No abstract

Ammonia titration methods were developed to discriminate and quantify Brønsted acid sites of different strength that compensate aluminum and boron heteroatoms incorporated within zeolite frameworks. Borosilicate and boroaluminosilicate MFI zeolites (B-Al-MFI) were synthesized with different Al contents and crystallite sizes, which are typically correlated structural properties in aluminosilicates synthesized hydrothermally, but independently varied here by incorporating boron as a second framework heteroatom and using ethylenediamine as a structure directing agent. Temperature-programmed desorption (TPD) of ammonia from B-Al-MFI samples saturated via liquid-phase NH4NO3 ion exchange resulted in quantifying the total number of Al and B heteroatoms. In contrast, TPD performed after NH4-form B-Al-MFI samples were purged in flowing helium (433 K), or after gas-phase NH3 adsorption (433 K) onto H-form B-Al-MFI samples, quantified only protons charge-compensating framework Al heteroatoms. Turnover rates for methanol dehydration to dimethyl ether, when measured in zero-order kinetic regimes that are sensitive predominantly to Brønsted acid strength, are dependent only on the number of protons compensating framework Al atoms in B-Al-MFI zeolites. The NH3 titration methods developed here are useful in rigorously normalizing turnover rates of Brønsted acid-catalyzed reactions in boroaluminosilicate zeolites, which have been recognized previously to be dependent solely on Al content. The incorporation of B heteroatoms into zeolite frameworks, which generate protons that are essentially unreactive, provides a strategy to influence crystallite sizes independently of Al content, especially relevant in cases where catalytic behavior is influenced by intracrystalline transport phenomena.

We report synthetic protocols to independently influence composition and crystallite sizes of MFI zeolites, properties that contribute to Thiele moduli, and the proximity of Al heteroatoms at fixed composition (Si/Al ∼ 50). Crystallite sizes decrease with the addition of non-catalytic B heteroatoms in B–Al–MFI zeolites. Using only tetra-n-propylammonium (TPA+) as a structure-directing agent (SDA) leads to occlusion of one TPA+ per channel intersection and a finite percentage of framework Al atoms (20–40%) in proximal configurations. In contrast, using mixtures of ethylenediamine (EDA) and TPA+ (EDA/TPA+ = 15) leads to incorporation of EDA that displaces some TPA+ and to nearly all (>95%) Al in isolated configurations. These findings indicate how adding B and EDA to zeolite synthesis solutions provides a route to crystallizing MFI zeolites of similar composition but with framework Al distributed at varying relative proximity, or with similar Al distributions but different diffusion path lengths.

Alkane dehydrogenation rates on acidic zeolites measured in the presence of co-fed H2 during initial contact with reactants solely reflect protolytic reactions at Brønsted acid sites, while rates measured without co-fed H2 and at later reaction times reflect additional contributions from an extrinsic dehydrogenation function derived from reactants and products. This extrinsic function consists of unsaturated organic residues that catalyze dehydrogenation turnovers by accepting H-atoms from alkanes and recombining them as H2. Such hydrogen transfer routes are inhibited by alkenes and H2 products and proceed with activation barriers much lower than for protolytic dehydrogenation at H+ sites, causing them to become more prevalent at lower temperatures and for zeolites with lower H+ densities. The number, composition, and reactivity of these extrinsic carbonaceous active sites depend on the local concentrations of reactants and products, which vary with alkane and H2 pressure, bed residence time, and axial mixing. These extrinsic catalytic moieties form within H2-deficient regions of catalyst beds but can be removed by thermal treatments in H2, which fully restore zeolite catalysts to their initial state. Carbonaceous deposits do not catalyze alkane cracking reactions; thus, cracking rate constants serve as a reporter of the state of proton sites, and their invariance with product pressure, residence time, and axial mixing confirms that protons remain unoccupied and undisturbed as extrinsic organic residues change in number, composition, and reactivity. The rates of the reverse reaction (alkene hydrogenation) under H2-rich conditions inhibit the formation and the reactivity of these organic residues, and taken together with formalisms based on nonequilibrium thermodynamics, they confirm that alkane dehydrogenation occurs solely via protolytic routes only at the earliest stages of reaction in the presence of added H2. These findings provide a coherent retrospective view of the root causes of the literature discord about alkane dehydrogenation turnover rates and activation barriers on acidic zeolites, variously attributed to extraframework Al or radical active sites and to turnovers limited by alkene desorption instead of protolytic steps. Importantly, these findings also prescribe experimental protocols that isolate the kinetic contributions of protolytic dehydrogenation routes, thus ensuring their replication, while suggesting strategies to deposit or remove extrinsic organocatalytic functions that mediate hydrogen transfer reactions.

No abstract