Machine-learning descriptor search on the density of states profile of bimetallic alloy systems and comparison with the d-band center theory

In this study, the electronic density of states (DOSs) calculated with density functional theory (DFT) were analyzed by the machine-learning techniques. More than 400 pure metal and bimetallic alloy systems were calculated with DFT, and obtained the surface DOSs and the CH3 adsorption energy (Ead). By fitting the Gaussian functions to the DOS, multiple descriptors, such as the Gaussian peak positions, heights, and widths were extracted. Several regression methods, such as the least absolute shrinkage of selection operator (LASSO), random-forest, gradient-boosting, and extra-tree were used to find the relationship between these descriptors and the Ead. The results show that the energy position of the peaks in the d-projected DOS is the most important descriptor, in agreement with the previously known d-band center theory. It was also shown that the peak position in d-projected DOS improves the regression model in addition to the d-band center, since it reduces the regression error.

Theoretical search for characteristic atoms in supported gold nanoparticles: a large-scale DFT study

The size and site dependences of atomic and electronic structures in isolated and supported gold nanoparticles have been investigated using large-scale density functional theory (DFT) calculations using multi-site support functions. The effects of the substrate on nanoparticles with diameters of 2 nm and several different shapes have been examined. First, isolated gold nanoparticles with diameters of 0.6 nm (13 atoms) to 4.5 nm (2057 atoms), which have comparable sizes to nanoparticles used in experiments, were considered. To analyse huge amounts of data obtained from large-scale DFT calculations, we performed principal component analysis (PCA), which helps systematically and efficiently clarify the electronic structures of large nanoparticles. The PCA results reveal the site dependence of the electronic structures. Notably, the atoms in the surface and subsurface have different electronic structures to those located in the inner layers, especially at the vertexes of the particles. The convergence of local electronic structures with respect to the particle size has also been demonstrated. For supported nanoparticles, PCA helps indicate which atoms are affected, and how much, by the substrate. The correlation between the PCA results and site dependence of reaction activity is also discussed herein.

Asymmetric Potential Model of Two-Dimensional Imine Covalent Organic Frameworks with Enhanced Quantum Efficiency for Photocatalytic Water Splitting

Two-dimensional (2D) covalent organic frameworks (COFs) assembled using building blocks have been widely employed in photocatalysis due to their customizable optoelectronic characteristics and porous structure, which facilitate charge carrier and mass movement. Nevertheless, the development of COF photocatalysts encounters a continuous obstacle in enhancing the efficiency of photocatalysis, impeded by a limited comprehension of the orbital interaction between molecular fragments and linkers. In this study, we present a model that examines the interaction between molecular fragments in an imine-based COF at the frontier molecular orbital level, enabling us to comprehend the impact of manipulating linkers on light adsorption, exciton efficiency, and catalytic activity. Our findings demonstrate that altering the connecting orientation of 14 R-C=N-R imine linkers in 2D COFs can enhance solar-to-hydrogen (STH) efficiency under visible light from 2.76% to 4.24%. This research has the potential to provide a valuable model for refining photocatalysts with enhanced photocatalytic performance.

Downshift of the Ni d band center over Ni nanoparticles in situ confined within an amorphous silicon nitride matrix

Herein, nanocomposites made of Ni nanoparticles in situ distributed in an amorphous silicon nitride (Ni/a-Si3N4) matrix, on the one hand, and within an amorphous silicon dioxide (Ni/a-SiO2) matrix, on the other hand, were synthesized from the same Ni-modified polysilazane precursor. In both compounds, the Ni/Si atomic ratio (0.06-0.07), average Ni nanocrystallite size (7.0-7.6 nm) and micro/mesoporosity of the matrix were rigorously fixed. Hydrogen (H2)-temperature-programmed desorption (TPD) profile analysis revealed that the activation energy for H2 desorption at about 100-130 °C evaluated for the Ni/a-Si3N4 sample (47.4 kJ mol-1) was lower than that for the Ni/a-SiO2 sample (68.0 kJ mol-1). Mechanistic study with X-ray photoelectron spectroscopy (XPS) analysis and density functional theory (DFT) calculations revealed that, at Ni nanoparticle/matrix heterointerfaces, Ni becomes more covalently bonded to N atoms in the a-Si3N4 matrix compared to O atoms in the a-SiO2 matrix. Therefore, based on experimental and theoretical studies, we elucidated that nickel-nitrogen (Ni-N) interactions at the heterointerface lead to remarkable Ni d band broadening and downshifting of the d band center relative to those generated by Ni-oxygen (Ni-O) interactions at the heterointerface. This facilitates H2 desorption, as experimentally observed in the Ni/a-Si3N4 sample.

Interpreting chemisorption strength with AutoML-based feature deletion experiments

The chemisorption energy of reactants on a catalyst surface, [Formula: see text], is among the most informative characteristics of understanding and pinpointing the optimal catalyst. The intrinsic complexity of catalyst surfaces and chemisorption reactions presents significant difficulties in identifying the pivotal physical quantities determining [Formula: see text]. In response to this, the study proposes a methodology, the feature deletion experiment, based on Automatic Machine Learning (AutoML) for knowledge extraction from a high-throughput density functional theory (DFT) database. The study reveals that, for binary alloy surfaces, the local adsorption site geometric information is the primary physical quantity determining [Formula: see text], compared to the electronic and physiochemical properties of the catalyst alloys. By integrating the feature deletion experiment with instance-wise variable selection (INVASE), a neural network-based explainable AI (XAI) tool, we established the best-performing feature set containing 21 intrinsic, non-DFT computed properties, achieving an MAE of 0.23 eV across a periodic table-wide chemical space involving more than 1,600 types of alloys surfaces and 8,400 chemisorption reactions. This study demonstrates the stability, consistency, and potential of AutoML-based feature deletion experiment in developing concise, predictive, and theoretically meaningful models for complex chemical problems with minimal human intervention.

Decorating Pd-Au Nanodots Around Porous In2 O3 Nanocubes for Tolerant H2 Sensing Against Switching Response and H2 S Poisoning

With the recently-booming hydrogen (H2 ) economy by green H2 as the energy carriers and the newly-emerged exhaled diagnosis by human organ-metabolized H2 as a biomarker, H2 sensing is simultaneously required with fast response, low detection limit, and tolerant stability against humidity, switching, and poisoning. Here, reliable H2 sensing has been developed by utilizing indium oxide nanocubes decorated with palladium and gold nanodots (Pd-Au NDs/In2 O3 NCBs), which have been synthesized by combined hydrothermal reaction, annealing, and chemical bath deposition. As-prepared Pd-Au NDs/In2 O3 NCBs are observed with surface-enriched NDs and nanopores. Beneficially, Pd-Au NDs/In2 O3 NCBs show 300 ppb-low detection limit, 5 s-fast response to 500 ppm H2 , 75%RH-high humidity tolerance, and 56 days-long stability at 280 °C. Further, Pd-Au NDs/In2 O3 NCBs show excellent stability against switching sensing response, and are tolerant to H2 S poisoning even being exposed to 10 ppm H2 S at 280 °C. Such excellent H2 sensing may be attributed to the synergistic effect of the boosted Pd-Au NDs' spillover effect and interfacial electron transfer, increased adsorption sites over the porous NCBs' surface, and utilized Pd NDs' affinity with H2 and H2 S. Practically, Pd-Au NDs/In2 O3 NCBs are integrated into the H2 sensing device, which can reliably communicate with a smartphone.

Enhancing Nitrate Reduction to Ammonia Through Crystal Phase Engineering: Unveiling the Hydrogen Bonding Effect in δ-FeOOH Electrocatalysis

Crystal phase engineering has emerged as a powerful tool for tailoring the electrocatalytic performance, yet its impact on nitrate reduction to ammonia (NRA) remains largely uncharted territory. Herein, density functional theory (DFT) calculations are performed to unravel the influence of the crystal phase of FeOOH on the adsorption behavior of *NO3 . Inspiringly, FeOOH samples with four distinct crystal phases (δ, γ, α, and β) are successfully synthesized and deployed as electrocatalysts for NRA. Remarkably, among all FeOOH samples, δ-FeOOH demonstrates the superior NRA performance, achieving a NH3 Faradic efficiency (FE NH 3 $\rm{FE} _ {\rm{NH_3}}$ ) of 90.2% at -1.0 V versus reversible hydrogen electrode (RHE) and a NH3 yield rate (Yield NH 3 $\rm{Yield} _ {\rm{NH_3}}$ ) of 5.73 mg h-1 cm-2 at -1.2 V. In-depth experiments and theoretical calculations unveil the existence of hydrogen bonding interaction between δ-FeOOH and *NOx , which not only enhances the adsorption of *NOx but also disrupts the linear relationships between the free energy of *NO3 adsorption and various parameters, including limiting potential, d-band center (εd ) and transferred charge from FeOOH to *NO3 , ultimately contributing to the exceptional NRA performance.

Feature-Assisted Machine Learning for Predicting Band Gaps of Binary Semiconductors

The band gap is a key parameter in semiconductor materials that is essential for advancing optoelectronic device development. Accurately predicting band gaps of materials at low cost is a significant challenge in materials science. Although many machine learning (ML) models for band gap prediction already exist, they often suffer from low interpretability and lack theoretical support from a physical perspective. In this study, we address these challenges by using a combination of traditional ML algorithms and the 'white-box' sure independence screening and sparsifying operator (SISSO) approach. Specifically, we enhance the interpretability and accuracy of band gap predictions for binary semiconductors by integrating the importance rankings of support vector regression (SVR), random forests (RF), and gradient boosting decision trees (GBDT) with SISSO models. Our model uses only the intrinsic features of the constituent elements and their band gaps calculated using the Perdew-Burke-Ernzerhof method, significantly reducing computational demands. We have applied our model to predict the band gaps of 1208 theoretically stable binary compounds. Importantly, the model highlights the critical role of electronegativity in determining material band gaps. This insight not only enriches our understanding of the physical principles underlying band gap prediction but also underscores the potential of our approach in guiding the synthesis of new and valuable semiconductor materials.

Ultrasmall metal alloy nanozymes mimicking neutrophil enzymatic cascades for tumor catalytic therapy

Developing strategies that emulate the killing mechanism of neutrophils, which involves the enzymatic cascade of superoxide dismutase (SOD) and myeloperoxidase (MPO), shows potential as a viable approach for cancer therapy. Nonetheless, utilizing natural enzymes as therapeutics is hindered by various challenges. While nanozymes have emerged for cancer treatment, developing SOD-MPO cascade in one nanozyme remains a challenge. Here, we develop nanozymes possessing both SOD- and MPO-like activities through alloying Au and Pd, which exhibits the highest cascade activity when the ratio of Au and Pd is 1:3, attributing to the high d-band center and adsorption energy for superoxide anions, as determined through theoretical calculations. The Au1Pd3 alloy nanozymes exhibit excellent tumor therapeutic performance and safety in female tumor-bearing mice, with safety attributed to their tumor-specific killing ability and renal clearance ability caused by ultrasmall size. Together, this work develops ultrasmall AuPd alloy nanozymes that mimic neutrophil enzymatic cascades for catalytic treatment of tumors.

Comparing the Catalytic Effect of Metals for Energetic Materials: Machine Learning Prediction of Adsorption Energies on Metals

Energetic materials (EMs) and metals are the important components of solid propellants, and a strong catalysis of metals on EMs could further enhance the combustion performance of solid propellants. Accordingly, the study on the adsorption of EMs such as octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX), hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX), and ammonium dinitramide (ADN) on metals (Ti, Zr, Fe, Ni, Cu, and Al) was carried out by density functional theory (DFT) to reveal the catalytic effect of metals. The deep dissociation of EMs on Ti and Zr represents a stronger interaction and corresponds to the rapid thermal decomposition behavior of the EMs/metal composite in the experiment. It is expected that DFT calculation can be selected instead of experiments to compare the catalytic effect of metals and preliminarily screen out potential high-performance metals. Based on the data set of the calculated adsorption energy, further machine learning (ML) was used to predict the adsorption energy of EMs on metals for a convenient comparison of the catalytic effect of metals, since a quite high adsorption energy value represents a more thorough dissociation. The kernel ridge regression (KRR) method shows the best performance on predicting adsorption energy and helps to choose the metals for efficiently catalyzing ammonium nitrate (AN) and hexanitrohexaazaisowurtzitane (CL-20). Such adsorption computation and ML not only reveal the decomposition mechanism of EMs on metals but also provide a simple underlying method to predict the catalytic effect of metals.

Insight into Controllable Metal-Support Interactions in Metal/Metal Electrocatalysts for Efficient Energy-Saving Hydrogen Production

Controllable metal-support interaction (MSI) modulations have long been studied for improving the performance of catalysts supported on metal oxides. However, the corresponding in-depth study for metal1-metal2 (M1-M2) composited configurations is rarely achieved due to the lack of reliable models and manipulation mechanisms of MSI modifications. We modeled ruthenium on copper support (Ru-Cu) metal catalysts with negligible interfacial contact potential (e0.06 V) and investigated MSI-dependent hydrogen evolution reaction (HER) catalysis kinetics induced by an electronic hydroxyl (HO-) modifier. Comprehensive simulations and characterizations confirmed that adjusting the HO- coverage can readily realize the tailorable improvement of MSI, facilitating charge migration at the Ru-Cu interface and optimizing the overall HER pathway on active Ru. As a result, a 5/10 monolayer (ML) HO-modified catalyst (5/10 ML) exhibits superior HER activity and durability owing to the relatively stronger MSI. This catalyst also ensured sustainable and efficient hydrogen generation in a urea electrolyzer with significant energy savings. Our work provides a valuable reference for optimizing the MSI-activity relationship in M1-M2 catalysts that target more than just HER.

Iterative Approach of Experiment-Machine Learning for Efficient Optimization of Environmental Catalysts: An Example of NOx Selective Reduction Catalysts

An iterative approach between machine learning (ML) and laboratory experiments was developed to accelerate the design and synthesis of environmental catalysts (ECs) using selective catalytic reduction (SCR) of nitrogen oxides (NOx) as an example. The main steps in the approach include training a ML model using the relevant data collected from the literature, screening candidate catalysts from the trained model, experimentally synthesizing and characterizing the candidates, updating the ML model by incorporating the new experimental results, and screening promising catalysts again with the updated model. This process is iterated with a goal to obtain an optimized catalyst. Using the iterative approach in this study, a novel SCR NOx catalyst with low cost, high activity, and a wide range of application temperatures was found and successfully synthesized after four iterations. The approach is general enough that it can be readily extended for screening and optimizing the design of other environmental catalysts and has strong implications for the discovery of other environmental materials.

Room temperature solid-state deformation induced high-density lithium grain boundaries to enhance the cycling stability of lithium metal batteries

Due to its high theoretical capacity and low anode potential advantages, lithium is becoming the ideal high-capacity anode of next generation batteries. Nevertheless, the satisfactory long-term cyclability of lithium metal batteries is still not achieved. Inspired by the intrinsic soft nature of the lithium metal, we have developed a simple room temperature solid-state deformation route to overcome the lithium dendrite issue, and the cycle life of the deformation treated lithium anode is 5 times that of the untreated lithium anode. It is demonstrated that microscale lithium grains are divided into nanoscale lithium grains by directional friction forces of solid-state deformation. The lithium grain boundaries are lithiophilic active sites towards Li ions, which regulate homogeneous deposition of Li ions to form a thin and stable SEI film, eventually overcoming the lithium dendrite issue and enhancing the cyclability of lithium batteries. Overcoming the challenges in conventional tedious chemical routes to grow high-density grain boundary active sites for catalysis, the room temperature solid-state deformation route will pave a new road to grow high-density grain boundaries for fuel cells and metal-based batteries.

The landscape of computational approaches for artificial photosynthesis

Artificial photosynthesis is an attractive strategy for converting solar energy into fuels, largely because the Earth receives enough solar energy in one hour to meet humanity's energy needs for an entire year. However, developing devices for artificial photosynthesis remains difficult and requires computational approaches to guide and assist the interpretation of experiments. In this Perspective, we discuss current and future computational approaches, as well as the challenges of designing and characterizing molecular assemblies that absorb solar light, transfer electrons between interfaces, and catalyze water-splitting and fuel-forming reactions.

Machine learning and density functional theory simulation of the electronic structural properties for novel quaternary semiconductors

In order to accelerate the application of quaternary optoelectronic materials in the field of luminescence, it is crucial to develop new quaternary semiconductor materials with excellent properties. However, faced with vast alternative quaternary semiconductors, traditional trial-and-error methods tend to be laborious and inefficient. Here, we combined machine learning (ML) with density functional theory (DFT) calculation to predict the bandgaps of 2180 quaternary semiconductors, most of which were undeveloped but environmentally friendly. The evaluation coefficient (R²) of the model using a random forest algorithm was up to 0.93 in ML. Four novel quaternary semiconductors with direct bandgaps: Ag₂InGaS₄, AgZn₂InS₄, Ag₂ZnSnS₄, and AgZn₂GaS₄, were selected from the ML model. Then their electronic structures and optical properties were further verified and studied by DFT calculations, which demonstrated that the four quaternary semiconductors had direct bandgaps, a small effective mass, and a large exciton binding energy and Stokes shift. Our calculation could significantly speed up the discovery of novel optoelectronic semiconductors and has a certain reference value for the study of luminescent materials and devices.

Machine Learning: A New Paradigm in Computational Electrocatalysis

Designing and screening novel electrocatalysts, understanding electrocatalytic mechanisms at an atomic level, and uncovering scientific insights lie at the center of the development of electrocatalysis. Despite certain success in experiments and computations, it is still difficult to achieve the above objectives due to the complexity of electrocatalytic systems and the vastness of the chemical space for candidate electrocatalysts. With the advantage of machine learning (ML) and increasing interest in electrocatalysis for energy conversion and storage, data-driven scientific research motivated by artificial intelligence (AI) has provided new opportunities to discover promising electrocatalysts, investigate dynamic reaction processes, and extract knowledge from huge data. In this Perspective, we summarize the recent applications of ML in electrocatalysis, including the screening of electrocatalysts and simulation of electrocatalytic processes. Furthermore, interpretable machine learning methods for electrocatalysis are discussed to accelerate knowledge generation. Finally, the blueprint of machine learning is envisaged for future development of electrocatalysis.

Computational investigation of van der Waals corrections in the adsorption properties of molecules on the Cu(111) surface

Here, we report a computational investigation on the role of the most common van der Waals (vdW) corrections (D2, D3, D3(BJ), TS, TS+SCS, TS+HI, and dDsC) employed in density functional theory (DFT) calculations within local and semilocal exchange-correlation functionals to improve the description of the interaction between molecular species and solid surfaces. For this, we selected several molecular model systems, namely, the adsorption of small molecules (CH₃, CH₄, CO, CO₂, H₂O, and OH) on the close-packed Cu(111) surface, which bind via chemisorption or physisorption mechanisms. As expected, we found that the addition of the vdW corrections enhances the energetic stability of the Cu bulk in the face-centered cubic structure, which contributes to increasing the magnitude of the mechanical properties (elastic constants, bulk, Young, and shear modulus). Except for the TS+SCS correction, all vdW corrections substantially increase the surface energy, while the work function changes by about 0.05 eV (largest change). However, we found substantial differences among the vdW corrections when comparing its effects on interlayer spacing relaxations. Based on bulk and surface results, we selected only the D3 and dDsC vdW corrections for the study of the adsorption properties of the selected molecules on the Cu(111) surface. Overall, the addition of these vdW corrections has a greater effect on weakly interacting systems (CH₄, CO₂, H₂O), while the chemisorption systems (CH₃, CO, OH) are less affected.

Applications of Machine Learning in Alloy Catalysts: Rational Selection and Future Development of Descriptors

At present, alloys have broad application prospects in heterogeneous catalysis, due to their various catalytic active sites produced by their vast element combinations and complex geometric structures. However, it is the diverse variables of alloys that lead to the difficulty in understanding the structure-property relationship for conventional experimental and theoretical methods. Fortunately, machine learning methods are helpful to address the issue. Machine learning can not only deal with a large number of data rapidly, but also help establish the physical picture of reactions in multidimensional heterogeneous catalysis. The key challenge in machine learning is the exploration of suitable general descriptors to accurately describe various types of alloy catalysts, which help reasonably design catalysts and efficiently screen candidates. In this review, several kinds of machine learning methods commonly used in the design of alloy catalysts is introduced, and the applications of various reactivity descriptors corresponding to different alloy systems is summarized. Importantly, this work clarifies the existing understanding of physical picture of heterogeneous catalysis, and emphasize the significance of rational selection of universal descriptors. Finally, the development of heterogeneous catalytic descriptors for machine learning are presented.

Propane to olefins tandem catalysis: a selective route towards light olefins production

On-purpose synthetic routes for propylene production have emerged in the last couple of decades in response to the increasing demand for plastics and a shift to shale gas feedstocks for ethylene production. Propane dehydrogenation (PDH), an efficient and selective route to produce propylene, saw booming investments to fill the so-called propylene gap. In the coming years, however, a fluctuating light olefins market will call for flexibility in end-product of PDH plants. This can be achieved by combining PDH with propylene metathesis in a single step, propane to olefins (PTO), which allows production of mixtures of propylene, ethylene and butenes, which are important chemical building blocks for a.o. thermoplastics. The metathesis technology introduced by Phillips in the 1960s and mostly operated in reverse to produce propylene, is thus undergoing a renaissance of scientific and technological interest in the context of the PTO reaction. In this review, we will describe the state-of-the-art of PDH, propylene metathesis and PTO reactions, highlighting the open challenges and opportunities in the field. While the separate PDH and metathesis reactions have been extensively studied in the literature, understanding the whole PTO tandem-catalysis system will require new efforts in theoretical modelling and operando spectroscopy experiments, to gain mechanistic insights into the combined reactions and finally improve catalytic selectivity and stability for on-purpose olefins production.

Accelerated mapping of electronic density of states patterns of metallic nanoparticles via machine-learning

Within first-principles density functional theory (DFT) frameworks, it is challenging to predict the electronic structures of nanoparticles (NPs) accurately but fast. Herein, a machine-learning architecture is proposed to rapidly but reasonably predict electronic density of states (DOS) patterns of metallic NPs via a combination of principal component analysis (PCA) and the crystal graph convolutional neural network (CGCNN). With the PCA, a mathematically high-dimensional DOS image can be converted to a low-dimensional vector. The CGCNN plays a key role in reflecting the effects of local atomic structures on the DOS patterns of NPs with only a few of material features that are easily extracted from a periodic table. The PCA-CGCNN model is applicable for all pure and bimetallic NPs, in which a handful DOS training sets that are easily obtained with the typical DFT method are considered. The PCA-CGCNN model predicts the R2 value to be 0.85 or higher for Au pure NPs and 0.77 or higher for Au@Pt core@shell bimetallic NPs, respectively, in which the values are for the test sets. Although the PCA-CGCNN method showed a small loss of accuracy when compared with DFT calculations, the prediction time takes just ~ 160 s irrespective of the NP size in contrast to DFT method, for example, 13,000 times faster than the DFT method for Pt147. Our approach not only can be immediately applied to predict electronic structures of actual nanometer scaled NPs to be experimentally synthesized, but also be used to explore correlations between atomic structures and other spectrum image data of the materials (e.g., X-ray diffraction, X-ray photoelectron spectroscopy, and Raman spectroscopy).

An Overview of Glycerol Electrooxidation Mechanisms on Pt, Pd and Au

In the most recent decade, glycerol electrooxidation (GEOR) has attracted extensive research interest for valorization of glycerol: the conversion of glycerol to value-added products. These reactions at platinum, palladium, and gold electrodes have a lot of uncertainty in their reaction mechanisms, which has generated some controversies. This review gathers many reported experimental results, observations and proposed reaction mechanisms in order to draw a full picture of GEOR. A particular focus is the clarification of two propositions: Pd is inferior to Pt in cleaving the C-C bonds of glycerol during the electrooxidation and the massive production of CO₂ at high overpotentials is due to the oxidation of the already-oxidized carboxylate products. It is concluded that the inferior C-C bond cleavability with Pd electrodes, as compared with Pt electrodes, is due to the inefficiency of deprotonation, and the massive generation of CO₂ as well as other C1/C2 side products is partially caused by the consumption of OH^- at the anodes, as a lower pH reduces the amount of carboxylates and favors the C-C bond scission. A reaction mechanism is proposed in this review, in which the generation of side products are directly from glycerol ("competition" between each side product) rather than from the further oxidation of C2/C3 products. Additionally, GEOR results and associated interpretations for Ni electrodes are presented, as well as a brief review on the performances of multi-metallic electrocatalysts (most of which are nanocatalysts) as an introduction to these future research hotpots.

Rationalization of promoted reverse water gas shift reaction by Pt3Ni alloy: Essential contribution from ensemble effect

Bimetallic alloys have attracted considerable attention due to the tunable catalytic activity and selectivity that can be different from those of pure metals. Here, we study the superior catalytic behaviors of the Pt₃Ni nanowire (NW) over each individual, Pt and Ni NWs during the reverse Water Gas Shift (rWGS) reaction, using density functional theory. The results show that the promoted rWGS activity by Pt₃Ni strongly depends on the ensemble effect (a particular arrangement of active sites introduced by alloying), while the contributions from ligand and strain effects, which are of great importance in electrocatalysis, are rather subtle. As a result, a unique Ni-Pt hybrid ensemble is observed at the 110/111 edge of the Pt₃Ni NW, where the synergy between Ni and Pt sites is active enough to stabilize carbon dioxide on the surface readily for the rWGS reaction but moderate enough to allow for the facile removal of carbon monoxide and hydrogenation of hydroxyl species. Our study highlights the importance of the ensemble effect in heterogeneous catalysis of metal alloys, enabling selective binding-tuning and promotion of catalytic activity.

Machine learned features from density of states for accurate adsorption energy prediction

Materials databases generated by high-throughput computational screening, typically using density functional theory (DFT), have become valuable resources for discovering new heterogeneous catalysts, though the computational cost associated with generating them presents a crucial roadblock. Hence there is a significant demand for developing descriptors or features, in lieu of DFT, to accurately predict catalytic properties, such as adsorption energies. Here, we demonstrate an approach to predict energies using a convolutional neural network-based machine learning model to automatically obtain key features from the electronic density of states (DOS). The model, DOSnet, is evaluated for a diverse set of adsorbates and surfaces, yielding a mean absolute error on the order of 0.1 eV. In addition, DOSnet can provide physically meaningful predictions and insights by predicting responses to external perturbations to the electronic structure without additional DFT calculations, paving the way for the accelerated discovery of materials and catalysts by exploration of the electronic space.

Computational catalysis would strongly benefit from general descriptors applicable for predicting adsorption energetics. Here the authors propose a machine-learning approach for adsorption energy predictions based on learning the relevant descriptors in a surface atom's density of states as part of the training.

Accelerating the Design of Automotive Catalyst Products Using Machine Learning

The design of catalyst products to reduce harmful emissions is currently an intensive process of expert-driven discovery, taking several years to develop a product. Machine learning can accelerate this timescale, leveraging historic experimental data from related products to guide which new formulations and experiments will enable a project to most directly reach its targets. We used machine learning to accurately model 16 key performance targets for catalyst products, enabling detailed understanding of the factors governing catalyst performance and realistic suggestions of future experiments to rapidly develop more effective products. The proposed formulations are currently undergoing experimental validation.

Insights into the Pt (111) Surface Aid in Predicting the Selective Hydrogenation Catalyst

The d-band center position of the metal catalyst is one of the most important factors for catalytic selective hydrogenation, e.g., the conversion of nitrostyrene to aminostyrene. In this work, we modulate the d-band center position of the Pt surface via H coverage manipulation in order to assess the highly efficient selective hydrogenation catalyst using density functional theory (DFT) calculation, which is validated experimentally. The optimal transition metal catalysts are first screened by comparing the adsorption energy values of two ideal models, nitrobenzene and styrene, and by correlating the adsorption energy with the d-band center positions. Among the ten transition metals, Pt nanoparticles have a good balance between selectivity and the conversion rate. Then, the surface hydrogen covering strategy is applied to modulate the d-band center position on the Pt (111) surface, with the increase of H coverage leading to a decline of the d-band center position, which can selectively enhance the adsorption of nitro groups. However, excessively high H coverage (e.g., 75% or 100%) with an insufficiently low d-band center position can switch the chemisorption of nitro groups to physisorption, significantly reducing the catalytic activity. Therefore, a moderate d-band center shift (ca. −2.14 eV) resulted in both high selectivity and catalytic conversion. In addition, the PtSn experimental results met the theoretical expectations. This work provides a new strategy for the design of highly efficient metal catalysts for selective hydrogenation via the modulation of the d-band center position.

Progress in Computational and Machine-Learning Methods for Heterogeneous Small-Molecule Activation

The chemical conversion of small molecules such as H₂ , H₂ O, O₂ , N₂ , CO₂ , and CH₄ to energy and chemicals is critical for a sustainable energy future. However, the high chemical stability of these molecules poses grand challenges to the practical implementation of these processes. In this regard, computational approaches such as density functional theory, microkinetic modeling, data science, and machine learning have guided the rational design of catalysts by elucidating mechanistic insights, identifying active sites, and predicting catalytic activity. Here, the theory and methodologies for heterogeneous catalysis and their applications for small-molecule activation are reviewed. An overview of fundamental theory and key computational methods for designing catalysts, including the emerging data science techniques in particular, is given. Applications of these methods for finding efficient heterogeneous catalysts for the activation of the aforementioned small molecules are then surveyed. Finally, promising directions of the computational catalysis field for further outlooks are discussed, focusing on the challenges and opportunities for new methods.

Biomass hydrodeoxygenation catalysts innovation from atomistic activity predictors

Circular economy emphasizes the idea of transforming products involving economic growth and improving the ecological system to reduce the negative consequences caused by the excessive use of raw materials. This can be achieved with the use of second-generation biomass that converts industrial and agricultural wastes into bulk chemicals. The use of catalytic processes is essential to achieve a viable upgrade of biofuels from the lignocellulosic biomass. We carried out density functional theory calculations to explore the relationship between 13 transition metals (TMs) properties, as catalysts, and their affinity for hydrogen and oxygen, as key species in the valourization of biomass. The relation of these parameters will define the trends of the hydrodeoxygenation (HDO) process on biomass-derived compounds. We found the hydrogen and oxygen adsorption energies in the most stable site have a linear relation with electronic properties of these metals that will rationalize the surface's ability to bind the biomass-derived compounds and break the C-O bonds. This will accelerate the catalyst innovation for low temperature and efficient HDO processes on biomass derivates, e.g. guaiacol and anisole, among others. Among the monometallic catalysts explored, the scaling relationship pointed out that Ni has a promising balance between hydrogen and oxygen affinities according to the d-band centre and d-band width models. The comparison of the calculated descriptors to the adsorption strength of guaiacol on the investigated surfaces indicates that the d-band properties alone are not best suited to describe the trend. Instead, we found that a linear combination of work function and d-band properties gives significantly better correlation. This article is part of a discussion meeting issue 'Science to enable the circular economy'.

High-throughput experimentation meets artificial intelligence: a new pathway to catalyst discovery

High throughput experimentation in heterogeneous catalysis provides an efficient solution to the generation of large datasets under reproducible conditions. Knowledge extraction from these datasets has mostly been performed using statistical methods, targeting the optimization of catalyst formulations. The combination of advanced machine learning methodologies with high-throughput experimentation has enormous potential to accelerate the predictive discovery of novel catalyst formulations that do not exist with current statistical design of experiments. This perspective describes selective examples ranging from statistical design of experiments for catalyst synthesis to genetic algorithms applied to catalyst optimization, and finally random forest machine learning using experimental data for the discovery of novel catalysts. Lastly, this perspective also provides an outlook on advanced machine learning methodologies as applied to experimental data for materials discovery.

Practical Deep-Learning Representation for Fast Heterogeneous Catalyst Screening

The binding site and energy is an invaluable descriptor in high-throughput screening of catalysts, as it is accessible and correlates with the activity and selectivity. Recently, comprehensive binding energy prediction machine-learning models have been demonstrated and promise to accelerate the catalyst screening. Here, we present a simple and versatile representation, applicable to any deep-learning models, to further accelerate such process. Our approach involves labeling the binding site atoms of the unrelaxed bare surface geometry; hence, for the model application, density functional theory calculations can be completely removed if the optimized bulk structure is available as is the case when using the Materials Project database. In addition, we present ensemble learning, where a set of predictions is used together to form a predictive distribution that reduces the model bias. We apply the labeled site approach and ensemble to crystal graph convolutional neural network and the ∼40 000 data set of alloy catalysts for CO₂ reduction. The proposed model applied to the data set of unrelaxed structures shows 0.116 and 0.085 eV mean absolute error, respectively, for CO and H binding energy, better than the best method (0.13 and 0.13 eV) in the literature that requires costly geometry relaxations. The analysis of the model parameters demonstrates that the model can effectively learn the chemical information related to the binding site.

Understanding the ML black box with simple descriptors to predict cluster–adsorbate interaction energy

Density functional theory (DFT) is currently one of the most accurate and yet practical theories used to gain insight into the properties of materials.

Predicting aggregation energy for single atom bimetallic catalysts on clean and O* adsorbed surfaces through machine learning models

Machine learning models are successfully developed for simultaneous prediction of stability and adsorption energy at single-atom bimetallic sites.

Pattern Learning Electronic Density of States

Electronic density of states (DOS) is a key factor in condensed matter physics and material science that determines the properties of metals. First-principles density-functional theory (DFT) calculations have typically been used to obtain the DOS despite the considerable computation cost. Herein, we report a fast machine learning method for predicting the DOS patterns of not only bulk structures but also surface structures in multi-component alloy systems by a principal component analysis. Within this framework, we use only four features to define the composition, atomic structure, and surfaces of alloys, which are the d-orbital occupation ratio, coordination number, mixing factor, and the inverse of miller indices. While the DFT method scales as O(N3) in which N is the number of electrons in the system size, our pattern learning method can be independent on the number of electrons. Furthermore, our method provides a pattern similarity of 91 ~ 98% compared to DFT calculations. This reveals that our learning method will be an alternative that can break the trade-off relationship between accuracy and speed that is well known in the field of electronic structure calculations.