本文构建重现 HfO$_2$ r2SCAN 势能面的神经网络力场，并利用可微 NPT 分子动力学模拟揭示其约 2000 K 和 3000 K 的相变行为，结果与实验衍射数据吻合。
摘要：
机器学习力场的进展为那些从头算 MD 资源消耗过高的化合物以及经典力场不足以描述的现象开启了分子动力学（MD）模拟。在此，我们描述了一种神经网络力场，其参数经过设置以重现 HfO$_2$ 的 r2SCAN 势能面。基于具有灵活晶胞涨落的等温-等压（NPT）系综的自动可微实现，我们研究了 HfO$_2$ 的相空间。我们发现该方法在晶格常数和实验 X 射线衍射数据方面具有出色的预测能力。在约 2000 K 的温度下，可以清楚地看到脱离单斜相的相变，这与现有实验数据和先前的计算结果一致。晶格常数的另一次突变发生在约 3000 K。虽然所得晶格常数更接近立方相，但它们表现出较小的四方畸变，并且没有伴随的体积变化。我们表明，这种高温结构与现有的高温衍射数据一致。
英文摘要：
The advances of machine-learned force fields have opened up molecular dynamics (MD) simulations for compounds for which ab-initio MD is too resource-intensive and phenomena for which classical force fields are insufficient. Here we describe a neural-network force field parametrized to reproduce the r2SCAN potential energy landscape of HfO$_2$. Based on an automatic differentiable implementation of the isothermal-isobaric (NPT) ensemble with flexible cell fluctuations, we study the phase space of HfO$_2$. We find excellent predictive capabilities regarding the lattice constants and experimental X-ray diffraction data. The phase transition away from monoclinic is clearly visible at a temperature around 2000 K, in agreement with available experimental data and previous calculations. Another abrupt change in lattice constants occurs around 3000 K. While the resulting lattice constants are closer to cubic, they exhibit a small tetragonal distortion, and there is no associated change in volume. We show that this high-temperature structure is in agreement with the available high-temperature diffraction data.
We have conducted a detailed thin film growth structure of oxygen engineered monoclinic HfO$_{2\pm x}$ grown by reactive molecular beam epitaxy (MBE). The oxidation conditions induce a switching between ($\bar{1}11$) and (002) texture of hafnium oxide. The band gap of oxygen deficient hafnia decreases with increasing amount of oxygen vacancies by more than 1 eV. For high oxygen vacancy concentrations, defect bands form inside the band gap that induce optical transitions and $p$-type conductivity. The resistivity changes by several orders of magnitude as a function of oxidation conditions. Oxyg…
Demonstrate of High-Performance Top-Gate ALD Crystalline In2O3 Transistor Enabled by Lattice-Matched HfO2 and In2O3 Heterostructure
National Key Laboratory of Advanced Micro and Nano Manufacture Technology and School of Information Science and Electronic Engineering, Shanghai Jiao Tong University（上海交通大学信息科学与工程学院及先进微纳制造技术国家重点实验室）
The discovery of ferroelectricity in HfO$_2$-based thin films opens up new opportunities for using this silicon-compatible ferroelectric to realize low-power logic circuits and high-density non-volatile memories. The functional performances of ferroelectrics are intimately related to their dynamic responses to external stimuli such as electric fields at finite temperatures. Molecular dynamics is an ideal technique for investigating dynamical processes on large length and time scales, though its applications to new materials is often hindered by the limited availability and accuracy of classica…
Defects such as oxygen vacancy are widely considered to be critical for the performance of HfO2-based devices, and yet atomistic mechanisms underlying various exotic effects such as wake-up and fluid imprint remain elusive. Here, guided by a lattice-mode-matching criterion, we systematically study the phase transitions between different polymorphs of hafnia under the influences of neutral and positively charged oxygen vacancies by mapping out the minimum energy pathways using a first-principles-based variable-cell nudged elastic band technique. We find that the positively charged oxygen vacanc…
In this publication we introduce SAMPLE, a structure search approach for commensurate organic monolayers on inorganic substrates. Such monolayers often show rich polymorphism with diverse molecular arrangements in differently shaped unit cells. Determining the different commensurate polymorphs from first principles poses a major challenge due to the large number of possible molecular arrangements. To meet this challenge, SAMPLE employs coarse-grained modeling in combination with Bayesian linear regression to efficiently map the minima of the potential energy surface. In addition, it uses ab in…
NMRPeak: a ready-to-use intelligent system for molecular structure elucidation enabled by synergistic cross-modal learning
Institute of Artificial Intelligence, Xiamen University（厦门大学人工智能研究院）；State Key Laboratory of Physical Chemistry of Solid Surfaces, iChEM, College of Chemistry and Chemical Engineering, Xiamen University（厦门大学化学化工学院）；DP Technology（星环科技）；Laboratory of AI for Electrochemistry (AI4EC), Tan Kah Kee Innovation Laboratory (IKKEM)（陈嘉庚创新实验室电化学人工智能实验室）；College of Chemistry and Molecular Engineering, Peking University（北京大学化学与分子工程学院）；Pen-Tung Sah Institute of Micro-Nano Science and Technology, Discipline of Intelligent Instrument and Equipment, iChEM, Xiamen University（厦门大学微纳科学技术研究院）；AI for Science Institute（人工智能科学研究机构）
Atomistic understanding of hydrogen bubble-induced embrittlement in tungsten enabled by machine learning molecular dynamics
University of Science and Technology Beijing（北京科技大学）；Fuzhou University（福州大学）；Beijing Advanced Innovation Center for Materials Genome Engineering, University of Science and Technology Beijing（北京科技大学材料基因组工程高等创新中心）；Aalto University（阿尔托大学）；Tel Aviv University（特拉维夫大学）
Carbon nanotubes (CNTs) are currently considered a successor to silicon in future nanoelectronic devices. To realize this, controlled growth of defect-free nanotubes is required. Until now, the understanding of atomic-scale CNT growth mechanisms provided by molecular dynamics simulations has been hampered by their short timescales. Here, we develop an efficient and accurate machine learning force field, DeepCNT-22, to simulate the complete growth of defect-free single-walled CNTs (SWCNTs) on iron catalysts at near-microsecond timescales. We provide atomic-level insight into the nucleation and …
We show that a crystal of mesoscopic Fe8 single molecule magnets is an experimental realization of the Quantum Ising Phase Transition (QIPT) model in a transverse field, with dipolar interactions. Quantum annealing has enabled us to explore the QIPT at thermodynamical equilibrium. The phase diagram and critical exponents we obtain are compared to expectations for the mean-field QIPT Universality class.
The discovery of very large piezo- and pyroelectric effects in ZrO2 and HfO2-based thin films opens up new opportunities to develop silicon-compatible sensor and actor devices. The effects are amplified close to the polar-orthorhombic to tetragonal phase transition temperature. Molecular dynamics is the preferred technique to simulate such effects, though its application has to solve the dilemma between sufficient accuracy and sufficient efficiency of the interatomic force field. Here we present a deep neural network-based interatomic force field of ZrO2 learned from ab initio data using a sys…
Recent molecular dynamics simulation methods have enabled thermal conductivity of bulk materials to be estimated. In these simulations, periodic boundary conditions are used to extend the system dimensions to the thermodynamic limit. Such a strategy cannot be used for nanostructures with finite dimensions which are typically much larger than it is possible to simulate directly. To bridge the length scales between the simulated and the actual nanostructures, we perform large-scale molecular dynamics calculations of thermal conductivities at different system dimensions to examine a recently deve…
We present a GPU implementation of LAMMPS, a widely-used parallel molecular dynamics (MD) software package, and show 5x to 13x single node speedups versus the CPU-only version of LAMMPS. This new CUDA package for LAMMPS also enables multi-GPU simulation on hybrid heterogeneous clusters, using MPI for inter-node communication, CUDA kernels on the GPU for all methods working with particle data, and standard LAMMPS C++ code for CPU execution. Cell and neighbor list approaches are compared for best performance on GPUs, with thread-per-atom and block-per-atom neighbor list variants showing best per…
Recent air pollution issues have raised significant attention to develop efficient air filters, and one of the most promising candidates is that enabled by nanofibers. We explore here selective molecular capture mechanism for volatile organic compounds in carbon nanotube networks by performing atomistic simulations. The results are discussed with respect to the two key parameters that define the performance of nanofiltration, i.e. the capture efficiency and flow resistance, which validate the advantage of carbon nanotube networks with high surface-to-volume ratio and atomistically smooth surfa…
In this article we demonstrate that a grating fabricated through nanoscale volumetric crosslinking of a liquid crystalline polymer enables remote polarization control over the diffracted channels. This functionality is a consequence of the responsivity of liquid crystal networks upon light stimuli. Tuning the photonic response of the device is obtained thanks to both a refractive index and a shape change of the grating elements induced by a molecular rearrangement under irradiation. In particular, the material anisotropy allows for nontrivial polarization state management over multiple beams. …
The influence of a staggered molecular field in frustrated rare-earth pyrochlores, produced via the magnetic iridium occupying the transition metal site, can generate exotic ground states, such as the fragmentation of the magnetization in the Ho compound. At variance with the Ising Ho$^{3+}$ moment, we focus on the behavior of the quasi isotropic magnetic moment of the Gd$^{3+}$ ion at the rare-earth site. By means of macroscopic measurements and neutron scattering, we find a complex situation where different components of the magnetic moment contribute to two antiferromagnetic non-collinear a…
I describe a semi-empirical molecular model of the surface attachment kinetics governing ice crystal growth from water vapor as a function of temperature, supersaturation, and crystal mesostructure. An important new hypothesis in this model is surface-energy-driven molecular diffusion enabled by a leaky Ehrlich-Schwoebel barrier. The proposed surface-diffusion behavior is sensitive to facet width and surface premelting, yielding structure-dependent attachment kinetics with a complex temperature dependence. By incorporating several reasonable assumptions regarding the surface premelting behavio…
