The role of the oxide component in the development of copper composite catalysts for methanol synthesis

Article abstract of DOI:10.1002/anie.201301419

The design of solid catalysts for industrial processes remains a major challenge in synthetic materials chemistry. Based on the investigation of the industrial Cu/ZnO/Al₂O₃ catalyst, a modular concept is introduced that helps to develop novel methanol synthesis catalysts that operate in different feed gas mixtures. SA=surface area, SMSI=strong metal-support interaction. Copyright

Full text of DOI:10.1002/anie.201301419

.
Angewandte
Communications
DOI: 10.1002/anie.201301419
Catalyst Design
The Role of the Oxide Component in the Development of Copper
Composite Catalysts for Methanol Synthesis**
Stefan Zander, Edward L. Kunkes, Manfred E. Schuster, Julia Schumann, Gisela Weinberg,
Detre Teschner, Nikolas Jacobsen, Robert Schlçgl, and Malte Behrens*
The development and optimization of industrially applied
high-performance catalysts is usually a continuous process
that is to a large extent based on the experience of the
manufacturer. The accumulated knowledge from the combi-
nation of empirical trial-and-error experimentation and
a subsequent structure–function-relationship-guided optimi-
zation approach within the boundary conditions of a feasible
and scalable synthesis often leads to very complex recipes,
which are sometimes generalized as the “black magic” of
catalyst preparation. In the last few years, we have worked on
better understanding of the synthesis and functionality of the
Cu/ZnO/Al2O3 methanol synthesis catalyst using the well-
documented industrially applied preparation route[1] as start-
ing point. As a result of this effort, we have elaborated
a model of the so-called “chemical memory”[2] of catalyst
preparation and of the active site in this catalyst system.[3]
Herein, we show how this knowledge can be applied to
develop a new family of copper-based catalysts.
a porous microstructure[2a,6] from a co-precipitated precursor
compound. In this context, ZnO acts as a geometrical spacer
between the Cu nanoparticles and helps to increase and
stabilize the Cu dispersion.[4b,7] Thus, ZnO has two functions
in the final catalyst: 1) As nanoparticles it acts as a physical
spacer between the Cu particles, stabilizing the porous
microstructure; and 2) as a thin layer at the surface of the
Cu particles it is an essential ingredient for the active site, and
its presence has been shown to affect the adsorption proper-
ties.[8] The work presented herein was guided by the idea of
separating these two effects.
A simplified scheme of the relevant properties of Cu/ZnO
methanol synthesis catalysts is shown in Figure 1a. Three
With the help of structure–performance relationships
observed within a series of functional powder catalysts and
DFT calculations, the active site of industrial methanol
synthesis could be identified as a combination of a surface
defect of Cu and the presence of partially reduced Zn species
at this defect,[3] explaining the widely studied Cu-ZnO
synergy.[4] In the industrial synthesis, high concentrations of
these sites can be realized by preparation of defective Cu
nanoparticles and migration of ZnOx species onto the Cu
surface as a result of a strong metal–support interaction
(SMSI)[3,5] and an intimate interface contact of both catalyst
components. At the same time, the total accessible Cu surface
area (SACu) is large, because the bulk catalyst is prepared with
Figure 1. a) Schematic representations of the necessary ingredients for
a high-performance methanol synthesis catalyst. b) The role of precur-
sor composition for the Cu dispersion in the final catalyst.
[*] Dr. S. Zander, Dr. E. L. Kunkes, Dr. M. E. Schuster, J. Schumann,
G. Weinberg, Dr. D. Teschner, Prof. R. Schlçgl, Dr. M. Behrens
Department of Inorganic Chemistry
prerequisites have to be fulfilled to generate a high-perfor-
mance catalyst. The material should have a high SACu to
expose a large number of active sites; the Cu phase must be
defective to achieve a high density of active sites at the
surface; and SMSI-induced synergetic effect of ZnO must be
present to activate the defect sites for methanol synthesis.
Only if all three factors come together (in the darkest shaded
region of Figure 1a) will the catalyst be highly active for
methanol synthesis. The defects are generated by the careful
and delicate preparation method yielding distorted Cu nano-
particles[9] in close contact to the oxide phase, while the other
two properties are governed by function (1) and (2) of the
ZnO component.
Fritz-Haber-Institut der Max-Planck-Gesellschaft
Faradayweg 4–6, 14195 Berlin (Germany)
E-mail: behrens@fhi-berlin.mpg.de
Dr. N. Jacobsen
Clariant Produkte (Deutschland) GmbH, BU Catalysts
Bruckmꢀhl (Germany)
[**] Edith Kitzelmann (XRD measurements), Achim Klein-Hoffmann
and Olaf Timpe (XRF), Gisela Lorenz (BET measurements) and
Nygil Thomas (help with catalytic measurements) are acknowl-
edged. Financial support was given by Clariant Produkte
(Deutschland) GmbH and the Bayerisches Wissenschaftsministe-
rium (NW-0810-0002, since 2010). We thank Martin Muhler and
Graham J. Hutchings for fruitful discussions.
The synthesis route for preparing Cu/ZnO/(Al2O3) cata-
lysts follows a multistep procedure including temperature-
and pH-controlled co-precipitation[2b] of aqueous Cu,Zn,Al
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201301419.
6536
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 6536 –6540

Angewandte
Chemie
nitrate solution with sodium carbonate solution followed by
aging,[10] washing, and drying to yield a hydroxide–carbonate
precursor. This material is calcined and finally activated by
reduction of CuO to Cu metal. Low amounts of Al2O3 acts as
a structural promoter in the industrial catalyst.[11]
The relevant precursor material has been identified as
thin needles of zincian malachite, (Cu,Zn)2(OH)2CO3.[2a] The
incorporation of Zn2+ into the cationic lattice of malachite
favors the nanostructuring of the CuO/ZnO aggregates
formed upon calcination owing to the perfect distribution of
both species in the joint crystal lattice of the precursor
compound. This can be understood as a purely geometric
effect, which is shown in Figure 1b and is the basis for the first
functionality of ZnO. Zn2+ is well-suited for this purpose
because it exhibits the same charge and an ionic radius similar
to Cu2+ favoring substitution in the precursor. However,
incorporation into the malachite lattice is limited to less than
30% owing to solid-state chemical constraints[12] that are most
likely due to the differences in the coordination environment
between the Jahn–Teller-ion Cu2+ (d9) and Zn2+ (d10). Mg2+ is
an interesting replacement for Zn2+, because its charge
matches and its ionic radius differs, alike Zn2+, by less than
2% from that of Cu2+. Furthermore, (Cu,Mg)2(OH)2CO3
crystallizes in the rosasite crystal structure, which is closely
related to that of malachite and therefore provides an
opportunity for a comparable precursor chemistry between
Cu,Zn and Cu,Mg compounds. Moreover, (Cu,M-
g)2(OH)2CO3 is naturally occurring as the mineral McGui-
nessite[13] that can incorporate even more Mg2+ than Cu2+,
which to date could not be achieved for synthetic zincian
malachite. Thus, an even more efficient dilution of the Cu2+
ions might be possible with Mg2+ compared to Zn2+ by
application of lower amounts of Cu to further promote the
nanostructuring and increase the Cu dispersion.
In this work we compare the classical zinc-containing
malachite-derived Cu/ZnO with a new Cu/MgO catalyst at
a fixed molar ratio of Cu to Zn and Mg, respectively, of 80:20.
At this ratio, Zn incorporation into malachite does not exceed
the critical Zn concentration in zinc-containing malachite to
assure synthesis of phase-pure precursor compounds, result-
ing in high comparability of the Cu,Zn and Cu,Mg systems
and in uniform catalysts whose properties can be easily traced
back to the phase-pure precursor compounds. Both precur-
sors were prepared from mixed nitrate solutions by controlled
co-precipitation with sodium carbonate solution and subse-
quent ageing in the mother liquor. They are denoted CZ and
CM in the following. Impregnation of the calcined CM with
5 wt% ZnO yielded a catalyst labeled CMZ that is discussed
later.
Figure 2. a) XRD patterns of the precursor materials of CZ (light gray)
and CM (dark gray). The reference pattern is malachite (black bar
graph; ICSD: 72-75). b) XRD patterns of the calcined samples CZ
(light gray), CM (dark gray), and CMZ (black). The reference pattern is
CuO (black bar graph; ICSD: 80-76). c) SEM images of CZ and d) of
CM.
Table 1: Properties of the CZ, CM, and CMZ catalysts.[a]
[e]
Sample
Cu:M
ratio
D[d] [nm]
prec./cal.
SABET [m2 gÀ1
prec./cal.
]
SACu
[m2 gcat
]
À1
CZ
CM
CMZ
80:20[b]
83:17[b]
79:16:5[c]
26.6/5.8
8.0/2.8
8.0/3.9
36/83
81/73
81/80
16.0
20.3
24.2
[a] prec.=precursor material, cal.=calcined material. [b] molar, deter-
mined by XRF, Æ1 mol%. [c] molar, estimated. [d] Crystallite domain
size of (Cu,M)2(OH)2CO3 (precursor) and CuO/MO (calcined), Æ0.2 nm
derived from XRD peak profiles. [e] Specific Cu surface area of the
reduced catalyst determined by N2O chemisorption, Æ1 m2 gÀ1
.
malachite structure in both samples. It is noted that CM
exhibits significantly broader XRD peaks indicative of
smaller crystallites. Also the particle morphology of CM
(Figure 2d) looks rather spongy compared to CZ, which
exhibits larger and well-separated particles (Figure 2c).
Accordingly, a larger BET surface area of the CM precursor
has been observed (Table 1).
Upon calcination at 603 K, poorly crystalline CuO is
formed in both samples, as evidenced by XRD (Figure 2b),
while the ZnO and MgO components are mostly X-ray
amorphous. Again CM exhibits a significantly smaller crys-
tallite size according to the XRD peak width, but a slightly
smaller specific surface area (Table 1). Furthermore, CM
yields a by more than 20% higher SACu than CZ after
reduction.
TEM investigation of the reduced catalysts showed that
the general microstructure of CZ (Figure 3a,b) and CM
(Figure 4a–c) is similar in both catalysts and characterized by
arrangements of round shaped Cu particles separated by
differently sized crystallites of ZnO or MgO, respectively. The
presence of larger Cu particles in CZ compared to CM is
consistent with the difference in Cu surface areas (Table 1).
X-Ray diffraction (XRD) of the precursors confirmed the
formation of single-phase materials with a crystal structure
similar to malachite (Figure 2a). In comparison to the
literature pattern of malachite, a pronounced shift of the
¯
201 peak is seen in both compounds. This is an indication for
the incorporation of non-Jahn–Teller cations in the lattice of
malachite, and from the similar angular position of the
reflections in both patterns a similar degree of substitution
can be estimated (Table 1), suggesting that the non-Jahn–
Teller ions have been completely incorporated into the
Angew. Chem. Int. Ed. 2013, 52, 6536 –6540
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6537

.
Angewandte
Communications
Figure 3. (HR-)TEM images of the reduced CZ catalyst. The insets
show power spectra of the neighboring particles and are used for
phase identification.
Figure 5. Surface and near-surface composition of the reduced cata-
lysts recorded with synchrotron XPS at a function of information
depth. a) CZ; b) CM; c) CMZ. Error bars represent estimated uncer-
tainty based on several fits with random variation of background fitting
parameters.
depth-sensitive XPS measurements (Figure 5a; Supporting
Information, Figure S2); a pronounced Zn enrichment at the
surface of the catalysts was observed. This deviation from the
nominal 80:20 ratio is slowly and continuously lowered if the
information depth is increased to 2 nm and beyond exceeding
the first surface layers. This trend is in agreement with
previous results obtained on Cu/ZnO-based catalysts[3,14] and
indicates the presence of (some of) the Zn species at the
surface in form of a thin overlayer. The situation is different
for CM, where beyond 1 nm Cu is already the most abundant
near-surface species, as expected from the nominal composi-
tion of CM. The measurements indicate Mg enrichment only
at the very outermost surface and a lower tendency of
overlayer formation, which is in agreement with the expected
weaker metal–oxide interaction in CM compared to CZ,
owing to the less reducible nature of MgO.
Figure 4. (HR-)TEM images of the reduced CM catalyst.
High-resolution TEM showed that the Cu particles in
both samples contain planar defects, which have been shown
to contribute to the methanol synthesis activity in Cu/ZnO
catalysts[3] (CZ: Figure 3c; CM: Figure 4d and Supporting
Information, Figure S1). Thus, the important defectiveness of
Cu is present in both catalysts CZ and CM and is probably
a result of the precursor decomposition approach common to
both catalysts, which leads to crystallization of distorted Cu
crystallites.
This result indicates that MgO is an intrinsically better
geometrical spacer compared to ZnO as even at the non-ideal
80:20 ratio, Cu particles can be obtained that with an average
size below 10 nm are similarly small as found in state-of-the-
art Cu/ZnO/Al2O3 catalysts.[6] Thus, the structurally promot-
ing role (1) of ZnO has been successfully replaced with MgO.
The formation of a thin ZnOx overlayer[3,5b] upon reduc-
tion of CZ was not directly observed by HRTEM, but the
surface coverage of the Cu particles was clearly probed by
6538
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 6536 –6540

Angewandte
Chemie
Cu/MgO aggregates (Supporting Information, Figure S3c).
Interestingly, the XPS results obtained on CMZ showed that
the concentration profile of Mg was hardly changed by the
impregnation and subsequent reduction (Figure 5c). How-
ever, the surface abundance of Cu was significantly lowered at
the expense of Zn, with a clear surface enrichment of the
latter as shown by the trends of the Cu and Zn depth profiles
(Figure 5c, inset). Thus, the dramatic change in the catalytic
performance appears to be associated with the formation of
a ZnOx overlayer in CMZ that has been absent in CM.
These results show that the functions of the oxide
component can be successfully separated in Cu-based meth-
anol synthesis catalysts. It was shown for a given catalyst
composition as a proof-of-principle that this approach enables
preparation of high-performance catalysts and leaves addi-
tional degrees of freedom for future optimization. In partic-
ular, Cu dispersion can be optimized within the proven
malachite-precursor method by increasing the Cu substitution
without being bound to the constraints of the Cu,Zn system.
Furthermore, the method of addition and optimal amount of
the synergistic promoter can be varied for a given highly
dispersed Cu/oxide system to switch on the production of
methanol from CO2 or synthesis gas.
Interestingly, the catalytic performance of the samples is
completely changed when a CO/H2 feed is used for methanol
synthesis. Here CM shows a very high methanol production
rate, which clearly exceeds that of CZ or CMZ in the other
feed gases (Figure 6c). This result is in line with previous
studies,[16] that have shown that MgO-supported Cu is
a powerful CO hydrogenation catalyst. Interestingly, while it
was a prerequisite for methanol production in CO2-containing
feeds, the addition of Zn to CM was detrimental in this
reaction possibly by partially covering of the active surface.
Thus, along with being a very powerful CO hydrogenation
catalyst, the CM and derived CMZ also represent a suitable
material basis for basic studies on the roles of synergy,
dispersion, and structural dynamics for methanol synthesis in
different feed gases. In particular, an investigation of the
carbon source for methanol on the ZnO-free Cu/MgO in CO/
CO2/H2 mixtures, which was found to be CO2 for an industrial
Cu/ZnO/Al2O3 catalyst,[17] seems interesting and will be
addressed in our future work.
In summary, the high comparability of the three concept-
catalysts, which is due to the similar general morphology
found by TEM investigation, allows the differences in activity
of the samples to be traced back to the influence of the oxide
phase(s) ZnO and/or MgO. These two oxides do not only act
as structural promoters, but also determine the preferred
pathway of methanol synthesis from CO2 or CO as carbon
source. We propose that the synthetic approach presented
here opens the door to exploit new forms of the traditional
Cu/ZnO/Al2O3 catalyst system that are based upon a solid
functional understanding of the respective components.
Furthermore, the presented materials show potential to
fertilize new progress in studies of the mechanism of
methanol synthesis by providing fundamental insight into
the role of different material components. Finally, with
respect to nonconventional feed compositions available for
methanol as an energy carrier molecule, new possibilities for
Figure 6. Results for the methanol synthesis with the CZ, CM, and
CMZ catalysts in different feed gas compositions at 30 bar and 503 K.
Both catalysts CZ and CM have been tested in methanol
synthesis with various feed gas compositions, that is, hydro-
genation of pure CO2, a CO2/CO mixture and pure CO
(Figure 6, Supporting Information, Table S1). In the hydro-
genation of pure CO2, CZ showed a much higher activity than
CM, showing clearly that the methanol synthesis rate is not
only a function of the exposed Cu surface area alone
(Figure 6a). Following the scheme in Figure 1a and in
accordance with the XPS results (Figure 5), the low activity
of CM can be explained with the absence of the synergetic
SMSI-effect as MgO is an irreducible oxide that does not
show the necessary SMSI in the relevant temperature regime.
The situation is similar if methanol is produced from a typical
synthesis gas mixture with CO2 and CO in the feed (Fig-
ure 6b). CZ shows a slightly lower rate of methanol produc-
tion compared to the CO2/H2 feed, while CM catalyzes the
reverse water gas shift reaction (rWGS), but remains
essentially inactive for methanol synthesis despite the large
exposed Cu surface area. These results strikingly confirm the
crucial synergetic role of the ZnO-promoter that has been
subject of many previous reports.
With the idea of “switching on” the lacking Cu-ZnO
synergy by addition of Zn (similar to that reported earlier for
model catalysts[11b,15] and physical mixtures),[4a,c] the catalyst
CMZ was prepared by impregnation of the calcined CM with
5 wt% ZnO. The procedure resulted in a catalyst that was
indeed able to convert CO2 and the synthesis gas mixture
much better than CM (Figure 6a,b). The weight-based
methanol production rate from synthesis gas of CMZ was
even higher than that of CZ, which is probably a result of the
higher Cu dispersion, as the intrinsic rates per SACu were
similar for CZ and CMZ in this experiment (Supporting
Information, Figure S4). Interestingly, only the formation of
methanol was promoted by the addition of ZnO not the
rWGS (Figure 6b,c), rendering the CMZ catalysts a very
promising material for selective CO2 hydrogenation.
(HR-)TEM images of CMZ are reported as supporting
information (Supporting Information, Figure S3a,b) and show
a similar general microstructure and particle morphology like
CM, but additionally confirmed the presence of ZnO at the
Angew. Chem. Int. Ed. 2013, 52, 6536 –6540
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6539

.
Angewandte
Communications
Tovar, R. W. Fischer, J. K. Nørskov, R. Schlçgl, Science 2012,
336, 893.
catalyst development occur through the present modular
approach composing the relevant functions of the catalyst
according to the chemical potential of the intended gas feed.
[4] a) Y. Kanai, T. Watanabe, T. Fujitani, T. Uchijima, J. Nakamura,
Catal. Lett. 1996, 38, 157; b) M. S. Spencer, Top. Catal. 1999, 8,
259; c) F. Liao, Y. Huang, J. Ge, W. Zheng, K. Tedsree, P. Collier,
X. Hong, S. C. Tsang, Angew. Chem. 2011, 123, 2210; Angew.
Chem. Int. Ed. 2011, 50, 2162.
Experimental Section
Hydroxide–carbonate precursors of CZ and CM were synthesized by
coprecipitation (T= 338 K) from Cu,Zn and Cu,Mg (80:20) nitrate
solutions and Na2CO3 solution as precipitating agent in an automated
lab reactor (LabMax, Mettler Toledo). The pH was set to 6.5 for CZ
and 9.0 for CM. The precipitates were aged (> 60 min), filtered,
washed, and dried. Calcination was carried out in air at 603 K
(2 KminÀ1) for 3 h. One part of the calcined CM was impregnated
with Zn citrate solution, dried, and calcined again at the same T.
Catalytic tests were carried out in a fixed-bed flow reactor with a feed
of 72% H2, 24% CO2, 4% Ar (internal standard) for CO2 hydro-
genation, 59% H2, 6% CO, 8% CO2, for syngas conversion and 14%
CO, 59% H2, 4% Ar for CO hydrogenation; balance was He. Online
analysis of products was performed with a gas chromatograph
(Agilent 7890A). After the start of the reaction, the catalysts were
allowed to stabilize for 6 h at 523 K. Measurements were done at
503 K and 30 bar. More details on the characterization and testing
methods used can be found in the Supporting Information.
[5] a) Y. Kanai, T. Watanabe, T. Fujitani, M. Saito, J. Nakamura, T.
Uchijima, Catal. Lett. 1994, 27, 67; b) J. D. Grunwaldt, A. M.
Molenbroek, N. Y. Topsoe, H. Topsoe, B. S. Clausen, J. Catal.
2000, 194, 452; c) R. N. dꢂAlnoncourt, X. Xia, J. Strunk, E.
Lçffler, O. Hinrichsen, M. Muhler, Phys. Chem. Chem. Phys.
2006, 8, 1525.
[6] I. Kasatkin, P. Kurr, B. Kniep, A. Trunschke, R. Schlçgl, Angew.
Chem. 2007, 119, 7465; Angew. Chem. Int. Ed. 2007, 46, 7324.
[7] T. Fujitani, J. Nakamura, Catal. Lett. 1998, 56, 119.
[8] P. R. Dennison, K. J. Packer, M. S. Spencer, J. Chem. Soc.
Faraday Trans. 1989, 85, 3537.
[9] T. Kandemir, F. Girgsdies, T. C. Hansen, K.-D. Liss, I. Kasatkin,
E. L. Kunkes, G. Wowsnick, N. Jacobsen, R. Schlçgl, M. Behrens,
Angew. Chem. 2013, 125, 5271; Angew. Chem. Int. Ed. 2013, 52,
5166.
[10] a) D. M. Whittle, A. A. Mirzaei, J. S. J. Hargreaves, R. W.
Joyner, C. J. Kiely, S. H. Taylor, G. J. Hutchings, Phys. Chem.
Chem. Phys. 2002, 4, 5915; b) S. Zander, B. Seidlhofer, M.
Behrens, Dalton Trans. 2012, 41, 13413.
Received: February 18, 2013
Published online: May 28, 2013
[11] a) M. V. Twigg, M. S. Spencer, Top. Catal. 2003, 22, 191; b) M.
Kurtz, N. Bauer, C. Buscher, H. Wilmer, O. Hinrichsen, R.
Becker, S. Rabe, K. Merz, M. Driess, R. A. Fischer, M. Muhler,
Catal. Lett. 2004, 92, 49.
[12] M. Behrens, F. Girgsdies, Z. Anorg. Allg. Chem. 2010, 636, 919.
[13] a) N. Perchiazzi, Z. Kristallogr. 2006, 505; b) M. Fleischer, L. J.
Cabri, Am. Mineral. 1981, 66, 1274.
Keywords: CO2 conversion · copper · magnesium ·
.
methanol synthesis · zinc
[1] a) C. Baltes, S. Vukojevic, F. Schꢀth, J. Catal. 2008, 258, 334;
b) D. Waller, D. Stirling, F. S. Stone, M. S. Spencer, Faraday
Discuss. Chem. Soc. 1989, 87, 107; c) J. L. Li, T. Inui, Appl. Catal.
A 1996, 137, 105; d) J. B. Hansen, P. E. H. Nielsen in Handbook
of Heterogeneous Catalysis, Vol. 6, 2nd ed. (Eds.: G. Ertl, G.
Knçzinger, F. Schꢀth, J. Weitkamp), Wiley-VCH, Weinheim,
2008, p. 2920; e) S. Schimpf, M. Muhler, Synthesis of Solid
Catalysts, Wiley-VCH, Weinheim, 2009, p. 329.
[2] a) M. Behrens, J. Catal. 2009, 267, 24; b) M. Behrens, D.
Brennecke, F. Girgsdies, S. Kißner, A. Trunschke, N. Nasrudin,
S. Zakaria, N. F. Idris, S. B. Abd Hamid, B. Kniep, R. Fischer, W.
Busser, M. Muhler, R. Schlçgl, Appl. Catal. A 2011, 392, 93.
[3] M. Behrens, F. Studt, I. Kasatkin, S. Kꢀhl, M. Hꢁvecker, F.
Abild-Pedersen, S. Zander, F. Girgsdies, P. Kurr, B.-L. Kniep, M.
[14] a) W. P. A. Jansen, J. Beckers, J. C. Van der Heuvel, A. W. D.
Van der Gon, A. Bliek, H. H. Brongersma, J. Catal. 2002, 210,
229; b) J. Sloczynski, R. Grabowski, P. Olszewski, A. Kozlowska,
J. Stoch, M. Lachowska, J. Skrzypek, Appl. Catal. A 2006, 310,
127.
[15] R. Becker, H. Parala, F. Hipler, O. P. Tkachenko, K. V.
Klementiev, W. Grꢀnert, H. Wilmer, O. Hinrichsen, M.
Muhler, A. Birkner, C. Wçll, S. Schꢁfer, R. A. Fischer, Angew.
Chem. 2004, 116, 2899; Angew. Chem. Int. Ed. 2004, 43, 2839.
[16] a) B. Denise, R. P. A. Sneeden, Appl. Catal. 1986, 28, 235;
b) J. C. J. Bart, R. P. A. Sneeden, Catal. Today 1987, 2, 1.
[17] G. C. Chinchen, P. J. Denny, D. G. Parker, M. S. Spencer, D. A.
Whan, Appl. Catal. 1987, 30, 333.
6540 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 6536 –6540