P. Kim, J. Van der Mynsbrugge, M. Head-Gordon, A. T. Bell, The Journal of Physical Chemistry C 2022, 126, 18744–18753. [ full text ]
Passive NOx adsorbers (PNAs) have been proposed for trapping NOx present in automotive exhaust during the period of cold start during which the three-way convertor is not yet hot enough to be effective for NOx reduction. Pd-exchanged chabazite (Pd/H–CHA) is a good candidate for passive NOx adsorption due to its ability to store NO and retain it to high temperatures (>473 K). Previous research suggests that NO adsorbs on both Pd2+ and Pd+ cations and that NO desorption from Pd2+ cations occurs at lower temperatures than from Pd+ cations. Since experimental evidence shows that Pd exchanges into CHA exclusively as Pd2+, it is not clear how these cations are reduced to Pd+. In this study we show through experiments and theoretical analysis that Pd+ cations can form via two processes, each of which involves water adsorbed on Brønsted-acid sites of the zeolite. The first of these processes is 1.5 NO + Pd2+Z–Z– + 0.5 (H2O)H+Z– → (NO)Pd+Z–H+Z– + 0.5 NO2 + 0.5 H+Z–. Experiments confirm that the ratio of the NO2 formed upon NO adsorption to the NO desorbing from Pd+ at elevated temperatures corresponds to 0.5. Pd2+ can also be reduced via the reaction 1.5 CO + Pd2+Z–Z– + 0.5 (H2O)H+Z– → (CO)Pd+Z–H+Z– + 0.5 CO2 + 0.5 H+Z–. Upon subsequent adsorption of NO, NO fully displaces CO from Pd+ to form (NO)Pd+Z–H+Z–. In this case, the amount of CO2 formed upon CO adsorption is 0.5 of the NO desorbing at elevated temperatures from Pd+. Gibbs free energy calculations for the above processes at various potential ion-exchange sites in the CHA framework indicate that these reactions are thermodynamically feasible. We also find that Pd+ is not formed in the absence of adsorbed water and is readily reoxidized to Pd2+ by trace amounts of O2.
P. Kim, J. Van der Mynsbrugge, H. Aljama, T. M. Lardinois, R. Gounder, M. Head-Gordon, A. T. Bell, Applied Catalysis B: Environmental 2022, 304, 120992. [ full text ]
This study investigates NO adsorption on Pd-exchanged chabazite (Pd/H-CHA), a promising passive NOx adsorber (PNA) for capturing cold-start NOx emissions of gasoline- and diesel-powered vehicles. Temperature-programed desorption (TPD) and IR spectroscopy are combined with theoretical calculations to elucidate how and where NO is stored, and how water and O2 affect this process. NO adsorption on Pd/H-CHA produces two TPD features, around 423 and 753 K, and IR bands centered at 1860 and 1810 cm−1. Calculated NO stretching frequencies and maximum-desorption temperatures reveal that Pd2+ and Pd+ sites are responsible for these low- and high-temperature features, respectively, and that while the IR feature at 1810 cm−1 is due to NO adsorption on Pd+, the 1860 cm−1 feature contains contributions from both weakly-bound NO on Pd2+ and more strongly bound NO on Pd+, consistent with experimentally observed effects of water and O2.
J. Van der Mynsbrugge, A. T. Bell, Journal of Catalysis 2021, 404, 832–849. [ full text ]
Zeolites are widely used as catalysts for the processing of petroleum to produce transportation fuels, the synthesis of a wide variety of chemicals, and for the abatement of automotive emissions. These applications have stimulated an interest in describing the mechanism and kinetics for zeolite-catalyzed reactions using theoretical methods. This Mini-review summarizes the author’s efforts towards this goal. It is shown that accurate predictions of adsorption and activation enthalpies and entropies requires that several criteria be met. The first is a correct description of the structure of the catalytically active center, as well as the portion of the zeolite framework immediately surrounding the active center and that located far from the active center. Second, the level of density functional theory (DFT) must be sufficiently high to account for the effects of dispersive interactions between the adsorbate, the active center, and the immediately surrounding zeolite atoms. Third, dispersive and coulombic interactions between the atoms in the vicinity of the active center and the balance of the zeolite framework must also be accounted for. It is shown that these conditions can be met using hybrid quantum mechanics/molecular mechanics (QM/MM) together with a high-level exchange-correlation functional and a large basis set. The success of our QM/MM approach is illustrated for reactions of light alkanes in H-MFI, as well as other protonated zeolites, and in Ga/H-MFI. We show that for low temperatures (< 400 K), the QM/MM approach gives good predictions of molecular adsorption enthalpies and activation enthalpies for elementary reactions. This is also true for higher temperatures (> 400 K) if the effects of configuration are considered using a correction obtained from configurationally biased Monte Carlo (CBMC) calculations. Calculations of the molecular adsorption entropy and the activation entropy for elementary reactions are more difficult to predict accurately. Application of the quasi-rigid rotor harmonic approximation overpredicts the loss of entropy of adsorption from the gas phase, particularly for zeolites containing large cavities and channels. CBMC corrections capture this deviation well for molecular adsorption and for early transition states resembling the adsorbed state but are inadequate for late transition states involving two loosely associated fragments.
J. Van der Mynsbrugge, M. Head-Gordon, A. T. Bell, Journal of Materials Chemistry A 2021, 9, 2161–2174. [ full text ]
The oxidation state (OS) of metals and ligands in inorganic complexes may be defined by carefully curated rules, such as from IUPAC, or by computational procedures such as the effective oxidation state (EOS) or localized orbital bonding analysis (LOBA). Such definitions typically agree for systems with simple ionic bonding and innocent ligands but may disagree as the boundary between ionic and covalent bonds is approached, or as the role of ligand noninnocence becomes nontrivial, or high oxidation states of metals are supported by heavy dative bonding, and so on. This work systematically compares IUPAC, EOS, and LOBA across a series of complexes where OS assignment is challenging. These systems include high-valent transition metal oxides, transition metal complexes with noninnocent ligands such as dithiolate and nitrosyl, metal sulfur dioxide adducts, and two transition metal carbene complexes. The differences in OS assignment by the three methods are carefully discussed, demonstrating the synergy between EOS and LOBA. In addition, a clarity index for LOBA OS assignments is introduced that provides an indication of whether or not its predictions are close to the ionic–covalent boundary.
This is a response to the paper published by S. A. Kadam, H. Li, R. F. Wormsbacher, A. Travert, Chem. Eur. J. 2018, 24, 5489. Key consistencies between our reported results and those reported in this work are also highlighted.
Catalytic alkene cracking on H-ZSM-5 involves a complex reaction network with many possible reaction routes and often elusive intermediates. Herein, advanced molecular dynamics simulations at 773 K, a typical cracking temperature, are performed to clarify the nature of the intermediates and to elucidate dominant cracking pathways at operating conditions. A series of C4–C8 alkene intermediates are investigated to evaluate the influence of chain length and degree of branching on their stability. Our simulations reveal that linear, secondary carbenium ions are relatively unstable, although their lifetime increases with carbon number. Tertiary carbenium ions, on the other hand, are shown to be very stable, irrespective of the chain length. Highly branched carbenium ions, though, tend to rapidly rearrange into more stable cationic species, either via cracking or isomerization reactions. Dominant cracking pathways were determined by combining these insights on carbenium ion stability with intrinsic free energy barriers for various octene β-scission reactions, determined via umbrella sampling simulations at operating temperature (773 K). Cracking modes A (3° → 3°) and B2 (3° → 2°) are expected to be dominant at operating conditions, whereas modes B1 (2° → 3°), C (2° → 2°), D2 (2° → 1°), and E2 (3° → 1°) are expected to be less important. All β-scission modes in which a transition state with primary carbocation character is involved have high intrinsic free energy barriers. Reactions starting from secondary carbenium ions will contribute less as these intermediates are short living at the high cracking temperature. Our results show the importance of simulations at operating conditions to properly evaluate the carbenium ion stability for β-scission reactions and to assess the mobility of all species in the pores of the zeolite.
J. Van der Mynsbrugge, A. Janda, L.-C. Lin, V. Van Speybroeck, M. Head‐Gordon, A. T. Bell, ChemPhysChem 2018, 19 (4), 341–358. [ full text ];
Recent experimental work has shown that variations in the confinement of n-butane at Brønsted acid sites due to changes in zeolite framework structure strongly affect the apparent and intrinsic enthalpy and entropy of activation for cracking and dehydrogenation. Quantum chemical calculations have provided good estimates of the intrinsic enthalpies and entropies of activation extracted from experimental rate data for MFI, but extending these calculations to less confining zeolites has proven challenging, particularly for activation entropies. Herein, we report our efforts to develop a theoretical model for the cracking and dehydrogenation of n-butane occurring in a series of zeolites containing 10-ring channels and differing in cavity size (TON, FER, -SVR, MFI, MEL, STF, and MWW). We combine a QM/MM approach to calculate intrinsic and apparent activation parameters, with thermal corrections to the apparent barriers obtained from configurational-bias Monte Carlo simulations, to account for configurational contributions due to global motions of the transition state. We obtain good agreement between theory and experiment for all activation parameters for central cracking in all zeolites. For terminal cracking and dehydrogenation, good agreement between theory and experiment is found only at the highest confinements. Experimental activation parameters, especially those for dehydrogenation, tend to increase with decreasing confinement. This trend is not captured by the theoretical calculations, such that deviations between theory and experiment increase as confinement decreases. We propose that, because transition states for dehydrogenation are later than those for cracking, relative movements between the fragments produced in the reaction become increasingly important in the less confining zeolites.
Catalytic cracking of alkenes takes place at elevated temperatures in the order of 773–833 K. In this work, the nature of the reactive intermediates at typical reaction conditions is studied in H-ZSM-5 using a complementary set of modeling tools. Ab initio static and molecular dynamics simulations are performed on different C4-C5 alkene cracking intermediates to identify the reactive species in terms of temperature. At 323 K, the prevalent intermediates are linear alkoxides, alkene π-complexes and tertiary carbenium ions. At a typical cracking temperature of 773 K, however, both secondary and tertiary alkoxides are unlikely to exist in the zeolite channels. Instead, more stable carbenium ion intermediates are found. Branched tertiary carbenium ions are very stable, while linear carbenium ions are predicted to be metastable at high temperature. Our findings confirm that carbenium ions, rather than alkoxides, are reactive intermediates in catalytic alkene cracking at 773 K.
Within the field of catalysis, theoretical modeling has also taken a prominent role. This chapter answers the question whether anno 2016 of a theoretical toolbox for a better catalytic understanding can be disposed. It focuses on heterogeneous catalysis and in particular on catalytic reactions taking place in the pores of a nanoporous material. An important issue when modeling catalytic processes in nanoporous materials is to select an appropriate model to simulate the extended molecular environment of the host material itself. The chapter concentrates on catalytic reactions taking place in two types of nanoporous materials: zeolites and metal-organic frameworks (MOFs) and also focuses on the determination of chemical kinetics of the catalytic reaction. Apart from modeling the reaction itself, computational spectroscopy has evolved as an indispensable characterization tool to identify intermediates during a reaction. In situ experiments deliver data, which evolve in time and vary with temperature and other operating conditions.
Adsorption of linear pentenes in H-ZSM-5 at 323 K is investigated using contemporary static and molecular dynamics methods. A physisorbed complex corresponding to free pentene, a π-complex and a chemisorbed species may occur. The chemisorbed species can be either a covalently bonded alkoxide or an ion pair, the so-called carbenium ion. Without finite temperature effects, the π-complex is systematically slightly more bound than the chemisorbed alkoxide complex, whereas molecular dynamics calculations at 323 K yield an almost equal stability of both species. The carbenium ion was not observed during simulations at 323 K. The transformation from the π-complex to the chemisorbed complex is activated by a free energy in the range of 33–42 kJ/mol. Our observations yield unprecedented insights into the stability of elusive intermediates in zeolite catalysis, for which experimental data are very hard to measure.
To optimally design next generation catalysts a thorough understanding of the chemical phenomena at the molecular scale is a prerequisite. Apart from qualitative knowledge on the reaction mechanism, it is also essential to be able to predict accurate rate constants. Molecular modeling has become a ubiquitous tool within the field of heterogeneous catalysis. Herein, we review current computational procedures to determine chemical kinetics from first principles, thus by using no experimental input and by modeling the catalyst and reacting species at the molecular level. Therefore, we use the methanol-to-olefin (MTO) process as a case study to illustrate the various theoretical concepts. This process is a showcase example where rational design of the catalyst was for a long time performed on the basis of trial and error, due to insufficient knowledge of the mechanism. For theoreticians the MTO process is particularly challenging as the catalyst has an inherent supramolecular nature, for which not only the Brønsted acidic site is important but also organic species, trapped in the zeolite pores, must be essentially present during active catalyst operation. All these aspects give rise to specific challenges for theoretical modeling. It is shown that present computational techniques have matured to a level where accurate enthalpy barriers and rate constants can be predicted for reactions occurring at a single active site. The comparison with experimental data such as apparent kinetic data for well-defined elementary reactions has become feasible as current computational techniques also allow predicting adsorption enthalpies with reasonable accuracy. Real catalysts are truly heterogeneous in a space- and time-like manner. Future theory developments should focus on extending our view towards phenomena occurring at longer length and time scales and integrating information from various scales towards a unified understanding of the catalyst. Within this respect molecular dynamics methods complemented with additional techniques to simulate rare events are now gradually making their entrance within zeolite catalysis. Recent applications have already given a flavor of the benefit of such techniques to simulate chemical reactions in complex molecular environments.
Framework-bound methoxides occur as intermediates in the stepwise mechanism for zeolite-catalyzed methylation reactions. Herein, the formation of methoxides from methanol or dimethyl ether in H-ZSM-5 is investigated by a combination of static and dynamic simulations, with particular focus on the effect of additional water and methanol molecules on the mechanism and kinetics. Metadynamics simulations allow partitioning the reaction path into distinct phases. Proton transfer from the zeolite to the reactants is found to be the rate-limiting phase in the methoxide formation. Additional methanol molecules only assist the proton transfer in the methoxide formation from methanol, whereas the reaction from dimethyl ether does not benefit from methanol assistance. Once formed, methoxides are found to be as reactive toward alkene methylation as methanol and dimethyl ether.
The methylation of arenes is a key step in the production of hydrocarbons from methanol over acidic zeolites. We performed ab initio static and molecular dynamics free energy simulations of benzene methylation in H-ZSM-5 to determine the factors that influence the reaction kinetics. Special emphasis is given to the effect of the surrounding methanol molecules on the methylation kinetics. It is found that for higher methanol loadings, methylation may also occur from a protonated methanol cluster, indicating that the exact location of the Brønsted acid site is not essential for the zeolite-catalyzed methylation reaction. However, methylations from a protonated methanol cluster exhibit higher free energy barriers than a methylation from a single methanol molecule. Finally, comparison with a pure methanol solvent reaction environment indicates that the main role of the zeolite during the methylation of benzene is to provide the acidic proton and to create a polar environment for the reaction. The metadynamics approach, which is specifically designed to sample rare events, allows exploring new reaction pathways, which take into account the flexibility of the framework and additional guest molecules in the pores and channels of the zeolite framework. This approach goes beyond the often applied static calculations to determine reaction kinetics.
The methylation of ethene, propene, and trans-2-butene on zeolites H-ZSM-58 (DDR), H-ZSM-22 (TON), and H-ZSM-5 (MFI) is studied to elucidate the particular influence of topology on the kinetics of zeolite-catalyzed reactions. H-ZSM-58 and H-ZSM-22 are found to display overall lower methylation rates compared to H-ZSM-5 and also different trends in methylation rates with increasing alkene size. These variations may be rationalized based on a decomposition of the free-energy barriers into enthalpic and entropic contributions, which reveals that the lower methylation rates on H-ZSM-58 and H-ZSM-22 have virtually opposite reasons. On H-ZSM-58, the lower methylation rates are caused by higher enthalpy barriers, owing to inefficient stabilization of the reaction intermediates in the large cage-like pores. On the other hand, on H-ZSM-22, the methylation rates mostly suffer from higher entropy barriers, because excessive entropy losses are incurred inside the narrow-channel structure. These results show that the kinetics of crucial elementary steps hinge on the balance between proper stabilization of the reaction intermediates inside the zeolite pores and the resulting entropy losses. These fundamental insights into their inner workings are indispensable for ultimately selecting or designing better zeolite catalysts.
The conversion of methanol to olefins (MTO) over a heterogeneous nanoporous catalyst material is a highly complex process involving a cascade of elementary reactions. The elucidation of the reaction mechanisms leading to either the desired production of ethene and/or propene or undesired deactivation has challenged researchers for many decades. Clearly, catalyst choice, in particular topology and acidity, as well as the specific process conditions determine the overall MTO activity and selectivity; however, the subtle balances between these factors remain not fully understood. In this review, an overview of proposed reaction mechanisms for the MTO process is given, focusing on the archetypal MTO catalysts, H-ZSM-5 and H-SAPO-34. The presence of organic species, that is, the so-called hydrocarbon pool, in the inorganic framework forms the starting point for the majority of the mechanistic routes. The combination of theory and experiment enables a detailed description of reaction mechanisms and corresponding reaction intermediates. The identification of such intermediates occurs by different spectroscopic techniques, for which theory and experiment also complement each other. Depending on the catalyst topology, reaction mechanisms proposed thus far involve aromatic or aliphatic intermediates. Ab initio simulations taking into account the zeolitic environment can nowadays be used to obtain reliable reaction barriers and chemical kinetics of individual reactions. As a result, computational chemistry and by extension computational spectroscopy have matured to the level at which reliable theoretical data can be obtained, supplying information that is very hard to acquire experimentally. Special emphasis is given to theoretical developments that open new perspectives and possibilities that aid to unravel a process as complex as methanol conversion over an acidic porous material.
V. Van Speybroeck, K. Hemelsoet, K. De Wispelaere, Q. Qian, J. Van der Mynsbrugge, B. De Sterck, B. M. Weckhuysen, M. Waroquier ChemCatChem 2013, 5 (1), 173–184. [ full text ]
Inside Back Cover: ChemCatChem 2013, 5 (1), 355–355. [ ]
The formation and nature of active sites for methanol conversion over solid acid catalyst materials are studied by using a unique combined spectroscopic and theoretical approach. A working catalyst for the methanol-to-olefin conversion has a hybrid organic–inorganic nature in which a cocatalytic organic species is trapped in zeolite pores. As a case study, microporous materials with the chabazite topology, namely, H-SAPO-34 and H-SSZ-13, are considered with trapped (poly)aromatic species. First-principle rate calculations on methylation reactions and in situ UV/Vis spectroscopy measurements are performed. The theoretical results show that the structure of the organic compound and zeolite composition determine the methylation rates: 1) the rate increases by 6 orders of magnitude if more methyl groups are added on benzenic species, 2) transition state selectivity occurs for organic species with more than one aromatic core and bearing more than three methyl groups, 3) methylation rates for H-SSZ-13 are approximately 3 orders of magnitude higher than on H-SAPO-34 owing to its higher acidity. The formation of (poly)aromatic cationic compounds can be followed by using in situ UV/Vis spectroscopy because these species yield characteristic absorption bands in the visible region of the spectrum. We have monitored the growth of characteristic peaks and derived activation energies of formation for various sets of (poly)aromatic compounds trapped in the zeolite host. The formation–activation barriers deduced by using UV/Vis microspectroscopy correlate well with the activation energies for the methylation of the benzenic species and the lower methylated naphthalenic species. This study shows that a fundamental insight at the molecular level can be obtained by using a combined in situ spectroscopic and theoretical approach for a complex catalyst of industrial relevance.
Benzene methylation by methanol is studied on acidic zeolites H-ZSM-5 (MFI) and H-beta (BEA) to investigate the influence of the catalyst topology on the reaction rate. Experimental kinetic measurements at 350 °C using extremely high feed rates to suppress side reactions show that methylation occurs considerably faster on H-ZSM-5 than on H-beta. Theoretical rate constants, obtained from first-principles simulations on extended zeolite clusters, reproduce a higher methylation rate on H-ZSM-5 and provide additional insight into the various molecular effects that contribute to the overall differences between the two catalysts. The calculations indicate this higher methylation rate is primarily due to an optimal confinement of the reacting species in the medium pore material. Co-adsorption of methanol and benzene is energetically favored in H-ZSM-5 compared with H-beta, to the extent that the stabilizing host–guest interactions outweigh the greater entropy loss upon benzene adsorption in H-ZSM-5 vs. in H-beta.
Since many industrially important processes start with the adsorption of guest molecules inside the pores of an acidic zeolite catalyst, a proper estimate of the adsorption enthalpy is of paramount importance. In this contribution, we report ab initio calculations on the adsorption of water, alcohols, and nitriles at the bridging Brønsted sites of H-ZSM-5, using both cluster and periodic models to account for the zeolite environment. Stabilization of the adsorption complexes results from hydrogen bonding between the guest molecule and the framework, as well as from embedding, i.e., van der Waals interactions with the pore walls. Large-cluster calculations with different DFT methods, in particular B3LYP(-D), PBE(-D), M062X(-D), and ωB97X-D, are tested for their ability to reproduce the experimental heats of adsorption available in the literature (J. Phys. Chem. B 1997, 101, 3811–3817). A proper account of dispersion interactions is found to be crucial to describe the experimental trend across a series of adsorbates of increasing size, i.e., an increase in adsorption enthalpy by 10–15 kJ/mol for each additional carbon atom. The extended-cluster model is shown to offer an attractive alternative to periodic simulations on the entire H-ZSM-5 unit cell, resulting in virtually identical final adsorption enthalpies. Comparing calculated stretch frequencies of the zeolite acid sites and the adsorbate functional groups with experimental IR data additionally confirms that the cluster approach provides an appropriate representation of the adsorption complexes.
The methanol-to-olefin (MTO) process, catalyzed by acidic zeolites such as H-ZSM-5, provides an increasingly important alternative to the production of light olefins from crude oil. However, the various mechanistic proposals for methanol-to-olefin conversion have been strongly disputed for the past several decades. This work provides theoretical evidence that the experimentally suggested ‘alkene cycle’, part of a co-catalytic hydrocarbon pool, offers a viable path to the production of both propene and ethene, in stark contrast to the often- proposed direct mechanisms. This specific proposal hinges on repeated methylation reactions of alkenes, starting from propene, which occur easily within the zeolite environment. Subsequent cracking steps regenerate the original propene molecule, while also forming new propene and ethene molecules as primary products. Because the host framework stabilizes intermediate carbenium ions, isomerization and deprotonation reactions are extremely fast. Combined with earlier joint experimental and theoretical work on polymethylbenzenes as active hydrocarbon pool species, it is clear that, in zeolite H-ZSM-5, multiple parallel and interlinked routes operate on a competitive basis.
The formation of cyclic hydrocarbons from smaller building blocks such as ethene and propene is investigated in protonated ZSM-5, using a 2-layered ONIOM(B3LYP/6-31+g(d):HF/6-31+g(d)) approach and an additional Grimme-type van der Waals dispersion correction term to account for the long-range dispersion interactions. These cyclic species form precursors for active hydrocarbon pool species and play a key role in activating the acidic zeolite host for successful methanol-to-olefin (MTO) conversion. Starting from trace amounts of ethene and propene that are formed during an initial induction period or during the active phase, dimerization reactions allow for rapid chain growth. The products of these reactions can be neutral alkenes, framework-bound alkoxide species or intermediate carbenium ions, depending on the zeolite environment taken into account. On the basis of rate constants for successive reaction steps, a viable route toward cyclization is proposed, which starts from the formation of a framework-bound propoxide from propene, followed by dimerization with an additional propene molecule to form the 2-hexyl carbenium ion which finally undergoes ring closure to yield methylcyclopentane. This cyclic species in turn forms a precursor for either an active hydrocarbon pool compound or for deactivating coke deposit.