Design of Modified Polymer Membranes Using Machine Learning.

Surface modification is an attractive strategy to adjust the properties of polymer membranes. Unfortunately, predictive structure-processing-property relationships between the modification strategies and membrane performance are often unknown. One possibility to tackle this challenge is the application of data-driven methods such as machine learning. In this study, we applied machine learning methods to data sets containing the performance parameters of modified membranes. The resulting machine learning models were used to predict performance parameters, such as the pure water permeability and the zeta potential of membranes modified with new substances. The predictions had low prediction errors, which allowed us to generalize them to similar membrane modifications and processing conditions. Additionally, machine learning methods were able to identify the impact of substance properties and process parameters on the resulting membrane properties. Our results demonstrate that small data sets, as they are common in materials science, can be used as training data for predictive machine learning models. Therefore, machine learning shows great potential as a tool to expedite the development of high-performance membranes while reducing the time and costs associated with the development process at the same time.

Substituting density functional theory in reaction barrier calculations for hydrogen atom transfer in proteins.

Hydrogen atom transfer (HAT) reactions are important in many biological systems. As these reactions are hard to observe experimentally, it is of high interest to shed light on them using simulations. Here, we present a machine learning model based on graph neural networks for the prediction of energy barriers of HAT reactions in proteins. As input, the model uses exclusively non-optimized structures as obtained from classical simulations. It was trained on more than 17 000 energy barriers calculated using hybrid density functional theory. We built and evaluated the model in the context of HAT in collagen, but we show that the same workflow can easily be applied to HAT reactions in other biological or synthetic polymers. We obtain for relevant reactions (small reaction distances) a model with good predictive power (R2 ∼ 0.9 and mean absolute error of <3 kcal mol-1). As the inference speed is high, this model enables evaluations of dozens of chemical situations within seconds. When combined with molecular dynamics in a kinetic Monte-Carlo scheme, the model paves the way toward reactive simulations.

Artificial design of organic emitters via a genetic algorithm enhanced by a deep neural network.

The design of molecules requires multi-objective optimizations in high-dimensional chemical space with often conflicting target properties. To navigate this space, classical workflows rely on the domain knowledge and creativity of human experts, which can be the bottleneck in high-throughput approaches. Herein, we present an artificial molecular design workflow relying on a genetic algorithm and a deep neural network to find a new family of organic emitters with inverted singlet-triplet gaps and appreciable fluorescence rates. We combine high-throughput virtual screening and inverse design infused with domain knowledge and artificial intelligence to accelerate molecular generation significantly. This enabled us to explore more than 800 000 potential emitter molecules and find more than 10 000 candidates estimated to have inverted singlet-triplet gaps (INVEST) and appreciable fluorescence rates, many of which likely emit blue light. This class of molecules has the potential to realize a new generation of organic light-emitting diodes.

3DSC - a dataset of superconductors including crystal structures.

Data-driven methods, in particular machine learning, can help to speed up the discovery of new materials by finding hidden patterns in existing data and using them to identify promising candidate materials. In the case of superconductors, the use of data science tools is to date slowed down by a lack of accessible data. In this work, we present a new and publicly available superconductivity dataset ('3DSC'), featuring the critical temperature TC of superconducting materials additionally to tested non-superconductors. In contrast to existing databases such as the SuperCon database which contains information on the chemical composition, the 3DSC is augmented by approximate three-dimensional crystal structures. We perform a statistical analysis and machine learning experiments to show that access to this structural information improves the prediction of the critical temperature TC of materials. Furthermore, we provide ideas and directions for further research to improve the 3DSC. We are confident that this database will be useful in applying state-of-the-art machine learning methods to eventually find new superconductors.

Accurate GW frontier orbital energies of 134 kilo molecules.

HOMO and LUMO energies are critical molecular properties that typically require high accuracy computations for practical applicability. Until now, a comprehensive dataset containing sufficiently accurate HOMO and LUMO energies has been unavailable. In this study, we introduce a new dataset of HOMO/LUMO energies for QM9 compounds, calculated using the GW method. The GW method offers adequate HOMO/LUMO prediction accuracy for diverse applications, exhibiting mean unsigned errors of 100 meV in the GW100 benchmark dataset. This database may serve as a benchmark of HOMO/LUMO prediction, delta-learning, and transfer learning, particularly for larger molecules where GW is the most accurate but still numerically feasible method. We anticipate that this dataset will enable the development of more accurate machine learning models for predicting molecular properties.

Integrated System Built for Small-Molecule Semiconductors via High-Throughput Approaches.

High-throughput synthesis of solution-processable structurally variable small-molecule semiconductors is both an opportunity and a challenge. A large number of diverse molecules provide a possibility for quick material discovery and machine learning based on experimental data. However, the diversity of the molecular structure leads to the complexity of molecular properties, such as solubility, polarity, and crystallinity, which poses great challenges to solution processing and purification. Here, we first report an integrated system for the high-throughput synthesis, purification, and characterization of molecules with a large variety. Based on the principle "Like dissolves like," we combine theoretical calculations and a robotic platform to accelerate the purification of those molecules. With this platform, a material library containing 125 molecules and their optical-electronic properties was built within a timeframe of weeks. More importantly, the high repeatability of recrystallization we design is a reliable approach to further upgrading and industrial production.

The Role of Experimental Noise in a Hybrid Classical-Molecular Computer to Solve Combinatorial Optimization Problems.

Chemical and molecular-based computers may be promising alternatives to modern silicon-based computers. In particular, hybrid systems, where tasks are split between a chemical medium and traditional silicon components, may provide access and demonstration of chemical advantages such as scalability, low power dissipation, and genuine randomness. This work describes the development of a hybrid classical-molecular computer (HCMC) featuring an electrochemical reaction on top of an array of discrete electrodes with a fluorescent readout. The chemical medium, optical readout, and electrode interface combined with a classical computer generate a feedback loop to solve several canonical optimization problems in computer science such as number partitioning and prime factorization. Importantly, the HCMC makes constructive use of experimental noise in the optical readout, a milestone for molecular systems, to solve these optimization problems, as opposed to in silico random number generation. Specifically, we show calculations stranded in local minima can consistently converge on a global minimum in the presence of experimental noise. Scalability of the hybrid computer is demonstrated by expanding the number of variables from 4 to 7, increasing the number of possible solutions by 1 order of magnitude. This work provides a stepping stone to fully molecular approaches to solving complex computational problems using chemistry.

Modeling Charge Transport in Organic Semiconductors Using Neural Network Based Hamiltonians and Forces.

The fewest switches surface hopping method has been widely used for the simulation of charge transport in organic semiconductors. In the present study, we perform nonadiabatic molecular dynamics (NAMD) simulations of hole transport in anthracene and pentacene. The simulations employ neural network (NN) based Hamiltonians in two different nuclear relaxation schemes, which utilize either a precalculated reorganization energy or site energy gradients additionally obtained from NN models. The performance of the NN models is evaluated in reproducing hole mobilities and inverse participation ratios in terms of both quality and computational cost. The results show that charge mobilities and inverse participation ratios obtained by models, which were trained on DFTB or DFT training data, are in very good agreement with the respective QM reference method for implicit relaxation and, where available, also for explicit relaxation. Reasonable agreement with experimental hole mobilities is achieved. Utilizing our models in NAMD simulations of charge transfer amounts to a reduction of the computational cost in a range of 1 to 7 orders of magnitude compared to DFTB and DFT. This proves neural networks as promising tools for the improvement of accuracy and efficiency of charge and potentially also exciton transport simulations in complex and large molecular systems.

High-Throughput Synthesis and Machine Learning Assisted Design of Photodegradable Hydrogels.

Due to the large chemical space, the design of functional and responsive soft materials poses many challenges but also offers a wide range of opportunities in terms of the scope of possible properties. Herein, an experimental workflow for miniaturized combinatorial high-throughput screening of functional hydrogel libraries is reported. The data created from the analysis of the photodegradation process of more than 900 different types of hydrogel pads are used to train a machine learning model for automated decision making. Through iterative model optimization based on Bayesian optimization, a substantial improvement in response properties is achieved and thus expanded the scope of material properties obtainable within the chemical space of hydrogels in the study. It is therefore demonstrated that the potential of combining miniaturized high-throughput experiments with smart optimization algorithms for cost and time efficient optimization of materials properties.

Graph neural networks for materials science and chemistry.

Machine learning plays an increasingly important role in many areas of chemistry and materials science, being used to predict materials properties, accelerate simulations, design new structures, and predict synthesis routes of new materials. Graph neural networks (GNNs) are one of the fastest growing classes of machine learning models. They are of particular relevance for chemistry and materials science, as they directly work on a graph or structural representation of molecules and materials and therefore have full access to all relevant information required to characterize materials. In this Review, we provide an overview of the basic principles of GNNs, widely used datasets, and state-of-the-art architectures, followed by a discussion of a wide range of recent applications of GNNs in chemistry and materials science, and concluding with a road-map for the further development and application of GNNs.

SELFIES and the future of molecular string representations.

Artificial intelligence (AI) and machine learning (ML) are expanding in popularity for broad applications to challenging tasks in chemistry and materials science. Examples include the prediction of properties, the discovery of new reaction pathways, or the design of new molecules. The machine needs to read and write fluently in a chemical language for each of these tasks. Strings are a common tool to represent molecular graphs, and the most popular molecular string representation, Smiles, has powered cheminformatics since the late 1980s. However, in the context of AI and ML in chemistry, Smiles has several shortcomings-most pertinently, most combinations of symbols lead to invalid results with no valid chemical interpretation. To overcome this issue, a new language for molecules was introduced in 2020 that guarantees 100% robustness: SELF-referencing embedded string (Selfies). Selfies has since simplified and enabled numerous new applications in chemistry. In this perspective, we look to the future and discuss molecular string representations, along with their respective opportunities and challenges. We propose 16 concrete future projects for robust molecular representations. These involve the extension toward new chemical domains, exciting questions at the interface of AI and robust languages, and interpretability for both humans and machines. We hope that these proposals will inspire several follow-up works exploiting the full potential of molecular string representations for the future of AI in chemistry and materials science.

On scientific understanding with artificial intelligence.

An oracle that correctly predicts the outcome of every particle physics experiment, the products of every possible chemical reaction or the function of every protein would revolutionize science and technology. However, scientists would not be entirely satisfied because they would want to comprehend how the oracle made these predictions. This is scientific understanding, one of the main aims of science. With the increase in the available computational power and advances in artificial intelligence, a natural question arises: how can advanced computational systems, and specifically artificial intelligence, contribute to new scientific understanding or gain it autonomously? Trying to answer this question, we adopted a definition of 'scientific understanding' from the philosophy of science that enabled us to overview the scattered literature on the topic and, combined with dozens of anecdotes from scientists, map out three dimensions of computer-assisted scientific understanding. For each dimension, we review the existing state of the art and discuss future developments. We hope that this Perspective will inspire and focus research directions in this multidisciplinary emerging field.

MOF Synthesis Prediction Enabled by Automatic Data Mining and Machine Learning.

Despite rapid progress in the field of metal-organic frameworks (MOFs), the potential of using machine learning (ML) methods to predict MOF synthesis parameters is still untapped. Here, we show how ML can be used for rationalization and acceleration of the MOF discovery process by directly predicting the synthesis conditions of a MOF based on its crystal structure. Our approach is based on: i) establishing the first MOF synthesis database via automatic extraction of synthesis parameters from the literature, ii) training and optimizing ML models by employing the MOF database, and iii) predicting the synthesis conditions for new MOF structures. The ML models, even at an initial stage, exhibit a good prediction performance, outperforming human expert predictions, obtained through a synthesis survey. The automated synthesis prediction is available via a web-tool on https://mof-synthesis.aimat.science.

Updated Calibrated Model for the Prediction of Molecular Frontier Orbital Energies and Its Application to Boron Subphthalocyanines.

A diverse range of computational methods have been used to calibrate against available data and to compare against the correlation for the prediction of frontier orbital energies and optical gaps of novel boron subphthalocyanine (BsubPc) derivatives and related compounds. These properties are of fundamental importance to organic electronic material applications and development, making BsubPcs ideal candidates in pursuit of identifying promising materials for targeted applications. This work employs a database of highly accurate experimental data from materials produced and characterized in-house. The models presented herein calibrate these properties with R2 values > 0.95. We find that computationally inexpensive semiempirical methods such as PM6 and PM7 outperform most density functional theory methods for calibration. We are excited to share these results with the field as it empowers the community to determine key physical properties of BsubPcs with confidence using free software and a standard laptop prior to the arduous synthesis and purification thereof. This study is a follow up to our previous work calibrating PM3, RM1, and B3LYP-6-31G(d), which used a smaller set of BsubPc derivatives at a past point when less data were available.

A Comprehensive Discovery Platform for Organophosphorus Ligands for Catalysis.

The design of molecular catalysts typically involves reconciling multiple conflicting property requirements, largely relying on human intuition and local structural searches. However, the vast number of potential catalysts requires pruning of the candidate space by efficient property prediction with quantitative structure-property relationships. Data-driven workflows embedded in a library of potential catalysts can be used to build predictive models for catalyst performance and serve as a blueprint for novel catalyst designs. Herein we introduce kraken, a discovery platform covering monodentate organophosphorus(III) ligands providing comprehensive physicochemical descriptors based on representative conformer ensembles. Using quantum-mechanical methods, we calculated descriptors for 1558 ligands, including commercially available examples, and trained machine learning models to predict properties of over 300000 new ligands. We demonstrate the application of kraken to systematically explore the property space of organophosphorus ligands and how existing data sets in catalysis can be used to accelerate ligand selection during reaction optimization.

Fast Generation of Machine Learning-Based Force Fields for Adsorption Energies.

Adsorption and desorption of molecules are key processes in extraction and purification of biomolecules, engineering of drug carriers, and designing of surface-specific coatings. To understand the adsorption process on the atomic scale, state-of-the-art quantum mechanical and classical simulation methodologies are widely used. However, studying adsorption using a full quantum mechanical treatment is limited to picoseconds simulation timescales, while classical molecular dynamics simulations are limited by the accuracy of the existing force fields. To overcome these challenges, we propose a systematic way to generate flexible, application-specific highly accurate force fields by training artificial neural networks. As a proof of concept, we study the adsorption of the amino acid alanine on graphene and gold (111) surfaces and demonstrate the force field generation methodology in detail. We find that a molecule-specific force field with Lennard-Jones type two-body terms incorporating the 3rd and 7th power of the inverse distances between the atoms of the adsorbent and the surfaces yields optimal results, which is surprisingly different from typical Lennard-Jones potentials used in traditional force fields. Furthermore, we present an efficient and easy-to-train machine learning model that incorporates system-specific three-body (or higher order) interactions that are required, for example, for gold surfaces. Our final machine learning-based force field yields a mean absolute error of less than 4.2 kJ/mol at a speed-up of ∼105 times compared to quantum mechanical calculation, which will have a significant impact on the study of adsorption in different research areas.

Phase-Property Diagrams for Multicomponent Oxide Systems toward Materials Libraries.

Exploring the vast compositional space offered by multicomponent systems or high entropy materials using the traditional route of materials discovery, one experiment at a time, is prohibitive in terms of cost and required time. Consequently, the development of high-throughput experimental methods, aided by machine learning and theoretical predictions will facilitate the search for multicomponent materials in their compositional variety. In this study, high entropy oxides are fabricated and characterized using automated high-throughput techniques. For intuitive visualization, a graphical phase-property diagram correlating the crystal structure, the chemical composition, and the band gap are introduced. Interpretable machine learning models are trained for automated data analysis and to speed up data comprehension. The establishment of materials libraries of multicomponent systems correlated with their properties (as in the present work), together with machine learning-based data analysis and theoretical approaches are opening pathways toward virtual development of novel materials for both functional and structural applications.

Automatic discovery of photoisomerization mechanisms with nanosecond machine learning photodynamics simulations.

Photochemical reactions are widely used by academic and industrial researchers to construct complex molecular architectures via mechanisms that often require harsh reaction conditions. Photodynamics simulations provide time-resolved snapshots of molecular excited-state structures required to understand and predict reactivities and chemoselectivities. Molecular excited-states are often nearly degenerate and require computationally intensive multiconfigurational quantum mechanical methods, especially at conical intersections. Non-adiabatic molecular dynamics require thousands of these computations per trajectory, which limits simulations to ∼1 picosecond for most organic photochemical reactions. Westermayr et al. recently introduced a neural-network-based method to accelerate the predictions of electronic properties and pushed the simulation limit to 1 ns for the model system, methylenimmonium cation (CH2NH2 +). We have adapted this methodology to develop the Python-based, Python Rapid Artificial Intelligence Ab Initio Molecular Dynamics (PyRAI2MD) software for the cis-trans isomerization of trans-hexafluoro-2-butene and the 4π-electrocyclic ring-closing of a norbornyl hexacyclodiene. We performed a 10 ns simulation for trans-hexafluoro-2-butene in just 2 days. The same simulation would take approximately 58 years with traditional multiconfigurational photodynamics simulations. We generated training data by combining Wigner sampling, geometrical interpolations, and short-time quantum chemical trajectories to adaptively sample sparse data regions along reaction coordinates. The final data set of the cis-trans isomerization and the 4π-electrocyclic ring-closing model has 6207 and 6267 data points, respectively. The training errors in energy using feedforward neural networks achieved chemical accuracy (0.023-0.032 eV). The neural network photodynamics simulations of trans-hexafluoro-2-butene agree with the quantum chemical calculations showing the formation of the cis-product and reactive carbene intermediate. The neural network trajectories of the norbornyl cyclohexadiene corroborate the low-yielding syn-product, which was absent in the quantum chemical trajectories, and revealed subsequent thermal reactions in 1 ns.

Analyzing Dynamical Disorder for Charge Transport in Organic Semiconductors via Machine Learning.

Organic semiconductors are indispensable for today's display technologies in the form of organic light-emitting diodes (OLEDs) and further optoelectronic applications. However, organic materials do not reach the same charge carrier mobility as inorganic semiconductors, limiting the efficiency of devices. To find or even design new organic semiconductors with higher charge carrier mobility, computational approaches, in particular multiscale models, are becoming increasingly important. However, such models are computationally very costly, especially when large systems and long timescales are required, which is the case to compute static and dynamic energy disorder, i.e., the dominant factor to determine charge transport. Here, we overcome this drawback by integrating machine learning models into multiscale simulations. This allows us to obtain unprecedented insight into relevant microscopic materials properties, in particular static and dynamic disorder contributions for a series of application-relevant molecules. We find that static disorder and thus the distribution of shallow traps are highly asymmetrical for many materials, impacting widely considered Gaussian disorder models. We furthermore analyze characteristic energy level fluctuation times and compare them to typical hopping rates to evaluate the importance of dynamic disorder for charge transport. We hope that our findings will significantly improve the accuracy of computational methods used to predict application-relevant materials properties of organic semiconductors and thus make these methods applicable for virtual materials design.

Machine-learned potentials for next-generation matter simulations.

The choice of simulation methods in computational materials science is driven by a fundamental trade-off: bridging large time- and length-scales with highly accurate simulations at an affordable computational cost. Venturing the investigation of complex phenomena on large scales requires fast yet accurate computational methods. We review the emerging field of machine-learned potentials, which promises to reach the accuracy of quantum mechanical computations at a substantially reduced computational cost. This Review will summarize the basic principles of the underlying machine learning methods, the data acquisition process and active learning procedures. We highlight multiple recent applications of machine-learned potentials in various fields, ranging from organic chemistry and biomolecules to inorganic crystal structure predictions and surface science. We furthermore discuss the developments required to promote a broader use of ML potentials, and the possibility of using them to help solve open questions in materials science and facilitate fully computational materials design.