Methanol-Assisted Autocatalysis in Catalytic Methanol Synthesis

Article abstract of DOI:10.1002/anie.202006921

Catalytic methanol synthesis is one of the major processes in the chemical industry and may grow in importance, as methanol produced from CO₂ and sustainably derived H₂ are envisioned to play an important role as energy carriers in a future low-CO₂-emission society. However, despite the widespread use, the reaction mechanism and the nature of the active sites are not fully understood. Here we report that methanol synthesis at commercially applied conditions using the industrial Cu/ZnO/Al₂O₃ catalyst is dominated by a methanol-assisted autocatalytic reaction mechanism. We propose that the presence of methanol enables the hydrogenation of surface formate via methyl formate. Autocatalytic acceleration of the reaction is also observed for Cu supported on SiO₂ although with low absolute activity, but not for Cu/Al₂O₃ catalysts. The results illustrate an important example of autocatalysis in heterogeneous catalysis and pave the way for further understanding, improvements, and process optimization of industrial methanol synthesis.

Full text of DOI:10.1002/anie.202006921

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Methanol assisted autocatalysis in catalytic methanol synthesis
Authors: Joachim Thrane, Sebastian Kuld, Niels Dyreborg Nielsen,
Anker Degn Jensen, Jens Sehested, and Jakob Munkholt
Christensen
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.202006921
Link to VoR: https://doi.org/10.1002/anie.202006921

10.1002/anie.202006921
Angewandte Chemie International Edition
Communication
Methanol assisted autocatalysis in catalytic methanol synthesis
Joachim Thrane[a], Sebastian Kuld[b], Niels D. Nielsen[a], Anker D. Jensen[a], Jens Sehested[b] and
Jakob M. Christensen*[a]
[a]
[b]
J. Thrane, N.D. Nielsen, Prof. A.D. Jensen, Assoc. Prof. J. M. Christensen
Department of Chemical and Biochemical Engineering
Technical University of Denmark
Søltofts Plads B229, 2800 Kgs. Lyngby (Denmark)
E-mail: jmc@kt.dtu.dk
S. Kuld, Dr. J. Sehested
Haldor Topsøe A/S
Nymøllevej 55, 2800 Kgs. Lyngby (Denmark)
Supporting information for this article is given via a link at the end of the document.
Abstract: Catalytic methanol synthesis is one of the major processes
in the chemical industry and may grow in importance, as methanol
produced from CO2 and sustainably derived H2 is envisioned to play
an important role as energy carrier in a future low CO2 emission
society. However, despite the widespread use, the reaction
mechanism and the nature of the active sites are not fully understood.
Here we report that methanol synthesis at commercially applied
conditions using the industrial Cu/ZnO/Al2O3 catalyst is dominated by
a methanol-assisted, autocatalytic reaction mechanism. We propose
that the presence of methanol opens for hydrogenation of surface
formate via methyl formate. Autocatalytic acceleration of the reaction
is also observed for Cu supported on SiO2 although with low absolute
activity, but not for Cu/Al2O3 catalysts. The results illustrate an
important example of autocatalysis in heterogeneous catalysis and
pave the way for further understanding, improvements and process
optimization of industrial methanol synthesis.
Here, we demonstrate that the origin of this obvious inconsistency
between theoretical understanding and experimental data is a
previously unknown, methanol-assisted, autocatalytic reaction
pathway. Our data confirm that this reaction pathway is
responsible for most of the turnovers to methanol under industrial
conditions, and we conclude that the autocatalytic mechanism
most likely is via a methyl formate intermediate.
Results and Discussion
A series of Cu catalysts supported on ZnO/Al2O3, ZnO, Al2O3 and
SiO2 were prepared, and the active Cu areas of the catalysts were
determined through the oxygen uptake (using O:Cusurf. 1:2 and
1.47∙1019 Cu atoms/m2)[11] from 1 mol% N2O/He at 333 K after
reduction in 5 vol% H2/N2 at atmospheric pressure and 523 K.
Cu/Al2O3 required that the 5% H2/N2 treatment was followed by a
100% H2 treatment at 523 K to obtain complete Cu reduction. The
catalytic activity of pre-reduced catalysts was measured in a flow
reactor at 523 K and 50 bar in syngas (CO/CO2/H2 = 29.6/2.8/67.6
mol%), and for each catalyst, the space velocity (SV) was varied
by 1-3 orders of magnitude to vary the level of conversion.
Turnover frequencies (TOF: methanol formation rate per Cu
surface atom) at the different conversion levels were determined
from the Cu surface areas and the methanol production rates, and
Fig. 1 shows the TOF as a function of the methanol concentration
Introduction
Heterogeneous catalysis is vital to modern chemical industry.
Understanding the nature of the active sites, reaction
mechanisms and the origin of metal-support/promoter
interactions is therefore not only fundamentally important, but also
significant due to the substantial economic impact of
improvements to large-scale industrial processes relying on
catalysis. Methanol is a major bulk chemical with a production
exceeding 70 million tons/year and is produced from syngas
(CO/CO2/H2) over a Cu/ZnO/Al2O3 catalyst[1]. Methanol synthesis
can proceed over Cu surfaces, but the reaction rate is strongly
enhanced by the presence of zinc in the catalyst[2-5]. However, the
exact nature of the active sites and the beneficial support effect
from ZnO in the industrial catalyst are still debated issues. By
contrast, there seems to be a more widespread agreement
concerning the general reaction mechanism for methanol
synthesis over Cu catalysts. Most theoretical and experimental
studies advocate that methanol is made directly from carbon
dioxide (CO2) and hydrogen (H2) via formation of surface formate
and its subsequent hydrogenation to methanol, and this reaction
mechanism explains most reported data in the literature. Yet,
such a mechanism is not able to explain some of the most
remarkable observations, where methanol synthesis rates under
at the exit of the reactor.
Cu/ZnO/Al2O3 ()
0.18
0.16
0.14
0.12
0.1
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
Cu/ZnO ()
Cu/Al2O3 ()
Cu/SiO2 ()
0.08
0.06
0.04
0.02
0
10-2
10-1
1
10-3
10
CH3OH concentration in effluent (mol%)
Fig. 1. The dependence of the TOF upon the CH3OH concentration in the
reactor effluent when varying the space velocity. Note the different scaling of
the Cu/SiO2 data. Dashed lines given as guides to the eye. Experimental
conditions: 50 bar, 523 K, H2/CO/CO2 = 67.6/29.6/2.8 mol%, 3·103-4.3·107
NL/kg/h. See Fig. S1 for a definition of the TOF and a comparison in terms of
space velocity. The equilibrium concentration of CH3OH at these conditions is
certain conditions are accelerated with increasing conversion[6, 7]
.
This acceleration is highly unexpected, as the methanol synthesis
has a relatively unfavourable equilibrium and experiences a
23 mol% based on data from Graaf and Winkelman[12]
.
strong kinetic inhibition from the accompanying water product[7-10]
.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202006921
Angewandte Chemie International Edition
Communication
This figure depicts, how the Cu catalysts respond to the changing
product concentrations, when the SV is varied. Under differential
conditions (below approximately 0.1 mol% methanol produced),
the TOF reaches a stable plateau for each catalyst type. Here
methanol formation through a mechanism directly from the
reactants must dominate. At these conditions there is a notable
support effect on the TOF, and the activity order of the supports
(ZnO/Al2O3 ≅ ZnO > Al2O3 > SiO2) is in good agreement with the
reaction-induced changes in CO2 concentration cannot affect the
experimental results. As diffusion limitations were ruled out by
experiments and calculations (see Supporting Information), it can
be ruled out that a local concentration rise within the catalyst
pores should affect this conclusion.
The increase in reaction rate in Fig. 1 must therefore arise from
the emergence of a faster pathway at higher conversions, and at
least two parallel reaction pathways in methanol synthesis are
needed to explain the results: a direct conversion of the reactants
dominating at differential conditions and a faster autocatalytic
route involving a reaction product dominant at higher conversions.
At the lower SV, where the autocatalytic pathway prevails, the
TOF for Cu/ZnO was observed to remain constant across an 8-
fold variation in Cu surface area, which corresponds to a linear
correlation between absolute activity and Cu surface area (Fig.
S3). This would imply that the rate limiting step in the autocatalytic
pathway occurs on the metal surface, but there is clearly also an
either direct or indirect involvement of the support that is of great
importance for the absolute activity.
observations in several previous studies[2-4]
.
Within the
uncertainty the same TOF is seen for Cu/ZnO/Al2O3 and Cu/ZnO
at identical conversion levels in Fig. 1, and these two systems are
therefore treated collectively in the remaining text. The TOF-
values are determined from the Cu surface areas obtained from
N2O titrations, and there are uncertainties in methods for
determination of metal area, as illustrated by the systematic and
carrier-dependent differences between areas from N2O and H2
titrations[11]
. However, with the applied pre-reduction, the
magnitude of these differences[11] is expected to be considerably
smaller than the differences between the TOF values of the
different catalyst systems, and the uncertainties on the area
should therefore not affect the conclusions. Consequently, this
issue is not discussed further in the following.
In the industrial process the syngas feed already contains
some methanol due to recirculation of unconverted reactants, and
the methanol concentration rises to significant percentages
Remarkable changes in TOF are observed in Fig. 1 when the
product concentrations are increased by lowering of the SV to
achieve finite conversions. Lowering the SV causes the TOF-
values for the ZnO(/Al2O3) and SiO2 containing catalysts to
increase more than three-fold. Such an increase with rising
product concentration is indicative of a significant autocatalytic
effect, whereby the product assists the formation of additional
product. The magnitude of the acceleration means that the
autocatalytic pathway is at least several times faster than the
direct pathway, and as a result the substantial majority of
turnovers in the industrial process must arise from the
autocatalytic pathway. This autocatalytic behaviour is clearly
support-dependent. The Cu/SiO2 catalyst has a low absolute
activity but shows an autocatalytic behaviour like that of
Cu/ZnO(/Al2O3), while reaction rates for Cu/Al2O3 are
independent of SV (see also Supporting Information Fig. S1d).
The heat generated in the exothermic reaction increases with
conversion/product concentration regardless of the support, but
here no significant temperature rise was observed, and the fact
that TOF for Cu/Al2O3 did not grow with conversion shows that the
acceleration is not due to a temperature rise. As CO2 is the
primary reactant in methanol synthesis over pure Cu and
Cu/ZnO(/Al2O3)[9, 13, 14] it is important to evaluate, if changes in CO2
concentration can arise from the variations in conversion. CO2 is
consumed via methanol synthesis (R1), but partly restored by the
water-gas shift reaction (R2).
through the reactor[15]
. Under such conditions, the faster
autocatalytic pathway will be responsible for the vast majority of
the turnovers and thus dominate the industrial process.
This conclusion raises the question of, which reaction product
that causes such an effect. As the highly selective methanol
synthesis produces only two major products, namely CH3OH and
H2O, these two products represent the most likely candidates.
Additional experiments were therefore conducted with the
Cu/ZnO/Al2O3 catalyst (Cu:Zn:Al = 6:3:1, 20 m2Cu/gcat.) to identify
the source and possible mechanism of the autocatalytic effect.
The methanol synthesis activity was measured for the
Cu/ZnO/Al2O3 catalyst at differential conditions (SV = 1.6∙106
NL/kg/h) far from equilibrium (< 1000 ppm CH3OH produced) with
low levels of water added to the syngas feed, and the results are
shown in Fig. 2.
H2O partial pressure in feed (kPa)
0
1
2
3
4
5
6
100
80
60
40
20
0
 493 K
 508 K
 523 K
ꢀꢁ2 + 3ꢂ2 ⇌ ꢀꢂ3ꢁꢂ + ꢂ2ꢁ
ꢀꢁ + ꢂ2ꢁ ⇌ ꢂ2 + ꢀꢁ2
R1
R2
0
500
1000
1500
H2O concentration in feed (ppmv)
The activity peak for Cu/ZnO(/Al2O3) in Fig. 1 (+372%) occurs at
2 mol% methanol in the effluent. If R2 is equilibrated at these
conditions there will be a 2.5% net conversion of the CO2.
However, the volume contraction due to the loss of molecules in
R1 was measured to be 4% at these conditions, and this leads to
a corresponding rise in the CO2 concentration. The consumption
of CO2 and the concentration rise due to volume contraction thus
nearly balance out, and the concentration of CO2 is therefore
essentially unaffected by the reaction (see Fig. S2). Consequently,
Fig. 2. The relative methanol production as a function of the water content
added to the syngas feed. Experimental conditions: Catalyst: Cu/ZnO/Al2O3, P
= 41 bar, H2/CO/CO2 = 67.6/29.6/2.8 mol%, 1.6·106 NL/kg/h. Dashed lines given
as guides to the eye. See Supporting Information for a definition of the error
bars.
The data in Fig. 2 illustrate that even minute amounts of water
lower the activity substantially, which strongly indicates that water
is not the source of the autocatalytic effect. Competitive
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202006921
Angewandte Chemie International Edition
Communication
adsorption of water or its dissociation products is a likely
explanation for this kinetic inhibition, as water adsorption isotherm
measurements[16, 17] on Cu/ZnO/Al2O3 suggest a high coverage on
the Cu surface at temperatures and H2O partial pressures similar
to those where major water inhibition is observed in Fig. 2. The
increase in TOF with rising product concentrations in Fig. 1 is
even more remarkable when considering the strong inhibition
caused by the co-produced water.
The effect of adding the second major product, methanol, to
the syngas feed is illustrated in Fig. 3. Two independent methanol
co-feeding experiments were performed in two very different
setups (described in detail in the Supporting Information), and the
results are depicted as Figs. 3a and 3b, respectively. Remarkably,
the net rate of methanol formation increases considerably, when
methanol is co-fed with the syngas feed, and on this basis we
deduce that methanol is the source of the autocatalytic
acceleration in methanol synthesis.
The net increase in reaction rate with increasing conversion
seen in a CO/CO2/H2 feed (Fig. 1) does not occur in CO2/H2 gas
mixtures as illustrated in Fig. S6. This can be rationalized through
the opposite effects of the two reaction products. If CO is present,
most of the produced water will be removed by R2, and the
beneficial effect of methanol will dominate over the adverse effect
of water. In a CO-free atmosphere, water is produced by both the
methanol synthesis and the reverse water-gas shift reaction, and
the greater concentration of inhibiting water causes the rate to
decline with increasing conversion. However, from a CO2/H2 feed,
only a moderate loss in activity of 6% was observed for a
Cu/ZnO/Al2O3 catalyst at a methanol effluent concentration of 0.4
mol% (see Fig. S6). At these conditions, an oxygen balance
suggested that 0.68 mol% water was co-produced, if only R1 and
R2 occur. Considering the inhibition expected from such a water
concentration (see Fig. 2), the autocatalytic acceleration seems
to partly compensate for the negative effect of water and thus also
play a major role in a CO2/H2 atmosphere. Hence, the net rate of
the reaction is determined by the balance between acceleration
from the autocatalytic pathway and inhibition from produced water,
which will depend on the reaction conditions such as the
composition of the syngas.
Liquid phase alcohols were previously found[18-20] to accelerate
the Cu-catalysed hydrogenation of CO2 to methanol at low
temperatures (≤ 443K) well below the industrial operating window.
However, the data in Fig. 3 imply that methanol synthesis at
commercial conditions is dominated by methanol mediated
autocatalysis.
We show here for the first time that methanol formation over
the industrially applied Cu/ZnO/Al2O3 catalysts is dominated by an
autocatalytic pathway that involves methanol. At high SV the
formed methanol is almost exclusively from CO2 with the industrial
Cu/ZnO/Al2O3 catalyst[9, 13, 14]. This is most likely also the case at
lower SV, where the rate from a CO2-containing gas (> 5
gMeOH/gcat./h in Fig. S1c) is also far higher than in a CO/H2 feed
(0.4 gMeOH/gcat./h at comparable conditions[9]). The autocatalytic
pathway at lower SV must therefore be expected to proceed from
CO2. At lower SV/higher conversion with greater water formation
more methanol is formed by the CO→CO2→CH3OH sequence of
R2 followed by R1, but the constancy in CO2 concentration across
the reactor (Fig. S2) suggests that R2 is fast, and consequently
R1 is solely rate limiting in methanol synthesis. Adsorbed formate
is present on the metal surface at methanol synthesis
conditions[14, 21, 22], which suggests that the autocatalytic pathway
could involve CO2-derived formate. Fig. 4a shows that CD3OOCH
was formed at low temperature, when Cu/ZnO/Al2O3 with pre-
adsorbed HCOO (from exposure to HCOOH/N2 at 303 K) was
subjected to temperature programmed reaction (TPR) in a flow of
CD3OD in He. This illustrates that methyl formate can be formed
from reaction between methanol and surface formate. A role of
molecularly adsorbed formic acid cannot be fully excluded here,
but previous IR studies[23, 24] also demonstrate that the ester is
easily formed by reaction of an alcohol with adsorbed formate at
typical methanol synthesis temperatures. A probable pathway for
the autocatalytic process is therefore via formation of the methyl
formate ester from methanol and surface formate (R3) followed
by hydrogenation of the ester (R4). However, methyl formate may
a
220
200
180
160
140
120
100
0
0.2
0.4
0.6
0.8
1
Methanol concentration in feed (mol%)
b
200
180
160
140
120
100
also be decarbonylated to methanol and CO (R5)[25]
.
0
0.2
0.4
0.6
0.8
1
1.2
Methanol concentration in feed (mol%)
CH3OH + HCOO ⇌ CH3OOCH + OH
CH3OOCH + 2H2 ⇌ 2CH3OH
CH3OOCH ⇌ CH3OH + CO
R3
R4
R5
Fig. 3. (a) Relative net productivity of methanol as a function of the methanol
concentration in the syngas feed at 50 bar. (b) Relative net productivity of
methanol as a function of the methanol concentration in the syngas feed at 41
bar. Other experimental conditions: Catalyst: Cu/ZnO/Al2O3, T = 523 K, syngas
before MeOH addition: H2/CO/CO2 = 67.6/29.6/2.8 mol%, 1.6·106 NL/kg/h.
Additional data in Figs. S4 and S5. See Supporting Information for a definition
of the error bars.
Reaction R3 followed by reaction R4 results in a net gain of
methanol, whereas the combination of R3 and R5 yields no net
gain. Effective autocatalysis via methyl formate therefore relies on
the rates of R3 and R4 being faster than direct CO2 hydrogenation
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202006921
Angewandte Chemie International Edition
Communication
and on a high selectivity to hydrogenation (R4) relative to
decarbonylation (R5) in the conversion of methyl formate. Fig. 4b
shows the results, when Cu/ZnO/Al2O3 pre-covered with surface
formate (from HCOOH) was subjected to a TPR in a flow of either
CH3OH/He or H2. Fig. 4b illustrates that the onset of methyl
formate formation in a CH3OH/He gas occurs at a lower
temperature than methanol formation in H2. Re-adsorption
phenomena can delay the appearance of a product in a TPH, but
the lower onset temperature does indicate that R3 is considerably
faster than direct hydrogenation of HCOO, and there are good
indications in the literature that this is also the case for both R4
and R5. Fakley and co-workers[26] used the data from Monti et
al.[25] to calculate that R4 should be of ample rate on Cu to account
for an acceleration of the methanol synthesis.
the Al2O3 support. By contrast, Cu/ZnO-systems[28] and Cu/SiO2-
systems[25] preserve a high selectivity to R4 versus R5 also at high
temperatures, possibly because these supports are less likely to
catalyse R5. A pathway via methyl formate would thus explain the
observed support dependence in Fig. 1: i.e. a net acceleration
with increasing conversion for the Cu/ZnO(/Al2O3) and Cu/SiO2
systems, where R4 dominates, whereas R5 dominates for
Cu/Al2O3 and interrupts the autocatalytic pathway. As previously
discussed, the net rate is governed by the competition between
the autocatalytic acceleration and the water inhibition.
Consequently, the conversion independent TOF of Cu/Al2O3 in
Fig. 1 does not rule out that the autocatalytic pathway proceeds
to some extent for Cu/Al2O3 and partly compensates for inhibiting
effects of water, but the efficacy of the autocatalytic pathway is
clearly lower for Cu/Al2O3 than for the Cu/ZnO(/Al2O3) and
Cu/SiO2 systems.
a
A consequence of the support dependence in the autocatalytic
pathway is that conclusions regarding the difference in TOF
between Cu/ZnO(/Al2O3) and Cu/Al2O3 are highly dependent on
the conditions. Fig. 1 shows that the ratio between their TOF-
values increases from a factor of three at differential conditions to
more than an order of magnitude at higher conversion. Such
variations illustrate the complexity that can emerge in catalytic
reactions due to the existence of multiple, parallel reaction
pathways with varying dependence on conditions and catalyst
composition. This new and more complex picture of the reaction
routes over Cu based methanol synthesis catalysts can possibly
help to reconcile some apparent disparities between previous
conclusions in the literature. Such complexity arising from the
existence of multiple reaction pathways should be a general
consideration for catalytic processes, since methanol synthesis
may not be the only reaction, where multiple, parallel pathways
coexist.
300
325
350
375
400
425
450
Temperature (K)
b
CH3OH
(m/z = 31)
from H2
CH3OOCH
(m/z = 60)
from CH3OH/He
Conclusion
The discovery of a previously unknown autocatalytic mechanism
dominating methanol synthesis at commercial reaction conditions
with the industrial catalyst satisfactorily explains the so far
unexplained observations[6, 7] of increasing methanol synthesis
rates with increasing conversion. This novel understanding
represents a paradigm shift in the mechanistic interpretation of
methanol synthesis and shows that important new discoveries
can be made even for well-studied and widely applied catalyst
systems. Finally, this breakthrough paves the way for further
understanding, improvements and process optimization of
industrial methanol synthesis.
300
325
350
375
400
Temperature (K)
Fig. 4. (a) The formation of methyl formate (CD3OOCH, m/z = 63) as a function
of temperature during TPR when HCOO-covered Cu/ZnO/Al2O3 was heated (1
K/min) at 1 atm in a 19 NmL/min flow of 4.1 mol% CD3OD in He, 0.5 g cat. See
Fig. S7 for details on peak shape. (b) Comparison of the onset of methyl formate
formation during TPR in 5.5 mol% CH3OH in He and the onset of methanol
formation during TPR in 100% H2 when formate covered Cu/ZnO/Al2O3 was
subjected to TPR at 1 atm. Ramp: 1 K/min, Flow: 40 NmL/min, 0.5 g cat.
Acknowledgements
Santiago et al.[27] co-fed methyl formate with the synthesis gas,
and their results showed that methyl formate was rapidly
converted over Cu/ZnO(/Al2O3) and Cu/Al2O3 catalysts at 473 K
and 65 bar. We infer from their data that the additional production
of methanol due to methyl formate co-feeding was almost twice
as high for Cu/ZnO/Al2O3 compared to Cu/Al2O3, which suggests
a preference for R4 on Cu/ZnO/Al2O3 and a preference for R5 on
Cu/Al2O3. A preference for R5 on Cu/Al2O3 is further supported by
the results of Lam et al.[23], who observed decarbonylation of
methyl formate over Cu/Al2O3 and found that R5 is catalysed by
This work was supported by a research grant (9455) from VILLUM
FONDEN. We thank Saint Gobain for providing the SiO2 support.
Keywords: Autocatalysis • Methanol • CO2 hydrogenation •
Copper • Reaction mechanisms
[1] J. Sehested, J. Catal. 2019, 371, 368-375.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202006921
Angewandte Chemie International Edition
Communication
[2] T. Fujitani, M. Saito, Y. Kanai, T. Kakumoto, T. Watanabe, J. Nakamura, T.
Uchijima, Catal. Lett. 1994, 25, 271-276.
[3] M. Saito, K. Murata, Catal. Survey Asia 2004, 8, 285-294.
[4] R. Burch, S. E. Golunski, M. S. Spencer, J. Chem. Soc. Faraday Trans. 1990,
86, 2683-2691.
[5] I. Nakamura, T. Fujitani, T. Uchijima, J. Nakamura, J. Vac. Sci. Technol. A
1996, 14, 1464-1468.
[6] J. S. Lee, S. H. Han, H. G. Kim, K. H. Lee, Y. G. Kim, Korean J. Chem. Eng.
2000, 17, 332-336.
[7] M. Sahibzada, I. S. Metcalfe, D. Chadwick, J. Catal. 1998, 174, 111-118.
[8] O. Cherifi, S. Monteverdi, M. M. Bettahar, M. Forissier, V. Perrichon, Bull.
Soc. Chim. Fr. 1985, 405-409.
[9] N. D. Nielsen, J. Thrane, A. D. Jensen, J. M. Christensen, Catal. Lett. 2020,
150, 1427-1433.
[10] Y. Yang, C. A. Mims, D. H. Mei, C. H. F. Peden, C. T. Campbell, J. Catal.
2013, 298, 10-17.
[11] R. Chatterjee, S. Kuld, R. van den Berg, A. Chen, W. Shen, J. M.
Christensen, A. D. Jensen, J. Sehested, Top. Catal. 2019, 62, 649-659.
[12] G. H. Graaf, J. G. M. Winkelman, Ind. Eng. Chem. Res. 2016, 55, 5854-
5864.
[13] G. C. Chinchen, P. J. Denny, D. G. Parker, M. S. Spencer, D. A. Whan,
Appl. Catal. 1987, 30, 333-338.
[14] N. D. Nielsen, A. D. Jensen, J. M. Christensen, Catal. Lett. in press, DOI:
https://doi.org/10.1007/s10562-020-03162-7.
[15] W.-H. Cheng, H. H. Kung (Eds.), Methanol production and use, CRC Press
1994.
[16] J. Słoczyński, R. Grabowski, J. Janas, J. Skrzypek, Chem. Eng. Sci. 1991,
46, 2599-2610.
[17] A. Tarasov, F. Seitz, R. Schlögl, E. Frei, ACS Catal. 2019, 9, 5537-5544.
[18] L. Fan, Y. Sakaiya, K. Fujimoto, Appl. Catal. A 1999, 180, L11-L13.
[19] N. Tsubaki, M. Ito, K. Fujimoto, J. Catal. 2001, 197, 224-227.
[20] P. Reubroycharoen, T. Yamagami, T. Vitidsant, Y. Yoneyama, M. Ito, N.
Tsubaki, Energy Fuels 2003, 17, 817-821.
[21] S. Lin, A. Oldfield, D. Klenerman, Surf. Sci. 2000, 464, 1-7.
[22] F. Le Peltier, P. Chaumette, J. Saussey, M. M. Bettahar, J. C. Lavalley, J.
Mol. Catal. A 1998, 132, 91-100.
[23] E. Lam, J. J. Corral-Perez, K. Larmier, G. Noh, P. Wolf, A. Comas-Vives,
A. Urakawa, C. Coperet, Angew. Chem. Int. Ed. 2019, 58, 13989-13996.
[24] R. Yang, Y. Zhang, N. Tsubaki, Catal. Commun. 2007, 8, 1829-1833.
[25] D. M. Monti, M. S. Wainwright, D. L. Trimm, N. W. Cant, Ind. Eng. Chem.
Prod. Des. Dev. 1985, 24, 397-401.
[26] M. E. Fakley, J. R. Jennings, M. S. Spencer, J. Catal. 1989, 118, 483-486.
[27] M. Santiago, K. Barbera, C. Ferreira, D. Curulla-Ferré, P. Kolb, J. Pérez-
Ramírez, Catal. Commun. 2012, 21, 63-67.
[28] G. Braca, A. M. R. Galletti, N. J. Laniyonu, G. Sbrana, E. Micheli, M. Di
Girolamo, M. Marchionna, Ind. Eng. Chem. Res. 1995, 34, 2358-2363.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202006921
Angewandte Chemie International Edition
Communication
Entry for the Table of Contents
Two ways to methanol
Autocatalytic
Pathway
Direct
hydrogenation
to methanol
Most methanol
produced in industrial
methanol synthesis
HCOO
CO/CO2/H2
It is shown that the methanol synthesis from syngas with the industrial Cu/ZnO/Al2O3 catalyst at commercially applied conditions is
dominated by a methanol-assisted, autocatalytic reaction mechanism. Of the two main reaction products H2O is strongly inhibiting,
whereas CH3OH is beneficial to the reaction rate. Methanol is proposed to open up a faster pathway of surface formate conversion
via methyl formate.
6
This article is protected by copyright. All rights reserved.