Computational Catalyst Design Research Articles

Computational Catalyst Design for Dry Reforming of Methane: A Review

Because of the recent global warming environmental issues, dry reforming of methane (DRM)─which converts greenhouse gases (CO2 and CH4) into syngases (CO and H2)─is receiving significant attention. Recently, density functional theory (DFT) calculations have been effectively used to obtain fundamental information on DRM reactions. The DFT calculations can provide valuable theoretical knowledge in various heterogeneous catalyst systems, which is difficult to derive from experiments alone due to the complexity of reaction pathways. This work introduces theoretical studies concerning the most plausible reaction pathways, catalytic activities, and stabilities of Ni- and non-Ni-based metal catalysts. The review includes fundamental analyses of reaction mechanisms, catalytic activities, and several strategies to improve the catalytic properties of Ni- and non-Ni-based catalysts. Such strategies include doping, introducing promoters, forming bimetallics, and utilizing various catalyst supports. In addition, DFT-based descriptors provide guidelines for DRM catalyst design in terms of activity and stability. In conclusion, this review also suggests DFT-based analyses of catalysts based on morphological and compositional engineering, such as core–shell nanoparticles, single-atom catalysts, phosphides, and reduced solid solution catalysts. Finally, this review suggests a rational catalyst design for DRM by tuning catalytic properties based on engineered catalyst characteristics under a comprehensive understanding provided by DFT calculations and experiments.

Density functional theory and CCSD(T) evaluation of ionization potentials, redox potentials, and bond energies related to zirconocene polymerization catalysts.

Quantum-mechanical-based computational design of molecular catalysts requires accurate and fast electronic structure calculations to determine and predict properties of transition-metal complexes. For Zr-based molecular complexes related to polyethylene catalysis, previous evaluation of density functional theory (DFT) and wavefunction methods only examined oxides and halides or select reaction barrier heights. In this work, we evaluate the performance of DFT against experimental redox potentials and bond dissociation enthalpies (BDEs) for zirconocene complexes directly relevant to ethylene polymerization catalysis. We also examined the ability of DFT to compute the fourth atomic ionization potential of zirconium and the effect the basis set selection has on the ionization potential computed with CCSD(T). Generally, the atomic ionization potential and redox potentials are very well reproduced by DFT, but we discovered relatively large deviations of DFT-calculated BDEs compared to experiment. However, evaluation of BDEs with CCSD(T) suggests that experimental values should be revisited, and our CCSD(T) values should be taken as most accurate.

Experimental and Computational Synergistic Design of Cu and Fe Catalysts for the Reverse Water–Gas Shift: A Review

Experimental and Computational Synergistic Design of Cu and Fe Catalysts for the Reverse Water–Gas Shift: A Review

Computational Evaluation and Design of Polyethylene Zirconocene Catalysts with Noncovalent Dispersion Interactions

Computational Evaluation and Design of Polyethylene Zirconocene Catalysts with Noncovalent Dispersion Interactions

Three Grand Challenges for the Computational Design of Heterogeneous Catalysts

Three Grand Challenges for the Computational Design of Heterogeneous Catalysts

Rational design of heterogeneous catalysts by breaking and rebuilding scaling relations

Various scaling relations have long been established in the field of heterogeneous catalysis, but the resultant volcano curves inherently limit the catalytic performance of catalyst candidates. On the other hand, it is still very challenging to develop universal descriptors that can be used in various types of catalysts and reaction systems. For these reasons, several strategies have recently been proposed to break and rebuild scaling relations to go beyond the top of volcanoes. In this review, some previously proposed descriptors have been briefly introduced. Then, the strategies for breaking known and establishing new and more generalized scaling relations in complex catalytic systems have been summarized. Finally, the application of machine-learning techniques in identifying universal descriptors for future computational design and high-throughput screening of heterogeneous catalysts has been discussed.

Computational Design of Pd Nanoclusters and Pd Single-Atom Catalysts Supported on O-Functionalized Graphene

A crucial step in the preparation of supported metal catalysts is related to the choice of support, since it regulates the anchoring of the metal species in an environment of strong interactions. In this sense, the binding on the support, which may present defects, as well as induced modifications of the electronic structure, helps to determine efficient metal/support combinations for catalysis applications. In this work, first-principles studies have been carried out to model and describe the geometric and electronic properties of O-functionalized graphene as a model carbon support, in addition to a single Pd atom and Pd13 nanocluster (experimentally observed on various carbon supports) as supported metal catalysts. The carbon-based nanomaterial includes experimentally probed abundant oxygen functional groups and point defects, demonstrating high thermal stability at room temperature. The characterization confirms the structural motifs presented in the model, distinguishing the organic functionalities and various defects using spectroscopy techniques. Interestingly, the oxygenated carbon-based support presents a metallic character when the formal charges in the embedded Pd atom and nanocluster have been determined to be [Pd]+ and [Pd13]3+. The hydrogenation process of the adsorbed cluster is also presented, with the determination of a ratio between the adsorbed hydrides and the surface metal atoms close to unity at room temperature, thanks to ab initio molecular dynamics.

Deep reinforcement learning for predicting kinetic pathways to surface reconstruction in a ternary alloy

The majority of computational catalyst design focuses on the screening of material components and alloy composition to optimize selectivity and activity for a given reaction. However, predicting the metastability of the alloy catalyst surface at realistic operating conditions requires an extensive sampling of possible surface reconstructions and their associated kinetic pathways. We present CatGym, a deep reinforcement learning (DRL) environment for predicting the thermal surface reconstruction pathways and their associated kinetic barriers in crystalline solids under reaction conditions. The DRL agent iteratively changes the positions of atoms in the near-surface region to generate kinetic pathways to accessible local minima involving changes in the surface compositions. We showcase our agent by predicting the surface reconstruction pathways of a ternary Ni3Pd3Au2(111) alloy catalyst. Our results show that the DRL agent can not only explore more diverse surface compositions than the conventional minima hopping method, but also generate the kinetic surface reconstruction pathways. We further demonstrate that the kinetic pathway to a global minimum energy surface composition and its associated transition state predicted by our agent is in good agreement with the minimum energy path predicted by nudged elastic band calculations.

Toward E-selective Olefin Metathesis: Computational Design and Experimental Realization of Ruthenium Thio-Indolate Catalysts

The selective transformation of 1-alkenes into E-olefins is a long-standing challenge in olefin metathesis. Density functional theory (DFT) calculations predict high E-selectivity for catalysts incorporating a bidentate, dianionic thio-indolate ligand within a RuXX’(NHC)(py)(= CHR) platform (NHC = N-heterocyclic carbene; py = pyridine). Such complexes are predicted to yield E-olefins by favoring anti-disposed substituents in the transition state expected to be rate-determining: specifically, that for cycloreversion of the metallacyclobutane intermediate. Three pyridine-stabilized catalysts Ru21a-c were synthesized, in which the thio-indolate ligand bears a H, Me, or Ph substituent at the C2 position, and the NHC ligand is the unsaturated imidazoline-2-ylidene Me2IMes (which bears N-mesityl groups and methyl groups on the C4,5 backbone). Single-crystal X-ray diffraction analysis of Ru21c confirms the ligand orientation required for E-selective metathesis, with the thio-indolate sulfur atom binding cis to the NHC, and the indolate nitrogen atom trans to the NHC. However, whereas the new complexes mediated metathetic exchange of their 2-thienylmethylidene ligand in the presence of the common metathesis substrates styrene and allylbenzene, no corresponding self-metathesis products were obtained. Only small amounts of 2-butene (73% (Z)-2-butene) were obtained in self-metathesis of propene using Ru21a. Detailed DFT analysis of this process revealed that product release is surprisingly slow, limiting the reaction rate and explaining the low metathesis activity. With the barrier to dissociation of (Z)-2-butene being lower than that of (E)-2-butene, the calculations also account for the observed Z-selectivity of Ru21a. These findings provide guidelines for catalyst redesign in pursuit of the ambitious goal of E-selective 1-alkene metathesis.Graphic abstract

Evaluating the benefits of kinetic Monte Carlo and microkinetic modeling for catalyst design studies in the presence of lateral interactions

Popular computational catalyst design strategies rely on the identification of reactivity descriptors, which can be used along with Brønsted−Evans−Polanyi (BEP) and scaling relations as input to a microkinetic model (MKM) to make predictions for activity or selectivity trends. The main benefit of this approach is related to the inherent dimensionality reduction of the large material space to just a few catalyst descriptors. Conversely, it is well documented that a small set of descriptors is insufficient to capture the intricacies and complexities of a real catalytic system. The inclusion of coverage effects through lateral adsorbate-adsorbate interactions can narrow the gap between simplified descriptor predictions and real systems, but mean-field MKMs cannot properly account for local coverage effects. This shortcoming of the mean-field approximation can be rectified by switching to a lattice-based kinetic Monte Carlo (kMC) method using cluster expansion representation of adsorbate−adsorbate lateral interactions.Using the prototypical CO oxidation reaction as an example, we critically evaluate the benefits of kMC over MKM in terms of trend predictions and computational cost when using only a small set of input parameters. After confirming that in the absence of lateral interactions the kMC and MKM approaches yield identical trends and mechanistic information, we observed substantial differences between the two kinetic models when lateral interactions were introduced. The mean-field implementation applies coverage corrections directly to the descriptors, causing an artificial overprediction of the activity of strongly binding metals. In contrast, the cluster expansion in kMC implementation can differentiate among the highly active metals but it is very sensitive to the set of included interaction parameters. Considering that computational screening relies on a minimal set of descriptors, for which MKM makes reasonable trend predictions at a ca. three orders of magnitude lower computational cost than kMC, the MKM approach does provide a better entry point for computational catalyst design.

Perspective on computational reaction prediction using machine learning methods in heterogeneous catalysis.

Heterogeneous catalysis plays a significant role in the modern chemical industry. Towards the rational design of novel catalysts, understanding reactions over surfaces is the most essential aspect. Typical industrial catalytic processes such as syngas conversion and methane utilisation can generate a large reaction network comprising thousands of intermediates and reaction pairs. This complexity not only arises from the permutation of transformations between species but also from the extra reaction channels offered by distinct surface sites. Despite the success in investigating surface reactions at the atomic scale, the huge computational expense of ab initio methods hinders the exploration of such complicated reaction networks. With the proliferation of catalysis studies, machine learning as an emerging tool can take advantage of the accumulated reaction data to emulate the output of ab initio methods towards swift reaction prediction. Here, we briefly summarise the conventional workflow of reaction prediction, including reaction network generation, ab initio thermodynamics and microkinetic modelling. An overview of the frequently used regression models in machine learning is presented. As a promising alternative to full ab initio calculations, machine learning interatomic potentials are highlighted. Furthermore, we survey applications assisted by these methods for accelerating reaction prediction, exploring reaction networks, and computational catalyst design. Finally, we envisage future directions in computationally investigating reactions and implementing machine learning algorithms in heterogeneous catalysis.

Trends in computational molecular catalyst design.

Computational methods have emerged as a powerful tool to augment traditional experimental molecular catalyst design by providing useful predictions of catalyst performance and decreasing the time needed for catalyst screening. In this perspective, we discuss three approaches for computational molecular catalyst design: (i) the reaction mechanism-based approach that calculates all relevant elementary steps, finds the rate and selectivity determining steps, and ultimately makes predictions on catalyst performance based on kinetic analysis, (ii) the descriptor-based approach where physical/chemical considerations are used to find molecular properties as predictors of catalyst performance, and (iii) the data-driven approach where statistical analysis as well as machine learning (ML) methods are used to obtain relationships between available data/features and catalyst performance. Following an introduction to these approaches, we cover their strengths and weaknesses and highlight some recent key applications. Furthermore, we present an outlook on how the currently applied approaches may evolve in the near future by addressing how recent developments in building automated computational workflows and implementing advanced ML models hold promise for reducing human workload, eliminating human bias, and speeding up computational catalyst design at the same time. Finally, we provide our viewpoint on how some of the challenges associated with the up-and-coming approaches driven by automation and ML may be resolved.

Decoding the intricate network of molecular interactions of a hyperstable engineered biocatalyst.

Computational design of protein catalysts with enhanced stabilities for use in research and enzyme technologies is a challenging task. Using force-field calculations and phylogenetic analysis, we previously designed the haloalkane dehalogenase DhaA115 which contains 11 mutations that confer upon it outstanding thermostability (Tm = 73.5 °C; ΔTm > 23 °C). An understanding of the structural basis of this hyperstabilization is required in order to develop computer algorithms and predictive tools. Here, we report X-ray structures of DhaA115 at 1.55 Å and 1.6 Å resolutions and their molecular dynamics trajectories, which unravel the intricate network of interactions that reinforce the αβα-sandwich architecture. Unexpectedly, mutations toward bulky aromatic amino acids at the protein surface triggered long-distance (∼27 Å) backbone changes due to cooperative effects. These cooperative interactions produced an unprecedented double-lock system that: (i) induced backbone changes, (ii) closed the molecular gates to the active site, (iii) reduced the volumes of the main and slot access tunnels, and (iv) occluded the active site. Despite these spatial restrictions, experimental tracing of the access tunnels using krypton derivative crystals demonstrates that transport of ligands is still effective. Our findings highlight key thermostabilization effects and provide a structural basis for designing new thermostable protein catalysts.

Towards High-Performancepgm-Free ORR Electrocatalysts: Interplay between Theory and Experiment

Fuel cells are clean and sustainable energy conversion devices for power generation, and platinum and other platinum-group-metal (PGM) have been long known as the most effective catalystsfor the oxygen reduction reaction (ORR) in fuel cells. However, the scarcity and high costof the PGM-based ORR catalysts are main obstacles forcommercialization of the fuel cell technology. Therefore, searching high-performance PGM-free ORR catalysts has been a key scientific issue in the development of fuel cells. In this talk, we will present our recent progress on computational design of ORR catalysts based on two-dimensional nanomaterials, especially single-atom catalysts and metal-free catalysts. Notably our designed nanocatalysts can be experimentally realized by standard procedures, as evidenced by our recent collaborations with experimental peers. We are calling for more interplays between theory and experiment, which will greatly facilitate the catalyst discovery and innovation.

Search for Catalysts by Inverse Design: Artificial Intelligence, Mountain Climbers, and Alchemists.

In silico catalyst design is a grand challenge of chemistry. Traditional computational approaches have been limited by the need to compute properties for an intractably large number of possible catalysts. Recently, inverse design methods have emerged, starting from a desired property and optimizing a corresponding chemical structure. Techniques used for exploring chemical space include gradient-based optimization, alchemical transformations, and machine learning. Though the application of these methods to catalysis is in its early stages, further development will allow for robust computational catalyst design. This review provides an overview of the evolution of inverse design approaches and their relevance to catalysis. The strengths and limitations of existing techniques are highlighted, and suggestions for future research are provided.

Toward Computational Design of Catalysts for CO2 Selective Reduction via Reaction Phase Diagram Analysis

AbstractChemical conversion of carbon dioxide (CO2) into value‐added chemicals, for example, ethanol, has been recognized to be one of the most effective ways of carbon sources cycling. However, the selectivity controlling of CO2 reduction is still a challenging problem due to the formation of a variety of products, such as methanol, methane, formic acid, and carbon monoxide, associating with complicated reaction mechanisms. Herein, a simple and effective theoretical analysis of “reaction phases diagram” to understand the selectivity of CO2 reduction, based on the principle of reaction free energies with sequential priority is proposed. Using this principle, the fundamental understanding of the formation of different products in CO2 reduction is rationalized. On the basis of reaction free energies’ analysis, the selectivity trends of CO2 reduction to methane, methanol, and ethanol from a number of experiments are discussed. At the end, the selectivity trends with a computer algorithm of searching full reaction pathways, explaining well the effects from catalyst sizes, substrate observed in experiments, are confirmed.

Computational Design of Single-Molybdenum Catalysts for the Nitrogen Reduction Reaction

Starting from molybdenum (Mo) embedded in black phosphorus, 17 single-Mo catalysts with various combinations of ligands, including phosphorous (P), boron (B), nitrogen (N), sulfur (S), and carbon (C), have been computationally examined as catalysts for the nitrogen reduction reaction. Among them, Mo–PC2, Mo–PB2, and Mo–BC2 have been identified as the most promising catalysts, offering an overall overpotential less than 0.60 V. Mo–BC2 is particularly attractive as it also shows a high nitrogen reduction reaction selectivity over the hydrogen evolution reaction. Such high performance is originated essentially from the mediation of the ligands, which effectively shift the d-band center of the Mo atom toward the Fermi energy.

Kinetics-Based Computational Catalyst Design Strategy for the Oxygen Evolution Reaction on Transition-Metal Oxide Surfaces

Density functional theory was used to examine the oxygen evolution reaction on a large number of active sites formed by doping three different surfaces of Co3O4 with various 3d transition-metal atoms. By combining the scaling and Bronsted–Evans–Polanyi (BEP) relationships that are determined for these sites, it is shown that the activity of a site is controlled by the redox potential for oxidation of the site. On the basis of this, a kinetics-based design strategy is presented for identifying the optimal active site at a given electrode potential. This design strategy is shown to be valid regardless of whether the rate-limiting water addition step occurs electrochemically or nonelectrochemically, as long as certain conditions are met. Another finding is that the BEP relations are sensitive to the structure of the active site, with sites reacting through μ3-oxo species having the most favorable relations. Finally, the kinetics-based design strategy is compared with the commonly used thermodynamics-based de...

When Is Ligand p Ka a Good Descriptor for Catalyst Energetics? In Search of Optimal CO2 Hydration Catalysts.

We present a detailed study of nearly 70 Zn molecular catalysts for CO2 hydration from four diverse ligand classes ranging from well-studied carbonic anhydrase mimics (e.g., cyclen) to new structures we obtain by leveraging diverse hits from large organic libraries. Using microkinetic analysis and establishing linear free energy relationships, we confirm that turnover is sensitive to the relative thermodynamic stability of reactive hydroxyl and bound bicarbonate moieties. We observe a wide range of thermodynamic stabilities for these intermediates, showing up to 6 kcal/mol improvement over well-studied cyclen catalysts. We observe a good correlation between the p Ka of the Zn-OH2 moiety and the resulting relative stability of hydroxyl moieties over bicarbonate, which may be rationalized by the dominant effect of the difference in higher Zn-OH bond order in comparison to weaker bonding in bicarbonate and water. A direct relationship is identified between isolated organic ligand p Ka and the p Ka of a bound water molecule on the catalyst. Thus, organic ligand p Ka, which is intuitive, easy to compute or tabulate, and much less sensitive to electronic structure method choice than whole-catalyst properties, is a good quantitative descriptor for predicting the effect of through-bond electronic effects on relative CO2 hydration energetics. We expect this to be applicable to other reactions where is it essential to stabilize turnover-determining hydroxyl species with respect to more weakly bound moieties. Finally, we note exceptions for rigid ligands (e.g., porphyrins) that are observed to preferentially stabilize hydroxyl over bicarbonate without reducing p Ka values as substantially. We expect the strategy outlined here, to (i) curate diverse ligands from large organic libraries and (ii) identify when ligand-only properties can determine catalyst energetics, to be broadly useful for both experimental and computational catalyst design.

Computational Transition-State Design Provides Experimentally Verified Cr(P,N) Catalysts for Control of Ethylene Trimerization and Tetramerization

Computational design of molecular homogeneous organometallic catalysts followed by experimental realization remains a significant challenge. Here, we report the development and use of a density functional theory transition-state model that provided quantitative prediction of molecular Cr catalysts for controllable selective ethylene trimerization and tetramerization. This computational model identified a general class of phosphine monocyclic imine (P,N)-ligand Cr catalysts where changes in the ligand structure control 1-hexene versus 1-octene selectivity. Experimental ligand and catalyst synthesis as well as reaction testing quantitatively confirmed predictions.