(10f) Confinement and Acid Strength Effects in Catalysis By Microporous Acids | AIChE

(10f) Confinement and Acid Strength Effects in Catalysis By Microporous Acids

Authors 

Jones, A. J. - Presenter, University of California at Berkeley
Iglesia, E., Chemical Engineering
Zones, S., Chevron Research



Microporous solid acids, such as zeolites, contain pores on the order of molecular dimensions, which solvate molecules through van der Waals interactions. These interactions can stabilize reactants and transition states to different extents resulting in void size effects that influence turnover rates of monomolecular alkane cracking [1] and dimethyl ether (DME) carbonylation [2] in zeolites. Yet, the individual contributions of van der Waals interactions to catalytic rates are often misconstrued as acid strength differences because of the challenges of separating the stabilizing effects of acid strength and electrostatic forces from those of solvation by van der Waals interactions on the energies of transitions states and relevant precursors [3].

Here we interpret CH3OH dehydration rate and equilibrium constants (433 K) on zeolites in the context of the separate effects of framework composition and void size, which introduce independent consequences of acid strength and solvation on reactivity. CH3OH dehydration turnover rates are combined with infrared spectroscopy and density functional theory to show that CH3OH dehydration on zeolites proceeds via the formation of H-bonded methanol and protonated dimers and the kinetically-relevant elimination of water from the latter, consistent with unconstrained solid acids, such as Keggin polyoxometalates [4].

First-order CH3OH dehydration rate constants (kfirst) depend on the differences in free energy between the DME formation transition state and one H-bonded and one gaseous CH3OH molecule. Their value increases exponentially with the calculated adsorption enthalpy of n-C4H10 on zeolites with different pore topologies (FAU, BEA, MOR, MFI) [5]. This suggests that these rate constants are sensitive to the preferential stabilization of the transition state over H-bonded methanol precursors. Zero-order rate constants (kzero) reflect the difference in free energy between the DME formation transition state and a protonated CH3OH dimer species, and are similar between the samples. This suggests that kzero is insensitive to the preferential stabilization of the transition state over CH3OH dimer species, presumably because both species are stabilized to similar extents by changes in void shape and size.

Values of first-order and zero-order rate constants on a series of MFI zeolites with Al, Ga, Fe and B heteroatoms decrease exponentially with deprotonation energies (DPE), a measure of acid strength, calculated by density functional theory. This indicates that both rate constants are sensitive to the changes in electrostatic stabilizations indicated by DPE values because transition states are preferentially stabilized over reactive intermediates as a result of their higher charge. The sensitivity of these rate constants to acid strength together with the monotonic increases in kfirst but not kzero for zeolites of different pore topologies (FAU, BEA, MOR, MFI) with n-C4H10adsorption enthalpies suggests acid strength differences among these aluminosilicates are negligible.

Average van der Waals adsorption energies of DME formation transition states located at crystallographically unique O-atom sites, calculated from rigid structures with a Lennard-Jones type potential, accurately predict rate constant differences measured on a range of zeolites (FAU, BEA, MTW, MOR, SFH, CHA, MFI) and suggest that this calculation technique can be used to prescreen large numbers of zeolites for catalytic testing in general and CH3OH dehydration in particular. Calculated interaction energies depend on the location of the transition state at different O-sites within a zeolite, emphasizing the distribution of solvating environments present in each zeolite. Similar values of kfirst measured in BEA and 12-MR channels in MOR (within a factor of 1.03) indicate that reactions occur, and thus H+ are located, almost exclusively in the 12-MR channels of BEA and not at their intersections because these larger voids would result in lower values of kfirst. MFI samples with Si/Al ratios between 22 and 118 (6 samples in total with different Al densities and provenance) have values of kfirst that are similar to those in BEA and MOR (within a factor of 1.33), suggesting that H+ in these samples are located at 10-MR channel intersections where voids solvate transition states and reactants similar to those in 12-MR channels of BEA and MOR. At a higher Al density (Si/Al = 17), kfirst increases by a factor of 3.4 relative to that measured on samples with lower Al densities and is consistent with predicted rates for van der Waals energies calculated for H+in 10-MR channels rather than in intersection voids.

These results indicate that CH3OH dehydration turnover rate differences on zeolitic Brønsted acids are due primarily to differences in void shape and size, and not differences in acid strength, which depends on the identity of the heteroatom. The comparison of rate constant values with calculated van der Waals interaction energies at crystallographically unique O-atoms provides insight into the location of active sites within each zeolite structure. CH3OH dehydration rate constants and their mechanistic understanding indicate the specific contributions of heteroatom composition and void environment to catalytic turnovers.

[1]        B. Xu, C. Sievers, S.B. Hong, R. Prins, and J.A. van Bokhoven, J. Catal. 244 (2006) 163-168.

[2]        R. Gounder, and E. Iglesia, Acc. Chem. Res. 45 (2012) 229-238.

[3]        R.J. Gorte, Catal. Lett. 62 (1999) 1-13.

[4]        R.T. Carr, M. Neurock, and E. Iglesia, J. Catal. 278 (2011) 78-93.

[5]        B.A. De Moor, M.F. Reyniers, O.C. Gobin, J.A. Lercher, and G.B. Marin, J. Phys. Chem. C 115 (2011) 1204-1219.

Financial support from Chevron Energy Technology Company and supercomputing resources provided by the Molecular Graphics and Computation Facility in the College of Chemistry at the University California, Berkeley under NSF CHE-0840505y are gratefully acknowledged.

Topics