Diffusion of 57Co in Surface Layers of Nickel Oxide and Alumina

Diffusion of ⁵⁷Co isotope on NiO (110) and AI₂O₃ (0001) surfaces was investigated by using the edge source method. The surface diffusion parameter α D ,Δ, where α is the segregation factor, D , the surface diffusion coefficient, and Δ the thickness of the high-diffusivity layer, was determined for the temperature range 750°–1200°C. For calculation of experimental results, the Whipple solution was used. The Arrhenius plot shows a break at ∽900°C for the surface diffusion of ⁵⁷Co isotope on AI₂O₃ (0001) plane. Above this temperature, vapor transport seems to be the overriding diffusion mechanism. Below this temperature, ionic transport predominates. The apparent activation energy for the ionic transport was calculated to be 120 ± 12 kj/mol.
Ionic transport predominated in the surface diffusion of ⁵⁷Co on NiO over the entire investigated temperature range. This can be explained by the weak bond between Co-vapor species and the NiO surface. The results obtained suggest that the surface diffusion of Co²⁺ ion on NiO at 750°C is ∼7 orders and at 1200°C ∼5 orders of magnitude faster than volume diffusion. Activation energies are 139 and 227 kJ/mol, for surface and volume diffusion, respectively.     

Diffusion of 51Cr in Surface Layers of Magnesia, Alumina, and Spinel

Diffusion of ⁵¹Cr isotope on MgO (100), Al₂O₃ (0001, and MgAl₂O₄ (111) surfaces was investigated by using the edgesource method. The surface diffusion parameter, αD,δ, where α is the segregation factor, D , the surface diffusion coefficient, and δ the thickness of the high-diffusivity layer, was determined for the temperature region 650° to 1250°C. For calculation of experimental results Whipple's solution was used. Arrhenius plots show a break at ∼1050°C for MgO and at ∼900°C for MgAl₂O₄. Above these temperatures vapor transport seems to be the overriding diffusion mechanism. Below these temperatures ionic transport predominates. Ionic transport is the predominant mechanism for surface diffusion of ⁵¹Cr on Al₂0₃ over the entire investigated temperature region. This can be explained by the weak bond between Cr-vapor species and Al₂0₃ surface. The apparent activation energies for ionic transport of 51Cr are 110 ± 12, 121 ± 12, and 119 ± 12 kJ/mol on MgO, Al₂O₃, and MgAl₂O₄ surfaces, respectively. They include enthalpy of motion and binding enthalpy and are 2.5 to 2.8 times smaller than the activation energies for volume diffusion. Investigations of surfaces by LEED, Auger, and SIMS indicated structural nonperiodicity and surface segregation of impurities.     

Point Defects and Surface Diffusion in Iron Oxide and Nickel Oxide

Diffusion of the ⁵⁷Co isotope on Fe₃O₄ (110) and NiO (100) surfaces was investigated by the edge-source method. The surface diffusion parameter, α D _sδ, where α is the segregation factor, D _s the surface diffusion coefficient, and δ the thickness of the high-diffusivity layer, was determined at 750°C for different partial pressures of oxygen. In both cases point defects strongly influenced surface diffusion. For Fe₃O₄ the vacancy mechanism is dominant at high oxygen activities and interstitial (or interstitialcy) mechanism dominates at low oxygen activities. For NiO the surface diffusion at low oxygen activities is influenced by aliovalent impurities and at high oxygen activities the strong influence of the intrinsic defects, nickel vacancies, on surface diffusion was observed. It was concluded that similar mechanisms operate during surface diffusion in the near surface layer and during diffusion in the lattice.     

Grain Boundary Diffusion of Magnesium in Zirconia

This paper reports the transport kinetics of Mg in cubic yttria-stabilized zirconia (containing 10% mol of Y₂O₃ (10YSZ)) involving the bulk and the grain boundary diffusion coefficients. The diffusion-controlled concentration profiles of Mg were determined using secondary ion mass spectrometry (SIMS) in the range 1073–1273 K. The determined bulk diffusion coefficient and the grain boundary diffusion product may be expressed as the following functions of temperature, respectively: D = 5.7 exp[(−390 kJ/mol)/ RT ] cm²·s⁻¹ and D 'αδ= 3.2 × 10⁻¹⁵ exp[(−121 kJ/mol)/ RT ] cm³·s⁻¹, where α is the segregation enrichment factor and δ is the boundary layer thickness. The grain boundary enhancement factor decreases with temperature from 10⁵ at 1073 K to 10³ at 1273 K.     

Grain-Boundary Diffusion of 60Co and 51Cr in NiO

Grain-boundary diffusion of ⁶⁰Co and ⁵¹Cr in bicrystals and polycrystals of NiO was measured at temperatures <0.6 T_m at various equilibrium oxygen partial pressures, using a tracer-sectioning technique. Values of D 'δ were obtained using the Whipple and Suzuoka analyses. The results show that, for both ⁶⁰Co and ⁵¹Cr, grain-boundary-enhanced diffusion is observed and the temperature dependence of D 'δ is about the same as that of the respective bulk diffusivity,/). The effects of equilibrium oxygen partial pressure and the type of gram boundary on D 'δ are not detectable within experimental error.     

Thermal Expansion of the Hexagonal (6H) Polytype of Silicon Carbide

The thermal expansion of the hexagonal (6H) polytype of α-SiC was measured from 20° to 1000°C by the X-ray diffraction technique. The principal axial coefficients of thermal expansion were determined and can be expressed for that temperature range by second-order polynomials: α¹¹= 3.27 × 10^–6+ 3.25 × 10^–9T – 1.36 × 10^–12 T ² (1/°C), and ş₃₃= 3.18 × 10^–6+ 2.48 × 10^–9 T – 8.51 × 10^–13 T ² (1/°C). The σ₁₁ is larger than α₃₃ over the entire temperature range while the thermal expansion anisotropy, the δş value, increases continuously with increasing temperature from about 0.1 × 10^–6/°C at room temperature to 0.4 × 10^–6/°C at 1000°C. The thermal expansion and thermal expansion anisotropy are compared with previously published results for the (6H) polytype and are discussed relative to the structure.     

Grain-Boundary Diffusion of Cr in Pure and Y-Doped Alumina

The diffusive transport of chromium in both pure and Y-doped fine-grained alumina has been investigated over the temperature range 1250°–1650°C. From a quantitative assessment of the chromium diffusion profile in alumina, as obtained from electron microprobe analysis, it was found that yttrium doping retards cation diffusion in the grain-boundary regime by over an order of magnitude. The Arrhenius equations for the undoped and Y-doped samples were determined to be: δ D _b=(4.77±0.24) × 10⁻⁷ exp (−264.78±47.68 (kJ/mol)/RT)(cm³/s) and δD_b=(6.87±0.18) × 10⁻⁸ exp (−284.91±42.57 (kJ/mol)/RT)(cm³/s), respectively. Finally, to elucidate the mechanism for this retardation, the impact of yttrium doping on diffusion activation energies and prefactors was examined.     

Grain-Boundary Diffusion of 51Cr in MgO and Cr-Doped MgO

The grain-boundary diffusion product, D'δ , of ⁵¹Cr in MgO and Cr-doped MgO as a function of grain-boundary orientation and point-defect concentration was determined at T =1200° to 1450°C. A large degree of anisotropy was found in the grain-boundary diffusion behavior in MgO. The ratio of D'δ_|| parallel to D'δ_‖ perpendicular to the growth direction, D'_||/D'_‖ , is 10² for a 5° (100) tilt boundary, decreased to ∼2 in boundaries with tilt angles > 10°. The decrease in D'_||/D'_‖ is due to a large increase in D'_‖ with increasing tilt angle. The results indicate that grain-boundary diffusion in MgO is connected to the orientation of dislocations and the mechanism is one of dislocation pipe diffusion. The grain-boundary diffusion product D'δ increases with increasing Cr concentration in MgO and is ∼4 times larger for MgO containing 0.56 at. % Cr than for the undoped MgO. For all bicrystals studied, the activation energies are within 180 ± 20 kJ/mol which is 60% of the activation energy for ⁵¹Cr diffusion in undoped MgO.     

Deviation from Stoichiometry in Cr2O3 at High Oxygen Partial Pressures

The deviation from stoichiometry, δ, in Cr₂−δO₃ was measured by a tensivolumetric method in the high pO₂ range of ≊10⁴ to 10⁴ Pa at 1100°C. The value of δ, or chromium vacancy concentration, was≊9×10⁻⁵ mol/mol Cr₂O₃ in air for Cr₂O₃ with 99.999% purity. The chemical diffusion coefficient, DT, determined from equilibration data was ≊4.6× cm²·s⁻¹ at 1100°C for pO₂= 2.2 ×10¹ Pa. The self-diffusion coefficient of Cr ions was calculated from and δ and found to be≊1.6×10-17 cm²-s⁻¹, in good agreement with recently measured values.     

Carbon Thermal Diffusion in the UC2-C System

Alpha uranium dicarbide disks coated with pyrolytic carbon migrate, relative to the carbon, up a temperature gradient, leaving behind a layer of rejected graphite. A model for the migration process based on solution of carbon at the hot UC₂-C interface, transport across the UC₂ phase by thermal diffusion, and rejection as graphite at the cool UC₂-C interface was developed. The rate of migration may be described by: The apparent activation energy for migration is 76.5±9 kcal/mol, compared to a literature value for carbon self-diffusion in _α-UC₂ of 90.2 kcal/mol. This difference in temperature dependence may be considered preliminary evidence that the heat of transport decreases with increasing temperature. The average value of the heat of transport, calculated using literature values for carbon self-diffusion in _α-UC₂, is 92±27 kcal/mol.     

Oxygen Tracer Diffusion in Lanthanum-Doped Single-Crystal Strontium Titanate

Strontium titanate (SrTiO₃) is known as a good high-temperature resistive oxygen sensor material; its response time depends on oxygen bulk diffusion and surface exchange processes. In the present work, ¹⁸O diffusion has been investigated in lanthanum-doped SrTiO₃, single crystals in the temperature range 700° to 900°C by secondary ion mass spectrometry (SIMS). Oxygen tracer diffusivities between 2 × 10⁻¹⁵ and 1 × 10⁻¹³ cm²/s have been calculated from the SIMS results. Low surface enrichment of ¹⁸O compared to the ¹⁸O concentration in the gas atmosphere gives clear evidence for a surface exchange reaction.     

Knudsen Cell Studies of the Vaporization of Samarium Dicarbide

The vapor pressure of SmC₂ in equilibrium with graphite was measured by the Mnudsen effusion technique. Rates of weight loss from the cells were measured with an automatic recording balance. The apparent pressures varied with orifice size, and equilibrium pressures were calculated by extrapolation to zero orifice area. This work was combined with other studies to obtain log₁₀ P(atm) = - 13.869 × 10³/T + 3.752 (1300°-2050°K) for the Sm vapor pressure above SmC₂-C. Estimates of S°₂₉₈ and c_p were made for SmC₂, and δH°₂₉₈ was calculated to be 72.0 ± 2 kcal/mol for the reaction SmC₂(s) = Sm(g) + 2C(s). This value combined with δH°_{v, 298}= 48.6 kcal/mol for Sm gives a δ°_f298 for SmC₂ of 23.4 ± 2 kcal/mol.     

Ion-Probe Measurement of Oxygen Self-Diffusion in Single-Crystal Al2O3

Anion self-diffusion coefficients normal to (1102) were obtained for single-crystal Al₂O₃ in a 1.3 × 10 ³ N/m² (10⁻⁵ torr) vacuum at 1585° to 1840°C. Tracer was supplied from an initial 650 to 1300 A Al₂¹⁸O₃ layer produced by the oxidation of vapor-deposited Al metal films in an ¹⁸O₂ atmosphere at 520°C. Concentration gradients extended over depths of 3000 to 5000 A and were measured by mass spectrometry of material sputtered from the samples with a beam of Ar⁺ ions. Crystals which had not been preannealed to remove surface damage displayed enhanced diffusion. Diffusion coefficients from preannealed crystals may be described by D₀ =6.4×10⁵cm²/s, with an activation energy of 188 ± 7 kcal/mol. The diffusion is interpreted as an extrinsic vacancy mechanism.     

Stability of Bi2Sr2Ca1Cu2O8 ±δ Thick Films at Elevated Oxygen Pressures and Temperatures

Bi₂Sr₂Ca₂Cu₂O_8±δ-type compound thick films were exposed to oxygen-argon-gas mixtures (1% to 20% oxygen gas) at elevated pressures (up to 207 MPa) and temperatures (500° to 940°C) for times ranging from 5 to 96 h. At a sufficiently high oxygen fugacity and temperature, Bi₂Sr₂Ca₁Cu₂O_8±δ decomposed via a solid-state reaction. Room-temperature X-ray diffractometry and electron probe microanalysis of decomposed films revealed the presence of Bi₂(Sr,Ca)₂-Cu₁O_6±θ ro-type compound, Bi₂Sr₂,Ca₁O_8±δ-type compound, and CuO. Bi₂Sr₂Ca₁Cu₂O_8±δ decomposition was accompanied by a modest weight gain, which was consistent with an oxidation reaction. The solid-state decomposition reaction could be reversed by heat treatment of decomposed films at 860°C in pure, flowing oxygen at ambient pressure.     

Synthesis, Characterization, and Consolidation of Si3N4 Obtained from Ammonolysis of SiCl4

Thermal decomposition of silicon diimide, Si(NH)₂, in vacuum resulted in very-high-purity, fine-particle-size, amorphous Si₃N₄ powders. The amorphous powder was isothermally aged at 50° to 100° intervals from 1000° to 1500°C for phase identification. Examination of ir spectra and X-ray diffraction patterns indicated a slow and gradual transition from an amorphous material to a crystalline α-phase occurring at 1200°C for >4 h and/or 1300° to 1400°C for 2 h. As the temperature was increased to ≥1450°C for 2 h, the crystalline β-phase was observed. Phase nucleation and crystallite morphology in this system were studied by electron microscopy and electron diffraction combined with TG as functions of temperature for the inorganic polymer starting materials. Powders prepared in this manner with 4 wt% Mg₃N₂ added as a sintering aid were hot-pressed to high-density fine-grained bodies with uniform microstructures. The optimum hot-pressing condition was 1650°C for 1 h. Silicon concentration steadily increased as the hot-pressing temperature or time was increased. A method for chemical etching for high-density fine-grained Si₃N₄ is described. Electrical measurements between room temperature and ∼500°C indicated dielectric constant and tan δ values of 8.3±0.03 and 0.65±0.05×10⁻², respectively.     

The 1300°C Isothermal Section in the Ti–In–C Ternary Phase Diagram

In this paper we map out the 1300°C isothermal section in the Ti–In–C ternary system. Two ternary compounds exist: Ti₂InC_α and Ti₃InC_δ. At 1300°C TiC _x is in equilibrium with all phases, Ti₃InC_δ is in equilibrium with all the phases except C, and Ti₂InC is in equilibrium with all phases except Ti and C. The range of δ in Ti₃InC_δ varies from 0.95 to 0.8. A correlation was found between δ and the lattice parameters of this phase. The maximum solubilities of In and C in Ti are 14 and 4 at.%, respectively. Similarly, α in Ti₂InC_α varies from ≈0.85 to 1. The dissolution of ∼4±0.3 at.% In in TiC _x reduces the C concentration to ≈24 at.%. In the In-rich corner, a liquid region exists.     

Vaporization of Lead Zirconate-Lead Titanate Materials

A thermogravimetric investigation of the vaporization of cold-pressed Pb(Zr_0.65Ti_0.35)O₃ in vacuum from 690° to 1130°C shows that the vaporization occurs in two steps. An initial loss of 1 to 7 wt% proceeds by a mechanism with a logarithmic time dependence and represents the vaporization of the unreacted PbO that remains after the initial calcining. The second step proceeds by a slower, diffusion-controlled mechanism with a parabolic time dependence. It is shown that this rate-determining step is bulk diffusion across a thickening lead-depleted layer in the cold-pressed pellet, despite pore volumes of 24 to 42%. The rate constant, g/cm² sec½, for the parabolic weight loss process is expressed by: log K = (3.37 ± 0.14) - (8.46 ± 0.16) (10³/ T _K). The activation energy is 38.7 ± 2.0 kcal/mole.     

Control of Dielectric Constant and Loss in Alumina Ceramics

The dissipation factors and dielectric constants of alumina ceramics containing less than 100 ppm of impurities and of specimens doped with Si, Ti, Ca, Mg, Fe, and Cr ions were measured in the region 10² to 8.5 × 10⁹ cps and 25° to 875° C. Multiple regression analysis of the data at 500° C and 10⁶ cps showed a linear relation between impurity concentration and tan δ, with a correlation coefficient of 0.93. Si ions caused the greatest rise of tan δ, Mg and Ti were second, Ca third, and Cr and Fe had no significant influence. These effects diminished with rising frequency and became negligible in the microwave region. Activation energies of conduction for pure and doped alumina were estimated from measurements of δ at 10⁵ cps and 500°C. Values between 1.2 and 1.6 ev were calculated for all compositions except the one containing Mg²⁺, for which 2.0 ev was obtained. At low frequencies, the dielectric constant (k') rose exponentially with temperature, reflecting a similar rise in the number of free charge carriers contributing to interfacial polarization. At higher frequencies the temperature variation of k' fell to a shallow positive slope of about 120 ppm per °C. This coefficient was not influenced by low concentrations of impurities but could be effectively compensated without excessive loss by additions of 10 to 20% SrTi0₃.     

Temperature Dependence of Lattice Parameters and Anisotropic Thermal Expansion of Bismuth Oxide

Temperature dependence of lattice parameters of bismuth sesquioxide (Bi₂O₃) has been obtained between 26° and 778°C by the Rietveld method using neutron powder diffraction data. Lattice parameters     and unit-cell volume of α and δ phases increased, while the β angle of α phase decreased with an increase of temperature. Here the     denotes the lattice parameter a of the α phase on the basis of pseudo-fluorite lattice. The thermal expansion coefficients of α phase were 26.7, 6.6, and 9.0 (× 10⁻⁶°C⁻¹) for     and     axes, respectively, indicating a large anisotropy. The α- to -δ phase transition of Bi₂O₃ was confirmed between 721° and 760°C on heating. At the α–δ transition point, the lattice parameters and unit-cell volume discontinuously changed, indicating that the transition is of the first order.     

Optical Absorption Study of the Reduction Kinetics of Rutile Crystals

The early stages of the reduction kinetics of single-crystal rutile, TiO₂, were studied by optical absorption techniques for specimens re duced in the range 550° to 760°C at a constant vacuum level of 10⁻³ mm Hg. When reduced to a room-temperature resistivity value of ∼10³ ohm-cm, rutile shows a blue coloration as a re sult of a broad infrared absorption band centered at a wavelength of 1.2μ. The kinetics of growth of the optical absorption coefficient at 1.2μ fit a parabolic relation of the form (Δα)²= K't. The Arrhenius rate constant, K' , exhibited an activation energy of 111 ± 9 kcal/mole which is ∼50% too high for the diffusion of oxygen in rutile. It is postulated that the defect center is likely to be a Ti interstitial and that the kinetics reflect the self-diffusion of Ti as the rate-con trolling mechanism.