RESUMEN
Structural and vibrational properties of aqueous solutions of alkali hydroxides (LiOH, NaOH, and KOH) are computed using quantum molecular dynamics simulations for solute concentrations ranging between 1 and 10M. Element-resolved partial radial distribution functions, neutron and x-ray structure factors, and angular distribution functions are computed for the three hydroxide solutions as a function of concentration. The vibrational spectra and frequency-dependent conductivity are computed from the Fourier transforms of velocity autocorrelation and current autocorrelation functions. Our results for the structure are validated with the available neutron data for 17M concentration of NaOH in water [Semrouni et al., Phys. Chem. Chem. Phys. 21, 6828 (2019)]. We found that the larger ionic radius [rLi+
RESUMEN
We have developed an extension of the Neural Network Quantum Molecular Dynamics (NNQMD) simulation method to incorporate electric-field dynamics based on Born effective charge (BEC), called NNQMD-BEC. We first validate NNQMD-BEC for the switching mechanisms of archetypal ferroelectric PbTiO3 bulk crystal and 180° domain walls (DWs). NNQMD-BEC simulations correctly describe the nucleation-and-growth mechanism during DW switching. In triaxially strained PbTiO3 with strain conditions commonly seen in many superlattice configurations, we find that flux-closure texture can be induced with application of an electric field perpendicular to the original polarization direction. Upon field reversal, the flux-closure texture switches via a pair of transient vortices as the intermediate state, indicating an energy-efficient switching pathway. Our NNQMD-BEC method provides a theoretical guidance to study electro-mechano effects with existing machine learning force fields using a simple BEC extension, which will be relevant for engineering applications such as field-controlled switching in mechanically strained ferroelectric devices.
RESUMEN
The principle of least action is the cornerstone of classical mechanics, theory of relativity, quantum mechanics, and thermodynamics. Here, we describe how a neural network (NN) learns to find the trajectory for a Lennard-Jones (LJ) system that maintains balance in minimizing the Onsager-Machlup (OM) action and maintaining the energy conservation. The phase-space trajectory thus calculated is in excellent agreement with the corresponding results from the "ground-truth" molecular dynamics (MD) simulation. Furthermore, we show that the NN can easily find structural transformation pathways for LJ clusters, for example, the basin-hopping transformation of an LJ38 from an incomplete Mackay icosahedron to a truncated face-centered cubic octahedron. Unlike MD, the NN computes atomic trajectories over the entire temporal domain in one fell swoop, and the NN time step is a factor of 20 larger than the MD time step. The NN approach to OM action is quite general and can be adapted to model morphometrics in a variety of applications.
Asunto(s)
Simulación de Dinámica Molecular , Redes Neurales de la Computación , Fenómenos Biofísicos , TermodinámicaRESUMEN
Aramid fibers composed of poly(p-phenylene terephthalamide) (PPTA) polymers are attractive materials due to their high strength, low weight, and high shock resilience. Even though they have widely been utilized as a basic ingredient in Kevlar, Twaron, and other fabrics and applications, their intrinsic behavior under intense shock loading is still to be understood. In this work, we characterize the anisotropic shock response of PPTA crystals by performing reactive molecular dynamics simulations. Results from shock loading along the two perpendicular directions to the polymer backbones, [100] and [010], indicate distinct shock release mechanisms that preserve and destroy the hydrogen bond network. Shocks along the [100] direction for particle velocity Up < 2.46 km/s indicate the formation of a plastic regime composed of shear bands, where the PPTA structure is planarized. Shocks along the [010] direction for particle velocity Up < 2.18 km/s indicate a complex response regime, where elastic compression shifts to amorphization as the shock is intensified. While hydrogen bonds are mostly preserved for shocks along the [100] direction, hydrogen bonds are continuously destroyed with the amorphization of the crystal for shocks along the [010] direction. Decomposition of the polymer chains by cross-linking is triggered at the threshold particle velocity Up = 2.18 km/s for the [010] direction and Up = 2.46 km/s for the [100] direction. These atomistic insights based on large-scale simulations highlight the intricate and anisotropic mechanisms underpinning the shock response of PPTA polymers and are expected to support the enhancement of their applications.
RESUMEN
The static dielectric constant ϵ_{0} and its temperature dependence for liquid water is investigated using neural network quantum molecular dynamics (NNQMD). We compute the exact dielectric constant in canonical ensemble from NNQMD trajectories using fluctuations in macroscopic polarization computed from maximally localized Wannier functions (MLWF). Two deep neural networks are constructed. The first, NNQMD, is trained on QMD configurations for liquid water under a variety of temperature and density conditions to learn potential energy surface and forces and then perform molecular dynamics simulations. The second network, NNMLWF, is trained to predict locations of MLWF of individual molecules using the atomic configurations from NNQMD. Training data for both the neural networks is produced using a highly accurate quantum-mechanical method, DFT-SCAN that yields an excellent description of liquid water. We produce 280×10^{6} configurations of water at 7 temperatures using NNQMD and predict MLWF centers using NNMLWF to compute the polarization fluctuations. The length of trajectories needed for a converged value of the dielectric constant at 0°C is found to be 20 ns (40×10^{6} configurations with 0.5 fs time step). The computed dielectric constants for 0, 15, 30, 45, 60, 75, and 90°C are in good agreement with experiments. Our scalable scheme to compute dielectric constants with quantum accuracy is also applicable to other polar molecular liquids.
RESUMEN
Despite the growing success of machine learning for predicting structure-property relationships in molecules and materials, such as predicting the dielectric properties of polymers, it is still in its infancy. We report on the effectiveness of solving structure-property relationships for a computer-generated database of dielectric polymers using recurrent neural network (RNN) models. The implementation of a series of optimization strategies was crucial to achieving high learning speeds and sufficient accuracy: (1) binary and nonbinary representations of SMILES (Simplified Molecular Input Line System) fingerprints and (2) backpropagation with affine transformation of the input sequence (ATransformedBP) and resilient backpropagation with initial weight update parameter optimizations (iRPROP- optimized). For the investigated database of polymers, the binary SMILES representation was found to be superior to the decimal representation with respect to the training and prediction performance. All developed and optimized Elman-type RNN algorithms outperformed nonoptimized RNN models in the efficient prediction of nonlinear structure-activity relationships. The average relative standard deviation (RSD) remained well below 5%, and the maximum RSD did not exceed 30%. Moreover, we provide a C++ codebase as a testbed for a new generation of open programming languages that target increasingly diverse computer architectures.
Asunto(s)
Redes Neurales de la Computación , Polímeros , Algoritmos , Bases de Datos Factuales , Aprendizaje AutomáticoRESUMEN
Scandium-doped aluminum nitride, Al1-xScxN, represents a new class of displacive ferroelectric materials with high polarization and sharp hysteresis along with high-temperature resilience, facile synthesizability and compatibility with standard CMOS fabrication techniques. The fundamental physics behind the transformation of unswitchable piezoelectric AlN into switchable Al-Sc-N ferroelectrics depends upon important atomic properties such as local structure, dopant distributions and the presence of competing mechanism of polarization switching in the presence of an applied electric-field that have not been understood. We computationally synthesize Al1-xScxN to quantify the inhomogeneity of Sc distribution and phase segregation, and characterize its crystal and electronic structure as a function of Sc-doping. Nudged elastic band calculations of the potential energy surface and quantum molecular dynamics simulations of direct electric-field-driven ferroelectric switching reveal a crossover between two polarization reversal mechanisms-inhomogeneous nucleation-and-growth mechanism originating near Sc-rich regions in the limit of low applied fields and nucleation-limited-switching in the high-field regime. Understanding polarization reversal pathways for these two mechanisms as well as the role of local Sc concentration on activation barriers provides design rules to identify other combinations of dopant elements, such as Zr, Mg etc. to synthesize superior AlN-based ferroelectric materials.
RESUMEN
A thorough understanding of native oxides is essential for designing semiconductor devices. Here, we report a study of the rate and mechanisms of spontaneous oxidation of bulk single crystals of ZrSxSe2-x alloys and MoS2. ZrSxSe2-x alloys oxidize rapidly, and the oxidation rate increases with Se content. Oxidation of basal surfaces is initiated by favorable O2 adsorption and proceeds by a mechanism of Zr-O bond switching, that collapses the van der Waals gaps, and is facilitated by progressive redox transitions of the chalcogen. The rate-limiting process is the formation and out-diffusion of SO2. In contrast, MoS2 basal surfaces are stable due to unfavorable oxygen adsorption. Our results provide insight and quantitative guidance for designing and processing semiconductor devices based on ZrSxSe2-x and MoS2 and identify the atomistic-scale mechanisms of bonding and phase transformations in layered materials with competing anions.
RESUMEN
Reverse osmosis through a polyamide (PA) membrane is an important technique for water desalination and purification. In this study, molecular dynamics simulations were performed to study the biofouling mechanism (i.e., protein adsorption) and nonequilibrium steady-state water transfer of a cross-linked PA membrane. Our results demonstrated that the PA membrane surface's roughness is a key factor of surface's biofouling, as the lysozyme protein adsorbed on the surface's cavity site displays extremely low surface diffusivity, blocking water passage, and decreasing water flux. The adsorbed protein undergoes secondary structural changes, particularly in the pressure-driven flowing conditions, leading to strong protein-surface interactions. Our simulations were able to present water permeation close to the experimental conditions with a pressure difference as low as 5 MPa, while all the electrolytes, which are tightly surrounded by hydration water, were effectively rejected at the membrane surfaces. The analysis of the self-intermediate scattering function demonstrates that the dynamics of water molecules coordinated with hydrogen bonds is faster inside the pores than during the translation across the pores. The pressure difference applied shows a negligible effect on the water structure and content inside the membrane but facilitates the transportation of hydrogen-bonded water molecules through the membrane's sub-nanopores with a reduced coordination number. The linear relationship between the water flux and the pressure difference demonstrates the applicability of continuum hydrodynamic principles and thus the stability of the membrane structure.
RESUMEN
The light-induced selective population of short-lived far-from-equilibrium vibration modes is a promising approach for controlling ultrafast and irreversible structural changes in functional nanomaterials. However, this requires a detailed understanding of the dynamics and evolution of these phonon modes and their coupling to the excited-state electronic structure. Here, we combine femtosecond mega-electronvolt electron diffraction experiments on a prototypical layered material, MoTe2, with non-adiabatic quantum molecular dynamics simulations and ab initio electronic structure calculations to show how non-radiative energy relaxation pathways for excited electrons can be tuned by controlling the optical excitation energy. We show how the dominant intravalley and intervalley scattering mechanisms for hot and band-edge electrons leads to markedly different transient phonon populations evident in electron diffraction patterns. This understanding of how tuning optical excitations affect phonon populations and atomic motion is critical for efficiently controlling light-induced structural transitions of optoelectronic devices.
RESUMEN
Two-dimensional (2D) transition metal dichalcogenide (TMDC) heterostructures have been proposed as potential candidates for a variety of applications like quantum computing, neuromorphic computing, solar cells, and flexible field effective transistors. The 2D TMDC heterostructures at the present stage face difficulties being implemented in these applications because of lack of large and sharp heterostructure interfaces. Herein, we address this problem via a CVD technique to grow thermodynamically stable heterostructure of 2H/1T' MoSe2-ReSe2 using conventional transition metal phase diagrams as a reference. We demonstrate how the thermodynamics of mixing in the MoReSe2 system during CVD growth dictates the formation of atomically sharp interfaces between MoSe2 and ReSe2, which can be confirmed by high-resolution scanning transmission electron microscopy imaging, revealing zigzag selenium-terminated interface between the epitaxial 2H and 1T' lattices. Our work provides useful insights for understanding the stability of 2D heterostructures and interfaces between chemically, structurally, and electronically different phases.
RESUMEN
Two-dimensional transition metal dichalcogenides (TMDs) draw strong interest in materials science, with applications in optoelectronics and many other fields. Good performance requires high carrier concentrations and long lifetimes. However, high concentrations accelerate energy exchange between charged particles by Auger-type processes, especially in TMDs where many-body interactions are strong, thus facilitating carrier trapping. We report time-resolved optical pump-THz probe measurements of carrier lifetimes as a function of carrier density. Surprisingly, the lifetime reduction with increased density is very weak. It decreases only by 20% when we increase the pump fluence 100 times. This unexpected feature of the Auger process is rationalized by our time-domain ab initio simulations. The simulations show that phonon-driven trapping competes successfully with the Auger process. On the one hand, trap states are relatively close to band edges, and phonons accommodate efficiently the electronic energy during the trapping. On the other hand, trap states localize around defects, and the overlap of trapped and free carriers is small, decreasing carrier-carrier interactions. At low carrier densities, phonons provide the main charge trapping mechanism, decreasing carrier lifetimes compared to defect-free samples. At high carrier densities, phonons suppress Auger processes and lower the dependence of the trapping rate on carrier density. Our results provide theoretical insights into the diverse roles played by phonons and Auger processes in TMDs and generate guidelines for defect engineering to improve device performance at high carrier densities.
RESUMEN
First-principles molecular dynamics (FPMD) simulations are highly accurate, but due to their high calculation cost, the computational scale is often limited to hundreds of atoms and few picoseconds under specific temperature and pressure conditions. We present here the guidelines for creating artificial neural network empirical interatomic potential (ANN potential) trained with such a limited FPMD data, which can perform long time scale MD simulations at least under the same conditions. The FPMD data for training are prepared on the basis of the convergence of radial distribution function [g(r)]. While training the ANN using total energy and atomic forces of the FPMD data, the error of pressure is also monitored and minimized. To create further robust potential, we add a small amount of FPMD data to reproduce the interaction between two atoms that are close to each other. ANN potentials for α-Ag2Se were created as an application example, and it has been confirmed that not only g(r) and mean square displacements but also the specific heat requiring a long time scale simulation matched the FPMD and the experimental values. In addition, the MD simulation using the ANN potential achieved over 104 acceleration over the FPMD one. The guidelines proposed here mitigate the creation difficulty of the ANN potential, and a lot of FPMD data sleeping on the hard disk after the research may be put on the front stage again.
RESUMEN
Atomically thin layers of transition metal dichalcogenide (TMDC) semiconductors exhibit outstanding electronic and optical properties, with numerous applications such as valleytronics. While unusually rapid and efficient transfer of photoexcitation energy to atomic vibrations was found in recent experiments, its electronic origin remains unknown. Here, we study the lattice dynamics induced by electronic excitation in a model TMDC monolayer, MoSe2, using nonadiabatic quantum molecular dynamics simulations. Simulation results show sub-picosecond disordering of the lattice upon photoexcitation, as measured by the Debye-Waller factor, as well as increasing disorder for higher densities of photogenerated electron-hole pairs. Detailed analysis shows that the rapid, photoinduced lattice dynamics are due to phonon-mode softening, which in turn arises from electronic Fermi surface nesting. Such mechanistic understanding can help guide optical control of material properties for functionalizing TMDC layers, enabling emerging applications such as phase change memories and neuromorphic computing.
RESUMEN
Transition metal dichalcogenides (TMDC) like MoS2 are promising candidates for next-generation electric and optoelectronic devices. These TMDC monolayers are typically synthesized by chemical vapor deposition (CVD). However, despite significant amount of empirical work on this CVD growth of monolayered crystals, neither experiment nor theory has been able to decipher mechanisms of selection rules for different growth scenarios, or make predictions of optimized environmental parameters and growth factors. Here, we present an atomic-scale mechanistic analysis of the initial sulfidation process on MoO3 surfaces using first-principles-informed ReaxFF reactive molecular dynamics (RMD) simulations. We identify a three-step reaction process associated with synthesis of the MoS2 samples from MoO3 and S2 precursors: O2 evolution and self-reduction of the MoO3 surface; SO/SO2 formation and S2-assisted reduction; and sulfidation of the reduced surface and Mo-S bond formation. These atomic processes occurring during early stage MoS2 synthesis, which are consistent with experimental observations and existing theoretical literature, provide valuable input for guided rational synthesis of MoS2 and other TMDC crystals by the CVD process.
RESUMEN
Rotation of methylammonium (CH3NH3 or MA) molecules is believed to govern the excellent transport properties of photocarriers in the MA lead iodide (MAPbI3) perovskite. Of particular interest is its cubic phase, which exists in industrially important films at room temperature. In order to investigate the rotational behaviors of the MA molecules, we have performed ab initio molecular dynamics simulations of cubic-MAPbI3 at room temperature. There are two types of rotational motions of MA molecules in a crystalline PbI3 cage: reorientation of a whole molecule and intramolecular rotation around the C-N bond within MA molecules. Using a cubic symmetry-assisted analysis (CSAA), we found that the prominent orientation of the C-N bond is the crystalline ⟨110⟩ directions, rather than the ⟨100⟩ and ⟨111⟩ directions. Rapid rotation around the C-N bond is also observed, which easily occurs when the rotational axis is parallel to the ⟨110⟩ directions according to the CSAA. To explain the atomistic mechanisms underlying these CSAA results, we have focused on the relation between H-I hydrogen bonds and the orientation of an MA molecule. Here, the hydrogen bonds were defined by population analysis, and it has been found that, while H atoms in the CH3 group (HC) hardly interacts with I atoms, those in the NH3 group (HN) form at least one hydrogen bond with I atoms and their interatomic distances are in a wide range, 2.2-3.7 Å. Based on these findings, we have given a possible explanation to why the ⟨110⟩ directions are preferred. Namely, the atomic arrangement and interatomic distance between MA and surrounding I atoms are most suitable for the formation of hydrogen bonds. In addition to films, these results are potentially applicable to the rotational behaviors in bulk MAPbI3 as well, considering that the atomistic structure and time constants regarding the rotation of MA molecules statistically agree with bulk experiments.
RESUMEN
Hydrogen production from water using Al particles could provide a renewable energy cycle. However, its practical application is hampered by the low reaction rate and poor yield. Here, large quantum molecular dynamics simulations involving up to 16,611 atoms show that orders-of-magnitude faster reactions with higher yields can be achieved by alloying Al particles with Li. A key nanostructural design is identified as the abundance of neighboring Lewis acid-base pairs, where water-dissociation and hydrogen-production require very small activation energies. These reactions are facilitated by charge pathways across Al atoms that collectively act as a "superanion" and a surprising autocatalytic behavior of bridging Li-O-Al products. Furthermore, dissolution of Li atoms into water produces a corrosive basic solution that inhibits the formation of a reaction-stopping oxide layer on the particle surface, thereby increasing the yield. These atomistic mechanisms not only explain recent experimental findings but also predict the scalability of this hydrogen-on-demand technology at industrial scales.
RESUMEN
We introduce an extension of the divide-and-conquer (DC) algorithmic paradigm called divide-conquer-recombine (DCR) to perform large quantum molecular dynamics (QMD) simulations on massively parallel supercomputers, in which interatomic forces are computed quantum mechanically in the framework of density functional theory (DFT). In DCR, the DC phase constructs globally informed, overlapping local-domain solutions, which in the recombine phase are synthesized into a global solution encompassing large spatiotemporal scales. For the DC phase, we design a lean divide-and-conquer (LDC) DFT algorithm, which significantly reduces the prefactor of the O(N) computational cost for N electrons by applying a density-adaptive boundary condition at the peripheries of the DC domains. Our globally scalable and locally efficient solver is based on a hybrid real-reciprocal space approach that combines: (1) a highly scalable real-space multigrid to represent the global charge density; and (2) a numerically efficient plane-wave basis for local electronic wave functions and charge density within each domain. Hybrid space-band decomposition is used to implement the LDC-DFT algorithm on parallel computers. 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 × 10(6)-atom SiC system. As a test of production runs, LDC-DFT-based QMD simulation involving 16 661 atoms is performed on the Blue Gene/Q to study on-demand production of hydrogen gas from water using LiAl alloy particles. As an example of the recombine phase, LDC-DFT electronic structures are used as a basis set to describe global photoexcitation dynamics with nonadiabatic QMD (NAQMD) and kinetic Monte Carlo (KMC) methods. The NAQMD simulations are based on the linear response time-dependent density functional theory to describe electronic excited states and a surface-hopping approach to describe transitions between the excited states. A series of techniques are employed for efficiently calculating the long-range exact exchange correction and excited-state forces. The NAQMD trajectories are analyzed to extract the rates of various excitonic processes, which are then used in KMC simulation to study the dynamics of the global exciton flow network. This has allowed the study of large-scale photoexcitation dynamics in 6400-atom amorphous molecular solid, reaching the experimental time scales.
RESUMEN
The oscillatory retinal neuron (ORN) is a promising technology for achieving in-sensor cognitive image computing without external power. While its operation is based on photoinduced negative differential resistance (NDR) at a graphene/silicon interface to directly convert the incident optical signal into voltage oscillations, the optoelectronic mechanism of the NDR remains elusive. Here, nonadiabatic quantum molecular dynamics simulations show that the interplay of band alignment and charge transfer rates of photoexcited carriers at varying applied voltages gives rise to NDR at a graphene/silicon interface under illumination. Such intrinsic NDR at an interface, along with extrinsic circuit-level factors, could enable the much needed rational design of desired image computing functionality of ORN devices in the era of ubiquitous AI on edge devices.
RESUMEN
Diffusion in solids is a slow process that dictates rate-limiting processes in key chemical reactions. Unlike crystalline solids that offer well-defined diffusion pathways, the lack of similar structural motifs in amorphous or glassy materials poses great challenges in bridging the slow diffusion process and material failures. To tackle this problem, we propose an AI-guided long-term atomistic simulation approach: molecular autonomous pathfinder (MAP) framework based on deep reinforcement learning (DRL), where the RL agent is trained to uncover energy efficient diffusion pathways. We employ a Deep Q-Network architecture with distributed prioritized replay buffer, enabling fully online agent training with accelerated experience sampling by an ensemble of asynchronous agents. After training, the agents provide atomistic configurations of diffusion pathways with their energy profile. We use a piecewise nudged elastic band to refine the energy profile of the obtained pathway and the corresponding diffusion time on the basis of transition-state theory. With the MAP framework, we demonstrate atomistic diffusion mechanisms in amorphous silica with time scales comparable to experiments.