The Catalyst Review August 2019 - 15

EXPERIMENTAL
High-Quality Gasoline Directly from Syngas by Dual Metal Oxide-Zeolite (OX-ZEO) Catalysis...
Efforts to convert syngas into quality gasoline (low aromatics content and high-octane number) have been hampered by low process
selectivity. Herein, the authors demonstrate that zeolites with 1D ten-membered-ring (10-MR) channel structures such as SAPO-11
and ZSM-22 in combination with zinc- and manganese-based metal oxides (ZnaMnbOx) enable the selective synthesis of gasolinerange hydrocarbons C5-C11 directly from syngas.
It had previously been noted that the role of zeolites in the bifunctional oxide-zeolite (OX-ZEO) catalyst concept is to catalyze the
coupling of intermediates generated over the metal oxides. The authors therefore anticipated that this bottom-up stacking of small
intermediate molecules should allow the direct synthesis of high-quality gasoline from syngas with a low content of aromatics and
olefins by controlling C-C coupling within the shape-selective zeolite pores.
A co-precipitated binary Zn2Mn1Ox oxide was taken as the oxide component,
and it was mixed with zeolite SAPO-11 (denoted as Zn2Mn1Ox-SAPO-11)
for syngas conversion. The CO2 selectivity was around 50% in this case. Few
C12+ hydrocarbons were detected, and the carbon balance over the OX-ZEO
catalysts was over 95%. Therefore, the gasoline selectivity is reported as
the amount of C5-C11 among all hydrocarbons. Figure 1 shows that the
composite catalyst Zn2Mn1Ox-SAPO-11 gives a gasoline selectivity of up
to 76.7% at a CO conversion of 20.3% at 360°C, 4.0 MPa, GHSV= 1000 mLg-1h-1,
and H2/CO = 1. The CH4 selectivity is only 2.3%. In the gasoline, the isoparaffin
content amounts to 52%, while olefins and aromatic compounds amount
to 28% and only 16%, respectively (Figure 1B). Furthermore, the ratio of
iso/n-paraffins reaches 15, which is almost five times higher than that in the
commercial gasoline #92, while the aromatics content is only half of that of
#92.

Figure 1. Direct syngas conversion over Zn2Mn1Ox-SAPO-11.
A) Hydrocarbon distribution compared to the ASF distribution
at a chain growth probability of 0.76. B) Hydrocarbon
distribution in the gasoline range. Reaction conditions:
Zn2Mn1Ox/SAPO-11=2:1 (mass ratio), H2/CO=1, 360 °C,
4.0 MPa, gas hourly space velocity (GHSV)=1000 mL g−1 h−1.

Figure 2. The role of metal oxides in the bifunctional composite catalysts. A) CO conversion, with
The authors then studied the role of the
filled violet circles corresponding to coprecipitated ZnaMnbOx calcined at 550°C, filled navy circles
metal oxide component in the OX-ZEO
for physically mixed ZnO-s and MnOx-s, and open diamonds for Zn2Mn1Ox after thermal treatment.
B) XRD patterns of used ZnaMnbOx-SAPO-11 catalysts with different Zn/Mn ratios. C) XRD patterns
catalyst. Figure 2A shows that the activity
of used Zn2Mn1Ox-SAPO-11 with Zn2Mn1Ox subjected to thermal treatment. Note that the thermal
of ZnaMnbOx binary oxides increases with
treatment temperatures in °C are indicated in parentheses, and the diffraction patterns are calibrated
as in Figure S11. Reaction conditions: oxide/zeolite=2:1 (mass ratio), H2/CO=1, 400°C, 4.0 MPa,
the ZnO molar content as the CO conversion
GHSV=1000 mLg-1h-1.
reaches a maximum of 28.7% at 67% ZnO
molar content, corresponding to Zn2Mn1OxSAPO-11. Of interest is the fact that by
adjusting balance the activities of metal
oxide and zeolite, the OX-ZEO catalyst
concept turns syngas conversion into a
tandem catalysis process. For example,
the CO conversion increases from 21.9% t
o 29.7% when the oxide content in the
composite is increased from 33% to 75%.
Beyond that content, the overall activity decreases because of the insufficient number of active sites provided by the zeolite to
convert the intermediates. Source: Li N, Jiao F, Pan X, et al. (2019). Angew. Chem. Int. Ed., 58: 7400 -7404.

Machine Learning for Computational Heterogeneous Catalysis...
Advances in machine learning and data science hold great potential for the rapid computational screening of solid-state catalyst
materials candidates. To this end, this review focuses on the application of machine learning techniques in combination with
electronic structure simulations, such as DFT beginning with an outline of some important technical aspects, like data structures,
featurization, and surrogate machine learning models followed by an overview of recent work in the field and discuss future
opportunities and perspectives.
The main goal of catalysis research is to optimize catalysts in terms of activity, selectivity, stability, sustainability, and price. The
inherent complexity of catalytic reactions looks to machine learning methods to efficiently navigate through the high-dimensional
The Catalyst Review

August 2019

The Catalyst Review August 2019

Table of Contents for the Digital Edition of The Catalyst Review August 2019