Dynamic fracture of a two-dimensional MoWSe2 membrane is studied with molecular dynamics (MD) simulation. The system consists of a random distribution of WSe2 patches in a pre-cracked matrix of MoSe2. Under strain, the system shows toughening due to crack branching, crack closure and strain-induced structural phase transformation from 2H to 1T crystal structures. Different structures generated during MD simulation are analyzed using a three-layer, feed-forward neural network (NN) model. A training data set of 36,000 atoms is created where each atom is represented by a 50-dimension feature vector consisting of radial and angular symmetry functions. Hyper parameters of the symmetry functions and network architecture are tuned to minimize model complexity with high predictive power using feature learning, which shows an increase in model accuracy from 67% to 95%. The NN model classifies each atom in one of the six phases which are either as transition metal or chalcogen atoms in 2H phase, 1T phase and defects. Further t-SNE analyses of learned representation of these phases in the hidden layers of the NN model show that separation of all phases become clearer in the third layer than in layers 1 and 2.

Rapid transitions between semiconducting and metallic phases of transition-metal dichalcogenides are of interest for 2D electronics applications. Theoretical investigations have been limited to using thermal energy, lattice strain and charge doping to induce the phase transition, but have not identified mechanisms for rapid phase transition. Here, we use density functional theory to show how optical excitation leads to the formation of a low-energy intermediate crystal structure along the semiconductor-metal phase transition pathway. This metastable crystal structure results in significantly reduced barriers for the semiconducting-metal phase transition pathway leading to rapid transition in optically excited crystals.

Ultrafast atomic dynamics induced by electronic and optical excitation opens new possibilities for functionalization of two-dimensional and layered materials. Understanding the impact of perturbed valence band populations on both the strong covalent bonds and relatively weaker van der Waals interactions is important for these anisotropic systems. While the dynamics of strong covalent bonds has been explored both experimentally and theoretically, relatively fewer studies have focused on the impact of excitation on weak bonds like van der Waals and hydrogen-bond interactions. We perform non-adiabatic quantum molecular dynamics (NAQMD) simulations to study photo-induced dynamics in MoS2 bilayer. We observe photo-induced non-thermal contraction of the interlayer distance in the MoS2 bilayer within 100 femtoseconds after photoexcitation. We identify a large photo-induced redistribution of electronic charge density, whose Coulombic interactions could explain the observed inter-layer contraction.

Monolayers of semiconducting transitional metal dichalcogenides (TMDC) are emerging as strong candidate materials for next generation electronic and optoelectronic devices, with applications in field-effect transistors, valleytronics, and photovoltaics. Prior studies have demonstrated strong light-matter interactions in these materials, suggesting optical control of material properties as a promising route for their functionalization. However, the electronic and structural dynamics in response to electronic excitation have not yet been fully elucidated. In this work, we use non-adiabatic quantum molecular dynamics simulations based on time-dependent density functional theory to study lattice dynamics of a model TMDC monolayer of MoSe2 after electronic excitation. The simulation results show rapid, sub-picosecond lattice response, as well as finite-size effects. Understanding the sub-picosecond atomic dynamics is important for the realization of optical control of the material properties of monolayer TMDCs, which is a hopeful, straightforward tactic for functionalizing these materials.

Transition metal dichalcogenide (TMDC) monolayers like MoS2 are promising materials for future electronic applications. Large-area monolayer MoS2 samples for these applications are typically synthesized by chemical vapor deposition (CVD) using MoO3 reactants and gas-phase sulfur precursors. Recent experimental studies have greatly improved our understanding of reaction pathways in the CVD growth process. However, atomic mechanisms of sulfidation process remain to be fully elucidated. In this work, we present quantum-mechanically informed and validated reactive molecular dynamics (RMD) simulations for CVD synthesis of MoS2 layers using S2 precursors. Our RMD simulations clarify atomic-level reaction pathways for the sulfidation of MoO3 surfaces by S2, which is a critical reaction step for CVD synthesis of MoS2 layers. These results provide a better understanding of the sulfidation process for the scalable synthesis of defect-free MoS2 and other TMDC materials.

Vertical hetero-structures made from stacked monolayers of transition metal dichalcogenides (TMDC) are promising candidates for next-generation optoelectronic and thermoelectric devices. Identification of optimal layered materials for these applications requires the calculation of several physical properties, including electronic band structure and thermal transport coefficients. However, exhaustive screening of the material structure space using ab initio calculations is currently outside the bounds of existing computational resources. Furthermore, the functional form of how the physical properties relate to the structure is unknown, making gradient-based optimization unsuitable. Here, we present a model based on the Bayesian optimization technique to optimize layered TMDC hetero-structures, performing a minimal number of structure calculations. We use the electronic band gap and thermoelectric figure of merit as representative physical properties for optimization. The electronic band structure calculations were performed within the Materials Project framework, while thermoelectric properties were computed with BoltzTraP. With high probability, the Bayesian optimization process is able to discover the optimal hetero-structure after evaluation of only ∼20% of all possible 3-layered structures. In addition, we have used a Gaussian regression model to predict not only the band gap but also the valence band maximum and conduction band minimum energies as a function of the momentum.

This study uses ab initio quantum molecular dynamics (QMD) simulations to validate multimillion-atom reactive molecular dynamics (RMD) simulations, and predicts unexpected condensation of carbon atoms during high-temperature oxidation of silicon-carbide nanoparticles (nSiC). For the validation process, a small nSiC in oxygen environment is chosen to perform QMD simulation. The QMD results provide the number of Si-O and C-O bonds as a function of time. RMD simulation is then performed under the identical condition. The time evolutions of different bonds are compared between the QMD and RMD simulations. We observe the condensation of large number of C-cluster nuclei into larger C clusters in both simulations, thereby validating RMD. Furthermore, we use the QMD simulation results as an input to a multi-objective genetic algorithm to train the RMD force-field parameters. The resulting force field far better reproduces the ground-truth QMD simulation results.

Oxidation behavior of aggregated aluminum nanoparticles (Al-NPs), specifically the combustion propagation, is studied, when only part of the aggregated Al-NPs is heated to 1100 K and the rest of the system is kept at 300 K. Here, multi-million atoms molecular dynamics (MD) simulation reveals the sintering/coalescence phenomena for the different diameters (D = 26, 36 and 46 nm) aggregated systems. Various consuming rates of core aluminum are investigated for different layers and different diameters aggregated systems. The formation of Al2O3 fragments outside the shell (the largest covalently bonded aluminum-oxide cluster) structure is confirmed from AlO and AlO2 intermediates. The smaller size of Al-NPs results in faster trend of transition from Al-rich to O-rich for most outside small clusters. However, more core aluminum reacts with shell oxygen leads to faster decreasing of the ratio of O/Al in the shell fragment for larger Al-NPs system.

Multimillion-atom reactive molecular dynamics (RMD) and large quantum molecular dynamics (QMD) simulations are used to investigate structural and dynamical correlations under highly nonequilibrium conditions and reactive processes in nanostructured materials under extreme conditions. This paper discusses four simulations:

1. 1. RMD simulations of heated aluminum nanoparticles have been performed to study the fast oxidation reaction processes of the core (aluminum)-shell (alumina) nanoparticles and small complexes.

2. 2. Cavitation bubbles readily occur in fluids subjected to rapid changes in pressure. We have used billion-atom RMD simulations on a 163,840-processor Blue Gene/P supercomputer to investigate chemical and mechanical damages caused by shock-induced collapse of nanobubbles in water near silica surface. Collapse of an empty nanobubble generates high-speed nanojet, resulting in the formation of a pit on the surface. The gas-filled bubbles undergo partial collapse and consequently the damage on the silica surface is mitigated.

3. 3. Our QMD simulation reveals rapid hydrogen production from water by an Al superatom. We have found a low activation-barrier mechanism, in which a pair of Lewis acid and base sites on the Aln surface preferentially catalyzes hydrogen production.

4. 4. We have introduced an extension of the divide-and-conquer (DC) algorithmic paradigm called divide-conquer-recombine (DCR) to perform large QMD simulations on massively parallel supercomputers, in which interatomic forces are computed quantum mechanically in the framework of density functional theory (DFT). A benchmark test on an IBM Blue Gene/Q computer exhibits an isogranular parallel efficiency of 0.984 on 786,432 cores for a 50.3 million-atom SiC system. As a test of production runs, LDC-DFT-based QMD simulation involving 16,661 atoms was performed on the Blue Gene/Q to study on-demand production of hydrogen gas from water using LiAl alloy particles.

Understanding of combustion of metastable intermolecular composites, including the burning of aluminum nanoparticles, is critical for broad applications such as propulsion, explosives and other pyrotechnics. Aluminum nanorods (Al-NR) with oxidized shells are good candidates for stable fuel-oxidizer combinations. We investigate the oxidation dynamics of Al-NRs of different diameters (26, 36 and 46 nm) but the same aspect ratio using molecular dynamics simulations. We heat one end of the Al-NR to 1100 K and then study the oxidation reaction at the interface of the alumina shell and the Al core. We find: (1) heat produced by oxidation causes the melting of nanorods; (2) heat release is accelerated due to Al-O reaction at outside-shell and core-shell interfaces; and (3) the larger surface-to-volume ratio causes faster burning of thinner nanorods. We present results for the oxidation speed of nanorods.

Multimillion-atom molecular dynamics simulations are used to investigate burning behavior of a chain of three alumina-coated aluminum nanoparticles (ANPs), where particles one and three are heated above the melting temperature of pure aluminum. The mode and mechanism behind the heat and mass transfer from the hot ANPs (particles one and three) to the middle, cold ANP (particle two) are studied. The hot nanoparticles oxidize first, after which hot Al atoms penetrate into the cold nanoparticle. It is also found that due to the penetration of hot Al atoms, the cold nanoparticle oxidizes at a faster rate than in the initially heated nanoparticles. The calculated speed of penetration is found to be 54 m/s, which is within the range of experimentally measured flame propagation rates. As the atoms penetrate into the central ANP, they maintain their relative positions. The atoms from the shell of the central ANP form the first layer, which is followed by the atoms from the shell of the outer ANP making the second layer and lastly the atoms from the core of the outer ANPs form the third layer. In addition to heating the central ANP by convection, the ejected hot Al atoms from the outer ANPs initiate exothermic oxidation reactions inside the central ANP, leading to further heating within the central ANP. During 1 ns, all three ANPs fuse together, forming a single ellipsoidal aggregate.

Multimillion-atom reactive molecular dynamics simulations were used to investigate the mechanisms which control heat-initiated oxidation in aluminum nanoparticles. The simulation results reveal three stages: (1) confined burning, (2) onset of deformation, and (3) onset of small cluster ejections. The first stage of the reaction is localized primarily at the core-shell boundary, where oxidation reactions result in strong local heating and the increased migration of oxygen from the shell into the core. When the local temperature rises above the melting point of alumina (T=2330K), the melting of the shell allows deformation of the overall particle and an increase in heat production. Finally, once the particle temperature exceeds 2800-3000 K, small aluminum-rich clusters are ejected from the outside of the shell. The underlying mechanisms were explored using global and radial statistical analysis, as well as developed visualization techniques and localized fragment analysis.

The three-stage reaction mechanism found here provides insight into the controlling factors of aluminum nanoparticle oxidation, a topic of considerable importance in the energetic materials community.

Oxidation dynamics of three different sizes (26, 36 and 46 nm) of single aluminum nanoparticle (ANP) in oxygen environment are studied using multimillion-atom reactive molecular dynamics simulations. In the simulation, each aluminum nanoparticle is coated with an amorphous alumina shell of the same thickness (3 nm), and is ignited by heating the nanoparticle to 1100 K. The metallic aluminum and ceramic alumina are modeled by the Voter- Chen embedded atom model and the interatomic potential by Vashishta et al., respectively. Energy release rate and atomistic-level details of combustion of these single aluminum nanoparticles are investigated, along with the effect of nanoparticle size. The onset temperature of shell Al ejection is found to be independent of the ANP size, whereas the onset time of ejection and the time delay to the highest temperature change rate dT/dt depend on the size.

Thermal properties of amorphous silicon carbide (a-SiC) at nanometric scales are studied by molecular dynamics (MD) simulations based on an empirical interatomic potential. A scalable parallel MD algorithm is used for studying systems as long as 30nm. To validate the potential, phonon density of states and specific heat of a-SiC are first calculated. Size effects are studied, and errors are estimated for the temperature profile for different system sizes. Simulation time required to achieve steady temperature profiles is also determined. Finally the thermal conductivities of a-SiC at various temperatures are calculated. The results show that thermal conductivities of a-SiC at nanometric scale also agree with Slack's minimum thermal conductivity model.

The broader context of this discussion, based on a workshop where materials technologists and computational scientists engaged in a dialogue, is an awareness that modeling and simulation techniques and computational capabilities may have matured sufficiently to provide heretofore unavailable insights into the complex microstructural evolution of materials in extreme environments.As an example, this article examines the study of ultrahigh-temperature oxidation-resistant ceramics, through the combination of atomistic simulation and selected experiments.We describe a strategy to investigate oxygen transport through a multi-oxide scale—the protective layer of ultrahigh-temperature ceramic composites ZrB2-SiC and HfB2-SiC—by combining first-principles and atomistic modeling and simulation with selected experiments.

We investigate mechanisms of stress corrosion cracking in Si using a hybrid quantum-mechanical/molecular-dynamics simulation code developed recently for parallel computers. We perform the simulation for a cracked Si-model under tension (mode-I opening) with three H2O molecules around the crack front to investigate possible effects of both saturation of dangling bonds of Si with hydrogen atoms and environment molecules on the fracture initiation. Our results demonstrate existence of a path for an H2O molecule to react with Si-Si bonds at the crack front in contrast to a previous theoretical study based on the molecular orbital theory [W. Wong-Ng et al., Comp. Mater. Sci. 6, 63 (1996)].

Parallel molecular dynamics simulations are performed to investigate dynamic fracture in bulk and nanostructured silica glasses at room temperature and 1000 K. In bulk silica the crack front develops multiple branches and nanoscale pores open up ahead of the crack tip. Pores coalesce and then they merge with the advancing crack-front to cause cleavage fracture. The calculated fracture toughness is in good agreement with experiments. In nanostrucutred silica the crack-front meanders along intercluster boundaries, merging with nanoscale pores in these regions to cause intergranular fracture. The failure strain in nanostructured silica is significantly larger than in the bulk systems.

A hybrid molecular-dynamics/finite-element simulation scheme is applied to describe multiscale phenomena in nanodevices. The quality of both static and dynamic coupling between atomistic and continuum regions is studied. The hybrid scheme is used for the Si/Si3N4 interface problem (static coupling), and for the projectile impact on Si problem (dynamic coupling). Excellent agreement is found between hybrid and full molecular dynamics simulation results in the static case, and no wave reflections are found at the atomistic/continuum hand-shake in the dynamic case. The hybrid scheme is thus validated a powerful and cost effective method for performing multiscale simulations of nanodevices.

Variable-charge molecular dynamics simulation of 32 TiO2-nanoparticles with diameter 60Å is performed for 40 ps at 1 GPa and 1,400 K for both rutile and anatase phases, to investigate their phase-dependent sintering mechanisms. In the rutile case, the nanoparticles rotate around their centers during the first 20 ps. Varying degrees of neck formation between neighboring rutile-nanoparticles are found at ∼ 40 ps. In the anatase case, the nanoparticles maintain their original orientations. Similar degrees of neck formation are observed at contacting regions of the anatase nanoparticles.

A hybrid quantum mechanical/molecular dynamics simulation scheme is developed by embedding a quantum mechanical system described by the real-space density-functional theory in a classical system of atoms interacting via an empirical interatomic potential. A novel scaled position method for handshake atoms coupling the quantum and the classical systems is introduced. Hybrid simulation run for oxidation of Si (100) surface is performed to demonstrate seamless coupling of the quantum and the classical systems.