Infrared radiation, which couples with vibrations in molecules and solids, has long been considered as a promising way to selectively control chemical reactions, materials processing, and biomedical applications. The expectation is that particular frequencies couple to specific atomic motions, in contrast to heat which imparts energy indiscriminately. The promise has yet to be fulfilled, but recent successes and the development of new lasers have rekindled interest. We report calculations of the dynamics of solids and molecules under intense infrared radiation. In solids, using a model calculation, we find that narrowly-defined “windows of opportunity” exist for resonant enhancement of impurity diffusion at moderate intensities. In molecules, using a generalization of Car-Parrinello dynamics, we find that, for very intense fields, energy absorption at normal modes is not very efficient and that selective bond-breaking may occur at nonresonant frequencies by “concerted kicks”. Finally, time-dependent density functional theory is used to show that the effect of very intense infrared radiation on atoms and molecules is to produce very high harmonics as found experimentally.

In this paper we extend the quasi-continuum method to study equilibrium properties of defects at finite temperatures. We present a derivation of an effective energy function to perform Monte Carlo simulations in a mixed atomistic and continuum setting. It is shown that the free energy minimization technique can be easily incorporated into the quasi-continuum frame work, permitting a reduction of the full set of atomistic degrees of freedom even in the finite temperature setting. The validity of the proposed methods is demonstrated by computing the thermal expansion and the temperature dependence of the elastic moduli for Cu. We also employ the quasi-continuum free energy minimization method to study the finite temperarure structure of a dislocation core in Al.

We formulate a method for imposing continuous field boundary conditions on an MD simulation with little or no disturbance to its dynamics. Our approach combines the previously developed thermodynamic field estimator, which extracts macroscopic fields from particle data, with a novel procedure, optimal in the sense of least disturbance, for imposing prescribed continuum field boundary conditions on the atomistic system. By means of feedback control and assuming molecular chaos in fluids, that disturbance can further be eliminated entirely, thus providing an exact solution for general steady-state fluid problems, where the desired fields at the real boundary is achieved by adjusting actions in a region separated from the real boundary by a buffer zone.

A mixed atomistic/continuum technique, in the spirit of the quasicontinuum method, is formulated and used to solve boundary value problems in strained semiconductor structures in which the mechanical fields and the local electronic structure are fully coupled. The technique is implemented by means of a standard structural mechanics finite element package. Within each element in the mesh, tight binding calculations are made for mechanical properties based on the local electronic structure. The finite element program calculates equilibrium mechanical fields based on this atom-istic constitutive information; the underlying electronic properties can then be extracted at the element level. The technique is demonstrated by examining several simple plane strain boundary value problems for coherently strained silicon.

We describe our studies of two methods for generating bias potentials for use in hyperdynamics simulations. In the first method, first reported by Steiner, et al, the bias potential is the additional energy needed to keep the total potential energy above a fixed level. This potential exerts a negligible computational load and is very easy to code. The second technique involves computing the iterative Hessian-based bias potential as usual, but including in those calculations only the atoms in a small “active region” that surrounds the area in which a state transition is expected to occur. This “subspace hyperdynamics” method is less costly than full HD. The extra computational effort required for HD scales as the size of the active region(s), instead of with the size of the entire simulation domain.

We have carried out extensive tests of both methods on two problems, the diffusion of adatoms on silver surfaces and the migration of vacancies in bulk silver. Using the first method, to obtain large boosts the bias potential must be so high that many transitions are prevented from occurring. Our results for subspace hyperdynamics are promising; we obtain the same results as with full hyperdynamics, but with considerably less computational effort.

Laser ablation of organic solids is a complex collective phenomenon that includes processes occurring at different length and time scales. A mesoscopic breathing sphere model developed recently for molecular dynamics simulation of laser ablation and damage of organic solids has significantly expanded the length-scale (up to hundreds of nanometers) and the time-scale (up to nanoseconds) of the simulations. The laser induced buildup of a high pressure within the absorbing volume and generation of the pressure waves propagating from the absorption region poses an additional challenge for molecular-level simulation. A new dynamic boundary condition is developed to minimize the effects of the reflection of the wave from the boundary of the computational cell. The boundary condition accounts for the laser induced pressure wave propagation as well as the direct laser energy deposition in the boundary region.

We present a new method to calculate the response of dislocations to applied stress. This new method, called the dislocation treadmill, can be used to study the effect of vacancies, interstitials, stresses, strain rate, temperature etc., on the steady state velocity of the dislocation. We demonstrate the use of the method by calculating the response of a dislocation to a constant applied shear stress.

Following the Green-Kubo formalism in linear response theory, the lattice thermal conductivity of solid argon is determined by using classical molecular dynamics simulation to calculate the heat current correlation function. Comparing the absolute conductivities obtained using the Lennard-Jones potential with experiments, we find the predicted results to uniformly underestimate the measurements in magnitude, whereas the calculated temperature dependence corresponds well with the data. The temporal behavior of the heat current autocorrelation function shows that while a single exponential decay description is appropriate at elevated temperatures, below the half of the Debye temperature, the heat current relaxation clearly consists of two stages, an initial rapid decay associated with local dynamics followed by a slower component associated with the dynamics of lattice vibrations (phonons).

We are designing and developing a novel software environment - the Digital Material -to support materials modeling across many length and time scales, in order to develop improved descriptions of material structure and evolution. Software support is required for high-performance numerical kernels, lightweight infrastructures for prototyping, steering and analysis, information transfer across scales, coupling of disparate simulation modules, and collaboration between a multidisciplinary collection of researchers.

A new rate-independent constitutive model for plastic deformation of crystalline materials deforming by slip and twinning has been formulated, and implemented in a finite-element program. We have simulated three different structural levels by choosing representative volume elements (RVEs) as (i) a small part of a single crystal for the analysis of the heterogeneity of plastic deformation in single crystals, (ii) a whole single crystal for polycrystal simulations, and (iii) a group of crystals for a Taylor-type model of polycrystals. We show that the predictions for the texture and stress-strain response from the model are in reasonably good agreement with experiments in plane-strain compression for a different single crystal and polycrystalline f.c.c, materials.

The pressure-induced bcc-hcp transformation in iron is studied ab initio by following a constant-volume one-parameter transformation path. The calculations are performed in the generalized gradient approximation (GGA). It is shown that the analysis of both spin-polarized and non-spin-polarized states is essential in order to correctly obtain the energetics of the phase transformation. Our calculations make it possible to locate, with high precision, the transition configuration at which the bcc ferromagnetic structure transforms into the hcp nonmagnetic structure.

The bond-order potentials (BOP) have been constructed for Mo and Nb. These potentials are based on the real-space parametrized tight-binding method in which diagonalization of the Hamiltonian is avoided by direct calculation of the bond-order. In this scheme the energy consists of three parts: The bond part that comprises contributions of d electrons and introduces into the scheme the covalent character of bonding, the central-force many-body part that reflects the environmental dependence of sp overlap repulsion and a pair-wise contribution. The potentials were tested by calculation of energy differences between the bcc and several alternate structures and by investigating the trigonal deformation path. These calculations have been made in parallel using BOP and the full-potential linearized augmented plane-wave method. The central-force many-body Finnis-Sinclair type potentials have also been included into the study of the deformation path. This evaluation of BOP reveals that the potentials reproduce very closely the ab initio results and are, therefore, very suitable for atomistic studies of extended defects in the transition metals.

New embedded-atom potentials for Al and Ni have been developed by fitting to both experimental data and the results of ab initio calculations. The ab initio data were obtained in the form of energies of different alternative computer-generated crystalline structures of these metals. The potentials accurately reproduce basic equilibrium properties of Al and Ni such as the elastic constants, phonon dispersion curves, vacancy formation and migration energies, stacking fault energies, and surface energies. The equilibrium energies of various alternative structures not included in the fitting database are calculated with these potentials. The results are compared with predictions of total-energy tight-binding calculations for the same structures. The embedded-atom potentials correctly reproduce the structural stability trends, which suggests that they are transferable to different local environments encountered in atomistic simulations of lattice defects.

We apply a self-consistent charge tight-binding scheme to biomolecules. This method has been shown to give a reliable description of reaction energies, geometries and vibrational frequencies of small organic molecules. We discuss the performance of this method for model peptides and non-bonding interactions in biologically relevant molecular complexes. A comparison with semi-empirical methods and ab initio calculations will be given for DNA base pair H-bonding and stacking interactions.

A material model is proposed which describes single crystal creep behavior by evolution equations for dislocation densities on individual slip systems. An interaction matrix determines the influence from one glide system to the other. Assuming a face centered cubic crystal, allowing deformation on octahedral glide planes and cube glide planes with a Burgers vector of the type a/2 <110>, nine independent parameters of the interaction matrix can be distinguished. A parameter check of the nine independent parameters has been carried out, showing the influence of parameters on specific orientations of the load axis. If one assumes dislocation interaction of a glide system only with itself a smooth behavior is predicted with a maximum creep rate for [001] orientation, followed by [011] and [111]. If a strong interaction is assumed, the orientation dependent creep behavior is not at all smooth, instead it shows a sharp drop in creep rates mainly in symmetric positions of the standard orientation triangle. The orientations with highest creep rates are in this case those which favor single glide. Highly symmetric orientations, such as [001], [011] and [111] have strongly decreased stationary creep rates.

We combined empirical and ab initio methods to study structural and electronic properties of amorphous silicon nitride. For such study, we developed an interatomic potential to describe the interactions between silicon, nitrogen, and hydrogen atoms. Using this potential, we performed Monte Carlo simulations in a simulated annealing scheme to study structural properties of amorphous silicon nitride. Then this potential was used to generate relevant structures of a-SiNx:Hy which were input configurations to ab initio calculations. We investigated the electronic and structural role played by hydrogen incorporation in amorphous silicon nitride.

A modeling of nanoparticles and ab initio simulation of the scattered intensity from the Debye functions is used as a tool for an examination of the strain induced under high pressure in nanocrystalline silicon carbide and diamond. The analysis of the experimental intensity profiles includes a determination of the atomic structure and microstructure of the materials. The advantages of modeling over conventional methods of the analysis of powder diffraction data are discussed. Examples of using the modeling for determination of the shape and size and of one dimensional disordering in very small particles (2-4 nm), and development of internal strains in 10 nm SiC nanocrystals subjected to high pressures are given.

We have performed ab-initio Hartree-Fock molecular orbital calculations of local atomic structures and chemical bonding states in Si-N covalent amorphous ceramics. Solute elements such as boron, carbon and oxygen were considered in the Si-N network, and the bonding characteristics around the solute elements were analyzed. When a nitrogen atom is substituted by a carbon atom, it was found that Si-C bonds reinforce the Si-N network due to strong covalency.

Aluminum oxynitride (AlON) is a nitrogen-stabilized defective spinel phase of alumina (y-Al2O3). Its structure has not been well understood so far. Solid-state27 Al magic-angle spinning nuclear magnetic resonance and quantum chemistry calculation are used to characterize y-Al2O3 and AMON ceramics in the present study. The resonance spectra clearly show different units (e.g. [AlN4], [AlO4] and [AlO6]) and vary with composition. The calculation shows that the vacancy located at octahedral site is more stable than that located at the tetrahedral site, and nitrogen atoms preferentially replace oxygen atoms in the tetrahedral site, which is in agreement with the experimental analysis by nuclear magnetic resonance.

Structural characteristics, chemical bonds and thermoelectric properties of Si-doped boron carbides are studied through calculations of various structural unit models by using a self-consistent-field discrete variation Xα method. The calculations show that Si atom doped in boron carbide is in preference to substituting B or C atoms on the end of boron carbide chain, and then may occupy interstitial sites, but it is difficult for Si to substitute B or C atom in the centers of chain or in the icosahedra. A representative structural unit containing a Si atom is [C-B-Si]ε+ [B11C]ε-, while the structural unit without Si is [C-B-B(C)]δ--[B11C]δ+, and the coexistence of these two different structural units makes the electrical conductivity increases. As the covalent bond of Si-B or Si-C is weaker than that of B-B or B-C, the thermal conductivity decreases when Si is added into boron carbides. With the electrical conductivity increases and the thermal conductivity decreases, Si doping has significant effect on thermoelectric properties of boron carbides.