Advancing Physical Chemistry with Machine Learning

Solving Advanced Task-Specific Problems in Measurement Sciences with Generative AI

The Generative Pre-Trained Transformer known as ChatGPT-4 has undergone extensive pretraining on a diverse data set, enabling it to generate coherent and contextually relevant text based on the input it receives. This capability allows it to perform tasks from answering questions and has attracted significant interest in material sciences, synthetic chemistry, and drug discovery. In this work, we posed four advanced task-specific problems to ChatGPT, which were recently published in leading journals for topics in analytical chemistry, spectroscopy, bioimage super-resolution, and electrochemistry. ChatGPT-4 successfully implemented the four ideas after assigning the "persona" to the AI and posing targeted questions. We show two cases where "unguided" ChatGPT could complete the assignments with minimal human direction. The construction of a microwave spectrum from a free induction curve and super-resolution of bioimages was accomplished using this approach. Two other specific tasks, correcting a complex baseline with morphological operations of set theory and estimating the diffusion current, required additional input, e.g., equations and specific directions from the user. In each case, the MATLAB code was eventually generated by ChatGPT-4 even when the original authors did not provide any code themselves. We show that a validation test must be implemented to detect and correct mistakes made by ChatGPT-4, followed by feedback correction. These approaches will pave the way for open and transparent science and eliminate the black boxes in measurement science when it comes to advanced data processing.

Another string to your bow: machine learning prediction of the pharmacokinetic properties of small molecules

INTRODUCTION

Prediction of pharmacokinetic (PK) properties is crucial for drug discovery and development. Machine-learning (ML) models, which use statistical pattern recognition to learn correlations between input features (such as chemical structures) and target variables (such as PK parameters), are being increasingly used for this purpose. To embed ML models for PK prediction into workflows and to guide future development, a solid understanding of their applicability, advantages, limitations, and synergies with other approaches is necessary.

AREAS COVERED

This narrative review discusses the design and application of ML models to predict PK parameters of small molecules, especially in light of established approaches including in vitro-in vivo extrapolation (IVIVE) and physiologically based pharmacokinetic (PBPK) models. The authors illustrate scenarios in which the three approaches are used and emphasize how they enhance and complement each other. In particular, they highlight achievements, the state of the art and potentials of applying machine learning for PK prediction through a comphrehensive literature review.

EXPERT OPINION

ML models, when carefully crafted, regularly updated, and appropriately used, empower users to prioritize molecules with favorable PK properties. Informed practitioners can leverage these models to improve the efficiency of drug discovery and development process.

Advances in Computational Methods for Modeling Photocatalytic Reactions: A Review of Recent Developments

Photocatalysis is a fascinating process in which a photocatalyst plays a pivotal role in driving a chemical reaction when exposed to light. Its capacity to harness light energy triggers a cascade of reactions that lead to the formation of intermediate compounds, culminating in the desired final product(s). The essence of this process is the interaction between the photocatalyst's excited state and its specific interactions with reactants, resulting in the creation of intermediates. The process's appeal is further enhanced by its cyclic nature-the photocatalyst is rejuvenated after each cycle, ensuring ongoing and sustainable catalytic action. Nevertheless, comprehending the photocatalytic process through the modeling of photoactive materials and molecular devices demands advanced computational techniques founded on effective quantum chemistry methods, multiscale modeling, and machine learning. This review analyzes contemporary theoretical methods, spanning a range of lengths and accuracy scales, and assesses the strengths and limitations of these methods. It also explores the future challenges in modeling complex nano-photocatalysts, underscoring the necessity of integrating various methods hierarchically to optimize resource distribution across different scales. Additionally, the discussion includes the role of excited state chemistry, a crucial element in understanding photocatalysis.

Swarm Smart Meta-Estimator for 2D/2D Heterostructure Design

Two-dimensional (2D) semiconductors are central to many scientific fields. The combination of two semiconductors (heterostructure) is a good way to lift many technological deadlocks. Although ab initio calculations are useful to study physical properties of these composites, their application is limited to few heterostructure samples. Herein, we use machine learning to predict key characteristics of 2D materials to select relevant candidates for heterostructure building. First, a label space is created with engineered labels relating to atomic charge and ion spatial distribution. Then, a meta-estimator is designed to predict label values of heterostructure samples having a defined band alignment (descriptor). To this end, independently trained k-nearest neighbors (KNN) regression models are combined to boost the regression. Then, swarm intelligence principles are used, along with the boosted estimator's results, to further refine the regression. This new "swarm smart" algorithm is a powerful and versatile tool to select, among experimentally existing, computationally studied, and not yet discovered van der Waals heterostructures, the most likely candidate materials to face the scientific challenges ahead.

Machine-Learning Solutions for the Analysis of Single-Particle Diffusion Trajectories

Single-particle traces of the diffusive motion of molecules, cells, or animals are by now routinely measured, similar to stochastic records of stock prices or weather data. Deciphering the stochastic mechanism behind the recorded dynamics is vital in understanding the observed systems. Typically, the task is to decipher the exact type of diffusion and/or to determine the system parameters. The tools used in this endeavor are currently being revolutionized by modern machine-learning techniques. In this Perspective we provide an overview of recently introduced methods in machine-learning for diffusive time series, most notably, those successfully competing in the anomalous diffusion challenge. As such methods are often criticized for their lack of interpretability, we focus on means to include uncertainty estimates and feature-based approaches, both improving interpretability and providing concrete insight into the learning process of the machine. We expand the discussion by examining predictions on different out-of-distribution data. We also comment on expected future developments.

Theoretical prediction and shape-controlled synthesis of two-dimensional semiconductive Ni3TeO6

Current progress in two-dimensional (2D) materials explorations leads to constant specie enrichments of possible advanced materials down to two dimensions. The metal chalcogenide-based 2D materials are promising grounds where many adjacent territories are waiting to be explored. Here, a stable monolayer Ni3TeO6 (NTO) structure was computationally predicted and its stacked 2D nanosheets experimentally synthesized. Theoretical design undergoes featuring coordination of metalloid chalcogen, slicing the bulk structure, geometrical optimizations and stability study. The predicted layered NTO structure is realized in nanometer-thick nanosheets via a one-pot shape-controlled hydrothermal synthesis. Compared to the bulk, the 2D NTO own a lowered bandgap energy, more sensitive wavelength selectivity and an emerging photocatalytic hydrogen evolution ability under visible light. Beside a new 2D NTO with the optoelectrical and photocatalytic merits, its existing polar space group, structural specification, and design route are hoped to benefit 2D semiconductor innovations both in species enrichment and future applications.

Segmentation and Morphology Computation of a Spiky Nanoparticle Using the Hourglass Neural Network

Morphological measurements of nanoparticles in electron microscopy images are tedious, laborious, and often succumb to human errors. Deep learning methods in artificial intelligence (AI) paved the way for automated image understanding. This work proposes a deep neural network (DNN) for the automated segmentation of a Au spiky nanoparticle (SNP) in electron microscopic images, and the network is trained with a spike-focused loss function. The segmented images are used for the growth measurement of the Au SNP. The auxiliary loss function captures the spikes of the nanoparticle, which prioritizes the detection of spikes in the border regions. The growth of the particles measured by the proposed DNN is as good as the measurement in manually segmented images of the particles. The proposed DNN composition with the training methodology meticulously segments the particle and consequently provides accurate morphological analysis. Furthermore, the proposed network is tested on an embedded system for integration with the microscope hardware for real-time morphological analysis.

Machine learning of spectra-property relationship for imperfect and small chemistry data

Machine learning (ML) is causing profound changes to chemical research through its powerful statistical and mathematical methodological capabilities. However, the nature of chemistry experiments often sets very high hurdles to collect high-quality data that are deficiency free, contradicting the need of ML to learn from big data. Even worse, the black-box nature of most ML methods requires more abundant data to ensure good transferability. Herein, we combine physics-based spectral descriptors with a symbolic regression method to establish interpretable spectra-property relationship. Using the machine-learned mathematical formulas, we have predicted the adsorption energy and charge transfer of the CO-adsorbed Cu-based MOF systems from their infrared and Raman spectra. The explicit prediction models are robust, allowing them to be transferrable to small and low-quality dataset containing partial errors. Surprisingly, they can be used to identify and clean error data, which are common data scenarios in real experiments. Such robust learning protocol will significantly enhance the applicability of machine-learned spectroscopy for chemical science.

A Unified View of Vibrational Spectroscopy Simulation through Kernel Density Estimations

To date, vibrational simulation results constitute more of an experimental support than a predictive tool, as the simulated vibrational modes are discrete due to quantization. This is different from what is obtained experimentally. Here, we propose a way to combine outputs such as the phonon density of states surrogate and peak intensities obtained from ab initio simulations to allow comparison with experimental data by using machine learning. This work is paving the way for using simulated vibrational spectra as a tool to identify materials with defined stoichiometry, enabling the separation of genuine vibrational features of pure phases from morphological and defect-induced signals.

Uncertainty-driven dynamics for active learning of interatomic potentials

Machine learning (ML) models, if trained to data sets of high-fidelity quantum simulations, produce accurate and efficient interatomic potentials. Active learning (AL) is a powerful tool to iteratively generate diverse data sets. In this approach, the ML model provides an uncertainty estimate along with its prediction for each new atomic configuration. If the uncertainty estimate passes a certain threshold, then the configuration is included in the data set. Here we develop a strategy to more rapidly discover configurations that meaningfully augment the training data set. The approach, uncertainty-driven dynamics for active learning (UDD-AL), modifies the potential energy surface used in molecular dynamics simulations to favor regions of configuration space for which there is large model uncertainty. The performance of UDD-AL is demonstrated for two AL tasks: sampling the conformational space of glycine and sampling the promotion of proton transfer in acetylacetone. The method is shown to efficiently explore the chemically relevant configuration space, which may be inaccessible using regular dynamical sampling at target temperature conditions.

Massive Monte Carlo simulations-guided interpretable learning of two-dimensional Curie temperature

Monte Carlo (MC) simulation of the classical Heisenberg model has become the de facto tool to estimate the Curie temperature (T _C) of two-dimensional (2D) magnets. As an alternative, here we develop data-driven models for the five most common crystal types, considering the isotropic and anisotropic exchange of up to four nearest neighbors and the single-ion anisotropy. We sample the 20-dimensional Heisenberg spin Hamiltonian and conceive a bisection-based MC technique to simulate a quarter of a million materials for training deep neural networks, which yield testing R ² scores of nearly 0.99. Since 2D magnetism has a natural tendency toward low T _C, learning-from-data is combined with data-from-learning to ensure a nearly uniform final data distribution over a wide range of T _C (10-1,000 K). Global and local analysis of the features confirms the models' interpretability. We also demonstrate that the T _C can be accurately estimated by a purely first-principles-based approach, free from any empirical corrections.

LGB-Stack: Stacked Generalization with LightGBM for Highly Accurate Predictions of Polymer Bandgap

Recently, the Ramprasad group reported a quantitative structure-property relationship (QSPR) model for predicting the E _gap values of 4209 polymers, which yielded a test set R ² score of 0.90 and a test set root-mean-square error (RMSE) score of 0.44 at a train/test split ratio of 80/20. In this paper, we present a new QSPR model named LGB-Stack, which performs a two-level stacked generalization using the light gradient boosting machine. At level 1, multiple weak models are trained, and at level 2, they are combined into a strong final model. Four molecular fingerprints were generated from the simplified molecular input line entry system notations of the polymers. They were trimmed using recursive feature elimination and used as the initial input features for training the weak models. The output predictions of the weak models were used as the new input features for training the final model, which completes the LGB-Stack model training process. Our results show that the best test set R ² and the RMSE scores of LGB-Stack at the train/test split ratio of 80/20 were 0.92 and 0.41, respectively. The accuracy scores further improved to 0.94 and 0.34, respectively, when the train/test split ratio of 95/5 was used.

Low-cost prediction of molecular and transition state partition functions via machine learning

We have generated an open-source dataset of over 30 000 organic chemistry gas phase partition functions. With this data, a machine learning deep neural network estimator was trained to predict partition functions of unknown organic chemistry gas phase transition states. This estimator only relies on reactant and product geometries and partition functions. A second machine learning deep neural network was trained to predict partition functions of chemical species from their geometry. Our models accurately predict the logarithm of test set partition functions with a maximum mean absolute error of 2.7%. Thus, this approach provides a means to reduce the cost of computing reaction rate constants ab initio. The models were also used to compute transition state theory reaction rate constant prefactors and the results were in quantitative agreement with the corresponding ab initio calculations with an accuracy of 98.3% on the log scale.

Comparison of Catalysts with MIRA21 Model in Heterogeneous Catalytic Hydrogenation of Aromatic Nitro Compounds

The vast majority of research and development activities begins with a detailed literature search to explore the current state-of-the-art. However, this search becomes increasingly difficult as we go into the information revolution of 21st century. The aim of the work is to establish a functional and practical mathematical model of catalyst characterization and exact comparison of catalysts. This work outlines the operation of the MIskolc RAnking 21 (MIRA21) model through the reaction of nitrobenzene catalytic hydrogenation to aniline. A total of 154 catalysts from 45 research articles were selected, studied, characterized, ranked, and classified based on four classes of descriptors: catalyst performance, reaction conditions, catalyst conditions, and sustainability parameters. MIRA21 is able to increase the comparability of different types of catalysts and support catalyst development. According to the model, 8% of catalysts received D1 (top 10%) classification. This ranking model is able to show the most effective catalyst systems that are suitable for the production of aniline.

Multiscale molecular modelling: from electronic structure to dynamics of nanosystems and beyond

Important contemporary biological and materials problems often depend on interactions that span orders of magnitude differences in spatial and temporal dimensions. This Tutorial Review attempts to provide an introduction to such fascinating problems through a series of case studies, aimed at beginning researchers, graduate students, postdocs and more senior colleagues who are changing direction to focus on multiscale aspects of their research. The choice of specific examples is highly personal, with examples either chosen from our own work or outstanding multiscale efforts from the literature. I start with various embedding schemes, as exemplified by polarizable continuum models, 3-D RISM, molecular DFT and frozen-density embedding. Next, QM/MM (quantum mechanical/molecular mechanical) techniques are the workhorse of pm-to-nm/ps-to-ns simulations; examples are drawn from enzymes and from nanocatalysis for oil-sands upgrading. Using polarizable force-fields in the QM/MM framework represents a burgeoning subfield; with examples from ion channels and electron dynamics in molecules subject to strong external fields, probing the atto-second dynamics of the electrons with RT-TDDFT (real-time - time-dependent density functional theory) eventually coupled with nuclear motion through the Ehrenfest approximation. This is followed by a section on coarse graining, bridging dimensions from atoms to cells. The penultimate chapter gives a quick overview of multiscale approaches that extend into the meso- and macro-scales, building on atomistic and coarse-grained techniques to enter the world of materials engineering, on the one hand, and cell biology, on the other. A final chapter gives just a glimpse of the burgeoning impact of machine learning on the structure-dynamics front. I aim to capture the excitement of contemporary leading-edge breakthroughs in the description of physico-chemical systems and processes in complex environments, with only enough historical content to provide context and aid the next generation of methodological development. While I aim also for a clear description of the essence of methodological breakthroughs, equations are kept to a minimum and detailed formalism and implementation details are left to the references. My approach is very selective (case studies) rather than exhaustive. I think that these case studies should provide fodder to build as complete a reference tree on multiscale modelling as the reader may wish, through forward and backward citation analysis. I hope that my choices of cases will excite interest in newcomers and help to fuel the growth of multiscale modelling in general.

Stability Analysis of Substituted Cobaltocenium [Bis(cyclopentadienyl)cobalt(III)] Employing Chemistry-Informed Neural Networks

Cationic cobaltocenium derivatives are promising components of the anion exchange membranes because of their excellent thermal and alkaline stability under the operating conditions of a fuel cell. Here, we present an efficient modeling approach to assessing the chemical stability of substituted cobaltocenium CoCp₂⁺, based on the computed electronic structure enhanced by machine learning techniques. Within the aqueous environment, the positive charge of the metal cation is balanced by the hydroxide anion through formation of the CoCp₂⁺OH^- complexes, whose dissociation is studied within the implicit solvent employing the density functional theory. The data set of about 118 the CoCp₂⁺OH^- complexes based on 42 substituent groups characterized by a range of electron-donating (ED) and electron-withdrawing (EW) properties is constructed and analyzed. Given 12 carefully chosen chemistry-informed descriptors of the complexes and relevant fragments, the stability of the complexes is found to strongly correlate with the energies of the highest occupied and lowest unoccupied molecular orbitals, modulated by a switching function of the Hirshfeld charge. The latter is used as a measure of the electron-withdrawing-donating character of the substituents. On the basis of this observation from the conventional regression analysis, two fully connected, feed-forward neural network (FNN) models with different unit structures, called the chemistry-informed (CINN) and the quadratic (QNN) neural networks, together with a support vector regression (SVR) model are developed. Both deep neural network models predict the bond dissociation energies of the cobaltocenium complexes with mean relative errors less than 10.56% and average absolute errors less than 1.63 kcal/mol, superior to the conventional regression and the SVR model prediction. The results show the potential of QNN to efficiently capture more complex relationships. The concept of incorporating the domain (chemical) knowledge/insight into the neural network structure paves the way to applications of machine learning techniques with small data sets, ultimately leading to better predictive models compared to the classical machine learning method SVR and conventional regression analysis.

Family of Two-Dimensional Transition Metal Dichlorides: Fundamental Properties, Structural Defects, and Environmental Stability

A large number of novel two-dimensional (2D) materials are constantly being discovered and deposited in databases. Consolidated implementation of machine learning algorithms and density functional theory (DFT)-based predictions have allowed the creation of several databases containing an unimaginable number of 2D samples. As the next step in this chain, the investigation leads to a comprehensive study of the functionality of the invented materials. In this work, a family of transition metal dichlorides have been screened out for systematic investigation of their structural stability, fundamental properties, structural defects, and environmental stability via DFT-based calculations. The work highlights the importance of using the potential of the invented materials and proposes a comprehensive characterization of a new family of 2D materials.

Authentic Intelligent Machine for Scaling Driven Discovery: A Case for Chiral Quantum Dots

The scaling laws have long been used as evidence of science where many fundamental physics laws emerge. As emerging nanomaterials, quantum dots are also sensitive to scaling because of their strong size effect. In this work, we developed the chiral dielectric theory based on the exciton absorption mechanism to explain the increment of the dielectric constant from chirality via its dimensionality. To help researchers discover and develop scaling relevant theories, the Authentic Intelligent Machine (AIM) protocol was developed to generate and interpret experimental data in an analytical and scaling-oriented manner. We show how the AIM protocol interprets spectra such as transient absorption data of chiral quantum dots with theories, where discrepancies concerning the dielectric constant were discovered. Examples for applying the AIM protocol on other spectra, such as absorption spectra and photoluminescence spectra, are also given.

Applications of machine learning in computational nanotechnology

Machine learning (ML) has gained extensive attention in recent years due to its powerful data analysis capabilities. It has been successfully applied to many fields and helped the researchers to achieve several major theoretical and applied breakthroughs. Some of the notable applications in the field of computational nanotechnology are ML potentials, property prediction, and material discovery. This review summarizes the state-of-the-art research progress in these three fields. ML potentials bridge the efficiency versus accuracy gap between density functional calculations and classical molecular dynamics. For property predictions, ML provides a robust method that eliminates the need for repetitive calculations for different simulation setups. Material design and drug discovery assisted by ML greatly reduce the capital and time investment by orders of magnitude. In this perspective, several common ML potentials and ML models are first introduced. Using these state-of-the-art models, developments in property predictions and material discovery are overviewed. Finally, this paper was concluded with an outlook on future directions of data-driven research activities in computational nanotechnology.

Significance of the Chemical Environment of an Element in Nonadiabatic Molecular Dynamics: Feature Selection and Dimensionality Reduction with Machine Learning

Using supervised and unsupervised machine learning (ML) on features generated from nonadiabatic (NA) molecular dynamics (MD) trajectories under the classical path approximation, we demonstrate that mutual information with the NA Hamiltonian can be used for feature selection and model simplification. Focusing on CsPbI₃, a popular metal halide perovskite, we observe that the chemical environment of a single element is sufficient for predicting the NA Hamiltonian. The conclusion applies even to Cs, although Cs does not contribute to the relevant wave functions. Interatomic distances between Cs and I or Pb and the octahedral tilt angle are the most important features. We reduce a typical 360-parameter ML force-field model to just a 12-parameter NA Hamiltonian model, while maintaining a high NA-MD simulation quality. Because NA-MD is a valuable tool for studying excited state processes, overcoming its high computational cost through simple ML models will streamline NA-MD simulations and expand the ranges of accessible system size and simulation time.

Charge-Transfer Landscape Manifesting the Structure-Rate Relationship in the Condensed Phase Via Machine Learning

In this work, we develop a machine learning (ML) strategy to map the molecular structure to condensed phase charge-transfer (CT) properties including CT rate constants, energy levels, electronic couplings, energy gaps, reorganization energies, and reaction free energies which are called CT fingerprints. The CT fingerprints of selected landmark structures covering the conformation space of an organic photovoltaic molecule dissolved in an explicit solvent are computed and used to train ML models using kernel ridge regression. The ML models show high predictive power with R² > 0.97 and both mean absolute error and root-mean-square error within chemical accuracy. The CT landscape for millions of molecular dynamics sampled structures is thus constructed, which allows for instant prediction of CT rate properties, given any conformation of the molecule. We demonstrate some immediate utilities of the CT landscape such as calculating the ensemble-averaged CT rate constant and interpreting the effects of molecular structural features on the CT rate. The unprecedented CT landscape will be useful for investigating real-time CT dynamics in nanoscale- and mesoscale-condensed phase systems and for the optimal fabrication design for homogeneous and heterogeneous optoelectronic devices.

AI-driven prediction of SARS-CoV-2 variant binding trends from atomistic simulations

We present a novel technique to predict binding affinity trends between two molecules from atomistic molecular dynamics simulations. The technique uses a neural network algorithm applied to a series of images encoding the distance between two molecules in time. We demonstrate that our algorithm is capable of separating with high accuracy non-hydrophobic mutations with low binding affinity from those with high binding affinity. Moreover, we show high accuracy in prediction using a small subset of the simulation, therefore requiring a much shorter simulation time. We apply our algorithm to the binding between several variants of the SARS-CoV-2 spike protein and the human receptor ACE2.

First-Principles Multiscale Modeling of Mechanical Properties in Graphene/Borophene Heterostructures Empowered by Machine-Learning Interatomic Potentials

Density functional theory calculations are robust tools to explore the mechanical properties of pristine structures at their ground state but become exceedingly expensive for large systems at finite temperatures. Classical molecular dynamics (CMD) simulations offer the possibility to study larger systems at elevated temperatures, but they require accurate interatomic potentials. Herein the authors propose the concept of first-principles multiscale modeling of mechanical properties, where ab initio level of accuracy is hierarchically bridged to explore the mechanical/failure response of macroscopic systems. It is demonstrated that machine-learning interatomic potentials (MLIPs) fitted to ab initio datasets play a pivotal role in achieving this goal. To practically illustrate this novel possibility, the mechanical/failure response of graphene/borophene coplanar heterostructures is examined. It is shown that MLIPs conveniently outperform popular CMD models for graphene and borophene and they can evaluate the mechanical properties of pristine and heterostructure phases at room temperature. Based on the information provided by the MLIP-based CMD, continuum models of heterostructures using the finite element method can be constructed. The study highlights that MLIPs were the missing block for conducting first-principles multiscale modeling, and their employment empowers a straightforward route to bridge ab initio level accuracy and flexibility to explore the mechanical/failure response of nanostructures at continuum scale.

Universal Machine Learning Interatomic Potentials: Surveying Solid Electrolytes

We apply ab initio molecular dynamics (AIMD) with on-the-fly machine learning (ML) of interatomic potentials using the sparse Gaussian process regression (SGPR) algorithm for a survey of Li diffusivity in hundreds of ternary crystals as potential electrolytes for all-solid-state batteries. We show that models generated for these crystals can be easily combined for creating more general and transferable models which can potentially be used for simulating new materials without further training. As examples, universal potentials are created for Li-P-S and Li-Sb-S systems by combining the expert models of the crystals which contained the same set of elements. We also show that combinatorial models of different ternary crystals can be directly applied for modeling composite quaternary ones (e.g., Li-Ge-P-S). This hierarchical approach paves the way for modeling large-scale complexity by a combinatorial approach.

Ab Initio Machine Learning in Chemical Compound Space

Chemical compound space (CCS), the set of all theoretically conceivable combinations of chemical elements and (meta-)stable geometries that make up matter, is colossal. The first-principles based virtual sampling of this space, for example, in search of novel molecules or materials which exhibit desirable properties, is therefore prohibitive for all but the smallest subsets and simplest properties. We review studies aimed at tackling this challenge using modern machine learning techniques based on (i) synthetic data, typically generated using quantum mechanics based methods, and (ii) model architectures inspired by quantum mechanics. Such Quantum mechanics based Machine Learning (QML) approaches combine the numerical efficiency of statistical surrogate models with an ab initio view on matter. They rigorously reflect the underlying physics in order to reach universality and transferability across CCS. While state-of-the-art approximations to quantum problems impose severe computational bottlenecks, recent QML based developments indicate the possibility of substantial acceleration without sacrificing the predictive power of quantum mechanics.

Predicting Rate Constants of Hydroxyl Radical Reactions with Alkanes Using Machine Learning

The hydrogen abstraction reactions of the hydroxyl radical with alkanes play an important role in combustion chemistry and atmospheric chemistry. However, site-specific reaction constants are difficult to obtain experimentally and theoretically. Recently, machine learning has proved its ability to predict chemical properties. In this work, a machine learning approach is developed to predict the temperature-dependent site-specific rate constants of the title reactions. Multilayered neural network (NN) models are developed by training the site-specific rate constants of 11 reactions, and several schemes are designed to improve the prediction accuracy. The results show that the proposed NN models are robust in predicting the site-specific and overall rate constants.

Generative adversarial networks for transition state geometry prediction

This work introduces a novel application of generative adversarial networks (GANs) for the prediction of starting geometries in transition state (TS) searches based on the geometries of reactants and products. The multi-dimensional potential energy space of a chemical reaction often complicates the location of a starting TS geometry, leading to the correct TS combining reactants and products in question. The proposed TS-GAN efficiently maps the space between reactants and products and generates reliable TS guess geometries, and it can be easily combined with any quantum chemical software package performing geometry optimizations. The TS-GAN was trained and applied to generate TS guess structures for typical chemical reactions, such as hydrogen migration, isomerization, and transition metal-catalyzed reactions. The performance of the TS-GAN was directly compared to that of classical approaches, proving its high accuracy and efficiency. The current TS-GAN can be extended to any dataset that contains sufficient chemical reactions for training. The software is freely available for training, experimentation, and prediction at https://github.com/ekraka/TS-GAN.

Physico-chemical properties of selenium-tellurium alloys across the scales

Selenium and tellurium are both energy critical elements as defined by the American Physical Society and the Materials Research Society. When mixed together, both elements form an alloy. The size- and shape-dependent thermal and optical properties of this alloy are investigated in this manuscript by using nano-thermodynamics and machine learning techniques. This alloy is found to have particularly interesting properties for solar cell applications.

The Rise of Neural Networks for Materials and Chemical Dynamics

Machine learning (ML) is quickly becoming a premier tool for modeling chemical processes and materials. ML-based force fields, trained on large data sets of high-quality electron structure calculations, are particularly attractive due their unique combination of computational efficiency and physical accuracy. This Perspective summarizes some recent advances in the development of neural network-based interatomic potentials. Designing high-quality training data sets is crucial to overall model accuracy. One strategy is active learning, in which new data are automatically collected for atomic configurations that produce large ML uncertainties. Another strategy is to use the highest levels of quantum theory possible. Transfer learning allows training to a data set of mixed fidelity. A model initially trained to a large data set of density functional theory calculations can be significantly improved by retraining to a relatively small data set of expensive coupled cluster theory calculations. These advances are exemplified by applications to molecules and materials.

Infrared Spectroscopy of Chemically Diverse Carbon Clusters: A Data-Driven Approach

Carbon clusters exhibit a broad diversity of topologies and shapes, encompassing fullerene-like cages, graphene-like flakes, and more disordered pretzel-like and branched structures. Here, we examine computationally their infrared spectra in relation with these structures from a statistical perspective. Individual spectra for broad samples of isomers were determined by means of the self-consistent charge density functional-based tight-binding method, and an interpolation scheme is designed to reproduce the spectral features by regression on a much smaller subset of the sample. This interpolation proceeds by encoding the structures using appropriate descriptors and selecting them through principal component analysis, Gaussian regression or inverse distance weighting providing the nonlinear weighting functions. Metric learning is employed to reduce the global error on a preselected testing set. The interpolated spectra satisfactorily reproduce the specific spectral features and their dependence on the size and shape, enabling quantitative prediction away from the testing set. Finally, the classification of structures within the four proposed families is critically discussed through a statistical analysis of the sample based on iterative label spreading.

Understanding simple liquids through statistical and deep learning approaches

Statistical and deep learning-based methods are employed to obtain insights into the quasi-universal properties of simple liquids. In the first part, a statistical model is employed to provide a probabilistic explanation for the similarity in the structure of simple liquids interacting with different pair potential forms, collectively known as simple liquids. The methodology works by sampling the radial distribution function and the number of interacting particles within the cutoff distance, and it produces the probability density function of the net force. We show that matching the probability distribution of the net force can be a direct route to parameterize simple liquid pair potentials with a similar structure, as the net force is the main component of the Newtonian equations of motion. The statistical model is assessed and validated against various cases. In the second part, we exploit DeepILST [A. Moradzadeh and N. R. Aluru, J. Phys. Chem. Lett. 10, 1242-1250 (2019)], a data-driven and deep-learning assisted framework to parameterize the standard 12-6 Lennard-Jones (LJ) pair potential, to find structurally equivalent/isomorphic LJ liquids that identify constant order parameter [τ=∫₀ ^ξ_cf gξ-1ξ²dξ, where gξ and ξ(=rρ¹³) are the reduced radial distribution function and radial distance, respectively] systems in the space of non-dimensional temperature and density of the LJ liquids. We also investigate the consistency of DeepILST in reproducibility of radial distribution functions of various quasi-universal potentials, e.g., exponential, inverse-power-law, and Yukawa pair potentials, quantified based on the radial distribution functions and Kullback-Leibler errors. Our results provide insights into the quasi-universality of simple liquids using the statistical and deep learning methods.

A review on machine learning algorithms for the ionic liquid chemical space

There are thousands of papers published every year investigating the properties and possible applications of ionic liquids. Industrial use of these exceptional fluids requires adequate understanding of their physical properties, in order to create the ionic liquid that will optimally suit the application. Computational property prediction arose from the urgent need to minimise the time and cost that would be required to experimentally test different combinations of ions. This review discusses the use of machine learning algorithms as property prediction tools for ionic liquids (either as standalone methods or in conjunction with molecular dynamics simulations), presents common problems of training datasets and proposes ways that could lead to more accurate and efficient models.

Recent Advances in Sequential Infiltration Synthesis (SIS) of Block Copolymers (BCPs)

In the continuous downscaling of device features, the microelectronics industry is facing the intrinsic limits of conventional lithographic techniques. The development of new synthetic approaches for large-scale nanopatterned materials with enhanced performances is therefore required in the pursuit of the fabrication of next-generation devices. Self-assembled materials as block copolymers (BCPs) provide great control on the definition of nanopatterns, promising to be ideal candidates as templates for the selective incorporation of a variety of inorganic materials when combined with sequential infiltration synthesis (SIS). In this review, we report the latest advances in nanostructured inorganic materials synthesized by infiltration of self-assembled BCPs. We report a comprehensive description of the chemical and physical characterization techniques used for in situ studies of the process mechanism and ex situ measurements of the resulting properties of infiltrated polymers. Finally, emerging optical and electrical properties of such materials are discussed.

First-Principles Prediction of Two-Dimensional B3C2P3 and B2C4P2: Structural Stability, Fundamental Properties, and Renewable Energy Applications

The existence of two novel hybrid two-dimensional (2D) monolayers, 2D B₃C₂P₃ and 2D B₂C₄P₂, has been predicted based on the density functional theory calculations. It has been shown that these materials possess structural and thermodynamic stability. 2D B₃C₂P₃ is a moderate band gap semiconductor, while 2D B₂C₄P₂ is a zero band gap semiconductor. It has also been shown that 2D B₃C₂P₃ has a highly tunable band gap under the effect of strain and substrate engineering. Moreover, 2D B₃C₂P₃ produces low barriers for dissociation of water and hydrogen molecules on its surface, and shows fast recovery after desorption of the molecules. The novel materials can be fabricated by carbon doping of boron phosphide and directly by arc discharge and laser ablation and vaporization. Applications of 2D B₃C₂P₃ in renewable energy and straintronic nanodevices have been proposed.

Nanoporous Material Recognition via 3D Convolutional Neural Networks: Prediction of Adsorption Properties

Nanoporous materials can be effective adsorbents for various energy applications. Because of their abundant number, brute-force-based material discovery can, however, be challenging. Data-driven approaches can be advantageous for such purposes. In this study, we demonstrate for the first time the applicability of a 3D convolutional neural network (CNN) in material recognition for predicting adsorption properties. 2D CNNs have been widely applied to image recognition, where the CNN self-learns important features of images, without the need of handcrafting features that are subject to human bias. This study explores methane adsorption in zeolites as a case study, where ∼6500 hypothetical zeolites are utilized to train/validate our designed CNN model. The CNN model offers highly accurate predictions, and the self-learned features resemble the channel and pore-like geometry of structures. This study demonstrates the extension of computer vision to materials science and paves the way for future studies such as carbon capture.