The kinetics of short and long-range ordering in FeAlx alloys have been investigated using resistometry. The results were obtained in the A2 solid solution (x = 0.10 and 0.17), in the DO3 ordered phase (x = 0.23, 0.27, 0.30 and 0.33) and in the A2 + DO3 two-phase region (x = 0.19). The ordering kinetics curves are well described in a classical model with a lognormal distribution of relaxation times whose center varies with the temperature according to an Arrhenius law. One thus gets the thermodynamic quantities which drive the atomic mobility. The data display a strong decrease of the activation energy with Al content and with the formation of the D03 structure, as well as very short relaxation times. Such an abnormally high atomic mobility can be attributed to a decrease of the migration energy through a lattice dynamics effect, similar to that recently proposed by several authors to explain the enhanced diffusion in some bcc transition metals and alloys.

As an “archtypal” spin-glass system, the local atomic arrangement in Au-Fe alloys is still a controversial subject. In the present experiment, Au-Fe alloys with Fe content from 10.7 to 33% were studied as to microstructure evolution, through various transmission electron microscopic techniques. Information derived from aging experiments using real and reciprocal spaces appears to suggest the coexistence of short-range-order and Fe-clusters. At early stages of aging, lobe-like and/or rod-shaped strain contrast images, identified as clusters, were revealed after specimens were further cleaned by an ion-miller. The (1 1/2 0) special point diffuse reflections were prominent in the as-quenched condition and/or early stage aging, and the intensifies decrease gradually as aging proceeds. After a certain period, (1 1/2 0) diffuse reflections disappear but strain contrast images still remain similar. This suggests that the strain contrast images are not related to the (1 1/2 0) diffuse reflections. High resolution electron microscopy was also employed. The results are consistent with the arguments derived from aging experiments.

Alternate layers of pure copper and gold were vapor deposited on a sodium chloride substrate, the average concentration of the films being Cu-16at% Au and the layering periodicity (modulation wavelength) being 3.31 nm. The composition modulation gives rise to satellite diffraction peaks around the (200) Bragg reflections. Synchrotron radiation at SSRL was able to detect up to third order satellite intensity the evolution of which was measured as a function of annealing time at 515 K. Although the first order satellite intensity decayed as expected exponentially with time, intensities of both second and third order satellites decreased very rapidly at first, then increased before decaying exponentially. These results are in conformity with theoretical models of satellite evolution during annealing in a onedimensional modulated system governed by a nonlinear diffusion equation.

we report grazing-incidence x-ray scattering measurements of the order-disorder phase transition near a Au-rich Cu3Au (001) surface. The bulk transition temperature TB = 648 K is lower than for stoichiometric Cu3Au(001). Surface ordering persists above TB in a layer several monolayers thick up to Ts = 667 K. Kinetics of the ordering were studied by quenching from Ts + 9 K to final temperatures of TF = 661 K and 645 K. Short-range order is established in the bulk within the quench time and does not appear to change subsequently. The bulk short-range order differs from that reported for stoichiometric Cu3Au. For TB < TF = 661 K < TS, the (100) in-plane diffraction beam intensity grows approximately as t1/4 for t > 1000 s and the width continues to narrow. For TF = 645 K < TB, power law growth is not observed for the longest times studied. In both cases the growth is substantially slower than predicted.

Kinetics of long-range ordering was studied in the alloys Ni75Al25-xFex by means of in situ resistometry. The alloys under consideration exhibit a dramatic decrease of the "order-disorder" transition temperature Ttr with increasing X. After the existing theoretical models, the activation energy of long-range ordering should depend on one hand on the energy of ordering - related to Ttr -, and on the other hand, on the current value of the LRO degree. The performed measurements yield an experimental evidence of the above relationships. In addition, a considerable effect of magnetic order on LRO kinetics has also been registered. The results are qualitatively discussed in terms of the model of LRO kinetics in L12 structure proposed by Schoijet and Girifalco.

We show that small angle neutron scattering can sensitively detect microphase separation in solid solutions of the C30H62:C36D74 mixed paraffins. By ramping the sample temperature at a constant rate of about 0.5°C per minute, information complementary to that from differential heat capacity measurements is obtained from the total small angle scattering signal and its first temperature derivative. Because neutron wavelengths in the range from 0.8 < λ (Å) < 14.0 are used, scattering functions over an extended range (0.005 < q (Å−1) < 0.35) are obtained at the Argonne pulsed neutron source in a single experiment. This is important inasmuch as there is significant scattering in the lowest q region accessible due to the presence of cracks and/or voids, in addition to that from the microphase separation. The specific interfacial area as a function of temperature is reported from time slicing experiments on the pure C36D74 compound. Weak superlattice reflections are identifiable in the scattering profiles of the mixed paraffins due to the microphase separation.

Structure change during the reversion process in an Al−12at.%Li alloy above the metastable δ′ solvus was investigated using a time-resolved small-angle x-ray scattering technique with synchrotron radiation. Results showed that the reversion process started after a short incubation time and that the growth of the stable δ phase began before completion of the δ′ dissolution. The radius of gyration of the second phase particles showed little change in the initial stage of reversion, then increased with time, suggesting the presence of diffuse interfaces between the dissolving δ′ particles and the matrix. It is suggested that the undissolved δ′ particles serve as the nuclei of the more stable δ precipitates, which continue to grow with their radii of gyration showing a parabolic power law in the early stage of growth followed by the familiar coarsening kinetics.

It has been proposed that martensite is formed by heterogeneous nucleation and growth or evolves from a precursor. In order to examine these theories, high resolution X-ray diffraction experiments have been performed on In0.77Tl0.23. It is known that the transformation of the alloy is close to second order and tweed has been observed which could be a possible precursor structure. Our experimental results do show precursors but of a different nature: satellites around the 220 Bragg peak have been detected up to 15 degrees above the transformation temperature. They develop reversibly at the angular positions of the product phase. The results are not compatible with the “crest-riding-peridon” model but rather show an equilibrium two phase mixture above the transformation temperature in which the second phase is structurally equal to the product phase. We have also investigated the mechanical response of the high temperature phase and found that both shear ‘elastic constants’ soften. This result can be understood in terms of twin boundary or stacking fault motion.

Antiphase domain (APD) structures and associated composition analyses of the α l-bainite which forms during isothermal heat-treatment of a Cu-26.7wt%Zn-4.0wt%Al alloy have been examined by TEM dark field imaging and energy dispersive X-ray microanalysis. At an early stage both nearest-neighbor (NN) and next-nearest-neigbor (NNN) APD structures of the β-parent phase are inherited by the α1-bainite having an M18R structure. The EDX analyses indicate that there is a difference in composition between the α 1-bainite and the β-parent phase at this stage in spite of the inheritence of both NN and NNN APD structures. Local depletion of Zn concentration at dislocations is also observed in the parent phase. With increasing aging time, the α 1-bainite grows in a manner that the NN APD structure still remains inherited but the NNN APB structure changes into a NNN disordered structure at outer regions of the plate. In addition, Zn and Al rich regions also form adjacent to the interface in the parent phase. The observed growth of APD structures in association with compositional analyses strongly suggest that the initial bainitic transformation is by a shear mechanism at solute depleted defects although subsequent plate growth is diffusion dominated.

The kinetics of an icosahedral to approximant phase transformation were investigated in cast alloys in the composition range Al(63.65)Cu(20–25)Fe(12–15). Perfect icosahedral crystals are stable in the temperature range below the peritectic melting temperature and above approximately 700°C, below which they transform to a textured polycrystalline mixture of small microcrystals of an icosahedral approximant. The transformation is isocompositional and does not appear to occur by nucleation and growth. The kinetics of the transformation under continuous cooling and isothermal conditions were investigated using conventional and high resolution electron microscopy. The transformation is reversible, takes place over a range of compositions, temperatures and times and results in distinctive strain contrast in the electron microscope.

Solid-state amorphization rates have been measured for amorphous Ni1-cMc in contact with the crystalline terminal phase M, (M=Hf or Zr). The interdiffusion coefficient Dis found to depend on the composition of the amorphous phase, with higher Ni content resulting in a higher Dover the composition range Ni67Hf33 to Ni47Hf53. The common tangent composition at which the amorphous alloy is in metastable equilibrium with the terminal crystalline phase is found to be greater than 70 at.% Hf, which is considerably higher than previously reported values. This discrepancy is explained in terms of the interdiffusion coefficient's variation with composition.

The main purpose of this work was to investigate the possibility of solid state amorphization of Ni-B crystal powder. A Ni-B alloy powder was obtained by mechanical alloying of Ni and B powder using a conventional ball milling machine. X-ray diffraction and EXAFS measurements were made for the mechanically alloyed samples as a function of milling time. An amorphization reaction was observed in Ni64B36 at 400 h of milling time. The long range order of the crystalline Ni phase diminished with an increase of solution of B into the Ni matrix. This suggests that the solution of B atoms into the Ni crystalline phase, forced by mechanical alloying, plays an important role in the crystalamorphous transformation.

This study provides information regarding how an intermetallic phase nucleates and grows at an interface between elemental components during solid-state reactions. The reaction of Al/Ni multilayer films was chosen as a model case. The main experimental technique employed was differential scanning calorimetry (DSC), with assistance of other analytical tools such as crosssectional transmission electron microscopy/microanalysis (XTEM/STEM) and thin-film x-ray diffraction. Al/Ni multilayer films were prepared by electron-beam evaporation in high and ultrahigh vacuum systems. We show evidence that interdiffusion of Al and Ni precedes the formation of Al3Ni, forming metastable solid solutions. Using isothermal calorimetry and modeling, we find that Al3Ni subsequently nucleates in the interdiffused region at preferred sites, and that the nucleation sites are quickly consumed in the early stages of Al3Ni formation. The nucleation site density strongly depends on the grain sizes of the deposited films, suggesting that only certain types of intergranular defects can serve as nucleation sites. After coalescence into a continuous layer, Al3Ni thickens through a diffusion-limited process. A kinetic model is applied which yields calculated calorimetric traces in good agreement with experimental data. The results are discussed in light of the the calculated free energy-composition diagram for the Al-Ni system. The implications of our results to the phase selection during thin-film reactions are discussed.

Microstructural changes in the bcc-based water-quenched alloy β Cr40Ti60 have been investigated as a function of annealing time at 873 K. The as-quenched structure has been found to contain a non-ideal form of the ω structure, which has been detected and characterized on the basis of diffuse scattering to which it gives rise in electron diffraction patterns. Annealing at 873 K results initially in platelet precipitates of the metastable β″ phase, which is found to posses a bulk body centered tetragonal structure in which chemical ordering has been detected via electron diffraction and high resolution microscopy. After several hours of annealing at 873 K the equilibrium phases α-Ti and the C15 Laves phase Cr2Ti form as spherical polycrystalline aggregates. No evidence is found of reported annealing-induced amorphization [1]. It is suggested that amorphization cannot occur in this alloy via thermal activation, and that the reported findings may have resulted from impure specimens.

Multilayer films, consisting of alternating layers of crystalline Ni and Ti, have been prepared by RF sputter deposition over a range of modulation wavelengths corresponding to an overall composition of Ni50Ti50. These films have been characterized by xray diffraction and Rutherford backscattering measurements. The solid-state transformation by interdiffusional mixing of the individual layers has been directly studied by differential scanning calorimetry and correlated with structural measurements. These measurements indicate that the solid-state reaction of Ni and Ti multilayers proceeds through the formation of a metastable solid solution of Ti in Ni followed by the formation of intermetallic equilibrium compounds. No direct calorimetric or structural evidence for the formation of an amorphous Ni-Ti phase has been found in these samples.

Sputter deposited films of amorphous NixZr100-x, have been prepared over a wide and continuous range of compositions between x=20 and x=90 atomic percent Ni on liquid nitrogen cooled and room temperature substrates. These films have been characterized by xray and electron diffraction, transmission electron microscopy, and Rutherford backscattering spectrometry. The crystallization behavior of these films has been systematically studied by differential scanning calorimetry and compared with earlier measurements on amorphous liquid quenched alloys. Systematic differences were found between the thermal stability of amorphous alloys of the same nominal composition prepared at different substrate temperatures. Larger differences exist between vapor deposited amorphous films and liquid quenched ribbons.

The structural evolution of a mixture of elemental Ni and Ru powders during high energy ball milling has been studied as a function of milling time using X-ray diffraction and transmission electron microscopy. On the Ni-rich side of the phase diagram (Ni70Ru30) we observe, during the milling, the diffusion of Ru in the Ni and an expansion of the Ni fcc lattice. After about 20 hours of milling, all the Ru has diffused into the Ni and we are left with an oversaturated Ni(Ru) fcc structure with a lattice parameter about 3.6% larger than the one of pure Ni. The Ni does not seem to diffuse into Ru during the process. On the Ru-rich side (Ni43Ru57) we observe initially the same diffusion process taking place but more rapidly. After about 8 hours of milling, the incorporation of Ru in the Ni has driven the fcc Ni lattice to its limit of stability (about 3.9% expansion) and the fcc structure transforms, upon further milling, into a new hcp structure. In the later stage of the process the remaining Ru diffuses into this new phase.

Crystallization of metal-metalloid glasses is known to proceed by nucleation and growth processes. Using crystallization statistics in partially crystallized glasses, at temperatures below the glass transition temperature, time-dependent heterogeneous nucleation has been found to occur at a number of quenched-in nucleation sites [1]. Close to the glass transition temperature crystallization proceeds so rapidly that partially crystallized microstructures could not be obtained. Initial results from fully crystallized glasses exhibit evidence for a transient homogeneous nucleation process at higher temperatures [1,2].

An x-ray technique has been used to measure the diffusion of phosphorus in crystalline Ni/amorphous NiPx and amorphous NiPx/NiPy multilayer thin films produced by electrodeposition. The films have repeat lengths in the range 30–80Å and P contents x, y < 25at.%. A value for the interdiffusivity in amorphous NiPx is derived from measurements on fully amorphous films. The behaviour of partially crystalline films is described in terms of phosphorus diffusion into the nickel grain boundaries.

In situ observations of CIGM in CaCO3 bicrystals with a SrCO3 solute source were made. The change in boundary orientation and migration rate were compared with solute concentration. The liquid film model for coherency strain Induced migration was generalized to any non-cubic system and applied to CaCO3-SrCO3. The coherent layer was modeled as a thin film on an infinite half-space. The strain energy was found from solution of the Hooke's law expressions transformed to the appropriate coordinate system. For triclinic or monoclinic films the strain tensor was found by an eigenvector decomposition of the transformation matrix that defined the lattice parameter change with composition. High anisotropy of Vegard's law constants for CaCO3-SrCO3 caused (111) to have the lowest coherency strain per unit solute. Surfaces perpendicular to (111) in coherent equilibria were predicted to have half the solute concentration and three times the migration driving force of those perpendicular to (111). However, no correlation between solute concentration and boundary orientation was observed. Ambiguous and contradictory evidence for a relationship between solute concentration, boundary orientation, and migration rate was found. The self-stress state of a grain boundary in a solute diffusion field may be better modelled as hydrostatic rather than plane stress. Hydrostatic compression may interact with the boundary excess volume and cause a PV driving force for migration. Predictions based on coherent equilibrium at a surface have not been tested for that geometry in calcite; they should be tested before they are applied to grain boundaries.