Materials Genes of CO2 Hydrogenation on Supported Cobalt Catalysts: An Artificial Intelligence Approach Integrating Theoretical and Experimental Data.

Designing materials for catalysis is challenging because the performance is governed by an intricate interplay of various multiscale phenomena, such as the chemical reactions on surfaces and the materials' restructuring during the catalytic process. In the case of supported catalysts, the role of the support material can be also crucial. Here, we address this intricacy challenge by a symbolic-regression artificial intelligence (AI) approach. We identify the key physicochemical parameters correlated with the measured performance, out of many offered candidate parameters characterizing the materials, reaction environment, and possibly relevant underlying phenomena. Importantly, these parameters are obtained by both experiments and ab initio simulations. The identified key parameters might be called "materials genes", in analogy to genes in biology: they correlate with the property or function of interest, but the explicit physical relationship is not (necessarily) known. To demonstrate the approach, we investigate the CO2 hydrogenation catalyzed by cobalt nanoparticles supported on silica. Crucially, the silica support is modified with the additive metals magnesium, calcium, titanium, aluminum, or zirconium, which results in six materials with significantly different performances. These systems mimic hydrothermal vents, which might have produced the first organic molecules on Earth. The key parameters correlated with the CH3OH selectivity reflect the reducibility of cobalt species, the adsorption strength of reaction intermediates, and the chemical nature of the additive metal. By using an AI model trained on basic elemental properties of the additive metals (e.g., ionization potential) as physicochemical parameters, new additives are suggested. The predicted CH3OH selectivity of cobalt catalysts supported on silica modified with vanadium and zinc is confirmed by new experiments.

Data-Centric Heterogeneous Catalysis: Identifying Rules and Materials Genes of Alkane Selective Oxidation.

Artificial intelligence (AI) can accelerate catalyst design by identifying key physicochemical descriptive parameters correlated with the underlying processes triggering, favoring, or hindering the performance. In analogy to genes in biology, these parameters might be called "materials genes" of heterogeneous catalysis. However, widely used AI methods require big data, and only the smallest part of the available data meets the quality requirement for data-efficient AI. Here, we use rigorous experimental procedures, designed to consistently take into account the kinetics of the catalyst active states formation, to measure 55 physicochemical parameters as well as the reactivity of 12 catalysts toward ethane, propane, and n-butane oxidation reactions. These materials are based on vanadium or manganese redox-active elements and present diverse phase compositions, crystallinities, and catalytic behaviors. By applying the sure-independence-screening-and-sparsifying-operator symbolic-regression approach to the consistent data set, we identify nonlinear property-function relationships depending on several key parameters and reflecting the intricate interplay of processes that govern the formation of olefins and oxygenates: local transport, site isolation, surface redox activity, adsorption, and the material dynamical restructuring under reaction conditions. These processes are captured by parameters derived from N2 adsorption, X-ray photoelectron spectroscopy (XPS), and near-ambient-pressure in situ XPS. The data-centric approach indicates the most relevant characterization techniques to be used for catalyst design and provides "rules" on how the catalyst properties may be tuned in order to achieve the desired performance.

Effects of Silica Modification (Mg, Al, Ca, Ti, and Zr) on Supported Cobalt Catalysts for H2-Dependent CO2 Reduction to Metabolic Intermediates.

Serpentinizing hydrothermal systems generate H2 as a reductant and harbor catalysts conducive to geochemical CO2 conversion into reduced carbon compounds that form the core of microbial autotrophic metabolism. This study characterizes mineral catalysts at hydrothermal vents by investigating the interactions between catalytically active cobalt sites and silica-based support materials on H2-dependent CO2 reduction. Heteroatom incorporated (Mg, Al, Ca, Ti, and Zr), ordered mesoporous silicas are applied as model support systems for the cobalt-based catalysts. It is demonstrated that all catalysts surveyed convert CO2 to methane, methanol, carbon monoxide, and low-molecular-weight hydrocarbons at 180 °C and 20 bar, but with different activity and selectivity depending on the support modification. The additional analysis of the condensed product phase reveals the formation of oxygenates such as formate and acetate, which are key intermediates in the ancient acetyl-coenzyme A pathway of carbon metabolism. The Ti-incorporated catalyst yielded the highest concentrations of formate (3.6 mM) and acetate (1.2 mM) in the liquid phase. Chemisorption experiments including H2 temperature-programmed reduction (TPR) and CO2 temperature-programmed desorption (TPD) in agreement with density functional theory (DFT) calculations of the adsorption energy of CO2 suggest metallic cobalt as the preferential adsorption site for CO2 compared to hardly reducible cobalt-metal oxide interface species. The ratios of the respective cobalt species vary depending on the interaction strength with the support materials. The findings reveal robust and biologically relevant catalytic activities of silica-based transition metal minerals in H2-rich CO2 fixation, in line with the idea that autotrophic metabolism emerged at hydrothermal vents.

Hierarchical Symbolic Regression for Identifying Key Physical Parameters Correlated with Bulk Properties of Perovskites.

Symbolic regression identifies nonlinear, analytical expressions relating materials properties and key physical parameters. However, the pool of expressions grows rapidly with complexity, compromising its efficiency. We tackle this challenge hierarchically: identified expressions are used as inputs for further obtaining more complex expressions. Crucially, this framework can transfer knowledge among properties, as demonstrated using the sure-independence-screening-and-sparsifying-operator approach to identify expressions for lattice constant and cohesive energy, which are then used to model the bulk modulus of ABO_{3} perovskites.

Identifying Outstanding Transition-Metal-Alloy Heterogeneous Catalysts for the Oxygen Reduction and Evolution Reactions via Subgroup Discovery.

In order to estimate the reactivity of a large number of potentially complex heterogeneous catalysts while searching for novel and more efficient materials, physical as well as data-centric models have been developed for a faster evaluation of adsorption energies compared to first-principles calculations. However, global models designed to describe as many materials as possible might overlook the very few compounds that have the appropriate adsorption properties to be suitable for a given catalytic process. Here, the subgroup-discovery (SGD) local artificial-intelligence approach is used to identify the key descriptive parameters and constrains on their values, the so-called SG rules, which particularly describe transition-metal surfaces with outstanding adsorption properties for the oxygen-reduction and -evolution reactions. We start from a data set of 95 oxygen adsorption-energy values evaluated by density-functional-theory calculations for several monometallic surfaces along with 16 atomic, bulk and surface properties as candidate descriptive parameters. From this data set, SGD identifies constraints on the most relevant parameters describing materials and adsorption sites that (i) result in O adsorption energies within the Sabatier-optimal range required for the oxygen-reduction reaction and (ii) present the largest deviations from the linear-scaling relations between O and OH adsorption energies, which limit the catalyst performance in the oxygen-evolution reaction. The SG rules not only reflect the local underlying physicochemical phenomena that result in the desired adsorption properties, but also guide the challenging design of alloy catalysts. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11244-021-01502-4.

Learning Design Rules for Selective Oxidation Catalysts from High-Throughput Experimentation and Artificial Intelligence.

The design of heterogeneous catalysts is challenged by the complexity of materials and processes that govern reactivity and by the fact that the number of good catalysts is very small in comparison to the number of possible materials. Here, we show how the subgroup-discovery (SGD) artificial-intelligence approach can be applied to an experimental plus theoretical data set to identify constraints on key physicochemical parameters, the so-called SG rules, which exclusively describe materials and reaction conditions with outstanding catalytic performance. By using high-throughput experimentation, 120 SiO2-supported catalysts containing ruthenium, tungsten, and phosphorus were synthesized and tested in the catalytic oxidation of propylene. As candidate descriptive parameters, the temperature and 10 parameters related to the composition and chemical nature of the catalyst materials, derived from calculated free-atom properties, were offered. The temperature, the phosphorus content, and the composition-weighted electronegativity are identified as key parameters describing high yields toward the value-added oxygenate products acrolein and acrylic acid. The SG rules not only reflect the underlying processes particularly associated with high performance but also guide the design of more complex catalysts containing up to five elements in their composition.

Materials genes of heterogeneous catalysis from clean experiments and artificial intelligence.

ABSTRACT: The performance in heterogeneous catalysis is an example of a complex materials function, governed by an intricate interplay of several processes (e.g., the different surface chemical reactions, and the dynamic restructuring of the catalyst material at reaction conditions). Modeling the full catalytic progression via first-principles statistical mechanics is impractical, if not impossible. Instead, we show here how a tailored artificial-intelligence approach can be applied, even to a small number of materials, to model catalysis and determine the key descriptive parameters ("materials genes") reflecting the processes that trigger, facilitate, or hinder catalyst performance. We start from a consistent experimental set of "clean data," containing nine vanadium-based oxidation catalysts. These materials were synthesized, fully characterized, and tested according to standardized protocols. By applying the symbolic-regression SISSO approach, we identify correlations between the few most relevant materials properties and their reactivity. This approach highlights the underlying physicochemical processes, and accelerates catalyst design. IMPACT STATEMENT: Artificial intelligence (AI) accepts that there are relationships or correlations that cannot be expressed in terms of a closed mathematical form or an easy-to-do numerical simulation. For the function of materials, for example, catalysis, AI may well capture the behavior better than the theory of the past. However, currently the flexibility of AI comes together with a lack of interpretability, and AI can only predict aspects that were included in the training. The approach proposed and demonstrated in this IMPACT article is interpretable. It combines detailed experimental data (called "clean data") and symbolic regression for the identification of the key descriptive parameters (called "materials genes") that are correlated with the materials function. The approach demonstrated here for the catalytic oxidation of propane will accelerate the discovery of improved or novel materials while also enhancing physical understanding. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1557/s43577-021-00165-6.

Viewpoint: Atomic-Scale Design Protocols toward Energy, Electronic, Catalysis, and Sensing Applications.

Nanostructured materials are essential building blocks for the fabrication of new devices for energy harvesting/storage, sensing, catalysis, magnetic, and optoelectronic applications. However, because of the increase of technological needs, it is essential to identify new functional materials and improve the properties of existing ones. The objective of this Viewpoint is to examine the state of the art of atomic-scale simulative and experimental protocols aimed to the design of novel functional nanostructured materials, and to present new perspectives in the relative fields. This is the result of the debates of Symposium I "Atomic-scale design protocols towards energy, electronic, catalysis, and sensing applications", which took place within the 2018 European Materials Research Society fall meeting.

What Can We Learn from First Principles Multi-Scale Models in Catalysis? The Role of the Ni/Al2O3 Interface in Water-Gas Shift and Dry Reforming as a Case Study.

Computational first principles models based on density functional theory (DFT) have emerged as an important tool to address reaction mechanisms and active sites in metal nanoparticle catalysis. However, the common evaluation of potential energy surfaces for selected reaction steps contrasts with the complexity of reaction networks under operating conditions, where the interplay of adsorbate populations and competing routes at reaction conditions determine the most relevant states for catalyst activity and selectivity. Here, we discuss how the use of a multi-scale first principles approach combining DFT calculations at the atomistic level with kinetic models may be used to understand reactions catalyzed by metal nanoparticles. The potential of such an approach is illustrated for the case of Al2O3-supported Ni nanoparticle catalysts in the water-gas shift and dry reforming reactions. In these systems, both Ni nanoparticle (metal) as well as metal/oxide interface sites are available and may play a role in catalysis, which depends not only on the energy for critical reaction steps, as captured by DFT, but also on the reaction temperature and adsorbate populations, as shown by microkinetic modelling and experiments.

Electronic Structure-Reactivity Relationship on Ruthenium Step-Edge Sites from Carbonyl 13C Chemical Shift Analysis.

Ru nanoparticles are highly active catalysts for the Fischer-Tropsch and the Haber-Bosch processes. They show various types of surface sites upon CO adsorption according to NMR spectroscopy. Compared to terminal and bridging η1 adsorption modes on terraces or edges, little is known about side-on η2 CO species coordinated to B5 or B6 step-edges, the proposed active sites for CO and N2 cleavage. By using solid-state NMR and DFT calculations, we analyze 13C chemical shift tensors (CSTs) of carbonyl ligands on the molecular cluster model for Ru nanoparticles, Ru6(η2-µ4-CO)2(CO)13(η6-C6Me6), and show that, contrary to η1 carbonyls, the CST principal components parallel to the C-O bond are extremely deshielded in the η2 species due to the population of the C-O π* antibonding orbital, which weakens the bond prior to dissociation. The carbonyl CST is thus an indicator of the reactivity of both Ru clusters and Ru nanoparticles step-edge sites toward C-O bond cleavage.

Contrasting the Role of Ni/Al2O3 Interfaces in Water-Gas Shift and Dry Reforming of Methane.

Transition metal nanoparticles (NPs) are typically supported on oxides to ensure their stability, which may result in modification of the original NP catalyst reactivity. In a number of cases, this is related to the formation of NP/support interface sites that play a role in catalysis. The metal/support interface effect verified experimentally is commonly ascribed to stronger reactants adsorption or their facile activation on such sites compared to bare NPs, as indicated by DFT-derived potential energy surfaces (PESs). However, the relevance of specific reaction elementary steps to the overall reaction rate depends on the preferred reaction pathways at reaction conditions, which usually cannot be inferred based solely on PES. Hereby, we use a multiscale (DFT/microkinetic) modeling approach and experiments to investigate the reactivity of the Ni/Al2O3 interface toward water-gas shift (WGS) and dry reforming of methane (DRM), two key industrial reactions with common elementary steps and intermediates, but held at significantly different temperatures: 300 vs 650 °C, respectively. Our model shows that despite the more energetically favorable reaction pathways provided by the Ni/Al2O3 interface, such sites may or may not impact the overall reaction rate depending on reaction conditions: the metal/support interface provides the active site for WGS reaction, acting as a reservoir for oxygenated species, while all Ni surface atoms are active for DRM. This is in contrast to what PESs alone indicate. The different active site requirement for WGS and DRM is confirmed by the experimental evaluation of the activity of a series of Al2O3-supported Ni NP catalysts with different NP sizes (2-16 nm) toward both reactions.

Cooperativity and Dynamics Increase the Performance of NiFe Dry Reforming Catalysts.

The dry reforming of methane (DRM), i.e., the reaction of methane and CO2 to form a synthesis gas, converts two major greenhouse gases into a useful chemical feedstock. In this work, we probe the effect and role of Fe in bimetallic NiFe dry reforming catalysts. To this end, monometallic Ni, Fe, and bimetallic Ni-Fe catalysts supported on a MgxAlyOz matrix derived via a hydrotalcite-like precursor were synthesized. Importantly, the textural features of the catalysts, i.e., the specific surface area (172-178 m2/gcat), pore volume (0.51-0.66 cm3/gcat), and particle size (5.4-5.8 nm) were kept constant. Bimetallic, Ni4Fe1 with Ni/(Ni + Fe) = 0.8, showed the highest activity and stability, whereas rapid deactivation and a low catalytic activity were observed for monometallic Ni and Fe catalysts, respectively. XRD, Raman, TPO, and TEM analysis confirmed that the deactivation of monometallic Ni catalysts was in large due to the formation of graphitic carbon. The promoting effect of Fe in bimetallic Ni-Fe was elucidated by combining operando XRD and XAS analyses and energy-dispersive X-ray spectroscopy complemented with density functional theory calculations. Under dry reforming conditions, Fe is oxidized partially to FeO leading to a partial dealloying and formation of a Ni-richer NiFe alloy. Fe migrates leading to the formation of FeO preferentially at the surface. Experiments in an inert helium atmosphere confirm that FeO reacts via a redox mechanism with carbon deposits forming CO, whereby the reduced Fe restores the original Ni-Fe alloy. Owing to the high activity of the material and the absence of any XRD signature of FeO, it is very likely that FeO is formed as small domains of a few atom layer thickness covering a fraction of the surface of the Ni-rich particles, ensuring a close proximity of the carbon removal (FeO) and methane activation (Ni) sites.

Increased Back-Bonding Explains Step-Edge Reactivity and Particle Size Effect for CO Activation on Ru Nanoparticles.

Carbon monoxide is a ubiquitous molecule, a key feedstock and intermediate in chemical processes. Its adsorption and activation, typically carried out on metallic nanoparticles (NPs), are strongly dependent on the particle size. In particular, small NPs, which in principle contain more corner and step-edge atoms, are surprisingly less reactive than larger ones. Hereby, first-principles calculations on explicit Ru NP models (1-2 nm) show that both small and large NPs can present step-edge sites (e.g., B5 and B6 sites). However, such sites display strong particle-size-dependent reactivity because of very subtle differences in local chemical bonding. State-of-the-art crystal orbital Hamilton population analysis allows a detailed molecular orbital picture of adsorbed CO on step-edges, which can be classified as flat (η1 coordination) and concave (η2 coordination) sites. Our analysis shows that the CO π-metal dπ hybrid band responsible for the electron back-donation is better represented by an oxygen lone pair on flat sites, whereas it is delocalized on both C and O atoms on concave sites, increasing the back-bonding on these sites compared to flat step-edges or low-index surface sites. The bonding analysis also rationalizes why CO cleavage is easier on step-edge sites of large NPs compared to small ones irrespective of the site geometry. The lower reactivity of small NPs is due to the smaller extent of the Ru-O interaction in the η2 adsorption mode, which destabilizes the η2 transition-state structure for CO direct cleavage. Our findings provide a molecular understanding of the reactivity of CO on NPs, which is consistent with the observed particle size effect.

Benzene partial hydrogenation: advances and perspectives.

The partial hydrogenation of benzene to cyclohexene is an economically interesting and technically challenging reaction. Over the last four decades, a lot of work has been dedicated to the development of an exploitable process and several approaches have been investigated. However, environmental constraints often represent a limit to their industrial application, making further research in this field necessary. The goal of this review is to highlight the main findings of the different disciplines involved in understanding the governing principles of this reaction from a sustainable chemistry standpoint. Special emphasis is given to ruthenium-catalyzed liquid phase batch hydrogenation of benzene.

Sputtering-deposition of Ru nanoparticles onto Al2O3 modified with imidazolium ionic liquids: synthesis, characterisation and catalysis.

Well-distributed Ru nanoparticles (Ru-NPs) were produced over Al(2)O(3) supports modified with covalently anchored imidazolium ionic liquids (ILs) containing different anions and cation lateral alkyl chain lengths by simple sputtering from a Ru foil. These Ru-NPs were active catalysts for the hydrogenation of benzene. Furthermore, depending on the nature of the IL used to modify the support (hydrophilic or hydrophobic), different catalytic behaviours were observed. Turnover numbers (TON) as high as 27 000 with a turnover frequency (TOF) of 2.73 s(-1) were achieved with Ru-NPs of 6.4 nm supported in Al(2)O(3) modified with an IL containing the N(SO(2)CF(3))2(-) anion, whereas higher initial cyclohexene selectivities (ca. 20% at 1% benzene conversion) were attained for Ru-NPs of 6.6 nm in the case where Cl(-) and BF(4)(-) anions were used. Such observations strongly suggest that thin layers of ILs surround the NP surface, modifying the reactivity of these catalytic systems. These findings open a new window of opportunity in the development of size-controlled Ru-NPs with tuneable reactivity.

Metal nanoparticle/ionic liquid/cellulose: new catalytically active membrane materials for hydrogenation reactions.

Transition metal-containing membrane films of 10, 20, and 40 µm thickness were obtained by the combination of irregularly shaped nanoparticles with monomodal size distributions of 4.8 ± 1.1 nm (Rh(0)) and 3.0 ± 0.4 nm (Pt(0)) dispersed in the ionic liquid (IL) 1-n-butyl-3-methylimidazolium bis(trifluoromethane sulfonyl)imide (BMI·(NTf)(2)) with a syrup of cellulose acetate (CA) in acetone. The Rh(0) and Pt(0) metal concentration increased proportionally with increases in film thickness up to 20 µm, and then the material became metal saturated. The presence of small and stable Rh(0) or Pt(0) nanoparticles induced an augmentation in the CA/IL film surface areas. The augmentation of the IL content resulted in an increase of elasticity and decrease in tenacity and toughness, whereas the stress at break was not influenced. The introduction of IL probably causes an increase in the separation between the cellulose macromolecules that results in a higher flexibility, lower viscosity, and better formability of the cellulose material. The nanoparticle/IL/CA combinations exhibit an excellent synergistic effect that enhances the activity and durability of the catalyst for the hydrogenation of cyclohexene. The nanoparticle/IL/cellulose acetate film membranes display higher catalytic activity (up to 7353 h(-1) for the 20 µm film of CA/IL/Pt(0)) and stability than the nanoparticles dispersed only in the IL.