Digital discovery最新文献

Solid polymer electrolytes exhibit limitations in room-temperature ionic conductivity and electrochemical stability. While molecular simulations and electronic-structure theory are able to sample these key properties at the molecular scale, the field currently lacks integrated, automated tools for end-to-end assessment. We introduce polymer electrolyte modeling and discovery (PEMD), an open-source Python framework that unifies polymer construction, force field parameterization, multiscale simulation, and property analysis for polymer electrolytes. The comprehensive analysis suite spans transport properties, transport mechanisms, and electrochemical stability. PEMD achieves a 100% success rate in constructing a collection of 656 homopolymers. The automated molecular dynamics workflow reproduces experimental ionic conductivities for 18 reported systems (Spearman ρ = 0.819; MAE = 0.684 in log 10 (S cm⁻¹)). Specifically, for poly(ethylene oxide)/LiTFSI electrolytes, PEMD captures the canonical non-monotonic dependence of ionic conductivity on salt concentration with built-in default settings. The workflow is further applied at scale to compute ionic conductivities for 200 polymer electrolytes. Moreover, automated oxidation window screening on 15 representative polymer electrolytes recovers experimental rankings for the oxidation potential (Spearman ρ = 0.754; MAE = 0.473 V). With standardized protocols and traceable workflows, PEMD provides a reliable platform for high-throughput screening and data-driven design of solid polymer electrolytes.

The ability to prepare states for quantum chemistry is a promising feature of quantum computers, and efficient techniques for chemical state preparation is an active area of research. In this paper, we implement and investigate two methods of quantum circuit preparation for multiconfigurational states for quantum chemical applications. It has previously been shown that controlled Givens rotations are universal for quantum chemistry. To prepare a selected linear combination of Slater determinants (represented as occupation number configurations) using Givens rotations, the gates that rotate between the reference and excited determinants need to be controlled on qubits outside the excitation (external controls), in general. We implement a method to automatically find the external controls required for utilizing Givens rotations to prepare multiconfigurational states on a quantum circuit. We compare this approach to an alternative technique that exploits the sparsity of the chemical state vector and find that the latter can outperform the method of externally controlled Givens rotations; highly reduced circuits can be obtained by taking advantage of the sparse nature (where the number of basis states is significantly less than 2^n_q for n_q qubits) of chemical wavefunctions. We demonstrate the benefits of these techniques in a range of applications, including the ground states of a strongly correlated molecule, matrix elements of the Q-SCEOM algorithm for excited states, as well as correlated initial states for a quantum subspace method based on quantum computed moments and quantum phase estimation.

This research focuses on efficiently collecting CO₂ adsorption data using experimental metal–organic framework (MOF) porous materials from the scientific literature, addressing the challenges related to data classification and access to MOF synthesis methods. The aim is to organize, classify, and facilitate easy access to materials science information using artificial intelligence (AI). Using advanced large language models (LLMs), we developed a systematic approach to extract and sort MOF synthesis data for CO₂ adsorption in a structured format. Using this method, we collected data from over 433 published experimental research papers and created a specific dataset to analyze the effects of metals, ligands, and carbon adsorption conditions on CO₂ uptake performance. The correlations between the material structure, such as metal types, ligands, specific surface area, pore size, pore volume, synthesis conditions, and CO₂ adsorption, under various process conditions were examined using the final database. We applied ChatGPT 4o mini as an AI assistant to text-mine all MOF information from different PDF file references. In addition to revealing the impact of each parameter on CO₂ uptake and MOF structure before synthesis, the AI analysis findings indicated which ligand and metal groups should be altered to customize the MOF structure for improved CO₂ capture.

Large Language Models (LLMs) are increasingly utilized for large-scale extraction and organization of unstructured data owing to their exceptional Natural Language Processing (NLP) capabilities. Empowering materials design, vast amounts of data from experiments and simulations are scattered across numerous scientific publications, but high-quality experimental databases are scarce. This study considers the effectiveness and practicality of five representative AI tools (ChemDataExtractor, BERT-PSIE, ChatExtract, LangChain, and Kimi) to extract bandgaps from 200 randomly selected materials science publications in two presentations (arXiv and publisher versions), comparing the results to those obtained by human processing. Although the integrity of data extraction has not met expectations, encouraging results have been achieved in terms of precision and the ability to eliminate irrelevant papers from human consideration. Our analysis highlights both the strengths and limitations of these tools, offering insights into improving future data extraction techniques for enhanced scientific discovery and innovation. In conjunction with recent research, we provide guidance on feasible improvements for future data extraction methodologies, helping to bridge the gap between unstructured scientific data and structured, actionable databases.

One of the main bottlenecks for the wide adoption of atomistic simulation pipelines for computational materials design is the high complexity of the workflows which many times requires the use of a diverse set of specialized toolkits and libraries. Here, we introduce a multi-agent artificial intelligence (AI) framework that autonomously performs end-to-end atomistic simulations, i.e. molecular dynamics (MD), with automated input and associated full suite of analyses, using large language models (LLMs) and multiple specialized AI agents. Our system orchestrates the entire simulation pipeline, from structure generation via Atomsk and interatomic potential discovery through automated web mining, to simulation setup and execution using LAMMPS on high-performance computing (HPC) platforms. Post-simulation, our agentic framework performs automated data analysis and visualization with popular analysis tools like OVITO and Phonopy. Each expert agent operates within a defined role, equipped with domain-specific functions and a shared memory context for coordination. Using a diverse set of representative elemental and alloy systems, we demonstrate the capability of our framework to execute a range of static and dynamic materials modeling tasks, including lattice parameter and cohesive energy estimation, elastic constants computation, phonon dispersion analysis, as well as perform MD simulations to determine dynamical properties that aid estimation of melting point. The results produced by the agents show strong agreement with those obtained by a human expert, highlighting the reliability of the agentic approach. By combining automation, reproducibility, and human-in-the-loop control, our framework lowers the barrier to the widespread adoption of scalable, AI-driven discovery tools in materials science.

This review explores the impact of deep learning (DL) techniques on understanding and predicting electronic structures in two-dimensional (2D) materials. We highlight unique computational challenges posed by 2D materials and discuss how DL approaches – such as physics-aware models, generative AI, and inverse design – have significantly improved predictions of critical electronic properties, including band structures, density of states, and quantum transport phenomena. Through selected case studies, we illustrate how DL methods accelerate discoveries in emergent quantum phenomena, topology, superconductivity, and autonomous materials exploration. Finally, we outline promising future directions, stressing the need for robust data standardization and advocating for integrated frameworks that combine theoretical modeling, DL methods, and experimental validations.

The discovery of next-generation battery electrolytes increasingly involves complex, multicomponent formulations that demand high-throughput, systematic exploration. We present the Bayesian Robotic Investigator of Novel Electrolytes (BRINE), a cost-effective, self-driving laboratory (SDL) that autonomously prepares and tests mixed electrolyte solutions. BRINE combines an open-source liquid-handling robot with a potentiostat and custom-made electrodes to mix reagents and perform electrochemical measurements without human intervention. A Bayesian optimization routine navigates multidimensional composition spaces, allowing the platform to rapidly identify promising formulations. As a proof of concept, BRINE mapped ionic conductivity in two aqueous electrolyte spaces (i) aqueous mixtures of NaCl, KCl, MgCl₂, and CaCl₂, and (ii) battery-oriented mixtures containing ZnCl₂, KCl, NH₄Cl, NaCl, and EMIMCl, testing ≈230 unique compositions in under 20 hours and finding conductivities up to 32.13 S m⁻¹. These results demonstrate how closed-loop autonomous experimentation and optimization accelerate the identification of electrolytes with the highest conductivity across a large multicomponent composition space, while minimizing experimental variability. This work lays the foundation for broader electrochemical studies using the BRINE platform.

Machine Learning Interatomic Potentials (MLIPs) are a promising alternative to expensive ab initio quantum mechanical molecular simulations. Given the diversity of chemical spaces that are of interest and the cost of generating new data, it is important to understand how universal MLIPs generalize beyond their training distributions. In order to characterize and better understand distribution shifts in MLIPs—that is, changes between the training and testing distributions—we conduct diagnostic experiments on chemical datasets, revealing common shifts that pose significant challenges, even for large universal models trained on extensive data. Based on these observations, we hypothesize that current supervised training methods inadequately regularize MLIPs, resulting in overfitting and learning poor representations of out-of-distribution systems. We then propose two new methods as initial steps for mitigating distribution shifts for MLIPs. Our methods focus on test-time refinement strategies that incur minimal computational cost and do not use expensive ab initio reference labels. The first strategy, based on spectral graph theory, modifies the edges of test graphs to align with graph structures seen during training. Our second strategy improves representations for out-of-distribution systems at test-time by taking gradient steps using an auxiliary objective, such as a cheap physical prior. Our test-time refinement strategies significantly reduce errors on out-of-distribution systems, suggesting that MLIPs are capable of and can move towards modeling diverse chemical spaces, but are not being effectively trained to do so. Our experiments establish clear benchmarks for evaluating the generalization capabilities of the next generation of MLIPs. Our code is available at https://tkreiman.github.io/projects/mlff_distribution_shifts/.

The estimation of the strength of the bond of adsorbates on the surface is key to the design of novel materials for heterogeneous catalysis. Machine learning (ML) methodologies have proven effective in rapidly and accurately evaluating adsorption energies on transition metal surfaces. However, the complexity of metal oxides and their diverse adsorbate–catalyst interactions hinder the sound transfer of ML approaches to these catalytically relevant materials. To address this challenge, we have evaluated the transferability of GAME-Net, a graph neural network developed for transition metals, by following an approach of increasing complexity, leading to GAME-Net-Ox. A density functional theory dataset was built with organic molecules on conductive (IrO₂ and RuO₂) and semiconductive (TiO₂) rutile oxides to evaluate GAME-Net's transferability. While the original GAME-Net failed to directly generalize between metals and metal oxides, GAME-Net-Ox trained exclusively on oxides achieved high accuracy (MAE = 0.16 eV) and both families of materials can be treated in GAME-Net-Ox with the same accuracy (MAE = 0.16 eV). This work demonstrates the adaptability of the GAME-Net architecture, enabling the screening of adsorbates on metal oxides, materials with complex electronic properties.

Machine learning models are often used as scoring functions to predict the binding affinity of a protein–ligand complex. These models are trained with limited amounts of data with experimentally measured binding affinity values. A large number of compounds are labeled inactive through single-concentration screens without measuring binding affinities. These inactive compounds, along with the active ones, can be used to train binary classification models, while regression models are trained using compounds with binding affinities only. However, the classification and regression tasks are often handled separately, without sharing the learned feature representations. In this paper, we propose a novel model architecture that jointly performs regression and classification objectives, aiming to maximize data utilization and improve predictive performance by leveraging two complementary tasks. In our setup, the regression yields the binding affinity, whereas the classification task yields the label as active or inactive. We demonstrate our method using PDBbind, the standard 3D structure database, as well as a dataset of flavivirus protease compounds with binding affinity data. Our experiments show that the new joint training strategy improves the accuracy of the model, increasing applicability in various practical drug screening scenarios.