Search

Iridium nanoparticles are important catalysts for several chemical and energy conversion reactions. Studies of iridium nanoparticles have also been a key for the development of kinetic models of nanomaterial formation. However, compared to other metals such as gold or platinum, knowledge on the nature of prenucleation species and structural insights into the resultant nanoparticles are missing, especially for nanoparticles obtained from IrxCly precursors investigated here. We use in situ X-ray total scattering (TS) experiments with pair distribution function (PDF) analysis to study a simple, surfactant-free synthesis of colloidal iridium nanoparticles. The reaction is performed in methanol at 50 °C with only a base and an iridium salt as precursor. From different precursor salts─IrCl3, IrCl4, H2IrCl6, or Na2IrCl6─colloidal nanoparticles as small as Ir∼55 are obtained as the final product. The nanoparticles do not show the bulk iridium face-centered cubic (fcc) structure but show decahedral and icosahedral structures. The formation route is highly dependent on the precursor salt used. Using IrCl3 or IrCl4, metallic iridium nanoparticles form rapidly from IrxClyn- complexes, whereas using H2IrCl6 or Na2IrCl6, the iridium nanoparticle formation follows a sudden growth after an induction period and the brief appearance of a crystalline phase. With H2IrCl6, the formation of different Irn (n = 55, 55, 85, and 116) nanoparticles depends on the nature of the cation in the base (LiOH, NaOH, KOH, or CsOH, respectively) and larger particles are obtained with larger cations. As the particles grow, the nanoparticle structure changes from partly icosahedral to decahedral. The results show that the synthesis of iridium nanoparticles from IrxCly is a valuable iridium nanoparticle model system, which can provide new compositional and structural insights into iridium nanoparticle formation and growth.

The dynamic atomic-scale behaviour of metallic mono- and bimetallic nanocrystals under reactive gas environments show the direct effect of alloying and Cu mobility on the corresponding restructuring processes.

Compositionally complex materials such as high-entropy alloys and oxides have the potential to be efficient platforms for catalyst discovery because of the vast chemical space spanned by these novel materials. Identifying the composition of the most active catalyst materials, however, requires unraveling the descriptor-activity relationship, as experimental screening the multitude of possible element ratios quickly becomes a daunting task. In this work, we show that inferred adsorption energy distributions of *OH and *O on complex solid solution surfaces within the space spanned by the system Ag-Pd-Pt-Ru are coupled to the experimentally observed electrocatalytic performance for the oxygen reduction reaction. In total, the catalytic activity of 1582 alloy compositions is predicted with a cross-validated mean absolute error of 0.042 mA/cm2 by applying a theory-derived model with only two adjustable parameters. Trends in the discrepancies between predicted electrochemical performance values of the model and the measured values on thin film surfaces subsequently provide insight into the alloys’ surface compositions during reaction conditions. Bridging this gap between computationally modeled and experimentally observed catalytic activities, not only reveals insight into the underlying theory of catalysis but also takes a step closer to realizing exploration and exploitation of high-entropy materials.

Rational design of catalysts by tailoring specific surface sites with different elements could result in catalysts with high activity, selectivity and stability. In this work, we show that *CO on-top and O* on-top adsorption energies are good descriptors for catalysis of the CO oxidation reaction (COOR) on pure metals and binary alloys. The observed Brønsted-Evans-Polanyi (BEP) and scaling relations for COOR on different surfaces are incorporated into a predictive model that uses the binding strength of the four adjacent metal atoms making up the active site for COOR catalysis to estimate reaction and activation energies. The model is used to screen 234 multi-metallic catalyst candidates made by combining Ru, Pt, Pd, Cu and Au at these four sites. The screening and subsequent calculations suggest that Ru-Cu-Au and Ru-Pt-Cu alloys are good catalysts for COOR. Our study shows that it is possible to use information from pure metals and binary alloys to predict the catalytic behavior of more complex alloys, and hereby reduce the computational cost of identifying new catalyst candidates for COOR.

Iridium nanoparticles are important catalysts for several chemical and energy conversion reactions. Studies of iridium nanoparticles have also been key for the development of kinetic models of nanomaterial formation. However, compared to other metals such as gold or platinum, there is very limited knowledge on the actual formation pathway of iridium nanoparticles on the atomic and molecular level. Here, we use in situ X-ray total scattering experiments with pair distribution function analysis to study a simple, surfactant-free synthesis of colloidal iridium nanoparticles. The reaction is performed in methanol at 50 °C with only a base and an iridium salt as precursor. From different precursor salts - IrCl3, IrCl4, H2IrCl6, or Na2IrCl6 – colloidal nanoparticles as small as Ir55 are obtained as the final product. The nanoparticles do not show the bulk iridium face-centered cubic (fcc) structure, but decahedral and icosahedral structures. The formation route is highly dependent on the precursor salt used. Using IrCl3 or IrCl4, metallic iridium nanoparticles form rapidly from IrxCly complexes, whereas using H2IrCl6 or Na2IrCl6, the iridium nanoparticle formation follows a sudden growth after an induction period and the brief ap-pearance of a crystalline phase. With H2IrCl6, the formation of different Irn (n= 55, 55, 85, 116) nanoparticles depends on the nature of the cation in the base - LiOH, NaOH, KOH, or CsOH, respectively - and larger particles are obtained with larger cations. As the particles grow, the nanoparticle structure changes from partly icosahedral to decahedral. The presented results introduce a new iridium nanoparticle synthesis model system and provide new chemical insights into nanoparticle formation and growth.

High-entropy alloy (HEA) electrocatalysts offer a vast composition space that awaits exploration to identify interesting materials for energy conversion reactions. While attempts have been made to explore the composition space of HEA thin-film libraries and compare experimental and computational studies, no corresponding approaches exist for HEA nanoparticles. So far, catalytic investigations on HEA nanoparticles are limited to small sets of individual catalysts. Here, we report the experimental exploration of the composition space of carbon-supported Pt−Ru−Pd−Rh− Au nanoparticles for the H2/CO oxidation reaction by constructing a dataset using Bayesian optimization as guidance. Applying a surfactant-free synthesis platform, a dataset of 68 samples was investigated. By constructing machine learning models, the relationship between the concentrations of the constituent elements and the catalytic activity was analyzed and compared to density functional theory calculations. The machine learning models confirm findings from previous studies concerning the role of Ru in the H2/CO oxidation reaction. This has been achieved starting from a random set of compositions and without any prior assumptions for the reaction mechanism nor any in-depth design of the active site. In addition, by comparing the trends of the computational and experimental studies, it is seen that the “onset potentials” across the compositions can be correlated with the adsorption energy of *OH. The best correlation between the computational and experimental data is obtained when considering 5% of the most strongly *OH adsorbing sites.

High-entropy alloys (HEAs) represent near-equimolar points in the middle of a vast composition space of multi-metallic catalysts. Successful modeling of the catalytic activity of these complex materials allows to search this composition space for optimal catalysts. Previous models of HEA catalytic activity have been based on local and intricate descriptions of the atomic environment on the catalyst surface to predict accurate adsorption energies. These are subsequently used to model the catalytic activity. In this study, we show that by approximating the ligand effect of the surrounding atoms around an adsorption site with a mean-field perturbation corresponding to equimolar AgIrPdPtRu, the same trend in the predictions of the oxygen reduction reaction catalytic activities are obtained for a majority of the quinary Ag-Ir-Pd-Pt-Ru composition space. By comparing to models that consider the ligand effect locally, we show that the extent of such a mean-field approximation is valid up to and including equimolar ternary alloys, corresponding to 60.3 % of the quinary composition space. When extrapolating to make predictions far from near-equimolar compositions, such as for binary alloys, the mean field has been sufficiently perturbed to cause large discrepancies compared to the local models. Here, the intricate models thus prove more useful for discovering optimal catalysts.

High-entropy alloys (HEAs) have the potential to be a paradigm-shift for rational catalyst discovery but this new type of alloy requires a completely new approach to predict the surface reactivity. In addition to the ligand effect perturbing the surface-adsorbate bond, the random configuration of elements in the surface will also induce local strain effects due to the varying radii of neighboring atoms. Accurate modelling of HEA surface reactivity requires an estimate of this effect: To what degree is the adsorption of intermediates on these lattice distorted atomic environments affected by local strain? In this study, more than 3,500 density functional theory (DFT) calculated adsorption energies of *OH and *O adsorbed on the HEAs IrPdPtRhRu and AgAuCuPdPt are statistically analyzed with respect to the lattice constants of the alloys and the surfaces of each individual binding site. It is found that the inherent distortion of the lattice structure in HEAs releases the local strain effect on the adsorption energy as the atomic environment surrounding the binding atom(s) settles into a relaxed structure. This is even observed to be true for clusters of atoms of which the sizes deviate significantly from the atomic environment in which they are embedded. This elucidates an important aspect of binding site interaction with the neighboring atoms and thus constitutes a step towards a more accurate theoretical model of estimating the reactivity of HEA surfaces.

Complex solid solutions (“high entropy alloys”), comprising five or more principal elements, promise a paradigm change in electrocatalysis due to the availability of millions of different active sites with unique arrangements of multiple elements directly neighbouring a binding site. Thus, strong electronic and geometric effects are induced, which are known as effective tools to tune activity. With the example of the oxygen reduction reaction, we show that by utilising a data‐driven discovery cycle, the multidimensionality challenge raised by this catalyst class can be mastered. Iteratively refined computational models predict activity trends around which continuous composition‐spread thin‐film libraries are synthesised. High‐throughput characterisation datasets are then used as input for refinement of the model. The refined model correctly predicts activity maxima of the exemplary model system Ag‐Ir‐Pd‐Pt‐Ru. The method can identify optimal complex‐solid‐solution materials for electrocatalytic reactions in an unprecedented manner., Complex solid solutions (“high‐entropy alloys”) promise a paradigm change in electrocatalysis but expose the challenge of almost unlimited options in adjusting their compositions. We propose the utilisation of computational models, combined with high‐throughput experimentation for the verification of the model assumptions, which allows for model refinement in iterative loops, understanding of binding mechanisms, and discovery of the most active composition.

We present an approach for a probabilistic and unbiased discovery of selective and active catalysts for the carbon dioxide (CO2) and carbon monoxide (CO) reduction reactions on high-entropy alloys ...

High entropy alloys (HEA) comprise a huge search space for new electrocatalysts. Next to element combinations, the optimization of the chemical composition is essential for tuning HEA to specific catalytic processes. Simulations of electrocatalytic activity can guide experimental efforts. Yet, the currently available underlying model assumptions do not necessarily align with experimental evidence. To study deviations of theoretical models and experimental data requires statistically relevant datasets. Here, a combinatorial strategy for acquiring large experimental datasets of multi-dimensional composition spaces is presented. Ru–Rh–Pd–Ir–Pt is studied as an exemplary, highly relevant HEA system. Systematic comparison with computed electrochemical activity enables the study of deviations from theoretical model assumptions for compositionally complex solid solutions in the experiment. The results suggest that the experimentally obtained distribution of surface atoms deviates from the ideal distribution of atoms in the model. Leveraging both advanced simulation and large experimental data enables the estimation of electrocatalytic activity and solid-solution stability trends in the 5D composition space of the HEA system. A perspective on future directions for the development of active and stable HEA catalysts is outlined.

The search for better and cheaper electrocatalysts is vital in the global transition to renewable energy resources. High-entropy alloys (HEAs) provide a near-infinite number of different alloys with approximately continuous properties such as catalytic activity. In this work, the catalytic activity for the electrochemical oxygen reduction reaction as a function of molar composition of Ag-Ir-Pd-Pt-Ru HEA is treated as a landscape wherein it is shown that the maxima are connected through ridges. By following the ridges, it is possible to navigate between the maxima using a modified nudged elastic band (NEB) model integrated in a machine learning NEB algorithm. These results provide a new understanding of the composition space being similar to an evolutionary landscape. This provides a possible new search and design strategy for new catalysts in which the composition of known catalysts can be optimized by following ridges rather than exploring the whole alloy composition space.

In recent years, the development of complex multi-metallic nanomaterials like high entropy alloy (HEA) catalysts has gained popularity. Composed of 5 or more metals, the compositions of HEAs exhibit extreme diversity. This is both a promising avenue to identify new catalysts and a severe constraint on their preparation and study. To address the challenges related to the preparation, study and optimization of HEAs, machine learning solutions are attractive. In this paper, the composition of PtRuPdRhAu hydrogen oxidation catalysts is optimized for the CO oxidation reaction. This is achieved by constructing a dataset using Bayesian optimization as guidance. For this quinary nanomaterial, the best performing composition was found within the first 35 experiments. However, the dataset was expanded until a total of 68 samples were investigated. This final dataset was used to construct a random forest regression model and a linear model. These machine learned models were used to assess the relationships between the concentrations of the consituent elements and the CO oxidation reaction onset potential. The onset potentials were found to correlate with the composition dependent adsorption energy of *OH obtained from density functional theory. This study demonstrates, how machine learning can be employed in an experimental setting to investigate the vast compositional space of HEAs.

Summary High-entropy alloys (HEAs) provide a near-continuous distribution of adsorption energies. With a minority of sites having optimal properties that dominate the catalysis, the overall catalytic activity can increase. In this article, we focus on the oxygen reduction reaction (ORR). We present density functional theory (DFT) calculated *OH and *O adsorption energies on a random subset of available binding sites on the surface of the HEA IrPdPtRhRu. Employing a simple machine learning algorithm, we predict remaining adsorption energies, finding good agreement between calculated and predicted values. With a full catalog of available adsorption energies, an appropriate expression for predicting catalytic activity was used to optimize the HEA composition. The HEA then becomes a design platform for the unbiased discovery of new alloys by promoting sites with exceptional catalytic activity. Setting different optimization constraints led to a new HEA composition and binary alloy IrPt, showing significant enhancements over pure Pt(111).

Electrocatalysis on high-entropy alloys (HEAs) is a very young field. There are only a few articles on the theoretical understanding of catalysis on these surfaces. Activity descriptors and a ‘theory of catalysis’ exist for uniform surfaces. In this article, we address scaling relations and Bronsted–Evans–Polanyi relations on HEAs which are fundamental for finding activity descriptors. We find that some relations hold, whereas others disappear on HEAs. This reveals some of the future challenges in modeling electrocatalysis on HEAs.