Three-dimensional analyses of anisotropic stress effects during thermal oxidation of silicon are performed in this study using finite strain formulation. A reaction rate model that conjugates pressure to a crystalline direction dependent activation volume is employed for the modeling of facet formation in an STI MOSFET device. The numerical results agree well with TEM observations. The 3-D nature of stress evolution in fabrication process is demonstrated and the implication of stress relaxation on device performance in a strained SiGe based device is discussed.

Transmission electron microscopy (TEM) observation of irradiated ferritic materials reveals the existence of large, interstitial dislocation loops with Burgers vectors ½ 〈111〉 and 〈100〉. These loops cause hardening of the material by pinning dislocations and impeding their glide during deformation. However, numerous molecular dynamics simulations of collision cascades in α-Fe have evidenced the exclusive formation of small, highly mobile, ½ 〈111〉 clusters. Additionally, continuum dislocation theory and atomistic simulations have shown that ½ 〈111〉 loops are energetically favored. This introduces the need to explain the mechanisms of formation and growth of 〈100〉 loops from small, cascade-produced clusters. The understanding of the physics underlying these phenomena is important for the development of solid damage accumulation models in ferritic materials that are being considered for fusion applications. In this work we propose a comprehensive set of dislocation reactions that explain the nucleation of 〈100〉 loops from ½ 〈111〉 clusters. The growth up to TEM visible sizes of 〈100〉 loops through absorption of one-dimensionally migrating ½ 〈111〉 clusters is also assessed. Finally, a direct comparison of TEM experimental micrographs with atomistic simulation-derived images is presented to show an example of how to help close the gap that exists between modeling and experiments.

We conduct molecular dynamics computer simulations of fracture in silica glass using the van Beest, Kramer, and van Santen model. Stress is applied by uniaxial strain at different pulling rates. Comparisons with previous fracture simulations of silica that used the Soules force function are presented. We find that in both models stress is relieved by rotation of the (SiO4)-2 tetrahedrons, increasing Si-O-Si bonding angles, and only small changes in the tetrahedron dimensions and O-Si-O angles.

Sintering theory has been developed either as the application of complex diffusion mechanisms to a simple geometry or as the deformation and shrinkage of a continuum body. We present a model that can treat in detail both the evolution of microstructure and the sintering mechanisms, on the mesoscale, so that constitutive equations with detailed microstructural information can be generated. The model is capable of simulating vacancy diffusion by grain boundary diffusion, annihilation of vacancies at grain boundaries resulting in densification, and coarsening of the microstructural features. In this paper, we review the capabilities of this model and present a number of different problems that have been treated by the model. Finally, we discuss the limitations of this model.

During coalescence, the surface area of the particle decreases, resulting in a heat release associated with the resulting lower surface energy. In a growth process particle heating competes with heat transfer by conduction to the cooler carrier gas and radiation. This temperature increase can be extremely important and should be accounted for when modeling collision/coalescence processes. The heat release associated with particle coalescence may reduce the coalescence time by as much as a few orders of magnitude. In addition, under some conditions there is insufficient time for the particles to cool to the gas temperature before another collision event takes place. It is shown that accounting for energy release and heat transfer effects have a dramatic effect on primary particle formation and the onset of aggregate formation. The results of the work indicate that to grow the largest primary particles one should operate at low pressures and high volume loadings.

Modeling of phenomena is increasingly being used to obtain an understanding of important physical events as well as to predict properties that can be directly tied to experimental data. For systems with relatively low densities of particles, the Direct Simulation Monte Carlo (DSMC) method is well suited for modeling gases with non-equilibrium distributions, coupled gas-dynamic and reaction effects, emission and absorption of radiation. On the other hand, if the density of particles is large such as in dense gases or condensed matter, the DSMC method is not appropriate and techniques such as molecular dynamics (MD) simulations are employed. There are phenomena such as laser ablation, however, in which the system evolves from a condensed state appropriate to be studied with MD to an expanding rarified gas appropriate to be studied with DSMC.

The work presented here discusses the means of transferring information from a MD simulation of laser ablation to a DSMC simulation of the plume expansion. The presence of clusters in the MD output poses the main computational challenge. When the laser fluence is above the ablation threshold, the cluster size distribution is very broad (up to 10,'000's of particles per cluster) but there are relatively few of each cluster size. We have developed a method for statistical processing of the MD results and have represented the cluster size as a random variable. Various aspects of the coupling between the MD and DSMC models are discussed and several examples are presented.

Understanding the morphological changes of Pb-Sn solder alloys helps to improve their performance in electronic applications. The focus of our study is degeneration of lamellar structures at high temperatures. Microstructural evolution of the Pb-Sn eutectic lamellar structure is modeled numerically using Monte Carlo Potts approach. The initial structure consists of alternating layers of Pb-rich and Sn-rich phases, simulating the lamellar array in a near eutectic system. Faults are introduced to destabilize the system. After a short incubation period the shape of lamellae become irregular. The perturbations grow with time and eventually break the lamellae into nearly equiaxed grains. The grain size of the degenerated structure is 2-3 times the original lamellar spacing weakly depending on the spacing between the faults. This is consistent with the experimental observation of degeneration of Pb-62 wt% Sn solder. The duration of degeneration processes is comparable with the time it would take Ostwald ripening to produce grains of the same size. Eventually grain growth reaches the asymptotic regime of coarsening described by a power-law function of time.

It is known that by controlling microstructural development, desirable properties of materials can be achieved. The main objective of our research is to understand and control interface dominated material properties, and finally, to verify experimental results with computer simulations. We have previously showed a strong similarity between small-scale grain growth experiments and anisotropic three-dimensional simulations obtained from the Electron Backscattered Diffraction (EBSD) measurements [1]. Using the same technique, we obtained 5170-grain data from an Aluminum-film (120μm thick) with a columnar grain structure. Experimentally obtained starting microstructure and grain boundary properties are input for the three-dimensional grain growth simulation. In the computational model, minimization of the interface energy is the driving force for the grain boundary motion. The computed evolved microstructure is compared with the final experimental microstructure, after annealing at 550°C.

Guided self assembly of periodic arrays of quantum dots has recently emerged as an important research field not only to reduce component size and manufacturing cost but also to explore and apply quantum mechanical effects in novel nanodevices. The intention of this kinetic Monte Carlo (KMC) simulation study is to investigate self-organized nanopatterning on hexagonal surfaces for relaxed periodic surface strain fields applied to Pt(111) epitaxy. The KMC model is a full diffusion bond-counting model including nearest neighbor as well as second-nearest neighbor interactions with an event catalogue consisting of 8989 events modeling the effect of the biaxial surface strain field. The strain dependence of the fcc site and the saddle point for a Pt adatom migrating on top of the Pt(111) surface is calculated using the embedded atom method. Both the valley and the saddle point energies show an excellent linear dependence on the strain. These results are applied in the KMC model. The surface strain in this study is caused by a hexagonal network of dislocations at the interface between the substrate and a mismatched epitaxial layer. How the selforganization of deposited atoms is influenced by the surface strain will be addressed.

Incommensurate structure of Hexamethylenetetramine suberate (C6H12N4)(HOOC-(CH2)6-COOH) has been solved from single crystal x-ray diffraction data. A molecular dynamics simulation of this system was carried out from 15 K to 580K. A second-generation consistent forcefield (CFF) and a compensating pressure tensor field were used to describe the interactions between atoms and to account for deficiencies in the forcefield. Starting from the experimental 298K structure, the phase transitions were investigated over an extended temperature range. A high symmetry commensurate structure exists at temperatures between 410K and 290K. For temperatures lower than 290K, a new periodicity appears in the structure. The system reaches a low symmetry lock-in phase at about 150K. An incommensurate structure appears between the high and low symmetry phases between 290K and 150K. The new periodicity associated with the incommensurate modulation is due to the appearance of additional long range ordering of the carbon chains. The present simulation not only reproduces well the experimental x-ray diffraction results but also gives new insight into the origin of the incommensurate behavior.

In this study, the linear shock velocity (U) versus particle velocity (u), which is well suited for describing the Hugoniot curve of metals, is revisited from a thermodynamical point of view. In this way, it is possible to set-up a theoretical relation between the Grüneisen coefficient and the isentrope derivatives to third order at the initial state before jump, ensuring the linear U(u) relationship. When applied to the classical Morse potential describing the Equation of State (EoS) in the Mie-Grüneisen form, the theoretical result is in a good acceptance with the shock experimental results carried out on copper.

In order to develop a reliable interconnect integration scheme by using Cu in ultra large scale integrated devices (ULSI), the evolutions of the concentration profile of copper ions in SiO2 was simulated under bias temperature stress (BTS) test. Diffusion equation was solved numerically in two electric field modes. One is constant electric field mode where copper drift was simulated with the assumption that electric field is constant within SiO2 film. In variable electric field mode, simulation was carried out considering the variation of electric field in SiO2 due to copper ions. The diffusion of copper ions in variable electric field mode is slower than that in constant electric field mode, because copper ions in SiO2 reduce electric field near the interface between Cu and SiO2. Flatband voltage shift ) (ΔVFB increases parabolically as BTS time increases in constant electric field mode. However, it has linear relation with BTS time in variable electric field mode, which is typically observed in experiments.

We describe a model for numerical calculation of phonon spectrum in a three-dimensional regimented array of semiconductor quantum dots. Regimentation and possibility of carrier mini-band formation make this structure analogous to a crystal, e.g. quantum dot crystal. It is demonstrated that the acoustic phonon dispersion undergoes strong modification in such a structure leading to emergence of low-energy quasi-optical branches. Strong phonon spectrum modification is expected to affect carrier relaxation and transport properties.

We describe and discuss the similarites in two different kinds of physical systems: Chua's circuit (and, broady speaking, chaotic systems) and atoms in the Periodic Table. We anayze the regularities arising from the chaotic nature of the former against the regularities in the energy spectra of the second. We show the existing analogies among the Bohr's law and the bifurcation diagram building a map connecting each energy level in atoms to each point in the bifurcation diagram in the parameter space. We found a genera rule valid for many element of the Periodic Table by calculating the two-dimensional bifurcation diagram versus Bohr's law. The result is a one-to-one correspondence between atoms and nonlinear Chua's circuits. Finaly we discuss possible implications of the model when applied to more complicated systems like molecules or atomic clusters.

The influence of carbon black (CB) in a rubber formulation can be observed during the vulcanization process. However, looking through literature there seems to be a lack of information related to the role of CBs in vulcanization reaction. The present work is focused on the application of Model Compound Vulcanization (MCV) approach and Inverse Gas Chromatography (IGC) to predict the role of carbon black on vulcanization.

Squalene was chosen as the model molecule of natural rubber to undergo the vulcanization in the presence of carbon black. This methodology gives rise to several advantages. On one hand, the model allows following the accelerator, vulcanization intermediate compounds, and even the crosslinked model molecule. Moreover, the similarity between the model and the polymer chain allows further extrapolation to discover the real sample's performance. In order to complete the information required IGC analysis was used. IGC has been performed using both dispersive and specific probes to elucidate a wide range of CB surface active sites.

The application of this prediction tool to special grades revealed different effects that have been verified with real formulations in both cases.

The alumina-chromia system shows complete mutual solubility and is represented by an isomorphous phase diagram. However, the alumina-chromia system exhibits an asymmetric miscibility gap under 1300°C. Using existing data from the literature, the alumina-chromia system was assessed using thermodynamic modeling by Kim and Sanders [1]. Regular and subregular solution models for the liquid and solid phases were used to define the phase boundaries for the miscibility gap in this system. Using this thermodynamic representation of the miscibility gap to select temperatures of interest, 75 mole percent Al2O3samples were synthesized via combustion of powders, followed by pressing into pellets and heat-treated for various times and temperatures. Both X-ray and TEM analysis showed evidence of spinodal decomposition after heat-treatment. X-ray analysis showed that decreasing the heat-treatment temperature increases the compositional difference between the phases present. The experimentally observed microstructures exhibit lamella-like structures that vary in spacing from 8nm to 3nm as the heat-treatment temperature varies from 400°C to 800°C.

Micro-indentation test at scales on the order of sub-micron has shown that the measured hardness increases strongly with decreasing indent depth or indent size, which is frequently referred to as the size effect. Simultaneously, at micron or sub-micron scale, the material microstructure size also has an important influence on the measured hardness. This kind of effect, such as the crystal grain size effect, thin film thickness effect, etc., is called the geometrical effect. In the present research, in order to investigate the size effect and the geometrical effect, the micro-indentation experiments are carried out respectively for single crystal copper and aluminum, for polycrystal aluminum, as well as for a thin film/substrate system, Ti/Si3N4. The size effect and geometrical effect are displayed experimentally. Moreover, using strain gradient plasticity theory, the size effect and the geometrical effect are simulated. Through comparing experimental results with simulation results, the length-scale parameter appearing in the strain gradient theory for different cases is predicted. Furthermore, the size effect and the geometrical effect are interpreted using the geometrically necessary dislocation concept and the discrete dislocation theory.

Structural models of silicon imidonitride (SiNH) solids are studied using density functional methods. The porous models have densities between 1.6 and 2.4 g/cm3 with pore sizes between 4.5 Å and 7.5 Å. The networks consist of Si-N and N-H bonds only, with N-H and N-H2 groups located at the internal surface. Calculated Raman spectra compare very well with experimental results. We observed proton transfer reactions from an inserted malononitrile molecule to the SiNH-host during Car-Parrinello molecular dynamic simulations. The proton affinity of porous silicon imidonitride structures is confirmed by an energy gain for the proton transfer reaction to both N-H and N-H2 groups.

The micro FEM model, which was considered in the middle bottom of a four-point bending macro FEM model, was constructed from a SEM micrograph of an Al2O3-10vol%ZrO2 ceramic, in which the crystallographic information was given to each grain randomly. One micro-crack was introduced in the micro FEM model as a basic case to simulate the four-point bending strength. The strength was determined by the average stress in a one-grain size distance from the crack tip. Two types of micro-cracks were introduced in the model as a second micro-crack, and the strength was compared with the basic case. For the case in which the second crack is parallel to the first crack and apart from the first crack in horizontal direction, the strength was increasing with the second crack closer to the first crack. For the case in which the second crack is parallel to the first crack but apart from the first crack in vertical direction, the strength was decreasing with the second crack closer to the first crack.

Structures and properties of amorphous silicon carbonitride (a-SiCN) materials are studied using density functional methods. Topologicaly different models of a-SiCN with 112-156 atoms each were generated from continuous alternating random networks. The networks have distinct topologies which result in a different chemical bonding in the investigated structural models. A first model consists of Si-N and Si-C bonds only. For such materials with “anionic” carbon we found the highest bulk modulus. However, the network strain is largely due to the high SiC content, resulting in bond ruputre and redistribution in the amorphous structure. A second model consists of Si-N and C-N bonds only. We found that a tetrahedral coordination of “cationic” carbon in such a compound is unstable at elevated temperatures, as indicated by Car-Parrinello molecular dynamic (CPMD) simulations. A third model consists of amorphous Si3N4 as host structure and a segregation of graphitic C inside a pore. Such a model, although low in density, has an enlarged bulk modulus, comparable to the “anionic” model.