Strategies to mitigate deactivation in methanol-to-hydrocarbons catalysis over zeolites
Sukaran Arora, Andrew Hwang, Aditya Bhan
Department of Chemical Engineering and Materials Science, University of Minnesota, Twin Cities
1. Introduction
An increasing fraction of the annual 250 MMT production of ethylene and propylene is now derived from methanol-to-hydrocarbons (MTH) conversion over zeolitic solid acids1. The identity and distribution of hydrocarbon products in MTH is largely governed by the zeolite topology; methanol conversion over medium-pore MFI zeolites (e.g., HZSM-5) predominantly yields gasoline-range hydrocarbons while small-pore CHA zeotypes (e.g., HSAPO-34, HSSZ-13) primarily produce light olefins2. Selectivity trends in MTH can be rationalized based on a dual-cycle schematic that considers aromatic and olefin species confined in zeolite micropores as organic co-catalysts in a ‘hydrocarbon pool’ mechanism with distinct propagation events based on methylation/dealkylation of aromatic species (referred as the ‘aromatics cycle’) and oligomerization/cracking of olefinic species (referred as the ‘olefins cycle’); isotopic labeling studies reveal that ethylene is predominantly formed in the aromatics cycle2–4. This description of propagation reaction sequences however, proffers no guidance regarding molecular events that engender catalyst deactivation. Our recent reports5,6 inquiring the identity and involvement of plausible intermediates and steps in MTH deactivation provide evidence that formaldehyde, formed in trace quantities by the loss of hydrogen from methanol, facilitates these undesired transformations via electrophilic addition to olefins and electrophilic substitution reactions of aromatics as also corroborated by other reports in the recent literature7,8. We have exploited this mechanistic understanding to demonstrate that catalyst lifetime can be enhanced either by scavenging formaldehyde or by impeding the extent of formaldehyde-mediated condensation pathways leading to the formation of multi-ring aromatic species.
2. Experimental Details
Product selectivities and methanol/dimethyl ether conversions were measured during catalytic reactions of methanol with or without high-pressure H2 co-feeds over beds comprised of HSAPO-34 ((Al+P)/Si = 10, 0.85 mmol H+/g), HSSZ-13 (Si/Al = 9, 0.5-0.7 mmol H+/g), Y2O3 (0.03 mmolCO2-site/g), or intra-/inter- pellet mixtures thereof in a packed bed reactor. The intensive parameter incorporating effects of changes in the methanol space velocity and total number of H+ sites in the catalyst bed, total turnovers, calculated as the total amount of methanol/DME converted to hydrocarbon products observed in the effluent per H+ until the methanol conversion drops to zero, was used for the rigorous assessment of MTH lifetime.
3. Results
We demonstrate the
efficacy of a bifunctional strategy to improve the lifetime of CHA zeotypic
materials (4x increase in total turnovers) for the conversion of methanol to
light olefins via physical addition of Y2O3 without
disrupting the high selectivity to 
ethylene and propylene. We also reveal that the efficacy of this
strategy increases with increasing proximity between the rare-earth metal oxide
surface and acid sites inside the zeotypic material (see Fig. 1). This strategy,
relying on kinetic coupling made 
possible
by the presence of acid and base catalytic functions within molecular diffusion
distances, exploits: (i) The scavenging of formaldehyde, derived in methanol
dehydrogenation events, that reacts via alkylation chemistries to transform
active hydrocarbon pool species to unreactive multi-ring aromatic species; (ii)
The previously unrecognized ability of Y2O3 and more
generally of rare-earth oxides to selectively and catalytically decompose
formaldehyde to COx in presence of unconverted methanol/DME
reactants and methanol homologation products; and (iii) The crucial role of
formaldehyde transport between and within zeotypic domains in determining
lifetime for methanol conversion.
We demonstrate
that lifetime of HSAPO-34 for methanol conversion can be enhanced with
increasing efficacy (~3x to >70x increase in total turnovers) by co-feeding
H2 at increasing partial pressures (400 to 3000 kPa) in the influent
with methanol (13 kPa) as compared to the case of co-feeding He at ~100 kPa
without disrupting the high light olefins selectivity (see Fig. 2). Co-feeding
H2 is shown to improve lifetime of HSSZ-13 (~4x), the aluminosilicate
analog of HSAPO-34, and HZSM-5 (~3x), an 
aluminosilicate zeolite with MFI topology, evincing the general
applicability of the proposed strategy. This strategy exploits the use of a
very weak hydrogenation catalyst—the zeolite itself—with the underlying postulate
that its ineffectiveness in activating H2 relative to prolific
metal-based hydrogenation catalysts will result in selective hydrogenation of undesired
polyunsaturated organic species resulting from formaldehyde-mediated alkylation
reactions while not effecting hydrogenation of the desired products—ethylene
and 
propylene.
Our preliminary results from kinetic studies of hydrogenation of propylene and
butadiene over HSSZ-13 corroborate this postulate wherein the observed rates of
formation of the hydrogenated analogs of propylene and butadiene during
reactions with high-pressure H2 at 673 K are observed to be ~175x
higher in the case of butadiene relative to propylene feeds.
4. Conclusions
Here, we demonstrate the efficacy of (i) a bifunctional strategy exploiting the selective and catalytic decomposition of formaldehyde by Y2O39 and (ii) co-processing H2 to intercept undesired formaldehyde-mediated alkylation pathways to effect significant improvement in lifetime (>70x) of CHA materials for MTH conversion10. These results provide guidance on process conditions to improve lifetime of methanol-to-hydrocarbons catalysis over zeolites and open up the prospect of utilizing packed bed reactor configurations with infrequent regeneration cycles relative to the existing fluidized bed configurations for methanol conversion to desired hydrocarbon products.
5. References
(1) Tian, P.; Wei, Y.; Ye, M.; Liu, Z. ACS Catal. 2015, 5 (3), 1922–1938.
(2) Ilias, S.; Bhan, A. ACS Catal. 2013, 3 (1), 18–31.
(3) Hwang, A.; Prieto-Centurion, D.; Bhan, A. J. Catal. 2016, 337, 52–56.
(4) Svelle, S. et al. J. Am. Chem. Soc. 2006, 128, 14770–14771.
(5) Hwang, A.; Kumar, M.; Rimer, J. D.; Bhan, A. J. Catal. 2017, 346, 154–160.
(6) Arora, S. S.; Bhan, A. J. Catal. 2017, 356, 300–306.
(7) Müller, S. et al. J. Am. Chem. Soc. 2016, 138 (49), 15994–16003.
(8) Martinez-Espin, J. S. et al. Catal. Sci. Technol. 2017, 6, 2314–2331.
(9) Hwang, A.; Bhan, A. ACS Catal. 2017, 7 (7), 4417–4422.
(10) Arora, S. S.; Nieskens, D. L. S.; Malek, A.; Bhan, A. Nat. Catal. 2018, 1 (9), 666–672.