close

Вход

Забыли?

вход по аккаунту

?

Metal@MOF Loading of Highly Porous Coordination Polymers Host Lattices by Metal Organic Chemical Vapor Deposition.

код для вставкиСкачать
Angewandte
Chemie
Metal Nanoparticles
DOI: 10.1002/anie.200462515
Metal@MOF: Loading of Highly Porous
Coordination Polymers Host Lattices by Metal
Organic Chemical Vapor Deposition**
Stephan Hermes, Marie-Katrin Schrter,
Rochus Schmid, Lamma Khodeir, Martin Muhler,
Arno Tissler, Richard W. Fischer, and
Roland A. Fischer*
Standing out from the vast majority of metal organic
coordination polymers is the class of highly porous basic
zinc carboxylates developed by Yaghi and co-workers.[1] Its
prototype is MOF-5 (MOF = metal organic framework), in
which {Zn4O} building blocks are linked together by terephthalate bridges to form a zeolite-like, cubic framework.[2] The
extremely high specific surface area[2] of up to 4500 m2 g 1 and
a pore volume of 0.69 cm3 cm 3 (for MOF-177), which has not
been surpassed by any other crystalline substance, and
thermal stability (up to 350 8C) opens up fascinating perspectives for the supramolecular host–guest chemistry.[3] Applications for these materials in miniaturized fuel cells and
convenient gas-storage devices (for H2, CH4), as gas sensors
and for gas separation, as catalyst materials, and also for
molecular electronics are emerging.[4]
A report on the quantitative inclusion of C60 and large
polycyclic dye molecules (e.g. Astrazon Orange R) into the
cavities of MOF-177 single crystals attracted our attention.[5]
Could these MOF host lattices also be suitable to efficiently
and selectively absorb typical metal organic chemical vapor
deposition (CVD) precursors, provided these were volatile
(gas absorption) or very soluble in nonpolar hydrocarbons
and had matching size and shape to fit into the cavity? The
release of the metal atoms of the precursors imbedded in the
[*] S. Hermes, M.-K. Schrter, R. Schmid, Prof. Dr. R. A. Fischer
Lehrstuhl f+r Anorganische Chemie II
Organometallics & Materials Chemistry
Ruhr-Universit1t Bochum
44780 Bochum (Germany)
Fax: (+ 49) 234-32-14174
E-mail: roland.fischer@ruhr-uni-bochum.de
L. Khodeir, Prof. Dr. M. Muhler
Lehrstuhl f+r Technische Chemie
Ruhr-Universit1t Bochum
44780 Bochum (Germany)
Dr. A. Tissler, Dr. R. W. Fischer
S+d-Chemie AG
Waldheimer Strasse 13, 83052 Bruckm+hl (Germany)
[**] The authors thank the German Research Foundation (DFG) for
support within the framework of the Priority Programme 1119 “CVD
Materials” as well as for support within the Collaborative Research
Centre SFB 558 “Metal substrate interaction in heterogeneous
catalysis”. M.-K.S. is grateful to the Evangelische Studienwerk for a
PhD grant.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2005, 44, 6237 –6241
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6237
Communications
cavities should then lead to “naked” metal-nanoclusters of a
size regime otherwise difficult to access (1–2 nm) that are
trapped in a novel and presumably only weakly interacting
chemical environment.
When pure, freshly synthesized MOF-5 gently dried at
110 8C (to remove embedded solvents)[2] was exposed to the
vapor of the auburn palladium-precursor [(h5-C5H5)Pd(h3C3H5)] (1)[6] in static vacuum (1 Pa) at 293 K in a tightly sealed
Schlenk-tube, the originally colorless to light beige microcrystalline MOF-5 material turned dark red within 5 min.
Surprisingly, while the palladium adsorption is not reversible,
the corresponding charging with pentacarbonyl iron
[Fe(CO)5] is. The quantitative desorption occurs at 0.01 Pa
(dynamic vacuum, 298 K, IR control).
The three-dimensional crystalline order of the MOF-host
lattice remains unchanged after loading with 1, as shown by
the comparison of the powder X-ray diffraction (XRD)
pattern before and after adsorption. The analysis of the IR
and 13C magic-angle spinning (MAS) NMR solid-state spectra
(see Supporting Information) reveal the presence of intact
molecules of 1 in the MOF. The elemental analysis data
suggest a formal inclusion of exactly four molecules per cavity
(Figure 1, Table 1).
The molecular volume of the precursor can be calculated
from the structural data[6] with Gaussian98[7] as 196.6 G3. The
palladium precursors 1 thus fill 36.3 % of the elemental cell,
which amounts to 45.3 % of the pore volume. Analogously,
other typical metal organic precursors for metal deposition
Figure 1. MOF-5 cage (blue/yellow) with four incorporated [(h5C5H5)Pd(h3-C3H5)] (1) precursors (red). The elemental cell of the crystalline MOF-5 contains eight cavities of this kind.
are absorbed unchanged, for example, [(h5-C5H5)Cu(PMe3)]
(2)[8] and [(CH3)Au(PMe3)] (3)[9]
As expected, size and form selectivity are very high. For
example, with 2 which is only slightly more space demanding
than 1 and 3 only two instead of four embedded molecules are
found, even though then only 28 % of the pore volume is filled
with precursor molecules 2. The only slightly larger copper
precursor [Cu(OR)2] (R = CH(CH3)CH2NMe2)[10] with a
volume of 327.5 G3 (principal axes of the circumscribed
ellipsoids 8.3, 10.4, 6.4 G) is not absorbed by MOF-5 in which
the pore opening diameter is 8 G, whereas it will be absorbed
by isoreticular IR-MOF-8[4] in which the pore opening
diameter is approximately 9.5 G.
As the partial pressure of the precursors is comparatively
very low (< 1 Pa at 298 K) the question of maximum loading
of the cavities remains open. Solution penetration proved to
be far less efficient than loading the empty cavities through
the gas phase. The driving force for the exchange by diffusion
of the solvent molecules in the cavity against the precursor
molecules is weak. Note that highly reactive precursors such
as ZnEt2 or TiCl4 will destroy the MOF-5 material immediately upon contact at room temperature.
If the inclusion compound 1@MOF-5 is treated with H2
gas, the reddish powder will turn to black immediately at
35 8C, indicating the reduction to palladium. GC/MS
analysis of the components desorbed into the H2 stream
(293 K, 2 h) that are condensable at 77 K indicates the
presence of cyclopentane and propane as the expected byproducts (catalytic hydrogenation of the ligands). Furthermore, we found plenty of other species formed as a result of
C C couplings, C H activation, isomerization, and (partial)
hydrogenation of the ligands and their C C coupling
products. The obtained material Pd@MOF-5 is thus highly
reactive and extremely air sensitive (glowing/calcination).
The XRD pattern (Figure 2) of a capillary probe prepared
under argon displays a broad reflection (full width at half
maximum (FWHM) = 5.48) at 40.998 2q, which indicates
palladium nanocrystallites of the dimension 1.4 ( 0.1) nm
(profile analysis with Topas P 1.0, Pseudo Voigt). The shift of
the 2q angle to slightly bigger values, corresponding to a
shrinking of the Pd–Pd distance, is also characteristic for small
metal particles.[11] The particle size is also confirmed by TEM
data (Figure 2).
The characteristic reflections for the MOF-5 framework
in the XRD of the Pd@MOF-5 samples prepared by H2
reduction (Figure 3) are significantly reduced or completely
missing, whereas the typical high Langmuir surface areas of
approximately 1600 m2 g 1 are still obtained. A noteworthy
hydrogenation of the terephthalate ligands of the lattice
Table 1: Loading parameters of precursors 1–3 in MOF-5.
Precursor
1
2
3
Molecules per
cavity
4
2
4
Elemental analysis [measured/calculated]
M [%]
C [%]
H [%]
26.4/26.3
10.7/10.8
40.8/41.0
41.5/41.5
40.4/40.7
23.4/24.9
3.14/3.2
3.4/3.4
3.4/3.1
Volume precursor [G3][a]
Per molecule
Per elemental cell MOF-5[b]
196.6
242.6
159.3
6291
3882
5098
Allocation of
pore volume [%][c]
45.3
28.0
36.7
[a] Calculated with Gaussian98[14] (B3 LYP/SDD). [b] The elemental cell of MOF-5 consists of eight cavities. [c] The framework of MOF-5 occupies only
20 % of the volume of the elemental cell (V = 17 343.6 G3).[2]
6238
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6237 –6241
Angewandte
Chemie
Figure 2. Powder diffraction patterns of a) MOF-5, b) photolytically
generated Pd@MOF-5, c) Pd@MOF-5 (reduction by H2). The 2q
values characteristic for palladium are highlighted. The enlargement
shows the shift to higher 2q values, typical of small particles; d) TEM
image of photolytically generated Pd@MOF-5.
Figure 3. Powder diffraction patterns of a) MOF-5 b) photolytically generated Cu@MOF-5, c) Cu@MOF-5 (after methanol catalysis test,
reduction by H2), d) TEM image of Cu@MOF-5 (sample in (b)).
structure can be excluded on the basis of the IR data. We
assume that only the long-range order of the host lattice is
faulty. The existence of, for example, a defect or layer
structure with 2D-order could be possible. The compound
Pd@MOF-5 proved to be a moderately active catalyst for the
hydrogenation of cyclooctene (coe) which was chosen as test
reaction[12] (Table 2).
We also found catalytic activity for the material Cu@
MOF-5 which was obtained by reduction of 2@MOF-5 in
hydrogen stream at 150 8C (1 h). With 70 mmolMeOH gcat:1 h 1
methanol production from synthesis gas (rapid test at normal
pressure[13]) Cu@MOF-5 matches the standard of our recently
reported novel supported mesoporous catalysts Cu/ZnO@
MCM-41/48 (Table 2) prepared by metal organic CVD.[14] The
specific copper surface of approximately 6 m2 g 1 (at
13.8 wt % Cu) proved to be stable. According to the XRD
data (Figure 3) there are copper particles in the range of 3–
4 nm as well as an intact MOF-5 framework. The Langmuir
surface area was determined to be 1100 m2 g 1 (after catalysis
tests). This activity of Cu@MOF-5 is surprising as the
promotion of copper by zinc or ZnOx species[13, 15] normally
essential for catalysis is, in this case, apparently not necessary,
or is provided in a novel way by the MOF-5 framework
stabilizing the copper particles and not collapsing under the
catalysis conditions (220 8C, CO/H2). These observations
confirm our earlier assumption, namely that the complicated
metal–support interaction in the case of Cu/ZnO could imply
a local interfacial phenomenon of Cu-O-Zn species that
apparently does not require a well-developed (nano)crystalline ZnO phase.[14]
Thermal transformation of 3@MOF-5 (190 8C, 4 h, H2
stream) into Au@MOF-5 leaves the crystalline host lattice
completely intact (XRD and Langmuir) as also found for
Cu@MOF-5 but in contrast to Pd@MOF-5. TEM data
(Figure 4) show polydisperse gold particles in a range of 5
to 20 nm. Apparently the gold atoms or gold clusters (or
nuclei) primarily formed by decomposition of 3 are more
mobile in the open MOF structure than the copper or
palladium clusters and thus bigger aggregates are formed
within the pores and a diffusion of gold particles to the outer
surface also takes place, which accounts for the large gold
particles of around 20 nm found outside the pores. Analogously this phenomenon was found for Au@MCM-41 and, as
in our case, attributed to a very weak gold-particle–support or
wall interaction.[16]
The highly porous Au@MOF-5 material proved to be
inactive for catalytic CO oxidation. The gold nanoparticles
spread through the intact MOF-5 lattice and at the surface of
the MOF-crystallites apparently lack the strong metal–
support interaction or promotion (Au/TiO2, Au/ZnO)[17]
necessary for this catalytic effect.
Table 2: Catalytic performance of different metal@MOF-5 systems.
Catalyst
Production rate
1
cat:
1
Metal surface
2
Cu
1
2
Cu
1
cat:
wt % metal
[mmolMeOH g
70
19
130
h ]
Cu@MOF-5
Cu/ZnO@MCM-41
Cu/ZnO@MCM-48
[mmolMeOH m h ]
11.1
4.1
22.4
[m g ]
6.3
4.6
5.8
wt % Cu
13.8
6.9
10.6
Pd@MOF-5
[mmolCOA gcat:1 h 1][b]
47.34
[mmolCOA mcat:2 h 1][b]
1.42
[m2Pdgcat:1 ]
33.3
wt % Pd
35.6
Au@MOF-5
[mmolCO2 gcat:1 h 1]
–
[mmolCO2 mcat:2 h 1]
–
[m2Au gcat:1 ]
–
wt % Au
48
wt % Zn
29.0[a]
10.4
21.9
[a] Zinc content of the MOF lattice and not additionally incorporated into MOF-5. [b] COA = cyclooctane.
Angew. Chem. Int. Ed. 2005, 44, 6237 –6241
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6239
Communications
77.36 K. The copper surface area of Cu@MOF-5 and its methanol
synthesis activity was determined as described elsewhere in detail.[16]
A mixture of 72 % H2, 10 % CO, 14 % CO2, and 14 % He was used as
synthesis gas. For tests on Au-catalyzed CO oxidation see ref. [20].
For the determination of the Pd surface and details on cyclooctene
hydrogenation see ref. [21].
Further experimental details can be found in the Supporting
Information.
Received: November 4, 2004
Revised: March 18, 2005
Published online: August 30, 2005
Publication delayed at authorQs request
Figure 4. Powder diffraction patterns of the systems a) MOF-5
b) Au@MOF-5 (after reductive treatment with hydrogen at 190 8C)
c) TEM image of Au@MOF-5.
An alternative, and very gentle path to metal@MOF
materials with a persistant, crystalline host lattice is the UV
photolysis of the intermediate inclusion compounds 1–3@
MOF-5 at room temperature (water cooling) under inert gas
(Ar, He) or vacuum. GC/MS analysis of the gaseous byproducts in the case of photochemically generated Pd@
MOF-5 merely shows cyclopentadiene as well as three further
products of composition C8H10 and C10H12. For Cu@MOF-5
C10H12 (fulvalen) as well as PMe3 are found. TEM images
(Figure 2 and Figure 3) reveal very small palladium and
copper clusters (1–2 nm).
Our results show that metal organic chemical vapor
deposition offers novel perspectives for the host–guest
chemistry of porous coordination polymers. Nonaqueous
colloid chemistry of metal and semiconductor nanoparticles
is likewise based on metal organic precursor chemistry.[18, 19]
Thus, MOF-5 and related solid-state materials will have lots
to offer as moderately temperature stable “crystalline solvent
cages” for the analysis of the chemical and physical properties
of “naked” inorganic nanoparticles of all kinds, even beyond
the scope of heterogeneous catalysis.
Experimental Section
Metal@MOF-5: Thermal metal organic CVD (MOCVD) loading: a
sample of the freshly synthesized,[2, 4] pure solvent free MOF-5
template (50 mg) was placed in a Schlenk tube together with
precursor (1–3; 100.0 mg) in a separate glass boat and warmed in
static vacuum (1 Pa) for 3 h to 343 K (for 2 and 3), or left at room
temperature (for 1). Samples of the thus obtained well defined
intermediates 1–3@MOF-5 (40 mg) were then reduced under H2 at
23 8C (30 min) for Pd@MOF-5 or at 150 8C (1 h) for Cu@MOF-5 and
190 8C (2 h) for Au@MOF-5. Cooling in vacuum (10 3 mbar) to room
temperature (2 h) removed traces of the gaseous decomposition
products (control by IR and 13C-/31P-MAS-NMR).
Photo MOCVD loading: Samples of the intermediates 1–3@
MOF-5 (30 mg) synthesized as described above where photolysed in
an inert gas stream (Ar, He) for 2 h at 25–30 8C (Hg high-pressure
lamp, 500 W, Normag TQ 718) and traces of remaining ligand
fragments were removed in vacuum as described above.
The specific surface (SLangmuir) of the empty MOF-5 and the
samples metal@MOF-5 was determined by application of the
Langmuir surface model in the pressure range p/p0 = 0.1–0.3 at T =
6240
www.angewandte.org
.
Keywords: coordination polymers · chemical vapor deposition ·
heterogeneous catalysis · methanol synthesis · nanoparticles
[1] O. M. Yaghi, M. OQKeeffe, N. W. Ockwig, H. K. Chae, M.
Eddaoudi, J. Kim, Nature 2003, 423, 705 – 714.
[2] a) H. Li, M. Eddaoudi, M. OQKeeffe, O. M. Yaghi, Nature 1999,
402, 276 – 279; b) Surface determination was carried out using
nitrogen adsorption and adapting to the Langmuir surface
model.
[3] C. Janiak, Dalton Trans. 2003, 2781 – 2804.
[4] a) A. Stein, Adv. Mater. 2003, 15, 763 – 775; b) N. L. Rosi, J.
Eckert, M. Eddaoudi, D. T. Vodak, J. Kim, M. L OQKeeffe, O. M.
Yaghi, Science 2003, 300, 1127 – 1129; c) U. MSller, L. Lobree, M.
Hesse, O. M. Yaghi, M. Eddaoudi (BASF AG, The Regents of
the University of Michigan), US6624318, 2003 and
US2004081611, 2004.
[5] H. K. Chae, D. Y. Siberio-PUrez, J. Kim, Y.-B. Go, M. Eddaoudi,
A. J. Matzger, M. OQKeeffe, O. M. Yaghi, Nature 2004, 427, 523 –
527.
[6] a) Y. Zhang, Z. Yuan, R. J. Puddephatt, Chem. Mater. 1998, 10,
2293 – 2300; b) J. E. Gozum, D. M. Pollina, J. A. Jensen, G. S.
Girolami, J. Am. Chem. Soc. 1988, 110, 2688 – 2689; c) R. R.
Thomas, J. M. Park, J. Electrochem. Soc. 1989, 136, 1661 – 1666.
[7] Gaussian 98 (Revision A.11.1), Gaussian, Inc., Pittsburgh, PA,
2001, M. J. Frisch et al., see Supporting Information.
[8] a) H. Werner, H. Otto, Tri Ngo-Khac, Ch. Burschka, J. Organomet. Chem. 1984, 262, 123 – 136; b) M. J. Hampden-Smith, T. T.
Kodas, M. Paffett, J. D. Farr, H.-K. Shin, Chem. Mater. 1990, 2,
636 – 639; c) D. B. Beach, F. K. LeGoues, Ch.-K. Hu, Chem.
Mater. 1990, 2, 216 – 219.
[9] a) H. Schmidbaur, A. Shiotani, Chem. Ber. 1971, 104, 2821 –
2830; b) J. L. Davidson, P. John, P. G. Roberts, M. G. Jubber,
J. I. B. Wilson, Chem. Mater. 1994, 6, 1712 – 1718; c) H. Uchida,
N. Saito, M. Sato, M. Take, K. Ogi, Jpn. Kokai Tokkyo Koho
1995, 6.
[10] 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.
[11] R. Lamber, S. Wetjen, N. I. Jaeger, Phys. Rev. B 1995, 51, 10 968 –
10 971.
[12] X. Mu, U. Bartmann, A. Guraya, G. W. Busser, U. Weckenmann,
R. Fischer, M. Muhler, Appl. Catal. A 2003, 248, 85 – 95.
[13] 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 – 52.
[14] R. Becker, H. Parala, F. Hipler, A. Birkner, C. WVll, O.
Hinrichsen, O. P. Tkachenko, K. V. Klementiev, W. GrSnert, S.
SchWfer, H. Wilmer, M. Muhler, R. A. Fischer, Angew. Chem.
2004, 116, 2899 – 2903; Angew. Chem. Int. Ed. 2004, 43, 2839 –
2842.
[15] a) P. L. Hansen, J. B. Wagner, S. Helveg, J. R. Rostrup-Nielsen,
B. S. Clausen, H. Topsøe, Science 2002, 295, 2053 – 2055;
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 6237 –6241
Angewandte
Chemie
[16]
[17]
[18]
[19]
[20]
[21]
b) M. M. GSnter, T. Ressler, B. Bems, C. BSscher, T. Genger, O.
Hinrichsen, M. Muhler, R. SchlVgl, Catal. Lett. 2001, 71, 37 – 44;
c) T. Fujitani, J. Nakamura, Appl. Catal. A 2000, 191, 111 – 129.
a) F. SchSth, personal communication; b) R. Kumar, A. Ghosh,
C. R. Patra, P. Mukherjee, M. Sastry, Nanotechnol. Catal. 2004, 1,
111 – 136.
a) J. M. C. Soares, P. Morrall, A. Crossley, P. Harris, M. Bowker,
J. Catal. 2003, 219, 17 – 24; b) Y. Iizuka, T. Tode, T. Takao, K.-I.
Yatsu, T. Takeuchi, S. Tsubota, M. Haruta, J. Catal. 1999, 187,
50 – 58; c) F. Boccuzzi, A. Chiorino, S. Tsubota, M. Haruta, J.
Phys. Chem. 1996, 100, 3625 – 3631.
a) F. Dumestre, B. Chaudret, C. Amiens, Ph. Renaud, P. Fejes,
Science 2004, 303, 821 – 823; b) F. Dumestre, B. Chaudret, C.
Amiens, M. Respaud, P. Fejes, Ph. Renaud, P. Zurcher, Angew.
Chem. 2003, 115, 5371 – 5374; Angew. Chem. Int. Ed. 2003, 42,
5213 – 5216; c) S. Jansat, M. GYmez, K. Philippot, G. Muller, E.
Guiu, C. Claver, S. CastillYn, B. Chaudret, J. Am. Chem. Soc.
2004, 126, 1592 – 1593.
a) J. Hambrock, R. Becker, A. Birkner, J. Weiß, R. A. Fischer,
Chem. Commun. 2002, 68 – 69; b) J. Hambrock, M.-K. SchrVter,
A. Birkner, C. WVll, R. A. Fischer, Chem. Mater. 2003, 15, 4217 –
4222; c) M. K. SchrVter, L. Khodeir, E. LVffler, M. Muhler,
R. A. Fischer, Langmuir 2004, 20, 9453 – 9455.
J. Assmann, V. Narkhede, L. Khodeir, E. LVffler, O. Hinrichsen,
A. Birkner, H. Over, M. Muhler, J. Phys. Chem. B 2004, 108,
14 634 – 14 642.
J. E. Benson, H. S. Wang, M. Boudart, J. Catal. 1973, 30, 146 –
153.
Angew. Chem. Int. Ed. 2005, 44, 6237 –6241
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6241
Документ
Категория
Без категории
Просмотров
4
Размер файла
158 Кб
Теги
porous, mof, coordination, organiz, vapor, loading, lattices, polymer, chemical, metali, deposition, host, highly
1/--страниц
Пожаловаться на содержимое документа