A Highly Active Molybdenum Phosphide Catalyst for Methanol Synthesis from CO and CO2

Article abstract of DOI:10.1002/anie.201806583

Methanol is a major fuel and chemical feedstock currently produced from syngas, a CO/CO₂/H₂ mixture. Herein we identify formate binding strength as a key parameter limiting the activity and stability of known catalysts for methanol synthesis in the presence of CO₂. We present a molybdenum phosphide catalyst for CO and CO₂ reduction to methanol, which through a weaker interaction with formate, can improve the activity and stability of methanol synthesis catalysts in a wide range of CO/CO₂/H₂ feeds.

Full text of DOI:10.1002/anie.201806583

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Discovery of a highly active molybdenum phosphide catalyst for
methanol synthesis from CO and CO2
Authors: Melis Duyar, Charlie Tsai, Jonathan L. Snider, Joseph A.
Singh, Alessandro Gallo, Jong Suk Yoo, Andrew J. Medford,
Frank Abild-Pedersen, Felix Studt, Jakob Kibsgaard, Stacey
Bent, Jens K. Nørskov, and Thomas Jaramillo
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.201806583
Angew. Chem. 10.1002/ange.201806583
Link to VoR: http://dx.doi.org/10.1002/anie.201806583
http://dx.doi.org/10.1002/ange.201806583

10.1002/anie.201806583
Angewandte Chemie International Edition
COMMUNICATION
Discovery of a highly active molybdenum phosphide catalyst for
methanol synthesis from CO and CO2
Melis S. Duyar+, Charlie Tsai+, Jonathan L. Snider, Joseph A. Singh, Alessandro Gallo, Jong Suk Yoo,
Andrew J. Medford, Frank Abild-Pedersen, Felix Studt, Jakob Kibsgaard, Stacey F. Bent, Jens K.
Nørskov*, Thomas F. Jaramillo*
and hence develop
a new MoP catalyst demonstrating
[*].
significant methanol production from CO, CO2 and their mixtures,
irrespective of feed composition.
Abstract: Methanol is a major fuel and chemical feedstock currently
produced from syngas, a CO/CO2/H2 mixture. Here we identify
formate binding strength as a key parameter limiting the activity and
stability of known catalysts for methanol synthesis in the presence of
CO2. We present a molybdenum phosphide catalyst for CO and CO2
reduction to methanol, which through a weaker interaction with
formate, can improve the activity and stability of methanol synthesis
catalysts in a wide range of CO/CO2/H2 feeds.
To determine the requirements for a catalyst with optimal
activity and selectivity towards methanol production, we used a
density-functional theory (DFT) based descriptor analysis[5].
Linear scaling relations[6] among reaction intermediates allow the
energetics of all reaction steps to be determined as a function of
two parameters, the carbon and oxygen binding strengths[5],
which can then determine all elementary reaction rates in a
mean-field micro-kinetic model. CO and OH adsorption energies
are used to evaluate the catalytic activity of transition metal and
transition metal phosphide catalysts in this work. These
descriptors were chosen because they led to small errors in the
predictions using the scaling relationships and facilitated
comparison to the published literature (Supporting Information).
For CO conversion, the reaction mechanisms are the same as in
previous work[5], with the steps for CO2 hydrogenation
included[7]:
Methanol can be a clean, sustainable fuel to supply the world's
increasing energy demand[1] and its synthesis from CO2 using
renewable energy can have major impacts on closing the
anthropogenic carbon cycle[2]. Methanol is commercially
produced on the scale of 65 mega-tonnes/year (2013) at
moderately high temperature (200-300oC) and high pressure
(50-100 bar) from CO/CO2/H2 mixtures (syngas) over zinc-
promoted copper catalysts[1-3]. The state of the art catalyst for
methanol synthesis, Cu/ZnO/Al2O3, favors CO2 hydrogenation
over CO hydrogenation; hence Cu/ZnO/Al2O3 cannot be used for
methanol synthesis from CO/H2 and is subject to kinetic
limitations when operating in CO/CO2/H2 mixtures [4]. If a catalyst
could be found that maintains high activity for both CO and CO2
conversion, it would make methanol synthesis more amenable
to changing feed compositions, such as those resulting from
incorporation of captured CO2. We know today that the ZnO/Cu
system is characterized by a Zn/Cu interface that converts CO2,
though is inactive for direct CO hydrogenation. Pure Cu sites are
less active under a mixed CO/CO2 feed since they are poisoned
by CO2, which has a low rate of conversion[3c, 3d]. Herein, we
present a molybdenum phosphide (MoP) catalyst, which is
considerably more active than Cu and stable with respect to
changing CO/CO2 ratios in the feed. In this work we show the
excellent performance of the MoP catalyst, exhibiting
comparable surface area normalized activity to the commercial
catalyst (Cu/ZnO/Al2O3), while explaining the origins of its high
performance. We show that by stabilizing only the monodentate
configuration of formate during CO2 hydrogenation, the
phosphide is able to circumvent the (scaling) relationship
between the necessary ability of the catalyst to bind oxygen-
containing intermediates and catalyst deactivation due to high
formate coverage. We present experimental evidence verifying
our kinetic model and spectroscopic evidence for monodentate
formate on MoP under methanol synthesis conditions using
diffuse reflectance infrared Fourier transform spectroscopy
(DRIFTS). By stabilizing the monodentate formate intermediate,
we are able to unlock new mechanisms for CO2 hydrogenation
CO2 (g) + H* ↔ HCOO*
(1)
(2)
(3)
(4)
HCOO* + H* ↔ HCOOH* + *
HCOOH* + H* ↔ H2COOH* + *
H2COOH* + * ↔ H2CO* + OH*
Activity maps for the production rates of methane,
methanol, and ethanol as a function of the descriptors ∆ECO and
∆EOH are shown in Fig.1a. The feed consists of CO and H2 (7.75
bar and 23.25 bar, respectively). Due to the location of the
activity optimum, activity towards methanol could be improved if
a catalyst can be found that binds CO similarly to the pure
transition metals but binds OH more strongly. Once 50% CO2 is
introduced into the feed (total 7.75 bar of CO and CO2), the
production rate diminishes by 1 order of magnitude compared to
the peak position for a pure CO feed (bottom right panel of
Fig.1a), in agreement with prior work[3d]. The smaller new peak is
positioned at weaker ΔEOH. The rate of methanol production
increases at weaker ΔECO compared to the original peak, which
represents methanol production from CO2. The Cu-Zn model
surface[3d] for the Cu/ZnO industrial catalyst is positioned in this
region, which is in agreement with previous results showing that
the total rate is dominated by methanol formed from CO2 rather
than from CO[3d]. When the feed is 100% CO2 (Fig.1a), only the
new peak remains. Cu/Zn is predicted to maintain a 1 order of
magnitude higher rate than Cu, again in agreement with
previous work[3d]
.
The decrease in optimal activity when CO2 is introduced is
due to the scaling between the binding energy of formate and all
other species bound to the surface through an oxygen atom.
The correlation between adsorption strengths prevents metal
catalysts from optimally converting the CO in the feed to
alcohols without also strongly binding formate formed from CO2
hydrogenation. The decrease in the methanol rate is a direct
consequence of an increased coverage of formate species
(Supporting Information). If formate binding could be
independently weakened without affecting CO and OH binding
energies, then the methanol formation rate would be expected to
increase. Strongly-bound formate has been observed in CO2
hydrogenation experiments on copper-based catalysts,
[*] Dr. M.S. Duyar+, Dr. C. Tsai+, J. L. Snider, J.A. Singh, Dr. A. Gallo, Dr. J.S.
Yoo, Prof. A.J. Medford, Dr. F. Abild-Pedersen, Prof. Dr. F. Studt, Prof., J.
Kibsgaard, Prof. S.F. Bent, Prof. J.K. Nørskov, Prof. T.F. Jaramillo
SUNCAT Center for Interface Science and Catalysis
Stanford University, Stanford, CA 94305, USA
SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
E-mail: jaramillo@stanford.edu, norskov@stanford.edu
[+] These authors contributed equally to the work.
Supporting Information for this article and high resolution images can be found
in the online version of the article.
adsorbed in a bidentate configuration, with both oxygen atoms of
3d,
the molecule binding to the surface[3b,
8]. The unfavorable
stabilization of bidentate formate is partly due to its adsorption
This article is protected by copyright. All rights reserved.

10.1002/anie.201806583
Angewandte Chemie International Edition
COMMUNICATION
Figure 1: a) Activity maps for transition metal and MoP catalysts for methane, methanol and ethanol formation from syngas and syngas containing 50% and
100% CO2. Here, T = 503.15 K, p(H2) = 23.25 bar, and p(CO) + p(CO2) = 7.75 bar for a total of p = 31 bar. This model assumes formate is present on the
surface of catalysts in the presence of 50-100% CO2 b) Activity map taking into account the differences in formate binding between MoP and metal catalysts
(same conditions as (a), with 50% CO and 50% CO2) c) Monodentate formate over a MoP surface (yellow=P, blue=Mo) and bidentate formate over an fcc metal
step surface. For MoP, the oxygen atom is binding to a surface P atom.
geometry on stepped metal surfaces. In combined theoretical
and experimental studies, formate’s role as both an intermediate
and poison was confirmed during hydrogenation of CO/CO2
mixtures to methanol on copper-based catalysts[3d]. In general,
the formate intermediate is found to be most stable in a
bidentate geometry on transition metal surfaces[9]. Recently
reported In2O3 catalysts are suggested to contain active oxygen
vacancies that stabilize monodentate formate, but these
catalysts are inactive for CO hydrogenation despite being highly
active for CO2 hydrogenation[10]. This motivates the search for a
catalyst with favorable binding of all species for methanol
production from CO, while constraining formate exclusively to a
monodentate geometry; such a catalyst would be expected to
convert both CO and CO2 to methanol at high rates,
independent of feed composition.
CO2, the 3.25 Å distance between the active sites prevents the
two oxygen atoms in formate from simultaneously binding to the
surface in a bidentate configuration. Since the CO and OH
adsorption energies are optimal for CO conversion (Fig.1) and
formate binds through a single oxygen atom (monodentate),
MoP is predicted from this model to be highly active for both CO
and CO2 conversion, irrespective of feed ratio, unlike transition
metal surfaces.
To determine how much the production rates can improve
if formate binding is weakened, we introduce the formate
adsorption strength as an additional descriptor (Supporting
Information). In Fig.1b, the production rates are shown as a
function of ∆EHCOO and ∆ECH3O. To make the result more intuitive,
we replaced the OH* descriptor with CH3O* (OH* and CH3O*
scale linearly, so either of them are valid descriptors; CH3O*
directly indicates the ease of desorbing the final product,
methanol). This illustrates the combined effect of formate
binding and methoxy binding strengths on the expected
methanol production. For transition metals, formate scales
linearly with all other species binding through O. The phosphide
surfaces deviate from this linear scaling relation due to the
preferred monodentate geometry for formate adsorption. The
calculated production rates for methanol are considerably
improved, indicating that MoP surfaces should remain highly
active in the presence of CO2, selectively converting syngas to
methanol over varying ratios of CO and CO2 in a robust manner.
We note that Fig.1b is necessarily more qualitative since CO
does not generally scale well with OH (Supporting Information).
We prepared MoP catalysts using a synthesis route
involving potassium and silica support (MoP K/SiO2). K promoter,
known to increase oxygenate production from syngas[13b], is
used in this study as a structural promoter to stabilize the MoP
phase during synthesis. The resulting catalysts contained 10
wt.% Mo and MoP nanoparticles of 62 nm average crystallite
size as determined by X-Ray Diffraction (XRD) which
corresponds to a catalytic surface area of 0.9 m2 MoP/g catalyst.
We investigated catalysts with crystal structures that can
impose geometric constraints on formate. Transition metal
phosphides, primarily used for hydrotreating reactions[11] are
ideal candidates because they have active sites that are further
apart compared to transition metals[12]
.
Transition metal
phosphides have been experimentally explored to a very limited
extent for CO hydrogenation[13] but structure-activity
relationships have not previously been investigated in detail. We
identified MoP to be promising, with neighboring active sites
separated by 3.25 Å, much larger than the 2.58 Å between sites
on a Cu (211) step[12]. MoP is also expected to be stable under
methanol synthesis conditions (Supporting Information). Using
descriptor-based analysis, the MoP catalyst is evaluated under
five scenarios to represent the range of activities that may be
observed in a realistic system (Fig.1): with and without hydrogen
coverage (MoP(H0.5) and MoP respectively), with a vacancy
(MoP(vac)), with
a
bare Mo surface with all
P
removed(MoP(Mo)), and with K-promoter on the surface (MoP-
K). Under CO/H2 feed, the phosphide surface resides near the
optimal for methanol production (Fig. 1a). This suggests that it
should be among the most active catalysts. In the presence of
This article is protected by copyright. All rights reserved.

10.1002/anie.201806583
Angewandte Chemie International Edition
COMMUNICATION
Energy Dispersive Spectroscopy (EDS) confirmed a bulk Mo:P
ratio of 1.06 and MoP phase formation was verified through both
XRD (Fig. S3) and X-ray Photoelectron Spectroscopy (XPS) (Fig.
S9). XPS also revealed the presence of an oxide layer that is
reduced upon exposure to hydrogen before catalytic testing (Fig.
S9). When considering a MoP surface with a high concentration
of surface K (1 K atom per 4 P atoms), we find that the CO and
OH descriptors are comparable to pristine MoP, suggesting that
the effect on activity is small (Supporting Information). Fig.2
displays the TEM images showing MoP particles on the support
(a,b) and the activity and product selectivity under methanol
synthesis conditions (31 bar, 230oC) over varying CO/CO2/H2
feeds in a flow reactor(c). MoP K/SiO2 exhibits remarkable
activity in the full range of CO/CO2/H2 feed compositions, with
methanol as the main product, consistent with our theoretical
model. MoP K/SiO2 shows no sign of deactivation over 54 hours
on stream during the feed switching study shown in Fig.2.
Investigations over more industrially relevant time scales
discussed in recent publications[14] have also prompted us to
perform a longer (150 hours) test under hydrothermally sintering
conditions of 4.2% initial CO2 conversion. A detailed discussion
of these results is presented in the Supporting Information and
stable production of methanol is demonstrated (Fig.S15). Our
calculations (SI) show that phosphorus loss through phosphine
generation is thermodynamically unfavorable under these
conditions, in agreement with prior reports of MoP as a highly
stable compound of molybdenum[15], maintaining its chemical
Figure 3: IR bands corresponding to adsorbed formate species detected via
in-situ DRIFTS measurements over MoP K/SiO2. Conditions: P = 10 bar,
T=230oC, H2:CO2=3
We investigated ceria as a promoter based on promising results
reported on Cu/CeOx catalysts[17], by depositing 1wt.% CeO2 on
MoP K/SiO2. Ceria addition to MoP K/SiO2 causes a pronounced
increase in both methanol selectivity and productivity (Fig.4).
The selectivity of CeO2 MoP K/SiO2 is similar to a commercial
(and highly optimized) Cu/ZnO/Al2O3 catalyst tested under
identical conditions (Fig.4b). MoP and CeO2 promoted MoP also
outperform (on a surface area normalized activity basis) an
unpromoted copper catalyst (Cu/SiO2), included as reference to
provide a range of activity for copper, the optimum metal for
methanol synthesis (Fig.4).
structure up to 950oC in hydrogen[15b]
.
Figure 2. TEM images showing a) MoP particles on SiO2 and b) an exposed
(100) facet of MoP. c) MoP K/SiO2 activity and selectivity at varying CO/CO2
ratios in feed (selectivity excludes CO and CO2 in all conditions)
In an effort to experimentally probe formate binding during
CO2 hydrogenation, we employed in-situ DRIFTS under
methanol synthesis conditions (H2/CO2=3, 10 bar and 230oC)
over MoP K/SiO2. The most prominent bands from the IR
spectrum are displayed in Fig.3 and the full spectrum along with
a discussion of peak assignments is presented in the Supporting
Information. The bands at 1601 and 1354 cm-1 are assigned to
stretching modes of formate, the asymmetric (ν a(OCO)) and
symmetric (νs(OCO)) modes respectively[16]. The separation (Δν)
between these bands is consistent with monodentate formate[16]
.
Given the results from microkinetic modeling presented earlier,
along with these spectroscopy results and associated reactor
testing under identical conditions that shows considerable
methanol production (Fig.S12), we interpret monodentate
formate as the main surface species during CO2 hydrogenation
to methanol over MoP catalysts.
Figure 4: a) Surface area-normalized methanol formation rates and b) product
selectivities during CO2 hydrogenation over MoP and Cu based catalysts.
Conditions: P = 31 bar, T = 230oC H2/CO2 = 3 t=3h, WHSV=10.8 h-1. CO2
conversion is less than 2% for all catalysts. All catalysts pretreated under
identical conditions (400oC, H2).
This article is protected by copyright. All rights reserved.

10.1002/anie.201806583
Angewandte Chemie International Edition
COMMUNICATION
presence of MoO3. In the CeO2 promoted sample this shoulder
and the high energy minimum are suppressed while the lower
energy minimum upshift of 0.5 eV is consistent with the
presence of MoO2. Accordingly, the second derivative minimum
at 2522.5 eV characteristic of MoO2 overlaps with the minimum
due to MoP.
These XAS results show that CeO2 interacts only with an
underphosphided Mo phase while the MoP phase remains
unaffected, consistent with a site blocking promotion effect.
Detailed investigation of promoter effects on the MoP system is
the subject of ongoing work. In-situ DRIFTS during methanol
synthesis from CO2 (10 bar and 230oC) over the ceria-promoted
MoP catalyst showed no new surface species during the
reaction, consistent with our interpretation of the observed
promotion effect (Fig.S14). The existence of underphosphided
regions as determined by XAS is consistent with the
methanation activity of MoP K/SiO2 catalysts (Fig.2). We show in
Fig. 1 from theoretical calculations that methanation activity is
attributed to Mo terminated or Mo-rich sites. Our findings
highlight the complex nature of combining theoretical and
experimental approaches in catalyst discovery, and the
necessity of thorough characterization of catalysts, particularly
using operando techniques to pinpoint catalyst structure under
real operating conditions.
Herein, we have developed a new methanol synthesis
catalyst with
a
stable performance irrespective of the
composition of CO and CO2 in the feed. By combining theory
with experiments, we identified the role of formate in modulating
the catalytic activity, which allowed us to develop a set of MoP
catalysts with surface area normalized activities that surpass the
performance of conventional metal catalysts by circumventing
scaling limitations. This study illustrates how the unique
properties of transition metal phosphides can be exploited to
overcome scaling limitations of conventional metal catalysts.
Rational catalyst synthesis based on both theoretical and
experimental insight is instrumental for catalyst discovery within
the vast and largely unexplored research space of transition
metal compounds for catalysis.
Figure 5: a) Normalized spectra at the P K-edge and b) first derivative, from
top to bottom: MoP K/SiO2, CeO2 MoP K/SiO2, MoP c)Normalized spectra at
the Mo L3 edge and d)second derivative, from top to bottom: MoO2, MoO3,
CeO2 MoP K/SiO2, MoP K/SiO2, MoP. Air exposed samples and standard.
Acknowledgements
Primary support by the U.S. Department of Energy (DOE) Office
of Basic Energy Sciences to the SUNCAT Center for Interface
Science and Catalysis is gratefully acknowledged. Part of this
work was conducted at the Stanford Nano Shared Facilities
(SNSF), supported by the National Science Foundation under
award ECCS-1542152. Use of the Stanford Synchrotron
Radiation Lightsource (SSRL), SLAC National Accelerator
Laboratory, is supported by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences under
Contract No. DE-AC02-76SF00515. A.G. acknowledges the
Department of Energy, Laboratory Directed Research and
Development funding, under Contract No. DE-AC02-76SF00515.
M.S.D. and A.G. thank Matthew Latimer and Erik Nelson for
technical support at beamline 14-3 (SSRL).
While the methanol productivity and selectivity have
increased, the total conversion of CO2 decreases upon CeO2
deposition on the catalyst surface, indicative of a site blocking
effect. Theoretical calculations on MoP, MoP with vacancy sites,
and MoP with fully exposed Mo sites (all P removed) suggest
that catalysts with increasing number of Mo sites are intrinsically
more active for methanation and RWGS (Fig.1). The promotion
effect of CeO2 is hence attributed to blocking these sites.
XAS was performed to identify structural differences that
might reveal the nature of the CeO2 promotion effect (Fig.5).
The comparison of the normalized spectra and first derivative of
the P K edge XANES region in Fig. 5 (a,b) for MoP K/SiO2 and
CeO2 MoP K/SiO2 clearly shows that CeO2 does not affect the P
species present. The spectroscopic features for metal phosphide
(2140-2150 eV)[18] and phosphate (maximum at 2152.5 eV)[19]
are identical in both samples. The XANES region for the two
catalysts reveal the presence of two main Mo species[20]. The
position of the first inflection point of the white line at 2521.2 eV
and the first minimum at 2522.1 eV in the second derivative for
the samples are indicative of MoP. Higher energy features
characteristic of MoO3 and MoO2 are evidenced for MoPK/SiO2
and CeO2 MoP K/SiO2, respectively. In particular, the feature at
2528 eV in the spectrum of MoP K/SiO2 and the two minima at
2525.0 eV and at 2528.6 eV in the second derivative reveal the
Keywords: methanol synthesis • molybdenum phosphide •
formate • CO hydrogenation • CO2 hydrogenation
This article is protected by copyright. All rights reserved.

10.1002/anie.201806583
Angewandte Chemie International Edition
COMMUNICATION
References
[1]
[2]
G. A. Olah, Angewandte Chemie International Edition 2013, 52,
104-107.
aM. D. Porosoff, B. Yan, J. G. Chen, Energy & Environmental
Science 2016, 9, 62-73; bJ. A. Rodriguez, P. Liu, D. J. Stacchiola,
S. D. Senanayake, M. G. White, J. G. Chen, ACS Catalysis 2015,
5, 6696-6706.
[3]
aM. Behrens, F. Studt, I. Kasatkin, S. Kühl, M. Hävecker, F. Abild-
Pedersen, S. Zander, F. Girgsdies, P. Kurr, B.-L. Kniep, M. Tovar,
R. W. Fischer, J. K. Nørskov, R. Schlögl, Science 2012, 336, 893-
897; bL. C. Grabow, M. Mavrikakis, ACS Catalysis 2011, 1, 365-
384; cS. Kuld, M. Thorhauge, H. Falsig, C. F. Elkjær, S. Helveg, I.
Chorkendorff, J. Sehested, Science 2016, 352, 969-974; dF. Studt,
M. Behrens, E. L. Kunkes, N. Thomas, S. Zander, A. Tarasov, J.
Schumann, E. Frei, J. B. Varley, F. Abild-Pedersen, J. K. Nørskov,
R. Schlögl, ChemCatChem 2015, 7, 1105-1111.
[4]
[5]
O. Martin, J. Pérez-Ramírez, Catalysis Science & Technology
2013, 3, 3343-3352.
A. J. Medford, A. C. Lausche, F. Abild-Pedersen, B. Temel, N. C.
Schjødt, J. K. Nørskov, F. Studt, Topics in Catalysis 2014, 57, 135-
142.
[6]
[7]
[8]
[9]
F. Abild-Pedersen, J. Greeley, F. Studt, J. Rossmeisl, T. R. Munter,
P. G. Moses, E. Skúlason, T. Bligaard, J. K. Nørskov, Physical
Review Letters 2007, 99, 016105.
F. Studt, I. Sharafutdinov, F. Abild-Pedersen, C. F. Elkjær, J. S.
Hummelshøj, S. Dahl, I. Chorkendorff, J. K. Nørskov, Nature
Chemistry 2014, 6.
aJ. F. Edwards, G. L. Schrader, The Journal of Physical Chemistry
1984, 88, 5620-5624; bI. Nakamura, H. Nakano, T. Fujitani, T.
Uchijima, J. Nakamura, Surface Science 1998, 402–404, 92-95.
J. S. Yoo, F. Abild-Pedersen, J. K. Nørskov, F. Studt, ACS
Catalysis 2014, 4, 1226-1233.
[10]
[11]
[12]
J. Ye, C. Liu, D. Mei, Q. Ge, ACS Catalysis 2013, 3, 1296-1306.
S. T. Oyama, Journal of Catalysis 2003, 216, 343-352.
J. Kibsgaard, C. Tsai, K. Chan, J. D. Benck, J. K. Norskov, F.
Abild-Pedersen, T. F. Jaramillo, Energy & Environmental Science
2015, 8, 3022-3029.
[13]
[14]
aS. F. Zaman, K. J. Smith, Catalysis Communications 2009, 10,
468-471; bS. F. Zaman, K. J. Smith, Applied Catalysis A: General
2010, 378, 59-68.
aO. Martin, A. J. Martín, C. Mondelli, S. Mitchell, T. F. Segawa, R.
Hauert, C. Drouilly, D. Curulla‐Ferré, J. Pérez‐Ramírez,
Angewandte Chemie International Edition 2016, 55, 6261-6265; bT.
Lunkenbein, F. Girgsdies, T. Kandemir, N. Thomas, M. Behrens, R.
Schlögl, E. Frei, Angewandte Chemie 2016, 128, 12900-12904.
aP. Liu, J. A. Rodriguez, Catal Lett 2003, 91, 247-252; bZ. Yao, Z.
Lai, X. Zhang, F. Peng, H. Yu, H. Wang, Materials Research
Bulletin 2011, 46, 1938-1941.
aP. G. Lustemberg, M. V. Bosco, A. Bonivardi, H. F. Busnengo, M.
V. Ganduglia-Pirovano, The Journal of Physical Chemistry C 2015,
119, 21452-21464; bG. Busca, V. Lorenzelli, Materials Chemistry
1982, 7, 89-126; cA. Bandara, J. Kubota, A. Wada, K. Domen, C.
Hirose, The Journal of Physical Chemistry 1996, 100, 14962-
14968; dM. Calatayud, S. Collins, M. Baltanas, A. Bonivardi,
Physical Chemistry Chemical Physics 2009, 11, 1397-1405; eM.
Marwood, R. Doepper, A. Renken, Applied Catalysis A: General
1997, 151, 223-246; fC. Li, K. Domen, K.-i. Maruya, T. Onishi,
Journal of Catalysis 1990, 125, 445-455; gC. Binet, M. Daturi, J.-C.
Lavalley, Catalysis Today 1999, 50, 207-225.
[15]
[16]
[17]
J. Graciani, K. Mudiyanselage, F. Xu, A. E. Baber, J. Evans, S. D.
Senanayake, D. J. Stacchiola, P. Liu, J. Hrbek, J. F. Sanz, J. A.
Rodriguez, Science 2014, 345, 546-550.
[18]
[19]
[20]
A. P. Grosvenor, R. G. Cavell, A. Mar, Journal of Solid State
Chemistry 2007, 180, 2702-2712.
R. Franke, J. Hormes, Physica B: Condensed Matter 1995, 216,
85-95.
G. George, W. Cleland, J. Enemark, B. Smith, C. Kipke, S.
Roberts, S. P. Cramer, J. Am. Chem. Soc 1990, 112, 2541-2548.
This article is protected by copyright. All rights reserved.

10.1002/anie.201806583
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents
Layout 1:
COMMUNICATION
Melis S. Duyar+[a,b], Charlie Tsai+[a,b], Jonathan L.
Snider[a,b], Joseph A. Singh[c], Alessandro Gallo[a,b,c,e]
,
MoP is presented as a new catalyst for CO
and CO2 hydrogenation to methanol. By
stabilizing only the monodentate
configuration of formate, the phosphide is
able to circumvent the (scaling)
relationship between the necessary ability
of the catalyst to bind oxygen-containing
intermediates and catalyst deactivation
due to high formate coverage in the
presence of CO2.
Jong Suk Yoo[a], Andrew J. Medford[f], Frank Abild-
Pedersen[b], Felix Studt[g], Jakob Kibsgaard[e], Stacey
F. Bent[a], Jens K. Nørskov*[a,b], Thomas F.
Jaramillo*[a,b]
Page No. – Page No.
Discovery of a highly active molybdenum
phosphide catalyst for methanol synthesis
from CO and CO2
This article is protected by copyright. All rights reserved.