Phase stability in the calcium carbonate system was investigated as a simultaneous function of pressure and temperature up to 40 kbar and several hundred degrees Kelvin using micro-Raman techniques to interrogate samples constrained within a resistively heated diamond anvil cell. Measured spectra allow unequivocal identification of crystalline phases and are used to refine the P, T phase diagram. Calcium carbonate was found to exhibit both reversible and irreversible transformation phenomena among the four known phases which exist under these conditions. Time-dependent Raman intensity variations as the material is perturbed from its equilibrium state allow real-time kinetics measurements to be performed. Evidence suggests that the order of certain observed transformations may be pressure dependent. The utility of Raman spectroscopy to follow transformation phenomena and to estimate fundamental thermophysical properties from the stress dependence of vibrational mode frequencies is demonstrated.

Time resolved X-ray powder diffraction was used to follow the crystallization processes in the systems MgO-MgCl2-H2O and MgO-H2O. A Stoe X-ray powder diffractometer with a position sensitive detector was used. The crystalline reaction products are [Mg2(OH)3(H2O)2]C1.H2O and Mg(OH)2 dependent upon the MgC12 concentrations in the heterogeneous systems. Mg(OH)2 was the product in a 1.50 M MgCl2 solution and [Mg2(OH)3(H2O)3]Cl.H2O was formed in 3.00 and 5.17 M MgCl2 solutions at 20°C.

One method for the fabrication of the superconducting compound Nb3Sn involves interdiffusion of a surface coating of Sn alloyed with Cu on Nb containing Zr and O. In this study, the kinetics and microstructure associated with this reaction have been studied in detail. The results show that small Nb3Sn grains nucleate at the Nb3Sn/Nb interface, and that the Nb3Sn grains experience grain growth immediately after they are formed. ZrO2 precipitates are observed in the Nb3Sn at the Nb3Sn/Nb interface and throughout the Nb3Sn. The ZrO2 precipitates occur in the form of small partially-coherent spheres in the Nb3Sn. No ZrO2 precipitates are observed by TEM in the unreacted Nb. The grain boundaries in the Nb3Sn region are coated with a Sn-Nb-Cu alloy which would have been liquid at the diffusion/reaction temperature. The thickness of the Nb3Sn reaction layer formed during the isothermal diffusion anneal is proportional to time to the first power, indicating “reaction”-controlled rather than “diffusion”-controlled kinetics. The absence of diffusion-controlled kinetics can be explained by the presence of the liquid coating on the Nb3Sn grains. Diffusion of Sn in this liquid layer is apparently fast enough to not be the limiting kinetic step.

316 L stainless steel (nominally 18% Cr, 8% Ni, 2% Mo, 1% Mn, balance Fe) has been ion nitrided in a N2/H2 atmosphere. Studies of nitride layer depth, structure, composition and microhardness show uniform layer formation with a sharp layer-to-matrix interface and a hardness increase over the bulk value by 600%. The structures of the nitride layers vary with treatment temperature, falling into two regimes over the temperature range studied. Generally, parabolic layer growth was found, and the growth coefficients were determined at five temperatures.

We have used Rutherford backscattering spectroscopy (RBS) to study the kinetics of NiSi growth during low temperature annealing of a Ni(Cr) layer on Si. During the silicide growth, a Cr-rich amorphous layer was formed between the silicide and the Ni(Cr) layers. This amorphous layer produces a slow and uniform silicide growth rate; the NiSi thickness was 60 nm after a 2h anneal at 475°C. After prolonged annealing, the amorphous layer recrystallized, causing very rapid silicide growth. When a thin (0.5 nm) layer of Pt was deposited between the Si and Ni(Cr) layers, the amorphous layer was thinner, the silicide growth rate was correspondingly greater and the rapid growth stage occurred earlier. The RBS data agreed qualitatively with transmission electron microscopy results for the growth of both the amorphous layer and the silicide.

X-ray diffraction and transmission electron microscopy have been used to study the kinetics of phase transformations and the structure of Pd/a-Si, Pd/a-Ge and Pd/a-GeSi thin films deposited on Si substrates. Different kinds of amorphouis structures were used: a-Si:H:D, a-Si.:F, a-Ge:H:D, and a-GeSi:H:D. The first stage of phase transformation during heat treatment was palladium silicide (Pd2Si) and palladium germanide (Pd2Ge) formation at temperatures above 200°C. Annealing studies demonstrated that the presence of F in a-Si promotes the Pd2Si formation. The study of the Pd2Si crystallization process showed that: a) when the Pd layer and the a-Si layer are thin, then c-PdSi grows in a fractal-]ike form; b) when the Pd and a-Si both are thick, then c-Pd2Si grows in a globular structure; c) in both above mentioned cases a well-oriented [0011 texture forms. The growth of the silicide and germanide layers in the temperature range of 200-300°C was found to be controlled by a diffusion limited process. It was found that c-Pd2Ge transforms to c-PdGe above 200°C. The a-Ge,.,Si,. 5 alloy behaved similarly to a-Si forming only [001] textured c-Pd2(Ge,Si).

A scenario has been developed combining the glass transition (liquid to glass) with the crystal-to glass transition. If the “melting” point of a crystalline solid solution is reduced to the ideal glass transition temperature a triple point is predicted between crystal, liquid and glass. Based on extrapolations of measured specific heat data of undercooled liquid glass-forming Au- Pb-Sb alloys the excess entropy is found to vanish close to the glass transition. On the other hand, the amorphization reaction of crystalline Fe2Er-hydrides is characterized by a lambda-type anomaly in the specific heat. The logarithmic temperature-dependence of the specific heat results from local fluctuations in the crystalline phase, rather than thermally activated lattice defects. These results suggest that glass formation from the liquid as well as the crystalline state is characterized by an underlying instability.

Melting velocities in silicon were measured on a picosecond time scale by employing pump-probe and self-reflectivity measurements. The pump laser was a 40 picosecond XeCl excimer (308 am) and the probe a dye laser at 600 am. Samples were phosphorous doped silicon wafers. Average melt-in velocities as high as 800 m/s were observed. Best fit to the data was achieved with an asymmetric transient state theory model.

The congruent melting point, or To curve, of crystalline Si-As alloys has been measured in the range of 1.6 to 18.1 at. % arsenic by line source electron beam annealing. Alloys were created by ion implantation of As into 0.1mm Si-on-sapphire and crystallized by pulsed laser melting. To temperatures decrease from 1673±10K at 2.0 at.% As to 1516±30K at 18.1 at.% As. The results of these measurements are significantly higher than the previous results of studies using pulsed laser melting techniques. Advantages of the e-beam technique over previous techniques are discussed. Chemical free energy functions of the solid and liquid phases were calculated from existing thermodynamic data. The calculated To curve agrees with the measured values only in low concentration region (less than 8 at.%).

A transmission electron microscopy study has revealed that the banded microstructure produced in Cu-61.7 at. pct. Ag after surface laser melting and resolidification is not due to a cellular breakdown of the interface as previously thought, but consists of alternating regions of the extended metastable solid solution, y, and a coupled growth structure containing thin plates of y and the copper-rich phase β'. The spacing of this coupled growth structure is approximately 100 Å, which is less than half the minimum spacing yet observed for coupled growth from the melt in Ag-Cu.

In order to explain the banded microstructure a theoretical model has been developed based on a finite elements solution of the diffusion equation in the melt, the interface response functions for continuous growth derived by Aziz and Kaplan and a recently developed theory for coupled growth that includes far-from-equilibrium regimes by Yankov et al..

The structure of rapidly quenched Nb100-x-yPdxGey alloys has been investigated using x-ray diffraction. Niobium concentrations were varied between 100 and 45 at.% the remainder at each Nb concentration was composed of Pd and up to y = 15 at.% Ge. Germanium was found to suppress the nucleation rate of the fcc α-NbPd phase relative to that of the bcc α-Nb phase, thereby extending the single-phase bcc solubility range by ≈ 2 at.% Nb. High Ge content (y > 6) also induced quenching of the amorphous phase. These results can be understood from the standpoint of classical nucleation theory and from a consideration of the polymorphic phase diagram of Nb-Pd. The two approaches are consistent with Ge addition depressing the To line of the fcc phase more rapidly than it depresses the To line of the bcc phase.

Experimental observations of the melt/crystal interface in the directional solidification of a thin sample of a binary alloy are reported for conditions only slightly beyond those for the onset of morphological instability. Spectral analysis for long times shows dynamics leading to a band of most probable wavelengths with sharply selected peaks at wavelengths significantly below the most dangerous wavelength predicted by linear stability theory. These large-amplitude cells develop first in packets that spread to fill the interface.

At the solidification velocities observed during pulsed laser annealing, the planar interface between solid and liquid is stabilized by capillarity and nonequilibrium effects such as solute trapping. We used Rutherford backscattering and electron microscopy to determine the nonequilibrium partition coefficient and critical concentration for breakdown of the planar interface as a function of interface velocity for Sn-implanted silicon. This allows us to test the applicability of the Mullins- Sekerka stability theory to interfaces not in local equilibrium and to test the Coriell-Sekerka and other theories for oscillatory instabilities.

Partitioning during rapid solidification of dilute Al-Ge alloys has been investigated. Implanted thin films of Al have been pulsed-laser melted to obtain solidification at velocities in the range of 0.01 m/s to 3.3 m/s, as measured by the transient conductance technique. Previous and subsequent Rutherford Backscattering depth profiling of the Ge solute in the Al alloys has been used to determine the nonequilibrium partition coefficient k. A significant degree of lateral film growth during solidification confines determination of k to the placing of an upper bound of 0.22 on k for solidification velocities in this range. We place a lower limit of 10m/s on the “diffusive velocity,” which locates the transition from solute paritioning to solute trapping in the Continuous Growth Model.

The melting of a polycrystalline iron surface with short laser pulses induces a surface structure with an orientation and a morphology depending on the crystal face of the surface. Different crystal planes irradiated with the same laser pulses show different degrees of disorder. This points to a crystal face dependent surface melting temperature.

A three-dimensional Monte Carlo computer program is used to study the time-dependent coarsening of a precipitate population in a solid. The emphasis is on the influence of the escape rate of solute atoms from a precipitate on the coarsening behavior. Three models for the escape attempt frequency are compared. The time when coarsening is observed and the extent of coarsening depend on the model as does the percentage of solute precipitated in a given time. The qualitative results of the present calculations are compared with earlier findings.

The precipitation kinetics of an ordered intermetallic from a disordered matrix, which involves simultaneous ordering and decomposition, is studied by a computer simulation technique based on the microscopic diffusion theory. It is found that the precipitation starts from a congruent ordering transition, which may be continuous or nucleation and growth. This congruent ordering transition transforms the initially disordered state into a single phase nonstoichiometric ordered state with antiphase domains. The next stage is the decomposition which starts from the antiphase domain boundaries and then propagates into the ordered domains. And the final process is the coarsening of the order/disorder two-phase mixture. The predicted kinetics of precipitation is in excellent agreement with recent experimental observations in important alloy systems.

An approximate late time solution to the dynamics of phase separation for a nonconserved ordering order parameter (ø) coupled to a stable conserved field (c) is presented. In the Halperin Hohenberg(1) classification scheme this model is known as Model C with a symmetric coupling between nonconserved and conserved fields. The different time dependences of long (i.e., domain size lengths ∼ power law in time) and short wavelength (i.e., interfacial lengths ∼ exponential decay in time) fluctuations imply a simple relationship between the two fields. In essence ø controls the growth of the long wavelength fluctuations, and c modifies the interfacial profile. Asymptotically the dynamic structure factor (Sø(k,t)≡<Ø(k,t)Ø*(k,t)>) for the nonconserved field is shown to scale in the form Sø(k,t) = tdnfø(ktn), with n = 1/2. Similarly the structure factor for the conserved field (Sc(k,t)) is shown to obey the scaling law Sc(k,t) = tdn−1fc(ktn), with n = 1/2. Explicit expressions for the scaling functions fc(z) and fø(z) are presented for arbitrary dimension. These predictions can be tested through scattering experiments.

Rate equations for thermally activated atomic exchange were used to calculate chemical diffusion coefficients for an initially disordered simple-cubic lattice and for a body-centered lattice with equilibrium long-range order. The results show that the chemical enhancement as a function of the enthalpy of mixing saturates, in contrast with results of conventional mean-field theories. Simulations in a disordered alloy agree with the calculations. The existence of longrange order is shown to increase the apparent activation energy for diffusion. The present theory is also shown to provide a simple method of calculating ordering kinetics and equilibrium longrange order, and good agreement with previous experiments and simulations is achieved.

This paper presents an alternative derivation of a discrete lattice regular solution model for computing the composition profile across an interphase boundary in a three component system. The model is applied to the calculation of the composition profiles across interfaces separating Ni-rich and Ag-rich phases in several Ni-Ag-X ternary systems (where X stands for Cu, Pd or Au). The results show that segregation to a (001) Ni-Ag interface is expected to be strong in the case of Cu, moderate in the case of Pd, and weak in the case of Au. It is also found that the Gibbsian interfacial excess of these segregants decreases with increasing temperature.