We have characterized, through transmission electron microscopy (TEM), the ReSi2−x thin films grown by reactive deposition on (001) Si. ReSi2−x thin films exhibit a distorted body-centered tetragonal MoSi2-type structure, and have excellent epitaxy on (001) Si since the face diagonal of the Si unit cell is equal to the c lattice parameter of silicide. The Si-deficient composition in the “disilicide” may be accommodated by collapse and shear of missing Si planes to form planar faults. Kirkendall voids are also observed at the film-substrate interface. The engineering of the defect and interface structures of these complex, non-stoichiometric silicides for optimized optoelectronic properties are discussed.

The influence of the dose on the crystal growth and on the selectivity of the Al-CVD process has been studied. Doses were varied from 100 L to 18000 L. Selective deposition was observed, both on passivated Si vs. SiO2, as well as on bare Si vs. passivated Si. On oxide elongated crystals have been deposited with dimensions of 1200 × 60 × 60 nm.

The influence of the germanium fraction on the parameters of the solid phase crystallization of amorphous silicon-germanium has been studied. Layers of different compositions were deposited by LPCVD at 450 °C and crystallized at 525, 550, 575 and 600 °C. The crystalline fraction was determined as a function of the annealing time for each germanium fraction and temperature. The incubation and the characteristic crystallization times were obtained by fitting the crystalline fraction to Avrami's model. The values of the diffusion activation energy were obtained for different compositions. The dependence of the nucleation rate and the growth velocity on the germanium fraction and the temperature was studied. The three parameters decrease as the Ge fraction increases; though the stronger dependence corresponds to the nucleation rate. The dependences on the composition and the temperature were analyzed in the frame of a conventional nucleation and growth model and their relationship with the corresponding free energies was discussed.

Nucleation and growth properties of Si1-xGex (0≦x≦0.3) have been investigated by using ellipsometric spectroscopy. The incubation time and the crystallization time significantly decrease with increasing x; the estimated incubation time (and crystallization time) in Si0.7Ge0.3 is about 1/100 of that in Si. During the incubation time, the increase of the crystallinity has been observed which corresponds to the nucleation of crystal seeds. From the behavior of the crystallinity during the incubation and the crystallization time, we estimated the activation energy of the nucleation and the crystallization. Both energies monotonically decrease with increasing x. The saturated crystallinity after crystallization decreases with increasing x. It has been found that the effect of Ge reduces the crystallinity of Si1-xGex while significantly enhancing the crystallization.

Epitaxial Cu/TiN heterostructures were grown on hexagonal (6H)-SiC(0001) substrate by pulsed laser deposition using the domain epitaxy, where integral multiple of lattice constant or major planes match across the interface1. Such layers are needed for metallization of SiC bond integrated circuit devices. These Cu and TiN layers on SiC(0001) were grown at 600 degrees centigrade in a high vacuum (<10−6 Torr). This structure was characterized using X-ray diffraction technique and transmission electron microscopy. The X-ray diffraction recorded only (111) and (222) reflection of Cu and TiN. The full-width at half maximum of ω-rocking curve of (111) reflection of Cu (0.4 degree) and TEM results indicated a high epitaxial quality. The plan view transmission electron micrograph shows that Cu forms three-dimensional islands indicating that the Cu/TiN interface energy is very high. The island size varies from 0.2 to 2 μm. Analysis of selective aperture diffraction patterns and cross-sectional transmission electron microscopy, including high-resolution imaging, showed relationships Cu(111)//TiN(111)//6H-SiC(0001). The TiN/SiC an interface was locally atomically sharp and free from secondary phases or obvious interdiffusion. The typical defects in the TiN(111) layers consisted of threading domain boundaries. The mechanism of three-dimension growth of copper on TiN layers was discussed.

In the context of the electronic industry, thin films of aluminum are deposited on silica. However, the Al/SiO2 interface is unstable. A reaction leading to oxidation of Al and reduction of SiO2 occurs. We present an in situ study of this reaction, in thin films, at temperatures above 500°C. This experiment allows to understand the essential steps of the oxidation of Al by SiO2, and it shows that thermal cycles can be used to generate nucleation centers artificially. More generally, the morphological interpretation of out-of-equilibrium growth patterns gives important insight into the coupling between the thermodynamics and morphological aspects of a thin film chemical reaction.

The highest performance fluoride ion conductor, PbSnF4, has been applied for the fabrication of an ambient temperature amperometric oxygen sensor, where it is used in the polycrystalline form. However, the structure of this material is highly anisotropic, thus one could expect polycrystalline samples to give a performance strongly dependent on crystallite direction. We have shown that the tin electronic structure has a very strong influence on the local structure, which determines the preferred direction of crystal growth, which is itself responsible for the crystal shape. This, in turn, determines the direction of preferred orientation, which can dramatically alter the properties relative to a randomly oriented sample. The effect of minor impurities on the size of the crystallites and on the crystalline symmetry has been studied.

Al-Sn surface alloys with 2-3 at.% Sn have been made by ion implantation of Sn in Al. The microstructure of the alloys consists of dense distributions of nanoscale Sn inclusions embedded in the Al matrix. For implantations carried out at 425 K the inclusions have sizes in the range from about 2 to 15 nm. The structure of the inclusions is tetragonal - the white Sn structure – with lattice parameters of a = 0.583 nm and c = 0.318 nm respectively, i.e. identical to the lattice parameters of bulk Sn. The inclusions grow in preferred alignment with the matrix and the most commonly observed orientation relationships is given by (100)Sn ||(111)Al and [010]Sn || [211]Al. The shape of the inclusions is partly faceted and partly rounded with larger flat facets on the {100}Sn/{111}Al interfaces. Melting and solidification of the inclusions, which have been studied by in-situ transmission electron microscopy (TEM) and Rutherford backscattering spectrometry (RBS) in combination with channeling, shows a distinct hysteresis. Melting of the inclusions which is associated with a distinct premelting, takes place in the range from about 430 K to 485 K, i.e. significantly below the bulk melting point of 505 K. The premelting is size dependent and the smallest inclusions melt at the lowest temperatures. Solidification requires a substantial undercooling and takes place from around 400 K with a much weaker size dependence.

In this study, the effect of grain size distribution on the thermal stability of electrodeposited nanocrystalline nickel was investigated by pre-annealing material such that a limited amount of abnormal grain growth was introduced. This work was done in an effort to understand the previously reported, unexpected effect, of increasing thermal stability with decreasing grain size seen in some nanocrystalline systems. Pre-annealing produced a range of grain size distributions in materials with relatively unchanged crystallographic texture and total solute content. Subsequent thermal analysis of the pre-annealed samples by differential scanning calorimetry showed that the activation energy of further grain growth was unchanged from the as-deposited nanocrystalline nickel.

Effects of 0.5 at. % boron doping on Ni85Al γ/γ superalloys were investigated for the first time, yielding a number of observations not previously observed in the literature. First, the grain growth kinetics of both the doped and undoped alloys were found to disobey the simple Nielsen law d = Ctn but instead follow an equation of the type d = C ln(t) + C0. The constants C and C0 were found to be respectively 10.2 and 23.2 for the boron-free alloy, and 6.2 and 14.2 for the boron-doped alloy, i.e., the grain growth rate was retarded significantly upon boron doping. Such a retarding effect is thought to be due to the formation of a boron-nickel cosegregated zone observed at the grain boundaries of the doped alloy; the width of the zone, in μm's, is two to three orders of magnitude larger than the boron induced disordered layer found in nickel rich Ni3Al compounds doped with boron. Other associated effects of the cosegregated zone include a sharp increase in toughness, much better slip transmission across grains and reduced workhardening rates. Another intriguing point is that the γ precipitates were found to segregate to the grain boundaries in the boron-free alloy after cold rolling, but no such segregation of γ precipitates has been observed in the boron-doped alloy. The different deformation microstructures and the retarded grain growth rates upon boron doping will be discussed.

The electrochemical cell provides a potentially powerful means of altering morphology and islanding phenomena on metallic surfaces. Diffusion and attachment processes on terraces and near step and island edges are known to profoundly affect island sizes, shapes and coarsening kinetics. Using the surface-embedded-atom-model (SEAM) for describing metallic surfaces in the electrolytic environment, we calculate the dependence of the activation energies for the aforementioned diffusion processes on the deposited surface charge for the Ag(111) and Ag(100) surfaces in an electrolytic environment. While all these processes show some degree of dependence on the potential, the step-edge barrier and the edge diffusion processes are the most sensitive. Step-edge barriers increase (to over 1 eV) with large positive potential (0.85 V), while edge diffusion barriers monotonically decrease with positive surface charge on Ag(100) and Ag(111). We assess the effect these diffusion barriers have on island size/shapes and coarsening dynamics and discuss the implications on electrochemical tuning of islanding phenomena.

This paper presents a study of heating-induced size and shape change for pre-synthesized composite nanoparticles of ∼2 nm gold cores encapsulated with alkanethiolate monolayers. The results have demonstrated an evolution in size and shape of the nanoparticles towards monodispersed larger core sizes with well-defined and highly-faceted morphologies. The evolved particles were encapsulated with the thiolate shells. The morphological and structural evolutions were characterized using TEM, XRD, UV-Vis and FTIR spectroscopy. While temperature-driven crystal growth is known for non-encapsulated particles, the evolution of the thiolate-encapsulated nanoparticles in solutions into well-defined morphologies represents an intriguing example of temperature manipulations of nanoparticle monodispersity and shape.

A number of MgO doped (3wt%) polycrystalline alumina samples were prepared. The preparation of these samples varied; different regimes for annealing (1500 – 1600°C) and a range of dwell times were examined. Atomic force microscopy (AFM) was used to study surface features. Information on grain boundary dimensions (depth, width) is presented along with a study of grain particle size. The growth of the grain boundaries was found to contradict Mullins' theory of uniform growth with time. We also present evidence for Ostwalds' ripening and the preferential growth of [001] oriented grains.

The crystallization of TiO2 in monolithic SiO2-TiO2 (SiO2>TiO2) gels by annealing has been investigated. The temperatures of the crystallization of TiO2 to anatase and anatase-to-rutile phase transition were considerably raised with the addition of SiO2. It is supposed that the anatase-torutile phase transition is primarily controlled by the grain size of the crystals. TiO2(B), one of the TiO2 polymorphs, was formed as one of the first crystalline phase in SiO2-TiO2 gels at 800-900°C. Transmission electron microscope observations revealed that TiO2(B) nanocrystallites with a size of 5-10 nm were dispersed in amorphous SiO2 matrix. TiO2(B) nanocrystallites did not grow larger but transformed to anatase at higher temperatures. It is supposed that the nucleation and stability of TiO2(B) are enhanced with the presence of surrounding SiO2, presumably by a low interfacial energy.

During the initial stage of phase formation reactions nucleation conditions often apply a controlling influence on the phase selection, product phase number density and grain size. In most cases the operation of nucleation control is also associated with the development of significant undercooling or supersaturation in order to initiate crystallization. Under this constraint it is common to observe different forms of alloy metastability. In fact, a hierarchy of equilibria can be identified based upon the severity of the kinetic constraints that act during transformation. One consequence of the different levels of metastability is the development of precursor reactions before the initial crystallization. In other cases a kinetic competition between different reaction modes or structures can yield a change in the transformation pathway during reaction. The appearance of kinetic transitions is one consequence of the competition between concurrent reactions and provides a valuable opportunity for kinetics analysis and modeling. In cases where sequential reactions are separated by temperature or time intervals other kinetic constraints operate to expose metastability due to diffusional growth limitations. The identification of the reaction control provides the foundation for kinetics analysis and the development of alloy design strategies. These features can be considered in terms of the observed crystallization behavior during rapid solidification and solid state interface reactions.

In this paper, we present a numerical model that incorporates algorithms to simulate nucleation and growth in supercooled liquid in a manner that properly accounts for the stochastic nature of nucleation. The basis of our model relies on a discretization of space and time to address thermal evolution, rapid growth of the undercooled interface, and nucleation in supercooled liquid. The present formulation of nucleation permits the spatially and temporally random nature of the phenomenon to be manifested in the transformation and resultant microstructure. This is accomplished by (1) calculating the probability of nucleation in each and every liquid node during each time step using the Poisson expression, and (2) triggering nucleation if and only when the random number assigned to a node for the time step is less than the calculated nucleation probability. No empirical or deterministic conditions for nucleation are imposed; nucleation occurs spontaneously and solely based on nucleation kinetics. We demonstrate the effectiveness of the overall model by analyzing conditions similar to those encountered in pulsed laser-induced crystallization of thin Si films, and discuss the generality of the proposed stochastic formulation of nucleation.

According to classical nucleation theory the nucleation- and phase-selection behavior of undercooled meltallic melts is strongly dependent on the solid-liquid interfacial energy.

A structural approach to the modelling of the interfacial energy for simple melt-crystal interfaces (fcc, hcp, bcc) was developed a number of years ago [1-4]. This approach is extended to polytetrahedral phases using numerical simulation. Results of these calculations are presented for different polytetrahedral structures: the tetragonal σ-phase in Ni-V, the monoclinic phase λ-Al13Fe4, the orthorhombic phase μ-Al5Fe2 and the icosahedral quasicrystalline I-phase in Al-Pd-Mn. The numerically estimated values for the solid-liquid interfacial energy are compared with results from experiments on the undercooling- and phase-selection behavior.

The crystallization kinetics of Al-Gd-La-Ni metallic glasses to nanostructured phases are analyzed using differential scanning calorimetry and transmission electron microscopy. In a narrow alloy composition range near Al88Gd6La2Ni4, TEM reveals an amorphous phase separation that occurs upon annealing at low temperatures prior to crystallization. Al-enriched regions, typically 40 nm in diameter, bounded by rare-earth rich regions, are visible. Upon crystallization, α-Al forms preferentially at the interface between these phase separated regions. The relevance of this crystallization sequence to previous work in Al-RE-TM glasses and to the evolution of nanoscale microstructures common in the crystallization of other metallic glasses are discussed.

The shape of liquid Pb inclusions in Al was investigated by in-situ electron microscopy over a temperature interval from 300 to 500°C. The inclusion shape was found to depend on size and temperature. During isothermal annealing after melting, small inclusions rounded off while larger inclusions remained faceted until the temperature was raised to about 500°C. During subsequent cooling, inclusions refaceted, although less strongly than during heating. The shape hysteresis between heating and cooling cycles was found to be due to the barrier of ledge nucleation necessary to advance the faceted interfaces. The observations show that the step energy depends on temperature and its disappearance at about 540-600°C indicates a roughening transition.

We have developed a model of dendritic grain growth in the solidification process of metallic alloys in which the output of a nanoscopic molecular dynamics simulation, based on the use of many-body interatomic potentials, provides the input into the microscopic nucleation and growth model of Rappaz et al. The microscopic model itself is also extended by the introduction of a stochastic dynamics, via the Ito stochastic calculus, to account for the random fluctuations in the growth trajectory of the dendrite tip. The model is applied to the industrially important Sn-Pb alloy and its result is compared with that obtained with the Rappaz model.