Ethylene Dehydroaromatization over Ga-ZSM-5 Catalysts: Nature and Role of Gallium Speciation

Article abstract of DOI:10.1002/anie.202007147

Bifunctional catalysis in zeolites possessing both Br?nsted and Lewis acid sites offers unique opportunities to tailor shape selectivity and enhance catalyst performance. Here, we examine the impact of framework and extra-framework gallium species on enriched aromatics production in zeolite ZSM-5. We compare three distinct methods of preparing Ga-ZSM-5 and reveal direct (single step) synthesis leads to optimal catalysts compared to post-synthesis methods. Using a combination of state-of-the-art characterization, catalyst testing, and density functional theory calculations, we show that Ga Lewis acid sites strongly favor aromatization. Our findings also suggest Ga(framework)–Ga(extra-framework) pairings, which can only be achieved in materials prepared by direct synthesis, are the most energetically favorable sites for reaction pathways leading to aromatics. Calculated acid site exchange energies between extra-framework Ga at framework sites comprised of either Al or Ga reveal a site-specific preference for stabilizing Lewis acids, which is qualitatively consistent with experimental measurements. These findings indicate the possibility of tailoring Lewis acid siting by the placement of Ga heteroatoms at distinct tetrahedral sites in the zeolite framework, which can have a marked impact on catalyst performance relative to conventional H-ZSM-5.

Full text of DOI:10.1002/anie.202007147

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: Ethylene Dehydroaromatization over Ga-ZSM-5 Catalysts: Nature
and Role of Gallium Speciation
Authors: Yunwen Zhou, Hari Thirumalai, Scott K. Smith, Kenton H.
Whitmire, Jing Liu, Anatoly I. Frenkel, Lars C. Grabow, and
Jeffrey D. Rimer
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202007147
Link to VoR: https://doi.org/10.1002/anie.202007147

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
Ethylene Dehydroaromatization over Ga-ZSM-5 Catalysts: Nature
and Role of Gallium Speciation
[
a]
[a]
[b]
[c]
[d]
Yunwen Zhou , Hari Thirumalai , Scott K. Smith , Kenton H. Whitmire , Jing Liu , Anatoly I.
Frenkel[e],[f], Lars C. Grabow*,[a],[b],[g], Jeffrey D. Rimer*,[a],[b]
[
a]
Dr. Y. Zhou, Dr. H. Thirumalai, Prof. L.C. Grabow, Prof. J.D. Rimer
Department of Chemical and Biomolecular Engineering
University of Houston
Houston, TX 77204, USA
E-mails: grabow@uh.edu, jrimer@central.uh.edu
Dr. S.K. Smith, Prof. L.C. Grabow, Prof. J.D. Rimer
Department of Chemistry
University of Houston
Houston, TX 77204, USA
Prof. K.H. Whitmire
Department of Chemistry
Rice University
Houston, TX 77005, USA
Prof. J. Liu
Physics Department
Manhattan College
Riverdale, NY 10471, USA
Prof. A.I. Frenkel
Department of Materials Science and Chemical Engineering
Stony Brook University
Stony Brook, NY 11794, USA
Prof. A.I. Frenkel
Chemistry Division
[
[
[
[
[
[
b]
c]
d]
e]
f]
Brookhaven National Laboratory
Upton, NY 11973, USA
Prof. L.C. Grabow
g]
Texas Center for Superconductivity at the University of Houston (TcSUH)
University of Houston
Houston, TX 77204, USA
Supporting information for this article is given via a link at the end of the document.
Abstract: Bifunctional catalysis in zeolites possessing both Brønsted
and Lewis acid sites offers unique opportunities to tailor shape
selectivity and enhance catalyst performance. Here, we examine the
impact of framework and extra-framework gallium species on
enriched aromatics production in zeolite ZSM-5. In this study we
compare three distinct methods of preparing Ga-ZSM-5 and reveal
direct (single step) synthesis leads to optimal catalysts compared to
post-synthesis methods. Using a combination of state-of-the-art
characterization, catalyst testing, and density functional theory
calculations, we show that Ga Lewis acid sites strongly favor
aromatization. Our findings also suggest Ga(framework) – Ga(extra-
framework) pairings, which can only be achieved in materials
prepared by direct synthesis, are the most energetically favorable
sites for reaction pathways leading to aromatics. Calculated acid site
exchange energies between extra-framework Ga at framework sites
comprised of either Al or Ga reveal a site-specific preference for
stabilizing Lewis acids, which is qualitatively consistent with
experimental measurements. These findings indicate the possibility of
tailoring Lewis acid siting by the placement of Ga heteroatoms at
distinct tetrahedral sites in the zeolite framework, which can have a
marked impact on catalyst performance relative to conventional H-
ZSM-5.
Introduction
Benzene, toluene, and xylene (BTX) are important precursors in
the petrochemical industry for the production of polymers, fuel
additives, and other value-added products.[1] Zeolite ZSM-5 (MFI
type) has been demonstrated to be a superb catalyst for BTX
[
2]
production from different feedstocks, including syngas ,
[
3]
[4]
[5]
methanol , light alkanes and alkenes , due to its unique
medium pore (10-member ring) shape selectivity. Modified
Brønsted acidic H-ZSM-5 catalyst with metals (e.g. Ag, Ga, Zn)
forming Lewis acid sites have demonstrated exceptional
enhanced aromatics yield for non-oxidative dehydroaromatization
(DHA) of ethylene and propylene.[5b, 5c, 6] Moreover, Ga-ZSM-5
has been used in dehydrogenation reactions,[ including the
commercial UOP/BP Cyclar process to convert light alkanes to
aromatics.[8] Despite numerous studies of Ga-zeolites, a general
understanding of active sites and the potential synergy between
Lewis and Brønsted acids is lacking.
7]
In addition to Ga-ZSM-5, prior studies have also
characterized structure-performance relationships for Ga-
exchanged chabazite (CHA) and beta (*BEA) catalysts. There
are two conventional post-synthetic approaches to prepare Ga-
modified zeolites: (1) solution ion exchange with a gallium salt (e.g.
[
9]
Ga(NO
3
)
3
);
and
(2)
incipient-wetness
impregnation.[7a]
Characterization of Ga-zeolites prepared by these methods often
1
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
x y
reveals large fractions of Ga O deposits on external surfaces of
zeolite particles owing to the slow diffusion of hydrated Ga ionic
species into the nanopores.[10] To obtain more dispersed Ga
species within zeolite pores, it has been shown that Ga
reduced under H at high temperature (>700 K) to facilitate Ga
migration into pores and ion exchange with Brønsted acid sites.[
x y
O can be
2
11]
Price et al. investigated the effects of various treatment methods
on the state of Ga (in ZSM-5 catalysts) and the aromatization
activity in DHA reactions.[11b, 12] They observed that the optimal
x y
catalyst was prepared by reduction of Ga O to dispersed Ga
species and reoxidation to obtain Ga3+ sites within the 3-
dimensional pores of ZSM-5; however, it has been hypothesized
that achieving a uniform dispersion of Ga3+ sites by this method is
challenging.[10] Bell and coworkers prepared isolated Ga3+ cations
inside H-ZSM-5 using vapor-phase exchange with GaCl
3
,[7d, 11a]
and investigated the mechanism and kinetics of propane
dehydrogenation over Ga-ZSM-5 catalysts. They report [GaH]2
is the active site for both propane dehydrogenation and cracking
reactions.[7d] Lercher and coworkers used conventional wetness
impregnation to prepare well-characterized Ga-ZSM-5 catalysts
+
Scheme 1. Possible zeolite Brønsted and Lewis acid sites generated from
heteroatom framework (T = Al or Ga) and/or extra-framework species: (A)
for propane dehydrogenation wherein they reported Lewis-
Brønsted acid pairs are essential to catalyze the reaction.[13] They
3
+
+
+
+
+
framework Ga ; (B) Ga
[
x y
O
; (C) Ga ; (D) [GaO] ; (E) [GaH
2
] , [Ga(OH) ] , or
2
GaH(OH)] ; and (F) [GaH]2+.
+
proposed the existence of Ga+ sites that can be protonated by
2
+
proximal Brønsted acids to form [GaH] . Bell and Lercher both
claim [GaH]2+ is the active species for propane dehydrogenation,
though there are differences in their proposed reaction
mechanisms. This active site, however, is more probable for
zeolites with low Si/Al ratio as it requires two anionic framework
Al sites in close proximity.
solid-state nuclear magnetic resonance (NMR) spectroscopy and
X-ray absorption spectroscopy (XAS), they showed that each Ga
species has a tetrahedral coordination and resolved their location
on oxide supports. For Ga-zeolites, the number of potential Ga
species reported in literature (Scheme 1) includes the following:
Ga-ZSM-5 catalysts prepared via direct synthesis have been
shown to enhance aromatics selectivity in alkane
+
+
+
+
+
Ga
x
O
y
2
, Ga , [GaO] , [GaH
2
] , [Ga(OH)
2
+
] , [GaH(OH)] , and
+ [7d, 11b, 18]
[
GaH] .
With the exception of Ga these species contain
dehydrogenation reactions.[11c,
developed
14]
Jones and coworkers
method using synthesis mixtures with 3-
[7b]
Ga in its +3 oxidation state. For direct synthesis, Ga incorporated
in framework sites (Scheme 1A) may also lead to the generation
of Brønsted acid sites.[
a
mercaptopropyl-trimethoxysilane (MPS) to generate gallosilicates
with a high percentage of Ga Lewis acid sites occluded as extra-
framework species into zeolite pores without directly forming
covalent bonds with framework oxygens. The resulting
gallosilicate catalyst (i.e. Ga-MFI in the absence of Al) enhanced
rates of propane dehydrogenation compared to gallosilicates
prepared by conventional methods. Shu and coworkers evaluated
several Ga-ZSM-5 catalysts obtained via different methods (i.e.
direct synthesis, ion exchange, impregnation, and physical
mixing) on n-heptane aromatization.[15] They reported the Ga-
ZSM-5 catalyst synthesized via direct synthesis exhibited the
highest BTX selectivity, which they attribute to enhanced Lewis
acidity and the presumed mesoporosity incurred by the
incorporation of framework Ga. Moreover, they hypothesized that
both liquid-phase ion exchange and direct synthesis lead to Ga
substitution in framework sites. On the contrary, Hsieh et al.[16]
argued that Ga is less prone to be incorporated into the zeolite
framework (compared to Al) based on their analysis of [Ga, Al]-
ZSM-5 catalysts obtained via seeded synthesis.
19]
Herein, we investigate the speciation and siting of Ga in
ZSM-5 catalysts, and correlate the synergy of Brønsted and Lewis
acid sites on the ethylene DHA reaction. We systematically
examine five Ga-ZSM-5 zeolite catalysts prepared via different
treatments, including both direct synthesis and post-synthesis
exchange. Using a combination of experimental techniques,
including XAS and solid-state 27Al and 71Ga magic angle spinning
(
MAS) NMR spectroscopy, we distinguish Ga species with
different oxidation states and coordination numbers. The catalytic
performance of various Ga-ZSM-5 catalysts helps differentiate
the roles of (extra-)framework Ga species. Experimental
characterization and catalytic studies were coupled with density
functional theory (DFT) calculations to assess the relative
activities of different Ga and Al sites. Collectively, these findings
highlight the challenges of elucidating Ga speciation, and the
impact of Ga-ZSM-5 preparation on enhanced activity for the
promotion of aromatics (i.e. BTX products).
Ga species supported on microporous materials can be
difficult to characterize due to the complexity of zeolite topology
and interactions with strong Brønsted acid sites. To this end, Scott
and coworkers systematically examined nonporous catalysts (e.g.
Results and Discussion
Ga-ZSM-5 catalysts were prepared by three distinct methods
of gallium occlusion. The first method is a direct synthesis
adapted from a previous study[7b] that places Ga in either extra-
framework (EFW) or framework (FW) sites. Inductively coupled
plasma analysis in combination with optical emission
spectroscopy (ICP-OES) reveals the product of direct synthesis
Ga(i-Bu)
3
grafted onto -alumina and silica) for propane
dehydrogenation wherein they showed that mononuclear
alumina-supported Ga sites exhibit higher activity than silica-
[17]
supported Ga sites; and using a combination of high-field
2
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
contains a molar composition of Si:Al:Ga = 57:2.4:1.0. This
sample is referred to herein as Ga-Z1. The second and third
methods involved post-synthesis modifications to a commercial
ZSM-5 (Zeolyst CBV 5524G, referred to as Z-ref) with nearly
identical molar Si/Al ratio as that of Ga-Z1 (Table 1). In one
sites based on shifts in two characteristic MFI peaks, (501) and
(303), when a larger Ga atom replaces a smaller Al atom. For all
H-Ga-ZSM-5 samples in this study, these two peaks (at 2θ = 23.0°
and 23.8°, respectively) exhibit no observable shifts compared to
commercial zeolite, Z-ref (Figure S3). This seems to suggest the
majority of Ga may be present as extra-framework species.
X-ray absorption fine structure (XAFS) spectroscopy was
used to characterize the changes in the local structure and
electronic state of Ga between samples Ga-Z1 and Ga-Z2 (a
similar analysis of Ga-Z3 has been reported by Phadke et al.[
and is repeated in the Supporting Information). X-ray absorption
near edge structure (XANES) data at the Ga K-edge (Figure 1A)
show that the absorption edge energy is very similar between Ga-
Z1 and Ga-Z2, while the main peak region is more broadened for
Ga-Z2. The absorption edge energy is defined here as the first
inflection point of the XANES spectrum and is related to the
charge state of the probed atoms. The charge state of Ga atoms
in both samples is, effectively, +3 since the position of the
absorption edge energy is at ~10376 eV for both samples, similar
approach we used solution ion exchange with Ga(NO
produce a product, referred to as Ga-Z2. In the second approach
we used vapor-phase exchange with GaCl according to a
3 3
) to
3
protocol reported by Bell and coworkers[11a] (referred to as sample
Ga-Z3). The molar compositions of samples Ga-Z2 and Ga-Z3
are both equal to Si:Al:Ga = 54:1.8:1.0.
11a]
The solution ion exchange method reportedly leads to a large
fraction of Ga
x
O
y
species deposited on the external surfaces of
ZSM-5.[10, 20] Using a previous approach
[18a]
to obtain more
dispersed Ga species, samples Ga-Z1 and Ga-Z2 were treated in
at 848K and are referred to herein as Ga-Z1H and Ga-Z2H,
respectively (note that the protocol for Ga-Z3 also involves H
H
2
2
treatment at 823K). The aluminum speciation for each sample
was quantified using solid-state NMR (Figure S1) where peaks at
2
7
3+
[11a]
5
5 ppm and 0 ppm in Al MAS NMR spectra correspond to
to the value of the reported standard Ga species.
The
tetrahedral AlFW (framework Al) and octahedral AlEFW (extra-
framework Al), respectively.[21] The percentages of Al species for
each sample are listed in Table 1, showing that Al speciation is
variation between XANES features in these two H-Ga-ZSM-5
samples is evidence of differences in their local environments.
2
not altered by either solution ion exchange or H treatment;
however, there is a 9% reduction in AlFW (with concomitant
increase in AlEFW) during Ga impregnation by chemical vapor-
phase exchange, as observed for Ga-Z3 (compared to the parent
Z-ref). This is likely due to the release of HCl during the
decomposition of GaCl
the zeolite framework.
3
, which could facilitate dealumination of
Table 1. Elemental analysis of H-zeolite catalyst samples.
Composition[a]
Al content (%)[b]
Sample[d]
Sample Preparation
Si/Al
24
Si/Ga
57
AlFW
89
AlEFW
11
6
Ga-Z1
Ga-Z2
Ga-Z3
Ga-Z1H
Ga-Z2H
Z-ref
Direct synthesis
Solution ion exchange
Vapor phase exchange
30
54
94
30
54
85
15
9
Z1 with H
Z2 with H
2
2
treatment
treatment
24
57
91
31
54
94
6
Commercial zeolite
29[c]
----
94
6
[
a] Determined by ICP-OES; [b] Measured by 27Al MAS NMR spectra of
hydrated zeolite materials (Figure S1); [c] Value of commercial zeolite CBV
524G (Zeolyst) reported by Phadke et al.[11a]; [d] SEM images of samples
reveal a spheroidal morphology with average particle sizes in the range of 300
5
–
500 nm (Figure S2).
Powder X-ray diffraction (PXRD) patterns confirmed the
structural integrity (MFI framework) of all calcined H-Ga-ZSM-5
catalysts (Figure S3) and the absence of impurities in Ga-Z1. It
Figure 1. (A) Normalized Ga K-edge XANES spectra for H-Ga-ZSM-5 zeolites
[
10]
prepared by different methods: Ga-Z1 (direct synthesis) and Ga-Z2 (solution ion
has been reported that solution ion exchange can lead to the
formation of crystalline Ga that is not visible in the PXRD
2
exchange). (B) Fourier transform magnitudes of k -weighted Ga K-edge EXAFS
O
x y
-1
spectra of Ga-Z1 and Ga-Z2 (using the k range: 2-12 Å ). Peaks are labelled
pattern of Ga-Z2; however, samples prepared by ion exchange
with higher Ga concentrations (Table S1) clearly show the
with Ga species corresponding to the fit results in Tables 2 and S2. All the
measurements are taken at ambient temperature.
presence of crystalline Ga
x
O
y
by PXRD (i.e. peaks at 2θ = 21.6°
and 37.1° in Figure S4).[22] It has also been reported that PXRD
[16]
patterns can provide evidence of Ga incorporation in framework
3
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
2
Fourier transform magnitudes of the k -weighted extended X-
ray absorption fine structure (EXAFS) spectrum (Figure 1B) of
Ga-Z2 features a shoulder near the 2 Å distance that is absent for
the Ga-Z1 sample. Moreover, the first peak between 1 and 2 Å in
the spectrum of Ga-Z2 is slightly shifted to higher distance
compared to that of Ga-Z1. These observations, together with the
large signal in the region of 2-4 Å in the spectrum of Ga-Z2,
indicate the local environments around Ga atoms in these two
zeolites are different. In particular, we observe better defined
structure of more distant atomic species around Ga in Ga-Z2
159 and 102 ppm, respectively. Since EXAFS suggests 80% of
the Ga sites in Ga-Z2 are tetrahedrally coordinated, the peak at
102 ppm could be assigned to distorted tetrahedral sites within
extra-framework Ga or amorphous Ga
peak located at 102 ppm disappears due to the formation of
dispersed tetrahedral [Ga(X)
]+ (X = OH or H) species that can
x y 2
O . After H treatment, the
2
exchange with Brønsted acid sites. Indeed, the NMR spectrum for
Ga-Z2H (Figure 2B) is similar to that of Ga-Z3 (i.e. single peak
around 147 ppm), which is consistent with tetrahedrally-
coordinated Ga reported by Bell.[11a] Direct comparison of these
two spectra with that of Ga-Z1 provides evidence for the presence
of framework Ga in the latter (although the exact percentage
cannot be quantified). The presence of framework Ga in sample
Ga-Z1 (and Ga-Z1H) is inferred from the large difference in signal-
to-noise ratio compared to Ga-Z2 and Ga-Z3, despite all samples
having nearly identical Ga concentration (Table 1). The presence
of tetrahedrally-coordinated GaFW leads to higher local symmetry
x y
compared to Ga-Z1, consistent with the hypothesis that Ga O
species dominate in the former sample, while Ga is in the zeolite
framework in the latter one. To extract detailed local structure
information in the both samples, quantitative analysis of their
EXAFS spectra was carried out. The fitting results (Table 2 and
Figure S5) highlight distinct differences in the local environments
around Ga atoms in these samples, in agreement with visual
observations of XANES and EXAFS data (vide supra). The
coordination number of the Ga-O bond in Ga-Z1 is close to four,
consistent with a tetrahedral-coordination of Ga atoms with an
effective Ga-O distance of 1.81 ± 0.01 Å. This value falls within
the range reported in literature[11a] and is close to the vales
measured for Ga-Z2 and Ga-Z3 (Table S2).The best fit for Ga-Z2
involves two distinct Ga-O shells with interatomic distances of
(i.e. higher peak intensity, Figure S6), consistent with previous
71Ga NMR spectra of gallosilicates,[24a,
asymmetrical environment of tetrahedral [Ga(X)
to lower signal-to-noise ratios.
26]
whereas the
]+ species leads
2
1.90 ± 0.02 Å and 2.11 ± 0.08 Å. Such a split in the Ga-O
distances is likely due to the distribution of Ga atoms between
tetrahedral and octahedral sites.[23] We estimate the fraction of
tetrahedral Ga sites in Ga-Z2 to be 80% (refer to the Supporting
Information for detailed calculations). The peak at about 2.97 Å is
attributed to Ga-Ga pairs.
Table 2. Best fits of the structural parameters obtained from Ga K-edge EXAFS.
2
2
Sample
Pair
N
R (Å)
σ (Å )
Ga-Z1
Ga-Z2
Ga-O
4.2 ± 0.4
3.4 ± 0.9
1.3 ± 0.8
8.5 ± 3.8
1.81 ± 0.01
1.90 ± 0.02
2.11 ± 0.08
2.97 ± 0.03
0.005 ± 0.001
0.006 ± 0.004
0.006 ± 0.004
0.017 ± 0.005
Ga-O1
Ga-O2
Ga-Ga
The effect of Ga-ZSM-5 sample preparation on the
coordination (or local environment) of Ga species was assessed
by solid-state NMR. Previous studies of 71Ga MAS NMR spectra
generally assign peaks in the ranges 150-160 ppm and -7-12 ppm
to tetrahedral (framework) GaFW and octahedral (extra-
[
24]
framework) GaEFW (i.e. Ga
x
O
y
), respectively.
However, an
[
25]
alternative interpretation from Dai et al. is the signal at 150-160
ppm is a tetrahedral coordination attributed to either GaFW cations
3
+
+
or GaEFW species such as Ga , [Ga(OH)
2
] , or [GaO(OH)].
Moreover, Gao et al.[3a] assigned the peak at 58 ppm to highly
Figure 2. (A) 71Ga MAS NMR spectra of hydrated Ga-ZSM-5 zeolites prepared
via direct synthesis (Ga-Z1) and after H treatment (Ga-Z1H). (B) Spectra of
hydrated Ga-ZSM-5 zeolite samples via other preparation methods: Ga-Z2
solution ion exchange), Ga-Z2H (Z2 after H treatment), and Ga-Z3 (chemical
dispersed Ga
[
x
O
y
particles, and the peak at 190 ppm to cationic
2
GaO]+ or its hydrous state [Ga(OH)
]+. The 71Ga MAS NMR
2
(
2
spectra of all Ga-ZSM-5 samples in this study are shown in Figure
. Samples Ga-Z1 and Ga-Z1H exhibit a single peak around 159
vapor-phase exchange). See Figure S6 for an overlay of all spectra.
2
ppm (Figure 2A), indicating the majority of Ga species have a
tetrahedral coordination. For the remaining samples (Figure 2B),
the signal-to-noise ratio of spectra is relatively low due to the
strong broadening effect on the oxidic extra-framework Ga
species.[24c] The Ga-Z2 spectrum contains two bands located at
The 71Ga MAS NMR spectra of samples Ga-Z1, -Z1H, -Z2H,
and –Z3 indicate a predominance of tetrahedral Ga species,
which could be either [GaX
2
]+ (Scheme 1E) or [GaH]2+ (Scheme
4
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
1
F). The relatively high Si/Al ratio of H-ZSM-5 samples suggests
Figure 3A compares product selectivities for specific light
paraffins and olefins where the most notable differences observed
among the catalysts are with respect to propylene and C olefins.
4
there are few paired Al sites, which would limit the number of
[
GaH]2+ species; thus, we propose the majority of extra-
+
framework sites are [GaX
Ga(OH)
2
] . Direct synthesis of Ga-Z1 generates
Comparisons were made at a fixed space velocity, where we
observe nearly identical ethylene conversion at initial time on
stream (Figure 3B), indicating all six catalysts have similar activity.
Comparison of the rates of catalyst deactivation reveal two groups
of similar stability: those with lower rates of deactivation (Ga-Z2,
Ga-Z2H, and Z-ref) and those with 3.4-fold higher rates of
deactivation (Ga-Z1, Ga-Z1H, and Ga-Z3). With respect to the
accelerated rates of deactivation for Ga-Z1 and Ga-Z3, this may
be attributed in part to their higher percentages of extra-
framework Al (Table 1). For Ga-Z3 these species are the result of
[
2
]+ in the as-synthesized product. High temperature
hydrogen treatment of sample Ga-Z1H is expected to result in
sequential hydrogen exchange steps leading to the overall
reaction
+
+
[
퐺푎(푂퐻) ] ꢀ ꢁ퐻 ↔ [퐺푎퐻 ] ꢀ ꢁ퐻 푂
(1)
2
2
2
2
Comparison of XAS data for Ga-Z1 and Ga-Z1H (Figure S7)
reveals no shift in spectra. A similar analysis of Ga-Z2 and Ga-
Z2H (Figure S8) reveals only subtle differences in XAS spectra
which may reflect the loss of distorted tetrahedral sites (consistent
with 71Ga MAS NMR). Treatment of Ga-Z2 with H
potential to mobilize Ga
via a similar reaction,
2
dealumination during H treatment, which leads to framework
defects that are evident in XAS analysis (Figure S9).
2
also has the
to generate dispersed [GaH
]+ sites
O
x y
2
+
퐺푎 푂 ꢀ (ꢂ ꢀ ꢃ)퐻 ↔ ꢂ[퐺푎퐻 ] ꢀ ꢃ퐻 푂
(2)
푥
푦
2
2
2
An additional hydrogen exchange generates a Brønsted acid site
and GaH , which we consider to be the short-lived, mobile species
3
during hydrogen treatment. DFT calculated free energy changes
and activation barriers for the sequential transformation of
2 3
Ga[OH) are provided in Table S3. The final step of
]+ to GaH
[
treatment involves exposure to O
2
at high temperature, which
+
reoxidizes sites to [Ga(OH)
2
] . During catalyst testing, non-
2 2
oxidative DHA of ethylene produces H to regenerate [GaH
]+
species in situ, which we presume herein to be the active site for
Lewis acid catalysis.
The catalytic performance of all H-Ga-ZSM-5 catalysts were
compared in the ethylene DHA reaction using a packed bed
reactor with a fixed space velocity (3000 ml g-1 h-1) and ethylene
partial pressure (12.7 kPa). Reactions were performed at sub-
complete ethylene conversion (ca. 80%) to assess catalyst
deactivation and selectivity towards light olefins and aromatics.
As shown in Figure 3A, isoconversion comparisons reveal the
following order of aromatic selectivity: Ga-Z1H > Ga-Z1 ≈ Ga-
Z2H > Ga-Z3 > Ga-Z2 > Z-ref. Commercial ZSM-5 (Z-ref), which
lacks GaEFW Lewis acid sites (LASs), exhibits the least aromatic
selectivity (16%) owing to the presence of Brønsted acid sites
(
1
BASs) that preferentially direct the formation of alkanes (e.g.
3% butane) and olefins (e.g. 23% propylene, 19% butylenes).
Among H-Ga-ZSM-5 catalysts, Ga-Z2 has the lowest aromatics
selectivity owing to the presence of distorted [GaH
]+ sites or
Ga species on the zeolite surface, which are reportedly inert
2
x y
O
[
10]
sites for aromatic formation. The Ga-Z3 catalyst obtained via
chemical vapor phase exchange produced an appreciable
amount of aromatics (45%), suggesting the existence of well-
+
dispersed GaEFW species (i.e. isolated [GaH
2
] sites substituting
[11a]
Brønsted acids as illustrated in Scheme 1E with T = Al) . The
catalyst prepared via direct synthesis (Ga-Z1) exhibits higher
aromatics selectivity (52%), which can be correlated to the
presence of tetrahedrally-coordinated Ga3+ species. Analyses by
XAS and 71Ga MAS NMR cannot distinguish the fraction of
tetrahedrally-coordinated Ga occupying framework and extra-
framework sites, which are illustrated in Scheme 1A and 1E,
respectively. One differentiating factor between samples Ga-Z3
and Ga-Z1 is that GaEFW in the latter can be associated with either
AlFW or GaFW in the framework. Another factor that must be
considered is the potential impact of framework Ga on the acidity
of BASs. Investigation of both scenarios by DFT modeling is
described later.
Figure 3. (A) Hydrocarbon product selectivity in the ethylene DHA reaction over
various H-form ZSM-5 catalysts (Ga-Z1, Ga-Z1H, Ga-Z2, Ga-Z2H, Ga-Z3, and
Z-ref) at 400 °C, 1 atm, and isoconversion of ethylene (ca. 80%). The space
-1
-1
velocity was fixed at 3000 ml g
h using a feed pressure of 12.7 kPa ethylene
and 88.6 kPa argon. Data were acquired after 10 min time on stream (TOS) with
a carbon balance > 96%, which is listed in Table S4 for each catalyst along with
detailed product selectivity for a broader class of compounds in the effluent
stream. Changes in product selectivity with increasing TOS are shown in Figure
S10. (B) Ethylene conversion for each catalyst at the same reaction conditions
as a function of TOS. Dashed lines are linear regression showing two groups
with lower and higher rates of deactivation (slopes with absolute values of 0.7 ±
-
1
0
.2 and 2.3 ± 0.5 h , respectively).
5
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
Hydrogen treatment leads to significant increases in
aromatics production for both Ga-Z1 and Ga-Z2. Prior studies
underlying ethylene DHA. To examine differences among
heteroatom sites, we explored the activation and oligomerization
of ethylene to butene (a characteristic product of the DHA
reaction) by means of periodic DFT calculations with van der
Waals (vdW) corrections. To this end, we selected the T12 site of
ZSM-5 as a representative BAS (and one of the most frequently
studied in literature), located at the intersection of straight and
sinusoidal channels in the MFI structure.[ The free energy
diagram depicting the energies of reaction and activation barriers
for each intermediate step for AlFW in orange and GaFW in blue at
400°C and standard pressure, are shown in Figure 5A.
Visualizations of the initial, transition, and final states along the
reaction coordinate are provided in Figure S11.
have shown that the reduction of Ga-zeolites in H
2
at high
temperature is essential to reduce Ga (Eq. 2) to more
x y
O
dispersed Ga species.[11b, 12] It has also been reported that
reoxidation to the Ga3+ phase after dispersion can further
enhance aromatic selectivity, indicating reoxidized Ga differs from
28]
the initial Ga
Ga-Z2 (containing Ga
than Ga-Z2H (after H
x
O
y
species. This is consistent with the observation of
species) producing fewer aromatics
treatment). Indeed, the BTX selectivity
x
O
y
2
increases from 28 to 52% (Figure 4), indicating a large shift to
well-dispersed Ga3+ species, consistent with its NMR spectrum
(
Figure 2B). It was anticipated that Ga-Z3, which also contains
3
+
2
Ga species (Figure S9 and Table S2) and was treated with H ,
would perform similarly to Ga-Z2H; however, we observe the
former produces 8% less BTX. This may be attributed, in part, to
the loss of framework Al in Ga-Z3 during chemical vapor phase
Figure 4. Comparison of aromatics products for H-Ga-ZSM-5 catalysts before
(
Ga-Z1, Ga-Z2) and after (Ga-Z1H, Ga-Z2H, Ga-Z3) H
the reference H-ZSM-5 catalyst (Z-ref). Here we report the BTX selectivities
histogram, left y-axis) and ethylene conversion (square symbols, right y-axis)
for the ethylene DHA reaction at 400 °C and 1 atm. The space velocity was fixed
2
treatment compared to
(
-
1
-1
at 3000 ml g h , and the feed pressure was composed of 12.7 kPa ethylene
and 88.6 kPa argon. Data were acquired after 10 min time on stream.
exchange. An even more unexpected observation was the fact
that H
2
treatment of Ga-Z1 raises its BTX selectivity from 52 to
6
8% (Figure 4, sample Ga-Z1H), the highest among all samples
7
1
Figure 5. Calculated properties of H+ Brønsted acid sites and [GaH ]+ Lewis
2
tested. The Ga MAS NMR spectra for samples before and after
treatment are identical (Figure 2A). Moreover, 27Al MAS NMR
reveals no loss of framework Al species with H
acid sites near an AlFW (orange) or GaFW (blue) heteroatom at the T12 site of
MFI. (A) Free energy diagram for ethylene activation and oligomerization to 2-
2
treatment, and
+
+
+
butene at an H site, and 1-butene at a [GaH ]
2
s
i
t
e
(
s
i
m
i
l
a
r
c
a
l
c
u
l
a
t
i
o
n
s
f
o
r
G
a
only a small (ca. 2%) reduction in the amount of extra-framework
Al (Table 1). The exact reasoning for the large increase in
aromatics yield is not fully understood, but herein we put forth a
hypothesis invoking preferential siting of BAS and LAS near AlFW
and GaFW, noting that Ga-Z1 is the only sample that features GaFW
sites.
are provided in Figure S12). (B) Binding free energies of unsaturated
hydrocarbons (i.e. olefins and aromatics) at [GaH
]+ Lewis acid sites in the T4
triangles) and T12 (circles) positions (corresponding calculations at Brønsted
acid sites are shown in Figure S13.). All free energies are reported at 400ºC and
standard pressure for each gas phase species.
2
(
Ethylene can potentially be activated on a BAS or LAS, [6d]
and we present the most likely pathways for each case at the T12
site in Figure 5A (additional pathways explored by DFT are
compiled in Table S5). Despite challenges in measuring[27] or
calculating[28] intrinsic acid strengths, it has been experimentally
and theoretically shown[29-30] that AlFW results in a BAS with
stronger intrinsic acidity than GaFW; therefore, the presence of
GaFW in sample Ga-Z1 is expected to weaken the intrinsic
strength of BASs, which can impact the catalytic mechanisms
At 400°C, ethylene is stabilized at the vicinity of the BAS
through vdW forces and π-bonds with a total free binding energy
of 0.06 eV for AlFW and 0.10 eV for GaFW. The activation of
ethylene at a BAS proceeds with the formation of a positively-
+) species (Figure S11, top) with free
charged ethyl (C H
2 5
activation barriers of 1.54 eV (AlFW) and 1.68 eV (GaFW). Next, a
second ethylene molecule attacks the unsaturated carbon atom
of the ethyl group and proceeds through the formation of a π-
+) complex (Figure
bonded transition state, forming a butyl (C H
4 9
6
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
S11, middle). The free activation barrier for this step was
calculated to be 1.07 eV (AlFW) and 1.45 eV (GaFW) and is followed
by an exergonic internal hydrogen transfer step. The formation of
suggest the presence of [GaH
2
]+ in sinusoidal channels alters
diffusion characteristics, which in turn can impact shape
selectivity by directing transport largely through less tortuous
straight channels.
2-butene and restoration of the BAS occurs by β-hydrogen
elimination with an activation barrier of 0.16 and 0.31 eV for AlFW
and GaFW, respectively (Figure S11, bottom). We also considered
the formation of 1-butene with AlFW, which was only marginally
more activated. Overall, each BAS provides an accessible
pathway to the activation and oligomerization of ethylene to
butene; however, the barriers of all elementary steps are
somewhat lower when the framework heteroatom is Al instead of
Ga, suggesting that the BAS bound to AlFW is more amenable to
oligomerization or hydrocarbon upgrade reactions. This result is
consistent with the higher intrinsic acidity of H-(Al)-ZSM-5
compared to H-(Ga)-ZSM-5.[30-31]
2
During H treatment we noted that extra-framework GaEFW
becomes mobile and can be atomically dispersed to increase
aromatics selectivity.[11] One of the unexplained observation in
Figure 4 is the dramatic enhancement of aromatics selectivity
over Ga-Z1 after H
enhanced BTX selectivity is due to a redistribution of extra-
framework [GaH
]+ among available framework AlFW and GaFW
2
treatment. Here we hypothesize that the
2
sites. While our DFT results are inconclusive regarding the role of
framework heteroatoms on selectivity, we aim to rationalize the
substantial selectivity improvement observed for Ga-Z1H by
calculating the free energies associated with swapping a proton
+
Zeolite samples Ga-Z1, -Z2, and -Z3 all contain GaEFW sites
with Lewis acidity that affects the overall catalytic activity and
product selectivity. To examine the role of GaEFW we built on the
at a GaFW framework site (labelled as T
i
) with [GaH
2
] at an Al site
(labelled as T
j
), which is expressed in the following reaction
work by Bell and coworkers[18e] to study the oligomerization of
퐻+ − 퐺푎퐹푊(푇 ) ꢀ [퐺푎퐻 ] − 퐴푙 ꢄ푇 ꢅ →
+
(3)
푖
+
2
퐹푊
+
푗
+
ethylene to 1-butene over [GaH
2
] sites bound to framework AlFW
[퐺푎퐻 ] − 퐺푎 (푇 ) ꢀ 퐻 − 퐴푙퐹푊ꢄ푇 ꢅ.
2
퐹푊
푖
푗
and GaFW anionic sites. The reaction free energies and activation
]+ to be the more likely GaEFW species
Here, we considered [GaH
2
under hydrogen treatment conditions. The ion-exchange energies
barriers of the accessible pathway for ethylene oligomerization for
+
[
GaH
2
] (dotted line) are visualized in Figure 5A. For comparison
for all possible T-site combinations are summarized in Figure 6,
+
we also include an analysis of Ga species, as well as various
other inaccessible pathways we considered for the activation of
ethylene in the Supporting Information (Figure S12 and Table S5).
+
where blue rectangles indicate a preference for a [GaH
2
]
LAS
located near framework GaFW and a strong BAS near AlFW. This
arrangement is generally preferred for GaFW in the T1, T2, T6, T7,
T8, and T11 positions. In contrast, the almost exclusively red
rectangles for GaFW at T10, and their prevalence in the rows for
GaFW at T5, T9, and T12 positions suggests that a BAS is more
likely on GaFW in these locations. The corresponding calculated
values are provided in Table S8.
The activation of ethylene on [GaH
intermediate, resembling the activation mechanism on the BAS,
but with a lower ethylene activation free energy on the [GaH
]+
2 2 5
]+ results in a [C -GaH]+
H
2
LAS. For instance, calculations with AlFW reveal an activation
barrier of only 0.62 eV at the LAS compared to 1.54 eV at the BAS.
In the case of GaFW, the free barrier at the LAS is only 0.61 eV,
but 1.68 eV at the BAS. For both framework heteroatoms, the
+
coupling to a second ethylene molecule to form [C
4
H
9
-GaH] is
exergonic and fast. Following this low energy pathway to C-C
+
coupling on [GaH
2
] is a -hydride elimination step restoring the
LAS and forming 1-butene. The free energy barriers we found
were 2.26 eV for AlFW and 2.10 eV for GaFW. Given the large
barrier of this final step, one might argue that it is rate-limiting. In
that case, our results suggest that GaFW is beneficial, given that
the barrier to restore the GaFW-BAS is 0.16 eV lower than
restoring the AlFW-BAS. This is again consistent with AlFW-BAS
having a higher intrinsic acidity, and therefore rendering GaFW the
stronger conjugated base.
Extra-framework GaEFW may play a critical role in increasing
the selectivity to aromatics,[6d] but since complete aromatization
pathways are complex and highly branched, it is not possible to
rigorously examine them with DFT. Instead, we have calculated
the binding free energies of various olefins and aromatics to BAS
+
]+) acid site exchange energies
2
Figure 6. Brønsted (H ) and Lewis ([GaH
between framework GaFW at T
i
sites and AlFW atoms at T
j
sites at 500°C
+
2
according to the reaction in Eq. 3. Blue rectangles indicate that the [GaH ]
Lewis acid is more stable near a GaFW and the Brønsted acid site forms near
AlFW. For combinations with white rectangles the preference for [GaH
near GaFW or AlFW is negligible.
and [GaH
2
]+ extra-framework sites located at AlFW and GaFW
+
2
]
to form
heteroatoms in the T12 and T4 positions (numerical values in
Tables S6 and S7) representing T sites of the MFI crystal
structure located within straight (i.e. intersection) and sinusoidal
channels, respectively. These unsaturated hydrocarbons shown
in Figure 5B have been detected in catalytic experiments (Figure
It is plausible that mobilizing GaEFW species during H
2
treatment can indeed alter the BAS/LAS site distribution, but
without knowledge of the exact distribution of framework AlFW and
GaFW sites in sample Ga-Z1, it is not possible to draw a firm
conclusion regarding the BAS/LAS distribution; however, the
calculated BAS/LAS site preferences along with the
demonstrated improvement of ethylene conversion and selectivity
3
), and their binding free energies show no significant sensitivity
to the identity of the framework heteroatom. We observed,
however, that binding to the [GaH
]+ site in the T4 position located
2
within the sinusoidal channel is highly unfavorable. In contrast,
the difference between T4 and T12 when only a BAS is present,
is much less pronounced (Figure S13). Hence, these calculations
2
to aromatics of Ga-Z1H support the hypothesis that H treatment
7
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
of Ga-Z1 leads to a beneficial redistribution of BAS/LAS sites. It
should also be noted that this is fully consistent with 71Ga MAS
NMR and XAS data (Figures 2A and S7, respectively) showing no
suggesting that these species are not the most active Lewis sites
for aromatics promotion.
We further explore the effect of H
Ga-Z2 and directly compare their selectivity with Ga-Z3, which
was obtained via chemical vapor-phase exchange involving H
2
treatment on Ga-Z1 and
apparent change in spectra before and after H
suggests that LAS speciation remains the same.
An underlying characteristic of catalysts Ga-Z1, -Z2H, and
Z3 is the presence of [GaH
]+ extra-framework LAS species,
2
treatment, which
2
treatment. For both treated Ga-Z1H and Ga-Z2H catalysts, we
observed increased BTX selectivity, but the improvement was
–
2
which is a likely explanation for their nearly identical performance
in DHA of ethylene. LAS/GaFW combinations can only be achieved
in Ga-Z1 (the direct synthesis method) given its distinguishing
attribute of having GaFW heteroatoms substituted at framework
sites. These sites may ultimately be responsible for improved BTX
selectivity, but conclusive evidence remains elusive. The notable
attributed to two different phenomena: For Ga-Z2 the H
2
treatment
+
disperses inactive Ga to active [GaH
the H
to AlFW and [GaH
2
] species, while for Ga-Z1
treatment is believed to redistribute Brønsted acid protons
]+ to select GaFW sites. The Brønsted and Lewis
2
2
acid site redistribution is supported by DFT simulations, which
+
also suggest that stronger H /AlFW sites lower barriers for ethylene
improvement of Ga-Z1 performance after H
Ga-Z1H) can be rationalized by the calculated differences in
stability for forming BAS and LAS sites near AlFW and GaFW sites.
2
treatment (sample
dimerization over BAS, while the stronger conjugated base sites
with GaFW may favor slow deprotonation steps. DFT calculations
also indicate possible changes in diffusion characteristics with
placement of GaEFW in sinusoidal channels, which may alter
shape selectivity. Jointly, these results would explain the
exceptional activity and selectivity to aromatics observed for the
Ga-Z1H sample. To our knowledge, few studies have considered
the possibility of directing Lewis acid siting in zeolites (i.e. the
majority focus on AlFW siting and distribution[34]). Our findings
indicate that if heteroatoms (such as Ga) can be selectively
placed at specific T sites within zeolites, this would afford the
opportunity to direct the spatial distribution of Lewis acid sites for
bifunctional catalysis.
It is plausible that the high temperature of H
2
treatment coupled
with the mobility of [GaH
2
]+ species can lead to this exchange of
sites; however, in order for this to be true, it implies that direct
synthesis of Ga-Z1 places an appreciable fraction of extra-
framework GaEFW at locations that are on average less
energetically favorable. Although the exact location and quantity
of framework and extra-framework Ga sites cannot easily be
resolved, it is frequently reported that zeolite synthesis produces
metastable, or kinetically trapped, structures; and with sufficiently
high temperature it is possible to promote interzeolite
transformations from the initial structure to one that is more
thermodynamically stable.[33] A similar phenomenon may occur
for GaEFW species during H
2
treatment to result in a more
Acknowledgements
energetically favorable distribution of sites. This hypothesis,
although unverifiable by the techniques used in this study, is
consistent with experimental and modeling results, and thus is a
viable explanation for the superior performance of Ga-Z1H.
JDR and LCG received financial support from the National
Science Foundation (Award CBET-1512224). JDR received
additional support from the Welch Foundation (Award E-1794).
XAS analysis work of AIF was supported in part by the Catalysis
Center for Energy Innovation, an Energy Frontier Research
Center funded by the U.S. Department of Energy, Office of
Science, Office of Basic Energy Sciences under Award No. DE-
SC0001004. JL was supported in part by the U.S. Department of
Energy, Office of Science, Office of Workforce Development for
Teachers and Scientists (WDTS) under the Visiting Faculty
Program (VFP). This Research used beamline 7-BM (QAS) of the
National Synchrotron Light Source II, a U.S. DOE Office of
Science User Facility operated for the DOE Office of Science by
Brookhaven National Laboratory (BNL) under Contract No. DE-
SC0012704. We acknowledge Dr. S. Ehrlich for collecting XAS
data at the QAS beamline and Dr. N. Marinkovic for help with
beamline measurements. Beamline operations were supported in
part by the Synchrotron Catalysis Consortium (U.S. DOE, Office
of Basic Energy Sciences, Grant No. DE-SC0012335). This work
used the Extreme Science and Engineering Discovery
Environment (XSEDE) Stampede2 at the Texas Advanced
Computing Center through allocation TG-CHE140109, which is
supported by National Science Foundation grant number ACI-
Conclusion
We present a direct synthesis method to introduce well-dispersed
extra-framework Ga3+ Lewis acid sites without the need for
additional post-synthetic treatments (i.e. solution ion exchange or
chemical vapor deposition).
A combination of advanced
characterization techniques such as XAS and solid-state NMR
spectroscopy indicates that framework GaFW and tetrahedrally-
coordinated extra-framework GaEFW sites can be simultaneously
generated via this direct synthesis method. We also investigate
the roles of Ga speciation in ZSM-5 prepared by three distinct
methods, direct synthesis (Ga-Z1), solution ion exchange (Ga-Z2),
and chemical vapor-phase exchange (Ga-Z3), to identify the
structure and catalytic function of different Ga sites. We have
shown that any type of GaEFW Lewis site can promote the
formation of BTX with a total selectivity as high as 57%, whereas
the aromatics selectivity is only 16% for commercial H-ZSM-5 at
comparable ethylene conversion. Isoconversion comparisons
further reveal the following order of aromatic selectivity: Ga-Z1 >
Ga-Z3 > Ga-Z2, indicating the Ga Lewis sites and the existence
of framework GaFW species in Ga-Z1 catalyst lead to superior
catalytic performance and promote the formation of aromatics.
XAS and NMR measurements both indicate Ga-Z2 prepared via
solution ion exchange contains a large amount of either distorted
1
548562.[35] We also used resources of the National Energy
Research Scientific Computing Center (NERSC), U.S.
a
Department of Energy Office of Science User Facility operated
under Contract No. DE-AC02-05CH11231, and the uHPC cluster
managed by the University of Houston and acquired through NSF
Award Number 1531814. Finally, we acknowledge the use of the
]+ or Ga
[
GaX O species on the external zeolite surface,
2 x y
8
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
M. Hensen, J. Catal. 2006, 239, 478-485; d) V. B. Kazansky, I.
R. Subbotina, R. A. van Santen, E. J. M. Hensen, J. Catal.
Maxwell/Opuntia/Sabine Cluster and HPC support from Keith
Crabb and Jeffrey Sarlo at the University of Houston.
2005, 233, 351-358; e) E. Mansoor, M. Head-Gordon, A. T.
Bell, ACS Catal. 2018, 8, 6146-6162.
Keywords: Lewis acid • zeolite synthesis • aromatics • acid
siting • bifunctional catalysis
[19] L. Meng, X. Zhu, B. Mezari, R. Pestman, W. Wannapakdee,
E. J. M. Hensen, ChemCatChem 2017, 9, 3942-3954.
[
20] I. Nowak, J. Quartararo, E. G. Derouane, J. C. Védrine, Appl.
Catal. A 2003, 251, 107-120.
[
1]
a) H.-G. Franck, J. W. Stadelhofer, Industrial Aromatic
Chemistry, Spring, 1988; b) K. W. H.-J. Arpe, Industrial
Organic Chemistry, Fourth Edition, Wiley-VCH, 2008.
X. Yang, X. Su, D. Chen, T. Zhang, Y. Huang, Chinese J.
Catal. 2020, 41, 561-573.
a) P. Gao, Q. Wang, J. Xu, G. Qi, C. Wang, X. Zhou, X. Zhao,
N. Feng, X. Liu, F. Deng, ACS Catal. 2018, 8, 69-74; b) P.
Gao, J. Xu, G. Qi, C. Wang, Q. Wang, Y. Zhao, Y. Zhang, N.
Feng, X. Zhao, J. Li, F. Deng, ACS Catal. 2018, 8, 9809-9820.
a) S. Ma, X. Guo, L. Zhao, S. Scott, X. Bao, J. Energy Chem.
[
21] J. Z. Hu, C. Wan, A. Vjunov, M. Wang, Z. Zhao, M. Y. Hu, D.
M. Camaioni, J. A. Lercher, J. Phys. Chem. C 2017, 121,
12849-12854.
[
[
2]
3]
[
22] a) C. Sun, J. Deng, L. Kong, L. Chen, Z. Shen, Y. Cao, H.
Zhang, X. Wang, IOP Conference Series: Materials Science
and Engineering 2017, 275, 012046; b) N. V. Golubev, E. S.
Ignat'eva, V. N. Sigaev, A. Lauria, L. De Trizio, A. Azarbod, A.
Paleari, R. Lorenzi, Phys. Chem. Chem. Phys. 2015, 17,
5141-5150.
[
[
4]
5]
[
[
23] a) K. Nishi, K.-i. Shimizu, M. Takamatsu, H. Yoshida, A.
2
2
013, 22, 1-20; b) P. Schwach, X. Pan, X. Bao, Chem. Rev.
017, 117, 8497-8520.
Satsuma, T. Tanaka, S. Yoshida, T. Hattori, J. Phys. Chem. B
1998, 102, 10190-10195; b) K.-i. Shimizu, M. Takamatsu, K.
a) Y. Ono, K. Osako, G. J. Kim, Y. Inoue, in Studies in Surface
Science and Catalysis, Vol. Volume 84 (Eds.: H. G. K. H. P. J.
Weitkamp, W. Hölderich), Elsevier, 1994, pp. 1773-1780; b) P.
Qiu, J. H. Lunsford, M. P. Rosynek, Catal. Lett. 1998, 52, 37-
Nishi, H. Yoshida, A. Satsuma, T. Tanaka, S. Yoshida, T.
Hattori, J. Phys. Chem. B 1999, 103, 1542-1549.
24] a) D. Ma, Y. Shu, C. Zhang, W. Zhang, X. Han, Y. Xu, X. Bao,
J. Mol. Catal. A 2001, 168, 139-146; b) B. Zheng, W. Hua, Y.
Yue, Z. Gao, J. Catal. 2005, 232, 143-151; c) A. Arnold, S.
Steuernagel, M. Hunger, J. Weitkamp, Micropor. Mesopor.
Mater. 2003, 62, 97-106.
42; c) V. R. Choudhary, P. Devadas, S. Banerjee, A. K.
Kinage, Micropor. Mesopor. Mater. 2001, 47, 253-267; d) Y.
Ono, T. Baba, Phys. Chem. Chem. Phys. 2015, 17, 15637-
1
5654.
a) L. A. Dufresne, R. Le Van Mao, Catal. Lett. 1994, 25, 371-
83; b) H. Coqueblin, A. Richard, D. Uzio, L. Pinard, Y.
[
[
[
[
[
[
[
25] W. Dai, L. Yang, C. Wang, X. Wang, G. Wu, N. Guan, U.
Obenaus, M. Hunger, L. Li, ACS Catal. 2018, 8, 1352-1362.
26] R. Fricke, H. Kosslick, G. Lischke, M. Richter, Chem. Rev.
[
[
6]
7]
3
Pouilloux, F. Epron, Catal. Today 2016; c) X. Chen, M. Dong,
X. Niu, K. Wang, G. Chen, W. Fan, J. Wang, Z. Qin, Chinese
J. Catal. 2015, 36, 880-888; d) M.-F. Hsieh, Y. Zhou, H.
Thirumalai, L. C. Grabow, J. D. Rimer, ChemCatChem 2017,
2000, 100, 2303-2406.
27] D. J. Parrillo, C. Lee, R. J. Gorte, D. White, W. E. Farneth, J.
Phys. Chem. 1995, 99, 8745-8749.
28] A. Ghorbanpour, J. D. Rimer, L. C. Grabow, Catal. Commun.
9
, 1675-1682.
a) A. Bhan, W. Nicholas Delgass, Catal. Rev. 2008, 50, 19-
51; b) S.-W. Choi, W.-G. Kim, J.-S. So, J. S. Moore, Y. Liu,
2
014, 52, 98-102.
29] A. Ghorbanpour, J. D. Rimer, L. C. Grabow, ACS Catal. 2016,
, 2287-2298.
30] A. J. Jones, R. T. Carr, S. I. Zones, E. Iglesia, J. Catal. 2014,
12, 58-68.
1
6
R. S. Dixit, J. G. Pendergast, C. Sievers, D. S. Sholl, S. Nair,
C. W. Jones, J. Catal. 2017, 345, 113-123; c) M. W.
Schreiber, C. P. Plaisance, M. Baumgärtl, K. Reuter, A.
Jentys, R. Bermejo-Deval, J. A. Lercher, J. Am. Chem. Soc.
3
31] C.-M. Wang, R. Y. Brogaard, B. M. Weckhuysen, J. K.
Nørskov, F. Studt, J. Phys. Chem. Lett. 2014, 5, 1516-1521.
32] R. Khare, D. Millar, A. Bhan, J. Catal. 2015, 321, 23-31.
33] M. Oleksiak, J. D. Rimer, Rev. Chem. Eng. 2014, 30, 1-49.
34] a) M. D. Oleksiak, K. Muraoka, M. F. Hsieh, M. T. Conato, A.
Shimojima, T. Okubo, W. Chaikittisilp, J. D. Rimer, Angew.
Chem. Int. Ed. 2017, 56, 13366-13371; b) R. Gounder, E.
Iglesia, Angew. Chem. Int. Ed. 2010, 49, 808-811; c) B. C.
Knott, C. T. Nimlos, D. J. Robichaud, M. R. Nimlos, S. Kim, R.
Gounder, ACS Catal. 2018, 8, 770-784.
2018, 140, 4849-4859; d) N. M. Phadke, E. Mansoor, M.
[
[
[
Bondil, M. Head-Gordon, A. T. Bell, J. Am. Chem. Soc. 2018.
C. D. Gosling, D. A. Hamm, Vol. US 5258563 A (Ed.: H. U.
LLC), Honeywell UOP LLC 1992.
a) A. A. Gabrienko, S. S. Arzumanov, A. V. Toktarev, D.
Freude, J. Haase, A. G. Stepanov, J. Phys. Chem. C 2011,
[
[
8]
9]
115, 13877-13886; b) S. Yasumura, M. Huang, X. Wu, C. Liu,
T. Toyao, Z. Maeno, K.-i. Shimizu, Catal. Today 2019.
[
[
10] J. F. Joly, H. Ajot, E. Merlen, F. Raatz, F. Alario, Appl. Catal.
A 1991, 79, 249-263.
11] a) N. M. Phadke, J. Van der Mynsbrugge, E. Mansoor, A. B.
Getsoian, M. Head-Gordon, A. T. Bell, ACS Catal. 2018, 8,
[
35] J. Towns, T. Cockerill, M. Dahan, I. Foster, K. Gaither, A.
Grimshaw, V. Hazlewood, S. Lathrop, D. Lifka, G. D.
Peterson, R. Roskies, J. R. Scott, N. Wilkins-Diehr, Comput.
Sci. Eng. 2014, 16, 62-74.
6106-6126; b) K. M. Dooley, C. Chang, G. L. Price, Appl.
Catal. A 1992, 84, 17-30; c) V. R. Choudhary, A. K. Kinage, C.
Sivadinarayana, S. D. Sansare, M. Guisnet, Catal. Lett. 1995,
33, 401-412.
[
[
12] G. L. Price, V. Kanazirev, J. Catal. 1990, 126, 267-278.
13] M. W. Schreiber, C. P. Plaisance, M. Baumgärtl, K. Reuter, A.
Jentys, R. Bermejo-Deval, J. A. Lercher, 2018.
[
14] a) N. Al-Yassir, M. N. Akhtar, S. Al-Khattaf, J. Porous Mater.
2012, 19, 943-960; b) V. R. Choudhary, S. A. R. Mulla, S.
Banerjee, Micropor. Mesopor. Mater. 2003, 57, 317-322; c) V.
R. Choudhary, P. Devadas, J. Catal. 1997, 172, 475-478.
15] M. Xin, E. Xing, X. Gao, Y. Wang, Y. Ouyang, G. Xu, Y. Luo,
X. Shu, Ind. Eng. Chem. Res. 2019, 58, 6970-6981.
[
[
[
16] C.-Y. Hsieh, Y.-Y. Chen, Y.-C. Lin, Ind. Eng. Chem. Res.
2018, 57, 7742-7751.
17] K. C. Szeto, Z. R. Jones, N. Merle, C. Rios, A. Gallo, F. Le
Quemener, L. Delevoye, R. M. Gauvin, S. L. Scott, M. Taoufik,
ACS Catal. 2018, 8, 7566-7577.
[
18] a) G. D. Meitzner, E. Iglesia, J. E. Baumgartner, E. S. Huang,
J. Catal. 1993, 140, 209-225; b) K. M. Dooley, G. L. Price, V.
I. Kanazirev, V. I. Hart, Catal. Today 1996, 31, 305-315; c) N.
Rane, A. R. Overweg, V. B. Kazansky, R. A. van Santen, E. J.
9
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.202007147
RESEARCH ARTICLE
Entry for the Table of Contents
Insert graphic for Table of Contents here.
Target in Site. Direct synthesis of Ga-ZSM-5 catalysts leads to dramatic enhancement in the production of aromatics from ethylene
compared to post-synthesis preparation methods. Combined experiments and modeling reveal an energetic preference for Ga extra-
framework sites to associate with select framework sites, suggesting the possibility of tailoring Lewis acid siting in zeolites by
controlled heteroatom substitution in framework tetrahedral sites.
Institute and/or researcher Twitter usernames:
@
@
@
@
@
UH_RimerGroup
TheGrabowGroup
UH_ChBE
UHEngineering
SBUArtsSciences
1
0
This article is protected by copyright. All rights reserved.