arXiv subjects

Arghya Bhowmik

Publications and source records attributed to Arghya Bhowmik.

28 records · Page 2Linked to original sources

arXiv2023

Wave function Ansatz (but Periodic) Networks and the Homogeneous Electron Gas

We design a neural network Ansatz for variationally finding the ground-state wave function of the Homogeneous Electron Gas, a fundamental model in the physics of extended systems of interacting fermions. We study the spin-polarised and paramagnetic phases with 7, 14 and 19 electrons over a broad range of densities from $r_s=1$ to $r_s=100$, obtaining similar or higher accuracy compared to a state-of-the-art iterative backflow baseline even in the challenging regime of very strong correlation. Our work extends previous applications of neural network Ansätze to molecular systems with methods for handling periodic boundary conditions, and makes two notable changes to improve performance: splitting the pairwise streams by spin alignment and generating backflow coordinates for the orbitals from the network. We illustrate the advantage of our high quality wave functions in computing the reduced single particle density matrix. This contribution establishes neural network models as flexible and high precision Ansätze for periodic electronic systems, an important step towards applications to crystalline solids.

physics.chem-ph↗

arXiv2022

NeuralNEB -- Neural Networks can find Reaction Paths Fast

Quantum mechanical methods like Density Functional Theory (DFT) are used with great success alongside efficient search algorithms for studying kinetics of reactive systems. However, DFT is prohibitively expensive for large scale exploration. Machine Learning (ML) models have turned out to be excellent emulators of small molecule DFT calculations and could possibly replace DFT in such tasks. For kinetics, success relies primarily on the models capability to accurately predict the Potential Energy Surface (PES) around transition-states and Minimal Energy Paths (MEPs). Previously this has not been possible due to scarcity of relevant data in the literature. In this paper we train state of the art equivariant Graph Neural Network (GNN)-based models on around 10.000 elementary reactions from the Transition1x dataset. We apply the models as potentials for the Nudged Elastic Band (NEB) algorithm and achieve a Mean Average Error (MAE) of 0.13+/-0.03 eV on barrier energies on unseen reactions. We compare the results against equivalent models trained on QM9 and ANI1x. We also compare with and outperform Density Functional based Tight Binding (DFTB) on both accuracy and computational resource. The implication is that ML models, given relevant data, are now at a level where they can be applied for downstream tasks in quantum chemistry transcending prediction of simple molecular features.

physics.comp-ph↗

arXiv2022

Transition1x -- a Dataset for Building Generalizable Reactive Machine Learning Potentials

Machine Learning (ML) models have, in contrast to their usefulness in molecular dynamics studies, had limited success as surrogate potentials for reaction barrier search. It is due to the scarcity of training data in relevant transition state regions of chemical space. Currently, available datasets for training ML models on small molecular systems almost exclusively contain configurations at or near equilibrium. In this work, we present the dataset Transition1x containing 9.6 million Density Functional Theory (DFT) calculations of forces and energies of molecular configurations on and around reaction pathways at the wB97x/6-31G(d) level of theory. The data was generated by running Nudged Elastic Band (NEB) calculations with DFT on 10k reactions while saving intermediate calculations. We train state-of-the-art equivariant graph message-passing neural network models on Transition1x and cross-validate on the popular ANI1x and QM9 datasets. We show that ML models cannot learn features in transition-state regions solely by training on hitherto popular benchmark datasets. Transition1x is a new challenging benchmark that will provide an important step towards developing next-generation ML force fields that also work far away from equilibrium configurations and reactive systems.

physics.chem-ph↗

arXiv2022

Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids

Electron density $ρ(\vec{r})$ is the fundamental variable in the calculation of ground state energy with density functional theory (DFT). Beyond total energy, features and changes in $ρ(\vec{r})$ distributions are often used to capture critical physicochemical phenomena in functional materials. We present a machine learning framework for the prediction of $ρ(\vec{r})$. The model is based on equivariant graph neural networks and the electron density is predicted at special query point vertices that are part of the message passing graph, but only receive messages. The model is tested across multiple data sets of molecules (QM9), liquid ethylene carbonate electrolyte (EC) and LixNiyMnzCo(1-y-z)O2 lithium ion battery cathodes (NMC). For QM9 molecules, the accuracy of the proposed model exceeds typical variability in $ρ(\vec{r})$ obtained from DFT done with different exchange-correlation functionals. The accuracy on all three datasets is beyond state of the art and the computation time is orders of magnitude faster than DFT.

physics.comp-ph↗

arXiv2022

Phase Separating Electrode Materials Chemical Inductors

We discover presence of chemical inductive effects in phase separating ion intercalation energy storage materials, specifically in lithium iron phosphate (LFP) and also lithium titanate oxide (LTO). These materials features fast (de)intercalation and slow diffusion relaxation phenomena which are prerequisites for observing such inductive effects. Presented finding is supported by the mechanistic model and analytical reasoning indicating that all equilibrium states that lay inside the miscibility gap of the phase separating material exhibit strong inductive response in the low frequency part of spectrum. We also explain why such inductive effects are not observed outside the miscibility gap. This letter presents the first mechanistic reasoning of previously reported electrode level experimental observation of inductance during impedance measurements at low currents.

cond-mat.mtrl-sci↗

arXiv2021

Calibrated Uncertainty for Molecular Property Prediction using Ensembles of Message Passing Neural Networks

Data-driven methods based on machine learning have the potential to accelerate computational analysis of atomic structures. In this context, reliable uncertainty estimates are important for assessing confidence in predictions and enabling decision making. However, machine learning models can produce badly calibrated uncertainty estimates and it is therefore crucial to detect and handle uncertainty carefully. In this work we extend a message passing neural network designed specifically for predicting properties of molecules and materials with a calibrated probabilistic predictive distribution. The method presented in this paper differs from previous work by considering both aleatoric and epistemic uncertainty in a unified framework, and by recalibrating the predictive distribution on unseen data. Through computer experiments, we show that our approach results in accurate models for predicting molecular formation energies with well calibrated uncertainty in and out of the training data distribution on two public molecular benchmark datasets, QM9 and PC9. The proposed method provides a general framework for training and evaluating neural network ensemble models that are able to produce accurate predictions of properties of molecules with well calibrated uncertainty estimates.

cs.LG↗

arXiv2021

Data Management Plans: the Importance of Data Management in the BIG-MAP Project

Open access to research data is increasingly important for accelerating research. Grant authorities therefore request detailed plans for how data is managed in the projects they finance. We have recently developed such a plan for the EU-H2020 BIG-MAP project - a cross-disciplinary project targeting disruptive battery-material discoveries. Essential for reaching the goal is extensive sharing of research data across scales, disciplines and stakeholders, not limited to BIG-MAP and the European BATTERY 2030+ initiative but within the entire battery community. The key challenges faced in developing the data management plan for such a large and complex project were to generate an overview of the enormous amount of data that will be produced, to build an understanding of the data flow within the project and to agree on a roadmap for making all data FAIR. This paper describes the process we followed and how we structured the plan.

cond-mat.mtrl-sci↗

arXiv2021

Automatic diffusion path exploration for multivalent battery cathodes using geometrical descriptors

Stable and fast ionic conductors for magnesium cathode materials have the prospect of enabling high energy density batteries beyond current Lithium-ion technologies. So far, only a few candidate materials have been identified leading to data only being scarcely available to the community. Here, we present a systematic study, in the framework of Density Functional Theory, including the estimation of the diffusion barrier for 16 materials through employing Nudged Elastic Band (NEB) calculations. By introducing a path finder algorithm based on the idea of Voronoi tessellations, we show that an estimate of the transition state configuration can be extracted automatically prior to running NEB-calculations. Using geometrical descriptors in combination with a principal component analysis it is possible to further sub-group the diffusion paths. This approach also extends to materials which are not part of the study, making it a viable approach to more efficiently explore crystal structures with distinguishable diffusion characteristics.

cond-mat.mtrl-sci↗

arXiv2020

DeepDFT: Neural Message Passing Network for Accurate Charge Density Prediction

We introduce DeepDFT, a deep learning model for predicting the electronic charge density around atoms, the fundamental variable in electronic structure simulations from which all ground state properties can be calculated. The model is formulated as neural message passing on a graph, consisting of interacting atom vertices and special query point vertices for which the charge density is predicted. The accuracy and scalability of the model are demonstrated for molecules, solids and liquids. The trained model achieves lower average prediction errors than the observed variations in charge density obtained from density functional theory simulations using different exchange correlation functionals.

physics.comp-ph↗

arXiv2016

Quantization of Hall Resistance at the Metallic Interface between an Oxide Insulator and SrTiO$_{3}$

The two-dimensional metal forming at the interface between an oxide insulator and SrTiO3 provides new opportunities for oxide electronics. However, the quantum Hall effect, one of the most fascinating effects of electrons confined in two dimensions, remains underexplored at these complex oxide heterointerfaces. Here, we report the experimental observation of quantized Hall resistance in a SrTiO3 heterointerface based on the modulation-doped amorphous-LaAlO$_{3}$/SrTiO$_{3}$ heterostructure, which exhibits both high electron mobility exceeding 10000 cm$^{2}$/Vs and low carrier density on the order of ~10$^{12}$ cm$^{-2}$. Along with unambiguous Shubnikov-de Haas oscillations, the spacing of the quantized Hall resistance suggests that the interface is comprised of a single quantum well with ten parallel conducting two-dimensional subbands. This provides new insight into the electronic structure of conducting oxide interfaces and represents an important step towards designing and understanding advanced oxide devices.

cond-mat.mtrl-sci↗