R^ is limited by the sponsivity is in good agreement with the desired input time constant Lj/Ri, where calculated value of 1.2X10-2 V/W. The calcu Li > 10"8 henries. When used as a magnetic lated NEP is .9<10_13 w/,/ITz. The experimental flux amplifier, Fig. 1(b), the asymmetric value was in fact limited by noise in the interferometer is voltage biased. The input voltage-measuring circuit rather than by noise flux produces a cuculating current desired time constant L0/R in the output loop. 5. JOSEPHSON JUNCTION AMPLIFIER John Clarke and James L. Paterson 6. A NEW TYPE OF FERMI SURFACE Last year we reported on a technique for MEASUREMENT achieving current gain in an asymmetric Josephson quantum interferometer. The cur John Clarke and Louis A. Schwartzkopf rent gain is adjustable from outside the cryostat to values between 1 and 100. The We have completed preliminary work on a preliminary devices employed lead-copper-lead new type of experiment to study the Fermi sur (SNS) junctions, which are unsuitable for face of a metal. In this experiment, two practical applications because of their low interlocking superconducting combs are eva normal state resistance of ~10"° SI. This low porated on the surface of a single crystal of resistance also results in high thermal cur a normal metal (see Fig. 1). A current is rent noise in the interferometer loop, thus limiting the current sensitivity to ~10"8 A. —i 1—50,0- Development of higher resistance (~1 ft)junc r~--f±------tions is continuing. Problems associated i-^Effi52Zffi2Z. with Notarys bridges, oxide tunnel junctions — Normol and shunted-oxide tunnel junctions have pre metal cluded their use. Thin foil, point contact ' Superconducting I'bios film -Jnsuloior '(SO*, £3 -pToo/T . Teeth o' comb ™<™ ® H w Normal Electron orbits "metol *•*(£)* Fig. 1. Asymmetric interferometer used as (upper) voltage amplifier, and (lower) magne Fig. 1. Comb structure, (a) Plan view, tic flux amplifier. V represents a constant showing interlocking combs; (b) side view, voltage bias. (XBL 714-779) showing electron orbits. (XBL 7112-4954) 185 passed between the combs, and a uniform mag superconducting transition temperatures will netic field applied parallel to the teeth. be expanded. We plan to investigate rhodium, When the magnetic field is adjusted so that magnesium, copper, silver, gold, and platinum, the diameter of the electron trajectories and possibly other materials. Theoretical equals the distance between two teeth, there predictions for the transition temperature of is an anomaly in the resistance between the most of these metals already exist, so that combs. This experiment provides a caliper an experimental determination is of great measurement of the Fermi surface of the nor interest. We plan to work on a more detailed mal metal. For a typical metal, the resis theoretical model of our effect. tance between the combs is about 10"^ £!, and the voltage developed with a current of More measurements will be made on semi 10 100 mA is about 10" V, a voltage easily conductors to see if they become superconduct measurable only with a Slug or with a similar ing at low temperatures. The scope of this superconducting quantum interference device. work depends largely on the availability of The first experimental requirement that suitable samples. must be met is that the electron mean free path in the metal be at least as long as the The Josephson junction bolometer appears to distance between the teeth; otherwise, the have great potential as an infrared detector. electron is scattered before it can complete We plan to improve the sensitivity of the one orbit. This requires that thu sample be device as fai as possible, and then to in a high-purity single crystal. It is fairly vestigate various methods of blackening it to straightforward to obtain an electron mean make it respond to infrared radiation. We free path of lOOu at 4.2°K in a number of have long-term plans to use the device in far metals. A second requirement is that there be infrared astronomy. a low boundary resistance between the super conducting film and the normal metal. This We are continuing our efforts to develop means that there must be low rates of inter- high resistance junctions suitable for use in diffusion and alloying, and that there must the Josephson junction amplifier. We are be little or no oxide on the surface of the presently concentrating on junctions with normal metal. We considered several combina oxide barriers, which are known to have very tions of superconducting film and normal metal. low inherent noise. We tried Pb-Zn, but found a large boundary resistance due to oxidation of the Zn. We The work on a new method of Fermi surface considered Pb-Au, since Au does not oxidize measurement will bo coro-inued, and extended at room temperature, but discarded this com to materials other than bismuth. A fully bination because of interdiffusicn and the automatic system is being developed to difficulty in achieving a long enough mean measure the resistance of the sample as a free path. function of magnetic field, thus greatly facilitating the taking of data. We are currently studying S'l-Bi, which seems a likely candidate for success. This system A new area of work is being started on non- fulfills the requirements, and Bi has a high equilibrium phenomena in superconductors. resistivity which reduced the importance of Techniques are being developed to obtain an any boundary resistance which may be present. inversion of the quasiparticle population, In addition, because of the small effective and ways of exploiting this effect will be mass of the electrons and holes in Bi, the investigated. resistance anomaly occurs at low fields (~10 G), which are easier to produce and work Measurements of 1/f noise in thin metal with. films will be made. We hope to investigate the noise developed at very high current Single crystals of Bi are grown by melting densities, where the resistance of the samples 99.999+4 pure Bi, then cooling it in a temper becomes nonlinear, in order to gain new in ature gradient. These_crystals are cleaved sight into the origins of J/f noise. along the (HI) or (HI) plane and the combs are evaporated on the freshly prepared surface. The edges of the comb are isolated from the sample by a thin layer of insulator, to keep them from shorting out the comb (see Fig. 1). 8. 1971 PUBLICATIONS AND REPORTS John Clarke and Associates 7. RESEARCH PLANS FOR CALENDAR YEAR 1972 Journals and book John Clarke 1. J. Clarke, Electric Transport Currents The work on the measurements of very low Across Normal-Superconducting Interfaces, in 186 Proceedings of 12th International Conference 8. S. M. Freake, Superconducting Density of on Low Temperature Physics, Kyoto, Japan,- States induced by the Proximity Effect, Sept. 4-lb, 1970, edited by Eizo Kanda Phil. Mag. 24, 319 (1971).* (Keigaku Publishing Co., Tokyo, Japan) p. 443; CUCRL-20312). Not supported by IMRD 2. J. Clarke, Electronics with Superconduct ing Junctions, Physics Today 2£, No. 8, 30 (1971); (UCRL-20582). 3. J. Clarke, W. E. Tennant, and D. Woody, Impedance Matching a Josephson Galvanometer Conferences by Means of a Superconducting Transformer, J. Appl. Phys. 42, 3859 (1971); CUCRL-20360). 1. J. Clarke and J. L. Paterson, Current Gain in an Asymmetric Quantum Interferometer, Bull. 4. J. Clarke, Finite-Voltage Behavior of Lead- Am. Phys. Soc. 16_, 399 (1971). Copper-Lead Junctions, Phys. Rev. B4, 2963 C1971); (UCRL-20548). 2. J. Clarke, Josephson Junctions as Detectors of Electromagnetic Radiation, talk at Confer 5. J. Clarke and J. L. Paterson, A Josephson ence on Applications of Josc-phson Junctions, Junction Amplifier, Appl. Phys. Letters 19, California Institute of Technology, June 1971. 469 (1971); (UCRL-20579). 6. J. Clarke, J. R. Waldram, and A. B. Pippard, A New Theory of the Current-Voltage Characteristics of Weak Links, Physica 55, 405 (1971); (UCRL-19067). LBL Report 7. S. M. Freake and T. L. Thorp, Shielding of 1. John Clarke, S. M. Freake, M. L. Rappaport Low Magnetic Field with Multiple Cylindrical and T. L. Thorp, A New Experimental Technique Shells, Rev. Sci. Instr. 42, 1411 (1971); for Determining the BCS Interaction Parameter (UCRL-20398). in Normal Metals, LBL-457, December 1971. 187 EXPERIMENTAL SOLID STATE PHYSICS AND QUANTUM ELECTRONICS y. Ron Shen, 'Principal Investigator propagation of hypersonic waves in a cholesteric medium at the liquid-to-liquid- 1. SELF-FOCUSING AND SMALL-SCALE FILAMENTS crystal transition. Contrary to the results OF LIGHT obtained by others, we have found no anomalous change in either the velocity or the attenua Michael M. T. Loy, George K. L. Wang, and tion of the hypersonic waves at the transi Y. Ron Shen tion. (See Figs. 1 and 2.) Our results can be well explained by the large domain size It is shown that the small-scale filaments of the sample we used and by the high fre generated by laser pulses in nonlinear media quency of the acoustic waves involved in the are the natural consequence of moving foci. experiment. Many controlled experiments performed with single-mode Q-switched laser pulses are Fig. 2. Brillouin line width of the choles 3. BRILLOUIN SCATTERING IN LIQUID CRYSTALS teric mixture of 344 cholesteryl nonanoate, 34t cholesteryl oleyl carbonate, and 32t Hal J. Rosen and Y. Ron Shen cholesteryl chloride as a function of tempera ture in the liquid crystal-to-liquid transi Brillouin scattering is used to study the tion region. (XBL 7111-1649) 188 4. RAMAN STUDY OF SMECTIC LIQUID CRYSTALS Nabil M. Amer, David Y. S. Lin, and Y. Ron Shen We have continued to USB Raman scattering as a probe to study the phase transitions in smectic liquid crystals. Changes in the position and the strength of many vibrational modes during phase transitions have been observed. In particular, intermolecular modes show up in some compounds but not in others. They disappear as the material goes into the liquid phase. Strong dipolar interaction bet ween dangling roots of neighboring molecules Fig. 1. The measured and the calculated is probably responsible for the intermolecular dR/RdE for Au. (XBL 722-2359) modes. More quantitative theoretical interpretation will be attempted. The same technique will be tried on biological molecules. 7. DISPERSION OF NONLINEAR OPTICAL SUSCEPTIBILITIES Irwin I. Ku, Arnold Schmidt, and Y. Ron Shen 5. FAR INFRARED GENERATION BY OPTICAL MIXING Measurements of dispersion of nonlinear James R. Morris and Y. Ron Shen optical susceptibilities XNL(2U) of semi conductors are in the preparation stage. A Possibility of far infrared generation by tunable dye laser pumped by a ruby laser optical mixing in magnetic crystal has been and its second hairnonics will be used as the investigated. Analysis shows that a few exciting source. The frequency 6. WAVELENGTH-MODULATION SPECTRA OF SOLIDS 8. THEORETICAL STUDY OF COHERENT OPTICAL PHONON GENERATION BY OPTICAL MIXING IN A Jeffrey A. Stokes, Stanley Kohn, and ONE-DIMENSIONAL SUPERLATTICE Y. Ron Shen Patrick K. H. Yang, Paul L. Richards, and The logarithmic derivative of the reflecti Y. Ron Shen vity is measured for Ag and Au for 2.0 < ho> < 6.0 eV at T = 5, 80, and 300°K. The results Optical and acoustic Umklapp processes can are compared with theoretical calculations. occur in a superlattice when its period is The agreement between the theoretical and comparable with the respective wavelength. the experimental spectra is good. (See, for With the help of the lattice momentum, phase- example, Fig. 1 for Au) The various struc matched generation of acoustic phonons with tures in the spectra are identified. frequencies higher than 10 GHz appears fea sible. We have investigated phonon genera Using the wavelength-modulation technique, tion by optical mixing in three types of we have also measured the stress dependence superlattice: (1) the medium is periodic of the EQ peak of Si under uniaxial compres for light, but not for sound, (2) periodic sion. The results suggest that the peak is for sound, but not for light, (3) periodic a composite peak, with one part arising from for both light and sound. A typical example 2 transitions in the region around A (4-5) and shows that two 10 MW/cm input laser beams the other part from transitions over an beating in a superlattice 15u thick, composed extended region around r. of 100 KCl-CdS layers, can generate about 189 5 mW/cm2 of acoustic power at 30 GHz. 6. M. M. T. Loy and Y. R. Shen, Correlation Calculations also indicate that the detrimen Between Backward Stimulated Raman Pulse and tal effect of imperfections in the periodic Moving Focus in Liquids, Appl. Phys. Letters structure of the superlattice increases with 19_, 285 (1971). the order of the Umklapp process. The generation of high frequency acoustic phonons 7. K. H. Yang, P. L. Richards, and Y. R. by direct absorption of millimeter or Shen, Generation of Far-Infrared Radiation submillimeter wave photons in a piezoelectric by Picosecond Light Pulses *;i LiNb03, Appl. superlattice will also be discussed. Phys. Letters 19, 320 (1C/1). 8. H. Rosen, Y. R. Shen, and F. Stenman, Raman Study of Iodine Complexes in Solution, 9. RESEARCH PLANS FOR CALENDAR YEAR 1972 Molecular Physics 22_, 33 (1971). Y. Ron Shen UCRL and LBL reports All programs except 2 and 3 described in 1. M. M. T. Loy and Y. R. Shen, Study of the previous sections will be continued. New Self-Focusing and Small-Scale Filaments of programs on optical properties of excitons Light in Nonlinear Media, LBL-431, Dec. 1971. and exciton drops will be initiated. Resonant Raman scattering in various materials will 2. J. Stokes, Y. R. Shen, Y. W. Tsang, also be looked into. M. L. Cohen, and C. Y. Fong, Wavelength Modulation Spectra of Single Crystals of Silver and Gold, LBL-451, Dec. 1971. 3. J. W. She.'ton and Y. R. Shen, Study of 10. 1971 PUBLICATIONS AND REPORTS Phase-Matched Normal and Umklapp Third- Harmonic Generation Processes in Choleseric Y. Ron Shen and Associates Liquid Crystals, LBL-161, Aug. 1971. Journals and book 4. H. Rosen and Y. R. Shen, Brillouin Scattering in a Cholesteric Liquid Crystal 1. Y. R. Shen, Nonlinear Optical Effects, near the Cholesteric-Isotropic Transition, in Proceedings of the Symposium on the LBL-452, Dec. 1971. Physics of Opto-Electronic Materials, edited by W. A. Albers (Plenum Press, New York, 5. J. Kbo, Y. R. Shen, and R. R. L. Zucca, 1971), p. 17. Effects of the uniaxial Stress on the E3 - Peak of Si, LBL-146, Aug. 1971. 2. J. W. Shelton and Y. R. Shen, Umklapp Optical Third Harmonic Generation in 6. N. M. Amer and Y. R. Shen, Raman Scatter Cnolesteric Liquid Crystals, Phys. Rev. ing from Nematic Liquid-Crystalline Letters 26, S38 (1971). Azoxynenzenes, UCRL-20564, April 1971. 3. D. W. Faries, P. L. Richards, Y. R. Shen, 7. M. M. Loy, Self-Focusing and Small-Scale and K. H. Yang, Tunable Far-Infrared Radia Filaments of Light in Liquids (Ph.D. thesis), tion Generated from the Difference Frequency UCRL-20585, May 1971. Between Two Ruby Lasers, Phys. Rev. A3, 2148 (1971). 8. H. J. Rosen, Light Scattering From Liquids and Liquid Crystals (Ph.D. thesis), 4. M. M. T. Loy and Y. R. Shen, Theoretical LBL-424, Nov. 1971. Interpretation of Small-Scale Filaments of Light Originating from Moving Focal Spots, 9. J. Shelton, Phase-Matched Third-Harmonic Phys. Rev. A3, 2099 (1971). Generation in Cholesteric Liquid Crystals (Ph.D. thesis), LBL-112, Aug. 1971. 5. J. R. Morris and Y. R. Shen, Far-Infrared Generation by Picosecond Pulses in Electro- 10. R. R. L. Zucca, Wavelength Modulation Optical Materials, Optics Comm. £, 81 (1971). Spectroscopy of Semiconductors (Ph.D. thesis), UCRL-20361, Jan. 1371. 190 F. SOLID STATE SPIN RESONANCE AND COHERENT LASER PULSE STUDIES Ertfin L. Hahn, Principal Investigator ly attenuated, can measure population dif ferences. The KC1 F-center absorption band 1. PICOSECOND OPTICAL PULSE NONLINEAR is found to be resonant with the laser second PROPAGATION EFFECTS harmonic (0.53 \i), and the emission band is resonant with the laser fundamental (1.06 u). Robert A. Fisher The F-center dynamics can be described by a 4-level system. Nonlinear saturation and gain This thesis research evolved into a number have been utilized to study the optical pump- of experimental and theoretical studies in band dynamics. In this investigation, com picosecond pulse nonlinear optical effects. plete inversion of the emission band has been Both photon echoes and self-induced trans generated by absorption of the second har parency had been predicted and observed for monic pulses , and the inverted emission band nanosecond duration laser pulses, and our is observed to amplify the fundamental laser earliest goal was to observe either of these pulses. To our knowlodge, this is the first two effects with the recently developed mode- observation of optical gain in color centers. locked neodymium-glass laser (which generates The 4-level model is consistent with all picosecond pulses). Although negative results observations reported in this thesis. An were obtained, an account is given of the attempt was made to measure the F-center search for these coherent optical effects. electron-phonon relaxation time (presumably The inability to measure and control the pulse a few picoseconds or less) but signal fluc shapes and frequency modulation and the tuations (due to fluctuations in the beam scarcity of suitable samples were major ob divergence) were larger than the signal stacles in the search. changes to be measured. A statistical meas urement was precluded because the laser mirror The experimental difficulties encountered damage occurred far too frequently. above led to a study of the generation, dis persion, and detection of frequency-modulated This thesis also describes the laser design picosecond pulses, and to the invention of a and setup. The laser utilizes split-image scheme presented in this thesis for temporally near-field photography in conjunction with compressing "normal" 5 picosecond pulses by internal spatial aperturing for accurate con a two-step process. The pulse is first passed trol of the transverse modes. through an optical Kerr material (such as CS2) and the pulse undergoes self-phase modulation. It is then passed through a dispersive delay line which is adjusted to compensate for the 2. 1971 PUBLICATIONS AND REPORTS frequency modulation. Compression by hundreds is possible. Erwin L. Hahn and Associates Pulse compression experiments performed elsewhere had suggested that the mode-locked Journal neodymium-glass laser pulses were linearly frequency swept. In this thesis it is sug 1. Jackson Koo and Yu-Nian Ifeieh, N»1 4 Nuclear gested that pulses might be more complicated. Quadrupole Coupling Constants in Imidazole, Since publication of these calculated counter examples by the author, new measurements Chem. Phys. Letters 9., 238 (1971). (performed elsewhere] were reported. Although these measurements aid the linear sweep pos sibility, it is pointed out that difficulties UCRL and LBL reports still exist in trying to find a model which fits all the data. 1. R. A. Fisher, J-P. E. Taran, R. Y. Chiao, T. K. Gustafson and P. L. Kelley, Effects on Although the properties of duration, the Propagation of Optical Beams Due to Self- temporal shape, and frequency modulation of Induced Changes in Index of Refraction, To be individual mode-locked neodymium-glass laser published in Proceedings of the 1969 Quantum pulses are quite uncertain, such pulses can Electronics Conf. Chania, Greece (Gordon and saturate broad transitions and, if sufficient Breach, publishers) 2. Robert A. Fisher, Picosecond Optical Pulse Robert A, Fisher (Ph.D. thesis), LBL-111, Nonlinear Propagation Effects, (Ph.D. thesis), Oct. 1971. LBL-111, Oct. 1971. • r V • , II. :-•••'',/-ci "•' '• ••// Q ,1/ Appendices '''f"^: * a- ;•'©.• 191 IMRD STAFF for 1971 Principal Postdoctoral Fellows Investigator and Other Scientists Graduate Students Neil Bartlett *C. Adams C. Kindle R. Mews K. Leary D. McKee B. Morrell T. Richardson Robert Bragg K. Ankra M. Biswal G. Kerhuel Leo Brewer K. Gingerich S. Chang *D. Jensen J. Stiga C. Kittel P. Yee John Clarke S. Freake G. Hoffer J. Paterson M. Rappaport L. Schwartzkopf R. Voss Marvin Cohen *R. Boyd *D. Chadi *J. Joannopolous *S. Sramek *C. Varea de Alvarez *J. Walter Robert Connick *R. ke H. Charles R. Dreyfuss W. Earl J. Neely T. Rowland Lee Donaghey R. Agnihotri *P. Cross *R. Dibble R. Headrick *A. Kay a S. Shaikh John Born John Morris, Jr. K. Murty A. Amin *M. Vaidya D. iaahn C. Liu F. Mohamed C. Syn Participant status deceased 192 Principal Postdoctoral Fe vs Investigator and Other Scientists Graduate Students Richard Fulrath J. Reed G. Bandyopadhyay *J. Shackelford J. Dih R. Holnan J. Holzgraf J. Masaryk B. Powell S. Kisbud J. Shackelford J. Sherohman C. Shumaker C. Young Erwin Hahn *J. Koo *R. Fisher *Y. Hsieh R. Macklin *P. Nordal Robert Hammond M. Collver B. Babu Leo Brewer *I. Charo Earl Parker M. Collver Victor Zackay B. Francis Milton Pickus Garrett Hartsough Hazra Hemachalam Krause R. Loop R. Lott K. Srinivasan T. Tom S. Wason Charles Harris M. Buckley *W. Breiland A. Francis M. Buckley R. Hoover M. Fayer W. Hughes *J. Koeit R. Panos P. Wilday Ronald Herm R. Behrens B. Earl L. Gundel S. Lin L. Loh C. Minis C. Sholeen A. Wong Ralph Hultgren *J. Cases J. Dickson P. Desai H. Yoon *D. Hawkins K. Kelley II. Yoon Participant status 193 Principal Postdoctoral Fellows Investigator and Other Scientists Graduate Students Harold Johnston R. Graham A. Harker D. Martin *G. Screit *J. Valdin *R. Young William Jolly P. Finn R. Dreyfuss *E. Jekel P. Finn H. Kuo J. Illige C. Riddle M. Mendelsohn *E. Staricco K. Strom F. Wang George Jura T. Chen B. Jurado C. Wu A. Yee Bruce Mahan K. Gillen M. Chiang C. Maltz J. Clark W. Dinpfl J. Fair R. Horton A. Werner J. Winn Marshal Merriam William Miller *J. Doll S. Augustin T. George S. Hornstein *W. Hunt C. Slocomb *C. Rankin D. Yuen Rolf Muller *J. Mathieu K. Beach C. Brown J. Cooper H. Gu K. Hellyar *F. McLarnon Rollie Myers J. Chang T. Hynes A. Jindo M. St. John D. Wilson R. Wilson J, Yarnell Participant status 194 Principal Postdoctoral Fellows Investigator and Other Scientists Graduate Students John Newman R. Homsy *V. Marathe C. Mohr K. NisanciogTu P. Sih N. Vahdat H. Yip Donald Olander J. Schwarz R. Acharya W. Siekhaus *C. Cann *J. Hovingh R. Jones V. Kruger A. Machiels *G. Moore *F. Nuh R. Olstad *D. Slaughter H. Tssi Earl Parker J. Ellis S. Friedrich R. Goolsby W. Horwood S. Jin R. Jones R. McCoy M. Perra E. Plaza W. Wcod Joseph Pask *C. Gorin I. Aksay R. Langston C. Dokko C. Hoge W. Snowden C. Tao B. Wong Norman Phillips B. Triplett S. Bader M. Conway T. Hall G. Schwartz Alan Portis D. Gordon K. Small Paul Richards *M. Werner R. Aurbu.1! *J. Mather Y. Taur W. Tennant D. Woody K. Yang Participant status 195 Principal Postdoctoral Fellows Investigator and Other Scientists Graduate Student Gene Rochlin P. Hansma M. Jack D. McBr.ide Song G. Williams Alan Searcy D. Beruto P. Bihuniak *A. Buchler R. Coyle D. Meschi D. Feather L. Fuke R. Galluzzo P. Mohazzabi J. Roberts Y. Ron Shen N. Amor A. Bromberg M. Loy *J. Diels *Y. Petroff S. Kohn *A. Schmidt I. Ku *P. Yu Y. Lin R. Zucca M. Loy J. Morris W. Nevins H. Rosen J. Shelton J. otokes K. Wong Gabor Senior jai *S. Berglund *S. Bernasek *G. Holland J. Gland R. Joyner D. Howlett B. Lang D. Kahn M. Martin «C. Megerle >K. Seikhaus R. Palmeri *L. West F. Szalkowski *J. Wasilczyk L. West Gareth Thomas W. Bell L. Chen *M. Bouchard R. Clark P. Fisher S. Das *M. Fukanachi R. Hall *R. Garwood D. Huang *W. MUller *R. Kovsky P. Okamoto R. Livak *D. Schwartz J. McMahon P. Swarm R. Gsiecki *C. Vercaemer M. Raghavan G. Stone T. Tan 0. Van der Biest Charles Tobias *D. Jacquot R. Acosta J. Selman *W. Blanton «M. Forbes J. Jorne K. Kojima U. Landau Participant status J. Selman 196 Principal Postdoctoral Fellows Investigator and Other Scientists Graduate Students Jack Washburn J. Narayan T. Astrup R. Jeanen Kosel Lira-Olivares Narayan Peterson Roberts Seshan Srivas Cava K. Stolt I. Tiraspolsky W. Wu Victor Zackay W. Gerberich H. Adkins *R. McCoy C, Atherton »M. Williams D. Atteridge A. Baghdasarian D. Bhandarkar J, Birat *S. Dong D. Griffiths W. Harris R. Horn *D. Leary *G. Sasaki C. St. John *R. Towari * Participant Status 197 VISITING SCIENTISTS Dr. A. W. Adamson, Department of Chemistry, University of Southern California, Los Angeles, California Dr. Jean Blaise, Chemistry Division of the National Center of Scientific Research, Orsay, France Dr. Dagfinn Bratland, Institute of Inorganic Chemistry, Technical University of Norway, Trondheim, Norway Professor E. Burstein, Department of Physics, University of Pennsylvania, Philadelphia, Pennsylvania Dr. A. Cassuto, Centre de Cindtique Physique et Chimique, Nancy, France Professor Steven Copley, Department of Metallurgy, University of Southern California, Los Angeles, California Professor Jack Dow, Department of Physics, Princeton University, Princeton, New Jersey Dr. P. Eisenberger, Bell Telephone Laboratory, Murrayhill, New Jersey Dr. Manfred Fink, Department of Physics, University of Texas, Austin, Texas Dr. Gideon Gilat, Professor of Physics at the Technion, Haifa, Israel Dr. John B. Goodenough, Lincoln Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts Professor R. R. Haering,* Department of Physics, Simon Fraser university, Burnaby, British Columbia, Canada Professor U. Heiland, Department of Physics, Technische Hochschule, Aachen, Germany Professor Keith Johnson,* Department of Metallurgy and Materials Science, Massachusetts Institute of Technology, Cambridge, Massachusetts Dr. Lars Kihlborg, Institute of Physical and Inorganic Chemistry, University of Stockholm, Stockholm, Sweden, and Visiting Professor of Chemistry, Arizona State University Dr. A. P. Lane,* Department of Chemistry, University of Clasgow, Glasgow, Scotland Dr. Roger Linford, Berkeley Nuclear Laboratory, Berkeley, England Dr. T. Novakov, Shell Development Company, Emeryville, California Professor D. F. Ollis,* Department of Applied Sciences, Princeton University, Princeton, New Jersey Professor L. D. Schmidt, Department of Chemical Engineering and Materials Science, University of Minnesota Professor Reinhard Schmutzler, Technical University, Braunschweig, West Germany Professor F. G. A. Stone, Department of Inorganic Chemistry, University of Bristol, Bristol, England Professor Daniel Trivitch, Department of Chemistry, Wayne State University, Detroit, Michigan Professor •). J. Turner, Department of Chemistry, Cambridge University, Cambridge, England Speakers 198 DEGREES AWARDED IN 1971 Ph.D. Degrees 1. Kmal E. Amin, Analysis of High-Tenperature Transient Creep of Pure Metals and Alloys, (LBL-163, August). 2. Kirk Watson Beach, Optical Methods for the Study of Convective Mass Transfer Boundary Layers on Extended Electrodes, (UCRL-20324, July). 3. Michael John Buckley, Optically Detected Magnetic Resonance of Molecules in Excited Triplet States, (UCRL-20390, February). 4. J. James Chang, Relaxation Time Measurements in Electron Paramagnetic Resonance, (UCRL-19691, February). 5. Herbert Lamar Charles, Jr., Nuclear Magnetic Resonance Study of the Transverse Relaxation of Oxygen 17 in Aqueous Solutions of Titanous Ion, (UCRL-20366, May). 6. Michael Moore Collver, Superconductivity in Amorphous Transition Metals, (LBL-167, November). 7. Donald R. Cropper, Creep of Lithium Fluoride Single Crystals at Elevated Temperatures, (UCRL-20350, January). 8. Santosh Kumar Das, Short Range Order, (LBL-J76, November). 9. Philip Dow, Tho Hydrogen-Nitric Oxide Spin Exchange Cross Section, (UCRL-20529, March). 10. Robert Michael Direyfuss, Reactions of Potassium Gennyl, (UCRL-20502, January). 11. Patricia Ann Finn, A Study of Diborane Dihydrate slid X-Ray Photoelectron Spectroscopy of Inorganic Compounds, (LBL-415, November). 12. Robert Alan Fisher, Picosecond Optical Pulse Nonlinear Propagation Effects, (LBL-111, October). 13. Theodore M. French, Low Energy Electron Diffraction Studies of Insulating Single Crystal Surfaces, (LBL-155, October). 14. William Warren Gerberich, Crack Growth Mechanisms in Fibrous Composities, (UCRL-20524, February). 15. Roger D. Goolsby, Relationships Between Microstructure and Fracture Toughness in A Secondary Hardening Steel, (LBL-405, November). 16. Larry D. Hartsough, Stability of A15 Type Phases as Affected by Electron Concentration, (LBL-119, September). 17. Robert L. Horton. Thermal Ion-Molecule Reactions in Gaseous Methane-Oxygen Mixtures, (LBL-114, August). 18. Akira Jindo, Single Crystal Studies of Hydrated Transition Metal Ions by Electron Paramagnetic Resonance, (UCRL-20386, January). 19. Russell H. Jones, Laves Phasu Precipitation in Fe-Ta Alloys, (LBL-179, October). 20. Chih-an Liu, Effect of Rhenium Additions on Anisotropic Slip in Single Molybdenum Crystals, (LBL-129, October). 21. Michael Ming-Tak Loy, Self-Focusing and Small-Scale Filaments of Light in Liquids, (UCRL-20585, May). 22. Jagdish Narayan, The Kinetics of Self Diffusion and Dislocation Glide in Magnesium Oxide, (LBL-406, November). 23. James Watson Neely, Oxyger.-17 Nuclear Magnetic Resonance Studies of the First Hydration Sphere of Diamagnetic Metal Ions in Aqueous Solution, (UCRI-20580, June). 24. Hal J. Rosen, Light Scattering From Liquids and Liquic Crystals, (LBL-424, November). 25. Charles Falding St. John, Theoretical and Experimental Studies of Flow and Fracture in Monocrystalline Silicon, (LBL-113, August). 26. Jan Robert Selman, Measurement and Interpretation of limiting Currents, (UCRL-20557, June). 19? 27. Janes Floyd Shackelford, The Solubility of Ga«es in Glass, (UCRL-20399, February). 28. John Sheltori, Phase-Matched Third-Harmonic Generation in Clwlesteric Liquid Crystals, (LBL-112, August). 29. Ping Huei Sih, Mass Transfer to the Res.r of Objects in Low Reynolds Clumber Flow, flOL-20509, toy). 30. Yeong-du Song, Transition from Bulk Behavior to losephson-Junction Like Beliavior in Superconducting Microbridges, (LBL-198, November). 31. Arthur Stephen Werner, Dynamic Isotope Effects in Ion-Molecule Reactions, (UCRL-2036:>, April). 32. Lloyd Albert West, Studies of Gas-Surface Interactions at the (100) Crystal Face 3f Platinum by Molecular Beam Scattering, Low Energy Electron Diffraction, w.d AugeT Electr n Spectroscopy, (LBL-107, November). 33. Che-Kuang Wu, fressuTe (up to 40 kbar) as u Parameter in the Study of Vibrational Exciton Levels and Crystalline Potentials of Molecular Crystals: Naphthalene ami N-Piraffins, (LBL-169, October). 34. Albert Fan Yoe, Heat Capacity of Iron as a Function o.r Pressure) and Temperature, (LBL-180, November). 35. Hong II Yoon, I. Lattice Heat Capacities -jf Alloys at low Temperatures: InPb, AuCd, Mg,Sb., II. Effect of Ordering on Lattice Iteat Capacity; AUCUT, (Lfil-101, August). 36. Ricardo R. L. ?>cca, Wavelength Modulation Spectroscopy of Semiconductors, (UCRL-20361, January). D.Eng. Degrees 1. Richard Harold Jones, Kinetics of the Reaction of Pyrolyt:.c Graphite with Molecular Oxygen Studied by Modulated Molecular Beam Mass Spectrometry, (LBL-104, November). 2. Robert Allyn McCoy, The Resistance of TRIP Steel to Hydrogen Embrittlement, (LBL-135, September). M. .3. Degrees 1. Charles Sheldon Atherton, The Relation of Toughness; to Microstructure in Sphcroidized Steel, (LCRL-20S40, March). 2. B. Naga Prakash Babu, Correlation of Thermo-Mechanically Varied Microstructures with Superconducting Properties of Nb.Sn Processed by a New Powder Metallurgy Technique, (LBL-437, December). 3. Madan Mohan Biswal, Effect of Polarized X-Ray Beam in Small Angle Scattering by Pyrolytic Graphite, (UCRL-20517, March). 4. J. Lester Dickson, Vapor Pressure of Lead and Activity Measurements of Lead-Platinum Alloys by the Torsion Effusion Method, (LBL-116, July). 5. Robert J. Galluzzo, Vaporization Kinetics of Zinc Oxide Single Crystals, (UCRL-20528, April). 6. Terry Nathaniel Garrett, Pulsed Magnetic Fields and the Critical Cu:rent Densities of Superconducting Nb,Sn Strips in These Fields, (LBL-448, December). 7. Hiram Gu, Ellipsometry of Surface Layers, (LBL-I65, December). 8. Subhas Hazra, Synthesis of Law Temperature Stable A15 Compounds in the Niobium System, (LBL-433, December). 9. Xenneth G. Hellyar, Cinematic Studies of the Anodic Dissolution of Metals at High Current Densities, (LBL-157, September). 10. Carl E. Hoge, The Effect of Atmosphere on Interfacial Reactions and Wetting Behavior of Glass-Metal Systems, (UCRL-20334, February). 11. Daniel Leonard Hewlett, Vacuum Vaporization Studies of Lithium Fluoride Single Crystals, (UCRL-20714, June). 12. Sungho Jin, Spheroidization of the Grain Boundary Precipitates in Fe-Ta-Cr Alloys, (LBL-443, December). 200 13. Guillaume Robert Kerhuel, Brittle Fracture in Filament-Reinforced Materials, (LBL-138, September). 14. Richard Bruce loop, The Structure of Vapor Depositee' ID Transition Metals, (LBL-177, September). 15. Barbara Kuc Morrel], Structural Studies of Xenon Fluoride Complexes, (1.BI.-1S6, September). 16. Richard Alan Osiecki, Interfacial Defects in allium Arsenide Devices, (UCRL-Z0591, June). 17. Bob R. Powell, Jr., Processing of Sodium-Potassium Niobate Ceramics, (T'BL-143, December). 18. Subhash Risbud, Influence of Surface Damage on the Strength of a Brittle-Matrix Composite, (LBL-137, September). 19. John William Sherohman, Jr., Sintering of High-Purity Subnicron Lead Titanate Containing Niobium, (UCRL-20S96, June). 20. William E. Snowden, High Temperature Mechanical Behavior of MgO-CaOMgO-SiO. Materials, (UCR1-20S67, June). 21. Glen A. Stone, On the Deformation Induced Martcnsite Transformation in Fc-Ni and Fe-Ni-Cr Single Crystals, (LBL-134, September). 22. Thomas Tom, Application of Powder Metallurgy Process in the Formation of NhJU Superconductor, (LBL-188, December). 23. John II. Turney, Profiles of Supermeniscus Electrolyte Films or Partially Submerged Gas Electrodes, (LBL-171, August). 24. David A. Yuen, 1. Optically Detected Nuclear Quadrupole Resonance and Transferred Hyperfine Coupling in Molecular Crystals, II. Electron-Atomic Hydrogen Scattering: Two Approaches, (UCRL-20702, June). 201 1971 RESEARCH SEMINARS Date Title Speaker 1/20 Crystallographic Anisotropics in Chemisorption Dr. L. D. Schmidt, Department of Chemical Engineering and Materials Science, University of Minnesota 2/2 Origin of Freckles in Castings Dr. Steven Copley, Associate Professor of Materials Science, University of Southern California, Los Angeles, California 2/24 Recent Studies of Clean Semiconductor Surfaces Prof. G. Heiland, Department of Physics, Technische Hochschule, Aachen, Germany 3/12 How Useful is Potential Theory to on Gases Dr. Manfred Fink, (A Comparison of Theory and Experiment) Department of Physics, University of Texas, Austin, Texas 5/24 Photoelectron Spectroscopy and its Application to the Dr. T. Novakov, Shell Study of Solid Materials Development Company, Emeryville, California 6/2 Polarized X-Ray Beams in Small Angle Scattering Madan Biswal (GS)* 6/2 Nuclear Magnetic Resonance Study of the Transverse Herbert Charles (GS)* Relaxation of Oxygen-17 in Aqueous Solution of Titanous Ion 6/2 Electron Microscope Studies of Vapor Deposited Richard Loop (GS)* 4D Transition Metals 6/22 Structure and Stoichiometry of Transition Metal Oxides Dr. Lars Kihlborg, Institute of Physical and Inorganic Chemistry, University of Stockholm and Visiting Professor of Chemistry, Arizona State University 7/21 Generation of Far Infrared Radiation by Picosecond Light Pulses in LiN, 0„ K. H. Yang (GS)* 7/21 Kinetics of the Reaction of Pyrolytic Graphite with Molecular Oxygen Studied by Modulated Molecular Beam Richard H. Jones (GS)* Mass Spectrometry 8/25 The Phonon Maser Professor R. R. Haering, Department of Physics, Simon Fraser University, Burnaby, B. C., Canada 8/26 Cinematic Studies of the Anodic Dissolution of Metals at High Current Densities Kenneth G. Hellyar (GS)* 8/26 Substrate Selectivity in Dehydrocyclization Reactivity + on Platinum Surfaces Richard Joyner (PD) 8/26 Experimental and Theoretical Studies of Flow and Fracture in Monocrystalline Silicon Charles St. John (GS)* 202 Date Title Speaker 10/19 CO and CL Coadsorption on Rhenium Dr. A. Cassuto, Centre de Cine*tique Physique et Chimique, Nancy, France 10/27 New Computational Approach to the Theory of the Dr. K. H. Johnson, Chemical Bond Department of Metallurgy and Materials Science, Massachusetts Institute of Technology 11/17 Auger Electron Spectroscopy Investigations of the Prof. G. A. Somoriai and Surfaces of Vanadium and its Oxides F. Szalkowski (GS)* 11/23 Investigations on Electrolytic Deposition of Dagfinn Bratland, Y-Al and Y-Mg Alloys Institute of Inorganic Chemistry Technical University of Norway, Trondheim, Norway 12/8 Thermodynamics of Binary Alloy Catalyst Particles Professor D. F. Ollis, Department of Applied Sciences, Princeton University, Princeton, New Jersey 12/10 Adhesion Between Clean Metal Surfaces Roger Linford, Berkeley Nuclear Laboratory, Berkeley, England 12/17 Polarized Raman Infrared Studies on Single Crystals Dr. A. P. Lane, of Some Halogenometallates Department of Chemistry, University of Glasgow, Scotland (GS) - graduate student '"(PD) = postdoctoral