(10a) Controlling Alkene Chain Growth Pathways On Solid Acid Catalysts

Emerging oxygenate feedstocks, such as
ethanol and isobutanol, can be produced from renewable resources and dehydrated
to alkenes by acid catalysts, which can undergo subsequent chain growth also on
catalysts. Product chain length depends on the relative rates of chain oligomerization,
hydride transfer, isomerization and β-scission reactions, which reflect,
in turn, effects of acid strength and confinement on the stability of the
respective transition states.

Weaker acids favor methylation over
isomerization or hydride transfer during dimethyl ether (DME) homologation,
leading to larger chains on zeolites than on stronger acids, such as Keggin-polyoxometalates
[1]. Light alkene conversion on acidic zeolites, however, forms a broad range of
products, consistent with concurrent oligomerization, isomerization, cracking,
hydrogen transfer and aromatization events. Small, undervalued alkanes can also
be incorporated into homologation and oligomerization processes through the use
of molecular hydride transfer co-catalysts, such as adamantane, or
dehydrogenation-hydrogenation metal sites. The relative rates of the
aforementioned events differ among zeolite frameworks with similar acid
strengths but different confining voids as is evident from near-linear
hydrocarbons produced from oligomerization on medium pore zeolites, like MFI
[2]. We probe here the effects of zeolite void structure on alkene
oligomerization to advance our mechanistic understanding about the consequences
of confinement for propagation and termination rates and chain growth
selectivity.

Rates of propene consumption are first
order in propene pressure indicating that at operating conditions, adsorbed
propene is the most abundant surface intermediate. The rate expression for
oligomerization thus simplifies to:

                                                                   (1)

where
k_oligo is the effective rate constant for propene consumption. H-TON
(10-membered ring, one-dimensional) exhibits turnover rates that are an order
of magnitude larger than H-MOR (12-membered ring, one-dimensional), which is
consistent with the dimeric transition state of oligomerization benefitting
from confinement. Comparable turnover rates (within 1.2) are observed on H-MFI
and H-BEA in accordance with similar diameters of the channel intersections in H-MFI
and 12-membered ring channels in H-BEA.

The relative rates of oligomerization
and cracking can be compared through a true oligomer selectivity (χ),
which is defined as the fraction of products observed as propene oligomers (C₆,
C₉, C₁₂). This χ value increases with
increasing alkene partial pressure on all zeolites, consistent with
oligomerization and β-scission proceeding as bimolecular and monomolecular
reactions, respectively (Eq. 2):

                                          
                (2)

Herek_β-n is the rate constant for a carbon species of length n
undergoing β-scission and K₃ and K_n are the
equilibrium constants for adsorption/desorption of a given alkene.

One-dimensional zeolites (H-TON, H-MTT,
H-MOR) exhibit χ values that are 2 times larger than three-dimensional
H-MFI at 77 kPa propene and values that are approximately 3.5 times larger at
26 kPa propene, consistent with Eq. 2. Similar trends are
also observed for three-dimensional H-BEA. In H-MFI and H-BEA, intersections of
the three-dimensional channel network create voids that are 0.1-0.2 nm larger
in diameter than the surrounding channels. These undulations in the diffusional
path increase selectivity to lower carbon numbers because large oligomers
created in these domains cannot diffuse out of the smaller channels and
subsequently undergo β-scission.
Transport through the one-dimensional channels in H-TON, H-MTT and H-MOR reduces
β-scission rates and forces the oligomers to terminate via isomerization
or desorption instead. These results demonstrate the control of alkene chain
growth on solid acids by preventing β-scission pathways during
oligomerization through confinement as a promising route for the production of
higher molecular weight hydrocarbons for fuels.

 

[1]
Simonetti,
D. A., Carr, R. T., Iglesia, E., J. Catal. 285, 19 (2012).

[2]
Garwood,
W. E., ACS Symp. Ser., 218, 383 (1983).

 

Financial
support from BP through the XC² Program and National
Science Foundation Graduate Research Fellowship Programs are gratefully
acknowledged.