10.1002/anie.201903802
Angewandte Chemie International Edition
COMMUNICATION
Monomeric Copper (II) Sites Supported on Alumina Selectively
Convert Methane to Methanol
Jordan Meyet, Keith Searles, Mark A. Newton, Michael Wörle, Alexander P. van Bavel, Andrew D. Horton, Jeroen A. van Bokhoven*,
Christophe Copéret*
Abstract: Monomeric CuII sites supported on alumina, prepared via
surface organometallic chemistry, selectively convert CH4 to CH3OH.
This reaction takes place via formation of CH3O surface species with
the concomitant reduction of two monomeric CuII sites to CuI,
according to mass balance analysis, IR, solid-state NMR, XAS and
EPR studies. This material contains a significant fraction of Cu active
sites (22 %) and displays a selectivity for CH3OH exceeding 83 %,
based on the number of electrons involved for this transformation.
These alumina-supported CuII sites show short reaction time and
demonstrate that C-H bond activation along with methoxy surface
species formation can occur on pairs of monomeric CuII sites.
A major challenge in these systems is associated with the
presence of a small fraction of active sites (often below 30%
even if the presence of 90% of active sites – 0.47 mol
CH3OH/mol Cu – have been recently reported),[18,19] rendering
unequivocal conclusions regarding active site structure
difficult.[13] Cu supported on non-zeolitic materials namely
amorphous silica, has also shown to be reactive for this
transformation, indicating that confinement is not required, but
the methanol yield does not exceed 3.6% CH3OH per Cu.[20,21]
Surface organometallic chemistry (SOMC)[22,23] combined
with thermolytic molecular precursors (TMP)[24] has emerged as
a powerful approach to generate supported isolated metal sites
with tailored nuclearity and oxidation state for a broad range of
metals.[25,26] This approach consists of grafting tailored molecular
precursors on supports with a controlled OH density, followed by
a post-treatment that removes organic ligands and generates
the desired isolated metal sites. Since nuclearity and local
environment are both key features for the selective CH4 to
CH3OH conversion on Cu sites, we reasoned that the
SOMC/TMP approach could be ideal to generate monomeric
CuII sites and to evaluate their reactivity for this reaction.
CH4, the main constituent of natural gas, is increasingly
produced by fracking and widely available at extraction sites for
fossil fuel. Its transport as liquefied natural gas is, however,
expensive and energy intensive, hence it is often flared despite
the highly negative environmental impacts and loss of resource.
Finding efficient routes to convert on-site CH4 to liquid fuels or
chemicals is, therefore, of significant economic and
environmental importance. As such, its direct conversion to
CH3OH is noteworthy since it provides a liquid that can be
directly used as chemical feedstock, fuel or fuel additive.
However, this process remains a grand challenge because over-
oxidation is favored due to the higher reactivity of CH3OH than
CH4.[1–5] In nature, bacterial enzymes, such as methane mono-
oxygenases, are able to oxidize CH4 to CH3OH in high selectivity
using O2 as primary oxidant.[6] These enzymes, containing Cu or
Fe metal centers, have been a source of inspiration to design
materials for the conversion of CH4 to CH3OH under mild
reaction conditions.[7] Attempts to perform this reaction
catalytically with tailored materials have been mostly
unsuccessful since at high conversion the selectivity towards
CH3OH is low. This hurdle can be in principle overcome by using
the concept of chemical looping, that consists in decoupling the
steps of oxidation of Cu sites and their reaction with CH4.[8] The
last step of the cycle consists of the desorption of the methoxy
species strongly bound to the copper sites using a protic solvent.
Of the various systems that have been investigated, Cu-
exchanged zeolites have shown potential for the selective
oxidation of CH4 to CH3OH using molecular oxygen as oxidant
Here, we report the synthesis of a monomeric CuII siloxide
molecular
precursor
[Cu(OSi(OtBu)3)2(TMEDA)]
(1;
tetramethylethylenediamine, TMEDA) and its use to generate
isolated CuII sites on alumina with the goal to preserve a silicon
and aluminum environment, as found in zeolites, without the
microporous environment (Figure 1b-c). We also prepared a
dimeric complex, [Cu(2,µ2-OSi(OtBu)3)(OSi(OtBu)3)]2 (2, Figure
1b), as a spectroscopic molecular probe for oxygen-bridged
dimeric Cu species and to evaluate its surface chemistry
towards the generation of CuII sites. Using a combination of
spectroscopic methods (UV-Vis, EPR, XAS), we show that
independently of the molecular precursors 1 or 2, mononuclear
CuII sites in an alumino-silicate environment are mostly formed
after thermal treatment at 700 ºC. These mononuclear sites
show unprecedented reactivity, activating CH4 within minutes at
200 °C and yielding CH3OH after hydrolysis in high selectivity
(>80%). CH4 activation takes place on a high fraction of proximal
monomeric CuII sites (22% of Cu) dispersed at the alumina
surface via a 2-electron oxidation process, indicating that di- or
tri-nuclear Cu-µ-oxo sites and/or micropores are not required for
the selective conversion of CH4 to CH3OH, thus contrasting the
expected situation in the corresponding zeolitic materials.
First, the dimeric complex (2) was prepared by reaction of
[Cu(OTf)2]x with 2 equiv. of NaOSi(OtBu)3 in THF and is further
converted to [Cu(OSi(OtBu)3)2(TMEDA)] (1) upon addition of
TMEDA. The complex (1) is isolated as light green single
crystals in 54% overall yield; it crystalizes in the P-1 space
group, where Cu occupies a distorted tetrahedral environment
(4’ = 0.65) as found in [Cu(OSi(OtBu)3)2(py)2].[27] This complex is
associated with (i) a typical EPR spectrum with g⊥ > g‖ ≈ ge and
a dz2 ground state (g⊥ = 2.27; g‖ = 2.00 A‖ ≈ 450 MHz, Fig. S1)
source.[9–13] The structure of the active sites is highly debated,
[9,14]
with evidence for both (µ-oxo) dicopper
and tris(µ-oxo)
trinuclear copper centers (Figure 1a);[15] recent theoretical
studies also suggest that monomeric Cu sites should not be
excluded.[16,17]
[*] J. Meyet, Dr. K. Searles, Dr. M. A. Newton, Dr. M. Wörle, Prof. Dr. C.
Copéret, Prof. Dr. J. A. van Bokhoven
Department of Chemistry and Applied Biosciences, ETH Zurich
Vladimir-Prelog-Weg 1-5, 8093 Zürich, Switzerland
E-mail: ccoperet@inorg.chem.ethz.ch
and (ii) an optical spectrum with characteristic max at 267 and
317 nm associated with the ligand to metal charge transfer
(LMCT) of the two terminal siloxides ligands and the two
nitrogen ligands bound to Cu, respectively (Fig. S2). For (2),
[Cu(2,µ2-OSi(OtBu)3) (OSi(OtBu)3)]2, isolated in 56% yield as
light green single crystals, the two Cu centers adopt a distorted
square planar geometry (4’ = 0.38 Cu1 and 0.27 Cu2),
Prof. Dr. J. A. van Bokhoven
Laboratory for Catalysis and Sustainable Chemistry, Paul Scherrer Institute
Villigen 5232, Switzerland
E-mail: jeroen.vanbokhoven@chem.ethz.ch
Dr. A. P. van Bavel, Dr. A. D. Horton
Shell Global Solutions International B.V.
Grasweg 31, 1031 HW Amsterdam, The Netherlands
This article is protected by copyright. All rights reserved.