Machine Learning for Electrocatalyst and Photocatalyst Design and Discovery.

Electrocatalysts and photocatalysts are key to a sustainable future, generating clean fuels, reducing the impact of global warming, and providing solutions to environmental pollution. Improved processes for catalyst design and a better understanding of electro/photocatalytic processes are essential for improving catalyst effectiveness. Recent advances in data science and artificial intelligence have great potential to accelerate electrocatalysis and photocatalysis research, particularly the rapid exploration of large materials chemistry spaces through machine learning. Here a comprehensive introduction to, and critical review of, machine learning techniques used in electrocatalysis and photocatalysis research are provided. Sources of electro/photocatalyst data and current approaches to representing these materials by mathematical features are described, the most commonly used machine learning methods summarized, and the quality and utility of electro/photocatalyst models evaluated. Illustrations of how machine learning models are applied to novel electro/photocatalyst discovery and used to elucidate electrocatalytic or photocatalytic reaction mechanisms are provided. The review offers a guide for materials scientists on the selection of machine learning methods for electrocatalysis and photocatalysis research. The application of machine learning to catalysis science represents a paradigm shift in the way advanced, next-generation catalysts will be designed and synthesized.

Synthesis of Layered Lead-Free Perovskite Nanocrystals with Precise Size and Shape Control and Their Photocatalytic Activity.

Halide perovskites have attracted enormous attention due to their potential applications in optoelectronics and photocatalysis. However, concerns over their instability, toxicity, and unsatisfactory efficiency have necessitated the development of lead-free all-inorganic halide perovskites. A major challenge in designing efficient halide perovskites for practical applications is the lack of effective methods for producing nanocrystals with precise size and shape control. In this work, a layered perovskite, Cs4ZnSb2Cl12 (CZS), is found from calculations to exhibit size- and facet-dependent optoelectronic properties in the nanoscale, and thus, a colloidal method is used to synthesize the CZS nanoparticles with size-tunable morphologies: zero- (nanodots), one- (nanowires and nanorods), two- (nanoplates), and three-dimensional (nanopolyhedra). The growth kinetics of the CZS nanostructures, along with the effects of surface ligands, reaction temperature, and time were investigated. The optoelectronic properties of the nanocrystals varied with size due to quantum confinement effects and with shape due to anisotropy within the crystals and the exposure of specific facets. These properties could be modulated to enhance the visible-light photocatalytic performance for toluene oxidation. In particular, the 9.7 nm CZS nanoplates displayed a toluene to benzaldehyde conversion rate of 1893 µmol g-1 h-1 (95% selectivity), 500 times higher than the bulk synthesized CZS, and comparable with the reported photocatalysts. This study demonstrates the integration of theoretical calculations and synthesis, revealing an approach to the design and fabrication of novel, high-performance colloidal perovskite nanocrystals for optoelectronic and photocatalytic applications.

Rational Atom Substitution to Obtain Efficient, Lead-Free Photocatalytic Perovskites Assisted by Machine Learning and DFT Calculations.

Inorganic lead-free halide perovskites, devoid of toxic or rare elements, have garnered considerable attention as photocatalysts for pollution control, CO2 reduction and hydrogen production. In the extensive perovskite design space, factors like substitution or doping level profoundly impact their performance. To address this complexity, a synergistic combination of machine learning models and theoretical calculations were used to efficiently screen substitution elements that enhanced the photoactivity of substituted Cs2 AgBiBr6 perovskites. Machine learning models determined the importance of d10 orbitals, highlighting how substituent electron configuration affects electronic structure of Cs2 AgBiBr6 . Conspicuously, d10 -configured Zn2+ boosted the photoactivity of Cs2 AgBiBr6 . Experimental verification validated these model results, revealing a 13-fold increase in photocatalytic toluene conversion compared to the unsubstituted counterpart. This enhancement resulted from the small charge carrier effective mass, as well as the creation of shallow trap states, shifting the conduction band minimum, introducing electron-deficient Br, and altering the distance between the B-site cations d band centre and the halide anions p band centre, a parameter tuneable through d10 configuration substituents. This study exemplifies the application of computational modelling in photocatalyst design and elucidating structure-property relationships. It underscores the potential of synergistic integration of calculations, modelling, and experimental analysis across various applications.

Membrane Permeating Macrocycles: Design Guidelines from Machine Learning.

The ability to predict cell-permeable candidate molecules has great potential to assist drug discovery projects. Large molecules that lie beyond the Rule of Five (bRo5) are increasingly important as drug candidates and tool molecules for chemical biology. However, such large molecules usually do not cross cell membranes and cannot access intracellular targets or be developed as orally bioavailable drugs. Here, we describe a random forest (RF) machine learning model for the prediction of passive membrane permeation rates developed using a set of over 1000 bRo5 macrocyclic compounds. The model is based on easily calculated chemical features/descriptors as independent variables. Our random forest (RF) model substantially outperforms a multiple linear regression model based on the same features and achieves better performance metrics than previously reported models using the same underlying data. These features include: (1) polar surface area in water, (2) the octanol-water partitioning coefficient, (3) the number of hydrogen-bond donors, (4) the sum of the topological distances between nitrogen atoms, (5) the sum of the topological distances between nitrogen and oxygen atoms, and (6) the multiple molecular path count of order 2. The last three features represent molecular flexibility, the ability of the molecule to adopt different conformations in the aqueous and membrane interior phases, and the molecular "chameleonicity." Guided by the model, we propose design guidelines for membrane-permeating macrocycles. It is anticipated that this model will be useful in guiding the design of large, bioactive molecules for medicinal chemistry and chemical biology applications.

Machine learning investigation of viscosity and ionic conductivity of protic ionic liquids in water mixtures.

Ionic liquids (ILs) are well classified as designer solvents based on the ease of tailoring their properties through modifying the chemical structure of the cation and anion. However, while many structure-property relationships have been developed, these generally only identify the most dominant trends. Here, we have used machine learning on existing experimental data to construct robust models to produce meaningful predictions across a broad range of cation and anion chemical structures. Specifically, we used previously collated experimental data for the viscosity and conductivity of protic ILs [T. L. Greaves and C. J. Drummond, Chem. Rev. 115, 11379-11448 (2015)] as the inputs for multiple linear regression and neural network models. These were then used to predict the properties of all 1827 possible cation-anion combinations (excluding the input combinations). These models included the effect of water content of up to 5 wt. %. A selection of ten new protic ILs was then prepared, which validated the usefulness of the models. Overall, this work shows that relatively sparse data can be used productively to predict physicochemical properties of vast arrays of ILs.

Design of Lipid-Based Nanocarriers via Cation Modulation of Ethanol-Interdigitated Lipid Membranes.

Short-chain alcohols (i.e., ethanol) can induce membrane interdigitation in saturated-chain phosphatidylcholines (PCs). In this process, alcohol molecules intercalate between phosphate heads, increasing lateral separation and favoring hydrophobic interactions between opposing acyl chains, which interpenetrate forming an interdigitated phase. Unraveling mechanisms underlying the interactions between ethanol and model lipid membranes has implications for cell biology, biochemistry, and for the formulation of lipid-based nanocarriers. However, investigations of ethanol-lipid membrane systems have been carried out in deionized water, which limits their applicability. Here, using a combination of small- and wide-angle X-ray scattering, small-angle neutron scattering, and all-atom molecular dynamics simulations, we analyzed the effect of varying CaCl2 and NaCl concentrations on ethanol-induced interdigitation. We observed that while ethanol addition leads to the interdigitation of bulk phase 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) bilayers in the presence of CaCl2 and NaCl regardless of the salt concentration, the ethanol-induced interdigitation of vesicular DPPC depends on the choice of cation and its concentration. These findings unravel a key role for cations in the ethanol-induced interdigitation of lipid membranes in either bulk phase or vesicular form.

Systematic Comparison of the Structural and Dynamic Properties of Commonly Used Water Models for Molecular Dynamics Simulations.

Water is a unique solvent that is ubiquitous in biology and present in a variety of solutions, mixtures, and materials settings. It therefore forms the basis for all molecular dynamics simulations of biological phenomena, as well as for many chemical, industrial, and materials investigations. Over the years, many water models have been developed, and it remains a challenge to find a single water model that accurately reproduces all experimental properties of water simultaneously. Here, we report a comprehensive comparison of structural and dynamic properties of 30 commonly used 3-point, 4-point, 5-point, and polarizable water models simulated using consistent settings and analysis methods. For the properties of density, coordination number, surface tension, dielectric constant, self-diffusion coefficient, and solvation free energy of methane, models published within the past two decades consistently show better agreement with experimental values compared to models published earlier, albeit with some notable exceptions. However, no single model reproduced all experimental values exactly, highlighting the need to carefully choose a water model for a particular study, depending on the phenomena of interest. Finally, machine learning algorithms quantified the relationship between the water model force field parameters and the resulting bulk properties, providing insight into the parameter-property relationship and illustrating the challenges of developing a water model that can accurately reproduce all properties of water simultaneously.

Machine learning approaches to understand and predict rate constants for organic processes in mixtures containing ionic liquids.

The ability to tailor the constituent ions in ionic liquids (ILs) is highly advantageous as it provides access to solvents with a range of physicochemical properties. However, this benefit also leads to large compositional spaces that need to be explored to optimise systems, often involving time consuming experimental work. The use of machine learning methods is an effective way to gain insight based on existing data, to develop structure-property relationships and to allow the prediction of ionic liquid properties. Here we have applied machine learning models to experimentally determined rate constants of a representative organic process (the reaction of pyridine with benzyl bromide) in IL-acetonitrile mixtures. Multiple linear regression (MLREM) and artificial neural networks (BRANNLP) were both able to model the data well. The MLREM model was able to identify the structural features on the cations and anions that had the greatest effect on the rate constant. Secondly, predictive MLREM and BRANNLP models were developed from the full initial set of rate constant data. From these models, a large number of predictions (>9000) of rate constant were made for mixtures of different ionic liquids, at different proportions of ionic liquid and molecular solvent, at different temperatures. A selection of these predictions were tested experimentally, including through the preparation of novel ionic liquids, with overall good agreement between the predicted and experimental data. This study highlights the benefits of using machine learning methods on kinetic data in ionic liquid mixtures to enable the development of rigorous structure-property relationships across multiple variables simultaneously, and to predict properties of new ILs and experimental conditions.

Review on the Use of Artificial Intelligence to Predict Fire Performance of Construction Materials and Their Flame Retardancy.

The evaluation and interpretation of the behavior of construction materials under fire conditions have been complicated. Over the last few years, artificial intelligence (AI) has emerged as a reliable method to tackle this engineering problem. This review summarizes existing studies that applied AI to predict the fire performance of different construction materials (e.g., concrete, steel, timber, and composites). The prediction of the flame retardancy of some structural components such as beams, columns, slabs, and connections by utilizing AI-based models is also discussed. The end of this review offers insights on the advantages, existing challenges, and recommendations for the development of AI techniques used to evaluate the fire performance of construction materials and their flame retardancy. This review offers a comprehensive overview to researchers in the fields of fire engineering and material science, and it encourages them to explore and consider the use of AI in future research projects.

Manipulating the Ordered Nanostructure of Self-Assembled Monoolein and Phytantriol Nanoparticles with Unsaturated Fatty Acids.

Mesophase structures of self-assembled lyotropic liquid crystalline nanoparticles are important factors that directly influence their ability to encapsulate and release drugs and their biological activities. However, it is difficult to predict and precisely control the mesophase behavior of these materials, especially in complex systems with several components. In this study, we report the controlled manipulation of mesophase structures of monoolein (MO) and phytantriol (PHYT) nanoparticles by adding unsaturated fatty acids (FAs). By using high throughput formulation and small-angle X-ray scattering characterization methods, the effects of FAs chain length, cis-trans isomerism, double bond location, and level of chain unsaturation on self-assembled systems are determined. Additionally, the influence of temperature on the phase behavior of these nanoparticles is analyzed. We found that in general, the addition of unsaturated FAs to MO and PHYT induces the formation of mesophases with higher Gaussian surface curvatures. As a result, a rich variety of lipid polymorphs are found to correspond with the increasing amounts of FAs. These phases include inverse bicontinuous cubic, inverse hexagonal, and discrete micellar cubic phases and microemulsion. However, there are substantial differences between the phase behavior of nanoparticles with trans FA, cis FAs with one double bond, and cis FAs with multiple double bonds. Therefore, the material library produced in this study will assist the selection and development of nanoparticle-based drug delivery systems with desired mesophase.

Discovery and Optimization of Materials Using Evolutionary Approaches.

Materials science is undergoing a revolution, generating valuable new materials such as flexible solar panels, biomaterials and printable tissues, new catalysts, polymers, and porous materials with unprecedented properties. However, the number of potentially accessible materials is immense. Artificial evolutionary methods such as genetic algorithms, which explore large, complex search spaces very efficiently, can be applied to the identification and optimization of novel materials more rapidly than by physical experiments alone. Machine learning models can augment experimental measurements of materials fitness to accelerate identification of useful and novel materials in vast materials composition or property spaces. This review discusses the problems of large materials spaces, the types of evolutionary algorithms employed to identify or optimize materials, and how materials can be represented mathematically as genomes, describes fitness landscapes and mutation operators commonly employed in materials evolution, and provides a comprehensive summary of published research on the use of evolutionary methods to generate new catalysts, phosphors, and a range of other materials. The review identifies the potential for evolutionary methods to revolutionize a wide range of manufacturing, medical, and materials based industries.

An Experimental and Computational Approach to the Development of ZnO Nanoparticles that are Safe by Design.

Zinc oxide nanoparticles have found wide application due to their unique optoelectronic and photocatalytic characteristics. However, their safety aspects remain of critical concern, prompting the use of physicochemical modifications of pristine ZnO to reduce any potential toxicity. However, the relationships between these modifications and their effects on biology are complex and still relatively unexplored. To address this knowledge gap, a library of 45 types of ZnO nanoparticles with varying particle size, aspect ratio, doping type, doping concentration, and surface coating is synthesized, and their biological effects measured. Three biological assays measuring cell damage or stress are used to study the responses of human umbilical vein endothelial cells (HUVECs) or human hepatocellular liver carcinoma cells (HepG2) to the nanoparticles. These experimental data are used to develop quantitative and predictive computational models linking nanoparticle properties to cell viability, membrane integrity, and oxidative stress. It is found that the concentration of nanoparticles the cells are exposed to, the type of surface coating, the nature and extent of doping, and the aspect ratio of the particles make significant contributions to the cell toxicity of the nanoparticles tested. Our study shows that it is feasible to generate models that could be used to design or optimize nanoparticles with commercially useful properties that are also safe to humans and the environment.

Modeling the Influence of Fatty Acid Incorporation on Mesophase Formation in Amphiphilic Therapeutic Delivery Systems.

Dispersed amphiphile-fatty acid systems are of great interest in drug delivery and gene therapies because of their potential for triggered release of their payload. The mesophase behavior of these systems is extremely complex and is affected by environmental factors such as drug loading, percentage and nature of incorporated fatty acids, temperature, pH, and so forth. It is important to study phase behavior of amphiphilic materials as the mesophases directly influence the release rate of the incorporated drugs. We describe a robust machine learning method for predicting the phase behavior of these systems. We have developed models for each mesophase that simultaneous and reliably model the effects of amphiphile and fatty acid structure, concentration, and temperature and that make accurate predictions of these mesophases for conditions not used to train the models.

Biophysical Characterization and Cryopreservation of Mammalian Cells Using Ionic Liquids.

Ionic liquids (ILs) are a diverse class of solvents which can be selected for task-specific properties, making them attractive alternatives to traditional solvents. To tailor ILs for specific biological applications, it is necessary to understand the structure-property relationships of ILs and their interactions with cells. Here, a selection of carboxylate anion-based ILs were investigated as cryoprotectants, which are compounds added to cells before freezing to mitigate lethal freezing damage. The cytotoxicity, cell permeability, thermal behavior, and cryoprotective efficacy of the ILs were assessed with two model mammalian cell lines. We found that the biophysical interactions, including permeability of the ILs, were influenced by considering the IL pair together, rather than as single species acting independently. All of the ILs tested had high cytotoxicity, but ethylammonium acetate demonstrated good cryoprotective efficacy for both cell types tested. These results demonstrate that despite toxicity, ILs may be suitable for certain biological applications. It also demonstrates that more research is required to understand the contribution of ion pairs to structure-property relationships and that knowing the behavior of a single ionic species will not necessarily predict its behavior as part of an IL.

Aqueous solubility prediction: do crystal lattice interactions help?

Aqueous solubility is a very important physical property of small molecule drugs and drug candidates but also one of the most difficult to predict accurately. Aqueous solubility plays a major role in drug delivery and pharmacokinetics. It is believed that crystal lattice interactions are important in solubility and that including them in solubility models should improve the accuracy of the models. We used calculated values for lattice energy and sublimation enthalpy of organic molecules as descriptors to determine whether these would improve the accuracy of the aqueous solubility models. Multiple linear regression employing an expectation maximization algorithm and a sparse prior (MLREM) method and a nonlinear Bayesian regularized artificial neural network with a Laplacian prior (BRANNLP) were used to derive optimal predictive models of aqueous solubility of a large and highly diverse data set of 4558 organic compounds over a normal ambient temperature range of 20-30 °C (293-303 K). A randomly selected test set and compounds from a solubility challenge were used to estimate the predictive ability of the models. The BRANNLP method showed the best statistical results with squared correlation coefficients of 0.90 and standard errors of 0.645-0.665 log(S) for training and test sets. Surprisingly, including descriptors that captured crystal lattice interactions did not significantly improve the quality of these aqueous solubility models.

Predicting the complex phase behavior of self-assembling drug delivery nanoparticles.

Amphiphilic lyotropic liquid crystalline self-assembled nanomaterials have important applications in the delivery of therapeutic and imaging agents. However, little is known about the effect of the incorporated drug on the structure of nanoparticles. Predicting these properties is widely considered intractable. We present computational models for three drug delivery carriers, loaded with 10 drugs at six concentrations and two temperatures. These models predicted phase behavior for 11 new drugs. Subsequent synchrotron small-angle X-ray scattering experiments validated the predictions.

Capturing the crystal: prediction of enthalpy of sublimation, crystal lattice energy, and melting points of organic compounds.

Accurate computational prediction of melting points and aqueous solubilities of organic compounds would be very useful but is notoriously difficult. Predicting the lattice energies of compounds is key to understanding and predicting their melting behavior and ultimately their solubility behavior. We report robust, predictive, quantitative structure-property relationship (QSPR) models for enthalpies of sublimation, crystal lattice energies, and melting points for a very large and structurally diverse set of small organic compounds. Sparse Bayesian feature selection and machine learning methods were employed to select the most relevant molecular descriptors for the model and to generate parsimonious quantitative models. The final enthalpy of sublimation model is a four-parameter multilinear equation that has an r(2) value of 0.96 and an average absolute error of 7.9 ± 0.3 kJ.mol(-1). The melting point model can predict this property with a standard error of 45° ± 1 K and r(2) value of 0.79. Given the size and diversity of the training data, these conceptually transparent and accurate models can be used to predict sublimation enthalpy, lattice energy, and melting points of organic compounds in general.

Accelerating the prediction of CO2 capture at low partial pressures in metal-organic frameworks using new machine learning descriptors.

Metal-Organic frameworks (MOFs) have been considered for various gas storage and separation applications. Theoretically, there are an infinite number of MOFs that can be created; however, a finite amount of resources are available to evaluate each one. Computational methods can be adapted to expedite the process of evaluation. In the context of CO2 capture, this paper investigates the method of screening MOFs using machine learning trained on molecular simulation data. New descriptors are introduced to aid this process. Using all descriptors, it is shown that machine learning can predict the CO2 adsorption, with an R2 of above 0.9. The introduced Effective Point Charge (EPoCh) descriptors, which assign values to frameworks' partial charges based on the expected CO2 uptake of an equivalent point charge in isolation, are shown to be the second most important group of descriptors, behind the Henry coefficient. Furthermore, the EPoCh descriptors are hundreds of thousands of times faster to obtain compared with the Henry coefficient, and they achieve similar results when identifying top candidates for CO2 capture using pseudo-classification predictions.

Quantum Chemistry-Machine Learning Approach for Predicting Properties of Lewis Acid-Lewis Base Adducts.

Synthetic design allowing predictive control of charge transfer and other optoelectronic properties of Lewis acid adducts remains elusive. This challenge must be addressed through complementary methods combining experimental with computational insights from first principles. Ab initio calculations for optoelectronic properties can be computationally expensive and less straightforward than those sufficient for simple ground-state properties, especially for adducts of large conjugated molecules and Lewis acids. In this contribution, we show that machine learning (ML) can accurately predict density functional theory (DFT)-calculated charge transfer and even properties associated with excited states of adducts from readily obtained molecular descriptors. Seven ML models, built from a dataset of over 1000 adducts, show exceptional performance in predicting charge transfer and other optoelectronic properties with a Pearson correlation coefficient of up to 0.99. More importantly, the influence of each molecular descriptor on predicted properties can be quantitatively evaluated from ML models. This contributes to the optimization of a priori design of Lewis adducts for future applications, especially in organic electronics.