Zeolites are versatile and fascinating materials which are vital for a wide range of industries, due to their unique structural and chemical properties, which are the basis of applications in gas separation, ion exchange and catalysis. Given their economic impact, there is a powerful incentive for smart design of new materials with enhanced functionalities to obtain the best material for a given application. Over the last decades, theoretical modeling has matured to a level that model guided design has become within reach. Major hurdles have been overcome to reach this point and almost all contemporary methods in computational materials chemistry are actively used in the field of modeling zeolite chemistry and applications. Integration of complementary modeling approaches is necessary to obtain reliable predictions and rationalizations from theory. A close synergy between experimentalists and theoreticians has led to a deep understanding of the complexity of the system at hand, but also allowed the identification of shortcomings in current theoretical approaches. Inspired by the importance of zeolite characterization which can now be performed at the single atom and single molecule level from experiment, computational spectroscopy has grown in importance in the last decade. In this review most of the currently available modeling tools are introduced and illustrated on the most challenging problems in zeolite science. Directions for future model developments will be given.

https://biblio.ugent.be/publication/6977178/file/8575390.pdf

Advances in theory and their application within the field of zeolite chemistry.

The active sites of heterogeneous catalysts can be difficult to identify and understand, and, hence, the introduction of active sites into catalysts to tailor their function is challenging. During the past two decades, scaling relationships have been established for important heterogeneous catalytic reactions. More specifically, a physical or chemical property of the reaction system, termed as a reactivity descriptor, scales with another property often in a linear manner, which can describe and/or predict the catalytic performance. In this Review, we describe scaling relationships and reactivity descriptors for heterogeneous catalysis, including electronic descriptors represented by d-band theory, structural descriptors, which can be directly applied to catalyst design, and, ultimately, universal descriptors. The prediction of trends in catalytic performance using reactivity descriptors can enable the rational design of catalysts and the efficient screening of high-throughput catalysts. Finally, we outline methods to break scaling relationships and, hence, to break the constraint that active sites pose on the catalytic performance. Recently, scaling relationships have been established between certain physical or chemical properties of heterogeneous catalytic reactions. These properties, or reactivity descriptors, can describe and predict catalytic performance, and thus enable the rational design of new catalysts.

Theory-guided design of catalytic materials using scaling relationships and reactivity descriptors

By combining metal nodes with organic linkers we can potentially synthesize millions of possible metal-organic frameworks (MOFs). The fact that we have so many materials opens many exciting avenues but also create new challenges. We simply have too many materials to be processed using conventional, brute force, methods. In this review, we show that having so many materials allows us to use big-data methods as a powerful technique to study these materials and to discover complex correlations. The first part of the review gives an introduction to the principles of big-data science. We show how to select appropriate training sets, survey approaches that are used to represent these materials in feature space, and review different learning architectures, as well as evaluation and interpretation strategies. In the second part, we review how the different approaches of machine learning have been applied to porous materials. In particular, we discuss applications in the field of gas storage and separation, the stability of these materials, their electronic properties, and their synthesis. Given the increasing interest of the scientific community in machine learning, we expect this list to rapidly expand in the coming years.

Big-Data Science in Porous Materials: Materials Genomics and Machine Learning.

Ensemble-averaged deprotonation energies (DPE) derived from periodic density functional theory models are insensitive to the location of isolated Al atoms and associated protons and similar among microporous aluminosilicates (i.e., zeolites) with different crystalline frameworks (MFI, BEA, FER, MOR, CHA, FAU). These DPE values are 1201 ± 11 kJ mol–1 after correcting for systematic artifacts of periodic DFT methods, which vary with framework density, and averaging over the four distinct proton locations at each Al atom. These energies rigorously reflect the strength of the acid sites in these important catalytic solids. Thus, the stability of the conjugate anions and the acid strength of these materials merely reflect the presence of Al atoms within the silicate framework, and not their specific siting or local confining environment. DPE values did not show any systematic trends with the vibrational frequency or length of O–H bonds, with Si–O–Al bond angles, or with NH3 adsorption enthalpies, properties th...

The Strength of Brønsted Acid Sites in Microporous Aluminosilicates

Little is known on the early stages of the methanol-to-olefin (MTO) conversion over H-SAPO-34, before the steady-state with highly active polymethylbenzenium cations as most important intermediates is reached. In this work, the formation and evolution of carbenium ions during the early stages of the MTO conversion on a H-SAPO-34 model catalyst were clarified via 1H MAS NMR and 13C MAS NMR. Several initial species (i.e., three-ring compounds, dienes, polymethylcyclopentenyl, and polymethylcyclohexenyl cations) were, for the first time, directly verified during the MTO conversion. Their detailed evolution network was established from theoretical calculations. On the basis of these results, an olefin-based catalytic cycle is proposed to be the primary reaction pathway during the early stages of the MTO reaction over H-SAPO-34. After that, an aromatic-based cycle may be involved in the MTO conversion for long times on stream.

Understanding the Early Stages of the Methanol-to-Olefin Conversion on H-SAPO-34

Recent work has reported the discovery of metal surface catalysts by employing a descriptor-based approach, establishing a correlation between a few well-defined properties of a material and its catalytic activity. This theoretical work aims for a similar approach in solid acid catalysis, focusing on the reaction between propene and methanol catalyzed by Bronsted acidic zeotype catalysts. Experimentally, the ammonia heat of adsorption is often used as a measure of the strength of acid sites. Using periodic DFT calculations, we show that this measure can be used to establish scaling relations for the energy of intermediates and transition states, effectively describing the reactivity of the acid site. This allows us to use microkinetic modeling to predict a quantitative relation between the ammonia heat of adsorption and the rate of propene methylation from first principles. We propose that this is the first step toward descriptor-based design of solid acid catalysts.

Chuan Ming Wang

Papers

Reactivity Descriptor in Solid Acid Catalysis: Predicting Turnover Frequencies for Propene Methylation in Zeotypes.