вход по аккаунту


MOCVD-Loading of Mesoporous Siliceous Matrices with CuZnO Supported Catalysts for Methanol Synthesis.

код для вставкиСкачать
Heterogeneous Catalysis
MOCVD-Loading of Mesoporous Siliceous
Matrices with Cu/ZnO: Supported Catalysts for
Methanol Synthesis**
Ralf Becker, Harish Parala, Frank Hipler,
Olga P. Tkachenko, Konstantin V. Klementiev,
Wolfgang Grnert, Hagen Wilmer, Olaf Hinrichsen,
Martin Muhler, Alexander Birkner, Christof W&ll,
Sven Sch(fer, and Roland A. Fischer*
The Cu/ZnO system is the basis of industrial methanol
synthesis[1] and an important component of fuel-cell technology.[2] Furthermore, it is a prototype for the exploration of
synergistic metal–support interactions in heterogeneous catalysis.[3] Recently, high-resolution in situ TEM studies
revealed dynamic structural changes of copper nanocrystallites (2–3 nm) deposited onto ZnO, which depended on the
redox potential of the gas phase.[4] The reducing conditions
(H2/CO) of methanol synthesis lead to a flattening of the
copper particles which increases the coverage of the ZnO
support. A positive correlation between the microstructural
strain of ZnO-supported copper nanoparticles and catalytic
activity has been described.[5] The promotion of Cu(111)
surfaces with zinc reveals that the formation of Cu/Zn alloys is
also of importance.[6] A common motif in discussion is the
presence of CuZnOx species at the Cu/ZnO interface, which is
in agreement with recent theoretical studies.[7] These findings
inspired us to look for new strategies in the synthesis of the
catalysts, which allow a molecular control of the Cu/ZnO
interface and offer new possibilities to its maximization.
Owing to their high specific surface area and precisely
controllable pore structures in the lower nm range, periodic
mesoporous silicates (PMS), such as MCM-41, MCM-48,
SBA-15, have proven to be excellent supports for numerous
catalytically active species. In particular the Cu/PMS and
CuOx/PMS have been studied extensively.[8] Less is known
about Cu/ZnO/PMS catalysts.[9] In addition to conventional
aqueous impregnation/calcinations procedures, metal–organic chemical vapor deposition (MOCVD) techniques can
be considered for the preparation of catalysts.[10]
Treating a sample (350 mg) of freshly calcined MCM-41[11]
(1BJH = 2.7 nm, SBET = 712 m2 g 1) under static vacuum
(0.1 Pa) with the vapor of the blue-violet copper precursor
[Cu(OCHMeCH2NMe2)2] (1, 1.0 g)[12] at 340 K in a sealed
Schlenk tube, causes the original white siliceous support to
turn light blue. Comparison of the IR data of loaded MCM-41
with that of pure 1 reveals that 1 is adsorbed intact (see
Supporting Information). The adsorption is strong, since 1
cannot be desorbed even at elevated temperatures (373 K) in
dynamic vacuum (0.1 Pa, 24 h). The interaction of 1 with the
pore walls is clearly through hydrogen bonds, this is indicated
by the absence of the bands at 3745 cm 1, which stem from
free silanol groups in the untreated material. Treating the
support, which is coated with 1, with diethyl zinc vapor (this is
done by placing both samples in a Schlenk tube and sealing it
after evacuation (0.5 g ZnEt2, 300 K, 0.1 Pa)), leads to a slow
change in color from light blue to red brown. The powder Xray diffraction (XRD) pattern of a sample prepared under
inert atmosphere shows a weak, broad reflection at 2 q =
44.708 which is assigned to the reflection of the (111) crystal
plane of small copper particles (Figure 1 and 2). Solid-state
C NMR spectroscopic studies reveal that [Zn(OCHMeCH2NMe2)2] (2) is formed as a byproduct (see Supporting
Information). The conversion taking place in the nanotubes of
the PMS corresponds to the solution-phase reaction outlined
in Scheme 1, in which metallic copper precipitates (XRD),
the zinc alkoxide 2 stays in solution (NMR spectroscopy), and
gaseous butane is evolved (GC-MS). It is observed, that
reducing the reactivity of the zinc alkyl used, for example, by
employing highly bulky alkyl groups, such as C(SiMe3)3, does
[*] Dr. R. Becker, Dr. H. Parala, Dr. F. Hipler, Prof. Dr. R. A. Fischer
Inorganic Chemistry II—Organometallics & Materials Chemistry
Ruhr-University Bochum, 44870 Bochum (Germany)
Fax: (+ 49) 234-321-4174
Dr. O. P. Tkachenko, Dr. K. V. Klementiev, Prof. Dr. W. GrBnert,
Dr. H. Wilmer, Priv.-Doz. Dr.-Ing. O. Hinrichsen, Prof. Dr. M. Muhler
Engineering Chemistry, Ruhr-University Bochum, 44870 Bochum
Dr. A. Birkner, Prof. Dr. C. WDll
Physical Chemistry I, Ruhr-University Bochum, 44870 Bochum
Dipl.-Chem. S. SchFfer
Inorganic Chemistry I—Cluster and Coordination Chemistry, RuhrUniversity Bochum, 44870 Bochum (Germany)
[**] The authors gratefully acknowledge the German science foundation
(DFG) for their generous support within the framework of the
collaborative research centre (SFB) 558 “metal –support interactions in heterogeneous catalysis”, and Prof. H. Gies for providing
high-quality MCM-48.
Supporting information for this article is available on the WWW
under or from the author.
Angew. Chem. Int. Ed. 2004, 43, 2839 –2842
Figure 1. X-ray powder diffractograms of the samples: a) the precursors Cu/[Zn(OCHMeCH2NMe2)2]/MCM-41 obtained at room temperature, b) the catalytically active sample Cu/ZnO/MCM-41 and the reference samples c) ZnO/MCM-41, and d) Cu/MCM-41. The (111), (200),
and (220) reflections of polycrystalline copper are marked at the
DOI: 10.1002/anie.200351166
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Low-angle powder X-ray diffractograms of a) empty calcined
MCM-41 and b) Cu/ZnO/MCM-41. The characteristic reduction of the
intensity of (b) compared to (a) is an indication for the loading of the
pores.[11] Inset: TEM image of Cu/ZnO/MCM-41 which clearly reveals
that the pore structure is intact, and that owing to their small size the
copper or zinc oxide particles can not be seen. The presence of copper
and zinc is verified by the corresponding EDX spectrum.
Scheme 1. Reaction of 1 with diethyl zinc in the nanotubes of the
MCM-41 support.
not lead to an alkyl/alkoxide metathesis and bimetallic alkyl
zinc/copper alkoxide complexes can be isolated and structurally characterized. To date, only microcrystalline and catalytically inactive Cu/ZnO materials were obtained by the solidstate pyrolysis of these bimetallic complexes.[13]
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Careful annealing of the Cu/[Zn(OCHMeCH2NMe2)2]/
PMS sample at 623 K in dynamic vacuum (0.1 Pa, 2 h)
provides a material which is free from CHx contamination,
but still exhibits a very broad Cu(111) reflection. On the basis
of XRD there is no evidence for the formation of ZnO
nanocrystallites (Figure 1). The specific copper surface area
of the Cu/ZnO/PMS samples is 5–6 m2Cu gcat1 both before and
after the catalytic tests (Table 1).[14] The methanol synthesis
performances, determined to be 19 to 130 mmol gcat1 h 1, are in
the same range as those of binary Cu/ZnO catalysts, prepared
by conventional coprecipitation/calcination techniques.
Because of its three dimensional (3D) pore structure,
MCM-48 offers more efficient diffusion than MCM-41.
Thus, the highest values in the test runs were obtained using
MCM-48 as the support. The reduction (with H2) of samples
that were fully oxidized (disappearance of the Cu(111)
reflection) by storage under ambient conditions regenerated
both the initial activity and the copper surface area. The two
samples Cu/MCM-41 (10–12 wt. %; 5–7 m2Cu gcat1) and ZnO/
MCM-41 used for comparison were synthesized by annealing
[M(OCHMeCH2NMe2)2]/MCM-41 (M = Zn, Cu), or by treating MCM-41 with diethyl zinc vapor followed by a calcination
step and were catalytically not active. Freshly prepared Cu/
ZnO/PMS samples exhibited remarkable high “apparent” Cu
surface areas of 50–60 m2Cu gcat1 during the first oxidation/
reduction cycle (N2O/H2), which decreased to the characteristic level of 5–6 m2Cu gcat1 within the next cycle. Clearly, this
discrepancy is not due to a sintering of the copper particles,
but is a result of the O-Zn-C2H5 groups (solid-state NMR, see
Supporting Information) bound to the pore walls of PMS. In
accordance with Scheme 1, these groups represent an excess
of Zn species and are oxidized by N2O.
EXAFS spectra (Figure 3) of Cu/ZnO/MCM-41 confirm
the existence of very small Cu aggregates. The calculated
coordination number of 5.8 and the relatively high Debye–
Waller factor of the first metal shell are indications of a high
amount of disorder. Assuming spherical, mono dispersed
particles, would lead to a particle diameter of 0.7 nm, which
corresponds to a cluster of 13 copper atoms.[15] Even though
an underestimation of the particle size has to be considered
because of the correlation of coordination number and
Debye–Waller factor, the typical dimension of the particles
must be below the pore diameter of 2–3 nm as higher
coordination shells are absent. There appears to be a particle
size distribution, hence X-ray diffraction can detect larger
particles (around 2 nm). The Cu–Cu separation of 2.50 H,
calculated from the Cu(111) reflex at 2 q = 44.358, resembles
that obtained from the EXAFS data (2.51 H) and differs from
the Cu–Cu separation in the bulk phase of 2.56 H. This
decrease in Cu–Cu separation is a typical effect of small
particles.[16] The origin of the Cu O coordination cannot be
explained unambiguously. Because of the medium particle
size it seems reasonable to assume that oxygen atoms from
the pore wall can be detected in the EXAFS spectrum.[17]
However, it cannot be excluded, that a small amount of
copper exists as Cu+. There is no similarity between the ZnK
spectra of the sample and those of ZnO and Zn (Figure 3 a).
The intensity of the first oxygen sphere is too weak and the
second sphere (Zn) is almost completely missing. By consid-
Angew. Chem. Int. Ed. 2004, 43, 2839 –2842
Table 1: Relevant catalytic properties of the Cu/ZnO/PMS samples.
Rate of production[a] in [mmolMeOH gcat1 h 1] at
4.6 m2Cu gcat1
6.2 m2Cu gcat1
5.8 m2Cu gcat1
mmolMeOH h 1 m2Cu
Weight % Cu
Weight % Zn
Cu surface area
5–7 m2Cu gcat1
[a] The copper surface area was determined using the N2O-RFC (reactive frontal chromatography) technique.[14] After a pretreatment with diluted H2
atmosphere (2 vol %), the catalyst was kept in flowing N2O (1 vol % N2O in He, 300 K) and the copper surface area was calculated from the produced
nitrogen (density of surface Cu atoms: 1.47 O 1019 m 2). The methanol-synthesis activity was examined under normal pressure at a temperature of
493 K. The synthesis gas used contained of a mixture of 72 % H2, 10 % CO, 4 % CO2, and 14 % He. The listed data were taken after a reaction time of
2 h. Because of the low turnover at ambient pressure, only methanol could be identified as a product. [b] Specific catalytic data of classically produced
catalysts, for which methanol-synthesis performances have been determined under the same conditions as those used for the Cu/ZnO/PMS samples.
By using regression analysis (rate of production dependent on the Cu surface area of catalysts of different metal loading), the production rates of the
classical systems were calculated according to the Cu surface area of the Cu/ZnO/PMS samples (interpolation for Cu/ZnO, and extrapolation for Cu/
ZnO/Al2O3). Data taken from ref. [21]. [c] not identified.
Figure 3. a) Znk and b) CuK EXAFS spectra (absolute value of the Fourier transform) of Cu/
ZnO/MCM-41 and reference samples, c) analysis of the Cu/ZnO/MCM-41 spectrum; model
parameters: neighbor Cu separation (d) = 2.512 0.002 Q, coordination number
(N) = 5.8 0.3, Debye–Waller factor (s2) = (9.6 0.4) O 10 3 Q2, neighbor O:
d = 1.86 0.04 Q, N = 0.3 0.1, s2 = (7 0.11) O 10 3 Q2.
Angew. Chem. Int. Ed. 2004, 43, 2839 –2842
ering Si and Zn as next-but-one neighbors,
we obtained a good alignment, which supports the absence of free silanol groups
(solid-state NMR and IR spectroscopy) and
indicates coverage of the pore walls with
ZnO. Therefore the degree of aggregation
of the ZnO component is infinitesimal.
ZnO/PMS materials that have been synthesized by aqueous impregnation/calcination
techniques exhibit similar properties.[9]
However, unequivocal evidence for
CuZnOx species, or Cu-O-Zn coordination
could not be obtained from the data.
Copper-based catalysts for methanol
synthesis can be divided into three groups:
the binary systems Cu/Al2O3 (I) and Cu/
ZnO (II), and the ternary Cu/ZnO/Al2O3
(III). Classically prepared materials show a
linear correlation between synthesis activity
and the copper surface area, which increases
from I to III.[18] The activity of our Cu/ZnO/
MCM-48 sample of 130 mmol gcat1 is much
higher than the expected value for a binary
Cu/ZnO catalyst of same copper surface
area and is found in the range of the ternary
systems III. Interactions of copper particles
with the siliceous matrix can be neglected,
since the Cu/PMS control sample was not
active and the walls of the active Cu/ZnO/
PMS sample appear to be covered with ZnO
(see above). It seems as if the simultaneous
high dispersion of the Cu and the ZnO
components leads to a new, positive effect.
An aggregated (nano)crystalline ZnO
phase, which is inevitably obtained from
coprecipitation/calcination methods, is not
necessary for the synergistic effects. This
observation corresponds to the cited formation of Cu/Zn alloys, and to our finding,
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
that a classically prepared Cu/Al2O3 catalyst can be efficiently
promoted by a short pulse of diethyl zinc vapor, which leads
to a catalyst with an activity higher than the ternary system.[18]
Using a special, large surface-area, high-defect, nanodispersed ZnO support (153 m2 g 1), synthesized from
[{(Me3SiO)ZnMe}4] by solid-state pyrolysis, a surprisingly
active, binary Cu/ZnO catalyst can be obtained.[18]
The perspectives for an MOCVD loading of PMS
supports to generate Cu/ZnO materials are clear. Not only
a variation of the dimension and framework of the pores (e.g.
MCM-41 vs. MCM-48), but also the exploitation of the
precursor chemistry offers a possibility to control the Cu/ZnO
interface. The MOCVD loading of PMS supports with the
ternary Cu/ZnO/Al2O3 system using appropriate Al2O3
precursors is an option. There is in principal no limit to the
simultaneous maximization of the specific copper surface
area and the Cu/ZnO interface area (or the ZnO dispersion)
and thus to an increase of the catalytic activity beyond that
which is currently possible.
Experimental Section
Cu/ZnO/MCM-41: Freshly synthesized,[11] calcined and dried MCM41 (350 mg) and a sample of 1 (ca. 1.0 g) were placed inside a Schlenk
tube in separate glass receptacles and evaporated in static vacuum
(0.1 Pa) to 340 K for 2 h. The blue product (200 mg) and diethyl zinc
(ca. 0.5 g) are placed together in a Schlenk tube in separate glass
receptacles and evaporated in static vacuum (0.1 Pa) over a period of
2 h at room temperature. The variation of time of vaporization,
temperature, initial amounts of the compounds, and differences in the
PMS material leads to different loadings (Table 1). Heating a sample
of Cu/[Zn(OCHMeCH2NMe2)2]/MCM-41 to 623 K (2 h) in dynamic
vacuum (0.1 Pa) provided ZnO. Other PMS-based materials, such as
MCM-48, were treated in the same manner.
Cu/MCM-41 and ZnO/MCM-41: Heating a sample of [Cu(OCHMeCH2NMe2)2]/MCM-41 in dynamic vacuum (0.1 Pa) at 523 K
(20 min) produced Cu/MCM-41. Pyrolysis of [Zn(OCHMeCH2NMe2)2]/MCM-41 at 623 K (0.1 Pa, 2 h), with 2 as a ZnO
precursor, gave ZnO/MCM-41. [Zn(OCHMeCH2NMe2)2]/MCM-41
was prepared by impregnation of MCM-41, in a solution of 2 (1 g)[19]
in pentane (40 mL), followed by several washing cycles of the
isolated, solid material. Alternatively, ZnO/PMS can be synthesized
by treating the supports with diethyl zinc vapor and a subsequent
calcination step.
Characterization: (see also Supporting Information) The powder
X-ray diffractograms (PXRD) were measured using a D8-Advance
Bruker AXS diffractometer equipped with a position sensitive
detector with CuKa radiation (l = 1.5418 H) in V-2V geometry
(capillary technique, inert gas). All diffractograms have been fitted
with the Profile Plus 2.0.1 software applying a pseudo-Voigt function.
TEM analyses were performed on a Hitachi H-8100 instrument at
200 kV with a tungsten filament (preparation under exclusion of air,
gold grids Plano, vacuum transfer holder). X-ray absorption spectra
were collected at the Hasylab (DESY, Hamburg) at station X1 using a
Si(311) double crystal monochromator in transmission (software
VIPER[20]). Nitrogen-adsorption measurements were performed
using a Quantachrome Autosorb-1 MP apparatus. The pore diameter
was calculated according to the Barrett–Joyner–Halenda (BJH)
method. The specific surface area (SBET) of empty, calcined MCM41 and that of CuOx/MCM-41 was obtained from the linear part of the
BET graph (p/p0 = 0.05–0.35).
Keywords: copper · CVD (chemical vapor deposition) ·
heterogeneous catalysis · mesoporous materials ·
methanol synthesis · zinc oxide
[1] J. B. Hansen in Handbook of Heterogeneous Catalysis, Vol. 4
(Eds.: G. Ertl, H. KnOzinger, J. Weitkamp), Wiley-VCH, New
York, 1997, pp. 1856 – 1876.
[2] J. R. Rostrup-Nielsen, Phys. Chem. Chem. Phys. 2001, 3, 283 –
[3] a) K. Klier, Adv. Catal. 1982, 31, 243 – 313; b) K. C. Waugh,
Catal. Today 1992, 15, 51 – 57.
[4] P. L. Hansen, J. B. Wagner, S. Helveg, J. R. Rostrup-Nielsen,
B. S. Clausen, H. Topsøe, Science 2002, 295, 2053 – 2055.
[5] M. M. GQnter, T. Ressler, B. Bems, C. BQscher, T. Genger, O.
Hinrichsen, M. Muhler, R. SchlOgl, Catal. Lett. 2001, 71, 37 – 44.
[6] T. Fujitani, J. Nakamura, Appl. Catal. A 2000, 191, 111 – 129.
[7] J. Greeley, A. A. Gokhale, J. Kreuser, J. A. Dumesic, H. Topsøe,
N.-Y. Topsøe, M. Mavrikakis, J. Catal. 2003, 213, 63 – 72.
[8] K. Hadjiivanov, T. Tsonscheva, M. Dimitrov, C. Minchev, H.
KnOzinger, Appl. Catal. A 2003, 241, 331 – 340, and references
[9] a) H. Gies, S. Grabowski, M. Bandyopadhyay, W. GrQnert, O. P.
Tkachenko, K. V. Klementiev, A. Birkner, Microporous Mesoporous Mater. 2003, 60, 31 – 42; b) S. Velu, L. Wang, M. Okazaki,
K. Suzuki, S. Tomura, Microporous Mesoporous Mater. 2002, 54,
113 – 126; c) M. Hartmann, S. Racouchot, C. Bischof, Microporous Mesoporous Mater. 1999, 27, 309 – 320; d) O. P. Tkachenko, K. V. Klementiev, E. LOffler, I. Ritzkopf, F. SchQth, M.
Bandyopadhyay, S. Grabowski, H. Gies, V. Hagen, M. Muhler,
Lianhai Lu, R. A. Fischer, W. GrQnert, Phys. Chem. Chem. Phys.
2003, 5, 4325 – 4334.
[10] a) R. Anwander, Chem. Mater. 2001, 13, 4419 – 4438; b) K.-B.
Lee, S.-M. Lee, J. Cheon, Adv. Mater. 2001, 13, 517 – 520; c) M.
Okumura, S. Tsubota, M. Iwamoto, M. Haruta, Chem. Lett. 1998,
4, 315 – 316; d) P. Serp, P. Kalck, R. Feurer, Chem. Rev. 2002, 102,
3085 – 3128.
[11] B. Marler, U. Oberhagemann, S. Vortmann, H. Gies, Microporous Mater. 1996, 6, 375 – 383.
[12] a) S. C. Goel, K. S. Kramer, M. Y. Chiang, W. E. Buhro, Polyhedron 1990, 9, 611 – 613.b) R. Becker, A. Devi, J. Weiß, U.
Weckenmann, M. Winter, C. Kiener, H.-W. Becker, R. A.
Fischer, Chem. Vap. Deposition 2003, 9, 149 – 156.
[13] R. Becker, J. Weiss, M. Winter, K. Merz, R. A. Fischer, J.
Organomet. Chem. 2001, 630, 253 – 262.
[14] a) G. C. Chinchen, C. M. Hay, H. D. Vanderwell, K. C. Waugh, J.
Catal. 1987, 103, 79 – 86; b) O. Hinrichsen, T. Genger, M.
Muhler, Chem. Eng. Technol. 2000, 11, 956 – 959.
[15] M. Borowski, J. Phys. IV 1997, 7, C2 259 – 260.
[16] R. Lamber, S. Wetjen, N. I. Jaeger, Phys. Rev. B 1995, 51, 10 968 –
10 971.
[17] D. E. Ramaker, J. de Graaf, J. A. R. van Veen, D. C. Koningsberger, J. Catal. 2001, 203, 7 – 17.
[18] a) M. Kurtz, N. Bauer, C. BQscher, H. Wilmer, O. Hinrichsen, R.
Becker, S. Rabe, K. Merz, M. Driess, R. A. Fischer, M. Muhler,
Catal. Lett., 2004, 92, 49 – 52; b) M. Kurtz, H. Wilmer, T. Genger,
O. Hinrichsen, M. Muhler, Catal. Lett. 2003, 86, 77 – 80.
[19] S. C. Goel, M. Y. Chiang, W. E. Buhro, Inorg. Chem. 1990, 29,
4646 – 4652.
[20] K. V. Klementiev, VIPER for Windows, freeware, http:// ~ klmn/viper.html
[21] T. Genger, Dissertation, Ruhr UniversitTt Bochum, 2000.
Received: February 12, 2003
Revised: October 29, 2003 [Z51166]
Publication delayed at authors request
2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2004, 43, 2839 –2842
Без категории
Размер файла
172 Кб
mocvd, siliceous, synthesis, mesoporous, matrices, cuzn, loading, supported, catalyst, methanol
Пожаловаться на содержимое документа