Dongjae Shin (신동재)

Catalyst design with good stability beyond simply having high activity is crucial for a variety of reactions. Here, we evaluate the ceria catalyst for CO oxidation as a model reaction to rationally design an ultrastable catalyst with high activity. The goal was achieved by co-doping with rare-earth (RE) and transition metals (TM) simultaneously. The RE dopant stabilized the catalyst by inhibiting sintering that could lead to catalyst deactivation. The TM dopant increased the activity by facilitating formation of surface defects. Consequently, ceria co-doped with RE (=La, Sm) and TM (=Cu) had increased catalytic activity as well as superior resistance to deactivation during 10 cycle measurement (1 cycle: 900 °C, 24 h → cooling at room temperature → target °C, 24 h) of ∼700 h, which is harsher than any other reported conditions. This approach will shed light on the design of ultrastable oxide materials for a wide range of catalytic reactions.

The water–gas shift reaction (WGSR) is important in industries because it can reduce the CO content of syngas to produce purified H2, which can be used as fuel or to make ammonia (NH3). Supported noble metal catalysts have been widely studied for the WGSR because they exhibit high reactivity. However, the role of a metal–support interface in the WGSR has not yet been revealed and remains elusive. Density functional theory (DFT) calculations were performed for a model system of Co3O4-supported Pd (Pd/Co3O4) catalysts. The presence of the interface was found to promote the H2O dissociation step, which is crucial for improving WGSR activity. Thus, the WGSR activity was predicted to be enhanced by an increased number of interfaces, which could be achieved by controlling the size of the supported Pd nanoparticles (NPs). Furthermore, electronic metal–support interactions (MSIs) were found to be a source of the promoted H2O dissociation at the interface. The DFT-predicted promotion of H2O dissociation was further experimentally validated using Pd/Co3O4 catalysts that were size-controlled with calcination temperatures, and the total length of the interface was shown to have a direct correlation with the WGSR rate. Theoretical insights into the role of the interface and the enhancement of WGSR activity due to increased interface sites, which can be achieved by size control, are believed to be useful in the design of efficient supported metal catalysts for the WGSR.

Oxidation of elemental Hg (Hg0) is catalytically and economically an efficient way to remove the harmful Hg contained in the flue gas from coal combustion facilities. Thus, the development of a highly active V2O5/TiO2 catalyst, which is active to the Hg oxidation, is very essential. Support modification can change the reactivity of V2O5/TiO2 by affecting the V2O5 active center which is critical to the surface–reactant interaction, so understanding the effects of support tuning methods, e.g., crystallographic phase control and reduction treatment, on the Hg oxidation activity is valuable. Herein, density functional theory calculations were performed to mechanistically investigate the change of Hg oxidation reactivity by the support tuning methods and to elucidate the change of the electronic environment at the active site. The phase control to the TiO2 support was found to improve the Hg oxidation activity, but the reduction treatment decreased the activity, which is attributed to the change of the charge density at V2O5. Furthermore, the origin of the reactivity change was elucidated within a Sabatier-like principle that the interaction between the V site and the surface Cl critically contributes to the change of the Hg oxidation reactivity by balancing the competition between two key reaction steps of HCl dissociation and HgCl2 desorption. Our results provide guidance to improve the activity of the VOx/TiO2 catalyst for various reactions such as Hg oxidation and selective catalytic reduction of NOx and so on.

Metal alloy catalysts have been known to enhance many essential reactions. Surface-adsorbate interaction of an alloy catalyst depends on the composition and nanostructure of the surface, which can be changed by a phenomenon called surface segregation. Thus, thermodynamic information on the surface segregation is critical to tune the surface-adsorbate interaction, and thus the reactivity of alloy catalysts. Collected 1,366 density functional theory-calculated segregation energies (Esegr) from literature were featurized by 19-dimensional inexpensive numerical fingerprint, which represents facet-, site-, and elemental-dependencies. Deep neural network (DNN) model was constructed based on the data. Reasonable interpolative and extrapolative prediction by the DNN was clearly demonstrated using principal component (PC) analysis. On top of that, impact of each feature on the Esegr was analyzed by an explainable artificial intelligence approach, giving useful insights for alloy catalyst design based on surface segregation. As an example of the applications, the DNN model was used to explain the formation thermodynamics of site-selective nanosegregated Pt alloy catalyst, and was also used to perform elemental screening to find impurity-host metal combinations feasible to form the nanosegregated alloy catalyst. In addition, the guidance for data collection to improve the DNN model was given by analyzing PC space and performing the weight analysis of PCs.

Machine learning (ML) and Artificial Intelligence (AI) are changing how catalysis science is performed, with the prospect of accelerating the discovery of catalysis knowledge. Despite the substantial efforts and resources invested in the development and application of AI/ML approaches, their performance still exhibits clear limitations due to the lack of usable data. Although decades of work have produced an abundance of catalysis data, much of it was generated without sufficient understanding of AI/ML or communitywide consensus on terminology and reporting standards. As a result, not all of this data is suitable for training models. It is therefore essential to define what constitutes AI-ready data and understand the efforts required to obtain it, in order to prepare for the era of AI-driven catalysis. In this perspective, we define the essential properties of AI-ready data that enable the construction of robust catalysis AI/ML models. Then, we discuss current practices in computational and experimental catalysis through the lens of data readiness for AI/ML applications. Lastly, we discuss community-level efforts needed to establish large-scale datasets ready for AI-driven catalysis research.