Mercaptosilane-assisted synthesis of metal clusters within zeolites and catalytic consequences of encapsulation

Article abstract of DOI:10.1021/ja102778e

We report here a general synthetic strategy to encapsulate metal clusters within zeolites during their hydrothermal crystallization. Precursors to metal clusters are stabilized against their premature colloidal precipitation as hydroxides during zeolite crystallization using bifunctional (3-mercaptopropyl)trimethoxysilane ligands. Mercapto (-SH) groups in these ligands interact with cationic metal centers, while alkoxysilane moieties form covalent Si-O-Si or Si-O-Al linkages that promote zeolite nucleation around ligated metal precursors. These protocols led to the successful encapsulation of Pt, Pd, Ir, Rh, and Ag clusters within the NaA zeolite, for which small channel apertures (0.41 nm) preclude postsynthesis deposition of metal clusters. Sequential treatments in O₂ and H₂ formed small (～1 nm) clusters with uniform diameter. Titration of exposed atoms with H ₂ or O₂ gave metal dispersions that agree well with mean cluster sizes measured from electron microscopy and X-ray absorption spectroscopy, consistent with accessible cluster surfaces free of mercaptosilane residues. NaA micropore apertures restrict access to encapsulated clusters by reactants based on their molecular size. The ratio of the rates of hydrogenation of ethene and isobutene is much higher on clusters encapsulated within NaA than those dispersed on SiO₂, as also found for the relative rates of methanol and isobutanol oxidation. These data confirm the high encapsulation selectivity achieved by these synthetic protocols and the ability of NaA micropores to sieve reactants based on molecular size. Containment within small micropores also protects clusters against thermal sintering and prevents poisoning of active sites by organosulfur species, thus allowing alkene hydrogenation to persist even in the presence of thiophene. The bifunctional nature and remarkable specificity of the mercapto and alkoxysilane functions for metal and zeolite precursors, respectively, render these protocols extendable to diverse metal-zeolite systems useful as shape-selective catalysts in demanding chemical environments.

Full text of DOI:10.1021/ja102778e

Published on Web 06/10/2010
Mercaptosilane-Assisted Synthesis of Metal Clusters within
Zeolites and Catalytic Consequences of Encapsulation
Minkee Choi, Zhijie Wu, and Enrique Iglesia*
Department of Chemical Engineering, UniVersity of California at Berkeley,
Berkeley, California 94720
Received April 4, 2010; E-mail: iglesia@berkeley.edu
Abstract: We report here a general synthetic strategy to encapsulate metal clusters within zeolites during
their hydrothermal crystallization. Precursors to metal clusters are stabilized against their premature colloidal
precipitation as hydroxides during zeolite crystallization using bifunctional (3-mercaptopropyl)trimethoxysilane
ligands. Mercapto (-SH) groups in these ligands interact with cationic metal centers, while alkoxysilane
moieties form covalent Si-O-Si or Si-O-Al linkages that promote zeolite nucleation around ligated metal
precursors. These protocols led to the successful encapsulation of Pt, Pd, Ir, Rh, and Ag clusters within
the NaA zeolite, for which small channel apertures (0.41 nm) preclude postsynthesis deposition of metal
clusters. Sequential treatments in O
2
and H
2
formed small (∼1 nm) clusters with uniform diameter. Titration
of exposed atoms with H or O gave metal dispersions that agree well with mean cluster sizes measured
2
2
from electron microscopy and X-ray absorption spectroscopy, consistent with accessible cluster surfaces
free of mercaptosilane residues. NaA micropore apertures restrict access to encapsulated clusters by
reactants based on their molecular size. The ratio of the rates of hydrogenation of ethene and isobutene
2
is much higher on clusters encapsulated within NaA than those dispersed on SiO , as also found for the
relative rates of methanol and isobutanol oxidation. These data confirm the high encapsulation selectivity
achieved by these synthetic protocols and the ability of NaA micropores to sieve reactants based on
molecular size. Containment within small micropores also protects clusters against thermal sintering and
prevents poisoning of active sites by organosulfur species, thus allowing alkene hydrogenation to persist
even in the presence of thiophene. The bifunctional nature and remarkable specificity of the mercapto and
alkoxysilane functions for metal and zeolite precursors, respectively, render these protocols extendable to
diverse metal-zeolite systems useful as shape-selective catalysts in demanding chemical environments.
2,13,14
Introduction
deposition.
These postsynthetic protocols are effective only
for large-pore zeolites (12-member ring (MR) structures), such
The encapsulation of metal or metal oxide clusters within
small voids in zeolites or other microporous solids can be used
to contain and protect active structures and to select reactants,
products, or transition states, thus providing diverse strategies
2,4-7,9-14
as FAU and MOR,
because smaller channels prevent
access to intracrystalline spaces by solvated or gaseous precur-
9
15
16,17
sors. Ru and Pt
clusters were recently encapsulated within
small-pore zeolites (8-MR apertures) during hydrothermal zeolite
1-3
for the design and synthesis of more effective catalysts.
The
synthesis; these clusters exhibited significant sulfur tolerance
small and uniform size of channels and voids within crystalline
microporous solids allows selective access of molecules to
intracrystalline active sites based on size or shape. The place-
ment of clusters within such constrained environments also tends
to stabilize small clusters against sintering and poisoning more
effectively than less restrictive mesoporous structures. Encap-
sulation within large-pore zeolites has been achieved by
introducing metal precursors after zeolite crystallization using
15,16
during hydrogenation reactions.
Small zeolite micropores
restrict access to active sites by larger organosulfur species and
allow H dissociation on protected metal clusters to form
2
hydrogen atoms that can be subsequently used to hydrogenate
15
unsaturated molecules either directly or on distant sites via
16-18
spillover.
1
,2,4-9
2,10-12
cation exchange,
impregnation,
or chemical vapor
(10) Guzman, J.; Gates, B. C. Dalton Trans. 2003, 3303.
(
(
11) Uzun, A.; Gates, B. C. Angew. Chem., Int. Ed. 2008, 47, 9245.
12) Fierro-Gonzalez, J. C.; Hao, Y.; Gates, B. C. J. Phys. Chem. C 2007,
111, 6645.
(
(
(
(
1) Sachtler, W. M. H. Acc. Chem. Res. 1993, 26, 383.
2) Guczi, L.; Kiricsi, I. Appl. Catal., A 1999, 186, 375.
3) Bein, T. ACS Symp. Ser. 1992, 499, 274.
(13) Okamoto, Y.; Inui, Y.; Onimatsu, H.; Imanaka, T. J. Phys. Chem.
1991, 95, 4596.
4) Ryoo, R.; Cho, S. J.; Pak, C.; Kim, J.-G.; Ihm, S.-K.; Lee, J. Y. J. Am.
Chem. Soc. 1992, 114, 76.
(14) Abdo, S.; Howe, R. F. J. Phys. Chem. 1983, 87, 1713.
(15) Zhan, B.-Z.; Iglesia, E. Angew. Chem., Int. Ed. 2007, 46, 3697.
(16) Yang, H.; Chen, H.; Chen, J.; Omotoso, O.; Ring, Z. J. Catal. 2006,
243, 36.
(
(
(
5) Ryoo, R.; Cho, S. J.; Pak, C. Catal. Lett. 1993, 20, 107.
6) Cho, I. H.; Cho, S. J.; Park, S. B.; Ryoo, R. J. Catal. 1995, 153, 232.
7) Sebastian, V.; Irusta, S.; Mallada, R.; Santamar ´ı a, J. Appl. Catal. A:
General 2009, 366, 242.
(17) Chen, H.; Yang, H.; Omotoso, O.; Ding, L.; Briker, Y.; Zheng, Y.;
Ring, Z. Appl. Catal., A 2009, 358, 103.
(
(
8) Li, X.; Iglesia, E. Chem. Commun. 2008, 594.
9) Djakovitch, L.; Koehler, K. J. Am. Chem. Soc. 2001, 123, 5990.
(18) Ohgoshi, S.; Nakamura, I.; Wakushima, Y. Stud. Surf. Sci. Catal. 1993,
77, 289.
10.1021/ja102778e  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 9129–9137 9 9129

A R T I C L E S
Choi et al.
Scheme 1. Schematic Representation of the Process for Mercaptosilane-Assisted Metal Encapsulation during Zeolite Crystallization
The direct encapsulation of metal precursors during hydro-
within LTA zeolite by chelating cations with protecting ethylene
diamine ligands.
1
5-19
20
thermal synthesis
involves the introduction of solvated
cationic complexes into an aluminosilicate gel and the assembly
of zeolite building units around such species promoted by
electrostatic or van der Waals interactions. The precipitation of
precursors as bulk metal hydroxides cannot be generally
synchronized with zeolite crystallization at the high pH required
for the latter process; as a result, direct hydrothermal syntheses
have not typically led to successful encapsulation using com-
monly used metal precursors. Many cationic precursors pre-
cipitate as bulk hydroxides in the alkaline media required for
zeolite crystallization (Supporting Information, Figure S1),
leading to the premature formation of colloidal particles too
Here, we report a general protecting strategy using bifunc-
tional ligands ((3-mercaptopropyl)trimethoxysilane) that simul-
taneously stabilize the metal precursors against precipitation and
promote the condensation of silicate structures around the ligated
precursors (Scheme 1). The mercapto group (-SH) in these
ligands binds strongly to late transition metals to form stable
metal-sulfur adducts via ligand-exchange; they have been used
2
1,22
to bind metal pollutants
and to prepare stable dispersions
These metal-sulfur adducts resist the
2
3,24
of colloidal metals.
formation of bulk metal hydroxides, even at the high pH required
for zeolite synthesis. The alkoxysilane moiety of the mercap-
tosilane ligand undergoes hydrolysis in alkaline media to form
covalent Si-O-Si or Si-O-Al bonds with nucleating zeolite
structures, thus creating linkages that enforce encapsulation
2
+
large for encapsulation. RuCl
3
and Pt(NH
3
)
4
, however, can
be encapsulated via direct methods because they do not form
1
5,19
bulk hydroxides during hydrothermal synthesis. Ru
and
1
6,17
Pt
LTA
recently found that Pd and Rh clusters can also be encapsulated
clusters have been shown to form within small-pore
1
5,16
17
19
and RHO, and large-pore FAU zeolites. We have
(
20) Wu, Z.; Choi, M.; Iglesia, E. unpublished result.
(21) Celis, R.; Carmenhermosi a¨ n, M.; Cornejo, J. EnViron. Sci. Technol.
2
000, 34, 4593.
(
22) Billinge, S. J. L.; McKimmy, E. J.; Shatnawi, M.; Kim, H. J.; Petkov,
V.; Wermeille, D.; Pinnavaia, T. J. J. Am. Chem. Soc. 2005, 127, 8492.
(
19) Zhan, B.-Z.; White, M. A.; Sham, T.-K.; Pincock, J. A.; Doucet, R. J.;
(23) Brust, M.; Fink, J.; Bethell, D.; Schiffrin, D. J.; Kiely, C. J. Chem.
Soc., Chem. Commun. 1995, 1655.
Rao, K. V. R.; Robertson, K. N.; Cameron, T. S. J. Am. Chem. Soc.
2
003, 125, 2195.
(24) Zheng, N.; Stucky, G. D. J. Am. Chem. Soc. 2006, 128, 14278.
9
130 J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010

Mercaptosilane-Assisted Synthesis of Metal Clusters
A R T I C L E S
Table 1. Metal Loading and Mean Diameter of Metal Clusters
Dispersed in NaA and SiO
2
a
b
D
c
d
sample
metal loading (wt %)
d
chem (nm)
d
TEM (nm)
e
e
Pt/NaA
1.1 (0.07)
0.92 (0.32)
0.61
0.65 (0.02)
0.38
0.69 (0.35)
0.74
0.83 (0.06)
0.60
1.2 (3.5)
1.9
1.7 (56)
2.9
1.5 (3.2)
1.3
1.3 (18)
1.8
1.3
2.4
1.9
3.1
1.4
1.4
1.3
2.1
1.3
4.7
Pt/SiO
Pd/NaA
2
0.79
1.0 (0.95)
0.55
0.98 (0.09)
1.1
1.0 (0.98)
1.1
Pd/SiO
Ir/NaA
2
Ir/SiO
Rh/NaA
Rh/SiO
Ag/NaA
2
2
0.97 (0.95)
1.0
0.99 (0.03)
0.26
1.2 (39)
4.5
Ag/SiO
2
a
Analyzed by inductively coupled plasma optical emission
b
spectroscopy. Metal dispersion estimated from H
2
or O
Mean cluster diameter estimated from metal dispersion and from dchem
C/D, where C is 1.13, 1.12, 0.991, 1.10, and 1.18 nm for Pt, Pd, Ir,
2
chemisorption.
c
)
2
8,29 d
Rh, and Ag, respectively.
Surface-area-weighted mean cluster
diameter determined using TEM analysis and calculated from dTEM
)
3
2
26,27 e
Σn
i
d
i
i
/Σn d
i
.
The values outside brackets are for the NaA samples
synthesized in the presence of mercaptosilane protecting ligand, while
those inside brackets are for the NaA samples synthesized using the
same condition but without mercaptosilane.
were 1.3-1.9 nm for all the metal/NaA samples treated in O
(Table 1).
Chemisorptive titrations of metal surfaces with H
Rh, and Ir) and O (Ag) gives metal dispersions between 0.65
2
/
H
2
2
(Pt, Pd,
2
Figure 1. TEM images for (a) 1.1 wt % Pt/NaA, (b) 0.98 wt % Ir/NaA,
and 0.99 for all samples (Table 1). These dispersions correspond
(
c) 1.0 wt % Rh/NaA samples synthesized by mercaptosilane-assisted
hydrothermal crystallization, and (d) 0.79 wt % Pt/SiO sample synthesized
by a conventional impregnation method. Surface-area-weighted cluster
to mean cluster diameters (dchem) of 1.2-1.7 nm, calculated by
2
assuming spherical crystallites and the density of each bulk
3
2
26,27
diameters, dTEM, were calculated from dTEM ) Σn
i
d
i
/Σn
i
d
i
.
28,29
metal.
These cluster diameters are consistent with those
measured by transmission electron microscopy (dTEM) (Table
). The excellent agreement between dchem and dTEM indicates
that residues from the organosulfur protecting ligands can be
2
5
during subsequent zeolite crystal growth. The dual function
of this ligand causes the selective and quantitative encapsulation
of most metals of catalytic interest (e.g., Pt, Pd, Ir, Rh, and
Ag) during hydrothermal zeolite crystallization, as shown by
the structural and functional characterization data reported here.
1
completely removed from cluster surfaces by sequential O
treatments, and that cluster surfaces are fully accessible to H
and O
2
/H
2
2
2
titrants, in spite of their location within the constrained
environment of NaA cages.
The formation of spherical clusters similar in diameter to the
Results and Discussion
NaA R-cage (1.1 nm) would correspond to metal dispersions
In the present work, we have directly crystallized NaA zeolite
LTA structure, 8-MR) containing Group VIIIB (Pt, Pd, Ir, Rh)
28,29
of ∼0.90 for all metals
(∼50 atoms per cluster). Dispersions
(
above 0.80 for Pt, Rh, and Ag/NaA catalysts reflect their
essentially complete encapsulation within zeolite micropores.
The slightly smaller dispersions measured for Pd (0.65) and Ir/
NaA (0.68) samples may reflect the formation of a small number
of larger metal clusters at zeolite external surfaces. Particle size
distributions from TEM show that more than 90% of the clusters
are smaller than 1.5 nm, even for Pd/NaA (Figure S2, Supporting
Information) and Ir/NaA (Figure 1a). X-ray diffractograms
and IB (Ag) metals in the presence of (3-mercaptopropyl)tri-
methoxysilane (see the Experimental Section for the synthesis
details). Inductively coupled plasma emission showed that >95%
of the metal precursor in the synthesis mixture was incorporated
into the final solid product. This remarkable selectivity reflects
the strong and selective binding of the mercapto groups to the
metal precursors and the preferential nucleation of the zeolite
framework around these ligated metal precursors (via Si-O-Al
or Si-O-Si bonds).
(
2 2
XRD) after O /H treatment showed sharp lines indicative of
crystalline NaA in all metal-containing samples, but no detect-
able lines corresponding to metal or oxide crystal structures
(Figure S3, Supporting Information), consistent with the absence
of bulk metal species and with the effective crystallization of
NaA structures during ligand-assisted encapsulation protocols.
The protecting mercapto ligands were successfully removed
by sequential treatments in O
at 573 K). This process forms clusters that chemisorb H
in amounts commensurate with the cluster size distributions
2
and H
2
at 623 K (Ag was treated
2
and
O
2
evident from transmission electron micrographs (TEM). High-
magnification TEM images (Figure 1 and Figure S2, Supporting
Information) show that all visible metal clusters in NaA zeolite
are uniform in diameter and distributed very narrowly around
(27) Surface-area-weighted mean cluster diameter determined using TEM
3
2
analysis and calculated from dTEM ) Σn
i i i i i
d /Σn d , where n is the
i
number of crystallites having a diamerter d
.
(
(
28) Boudart, M.; Djega-Mariadassou, G. The Kinetics of Heterogeneous
Catalytic Reactions, Princeton University Press, Princeton, NJ, 1984.
2
6,27
1
nm. The surface-area-weighted cluster diameters, dTEM,
29) Average particle diameter, dchem, is obtained by dchem ) f *Vsp/S
M
, where
is
f is a shape factor, Vsp is the specific volume of the metal, and S
M
(
(
25) Choi, M.; Cho, H. S.; Srivastava, R.; Venkatesan, C.; Choi, D.-H.;
Ryoo, R. Nat. Mater. 2006, 5, 718.
the specific surface area determined from chemisorption data. By
assuming spherical geometry (f ) 6) of metal clusters, dchem is
calculated by dchem ) C/D, where C is 1.13, 1.12, 0.991, 1.10 and
1.18 nm for Pt, Pd, Ir, Rh, and Ag, respectively.
26) Schneider, M.; Duff, D. G.; Mallat, T.; Wildberger, M.; Baiker, A. J.
Catal. 1994, 147, 500.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010 9131

A R T I C L E S
Choi et al.
of metal hydroxides (Figure S1, Supporting Information) or
formation of anionic species, respectively. Low-magnification
TEM images indicated that the absence of protecting mercaps-
tosilane ligands in these synthetic protocol leads to physical
mixtures of large Pd crystallites (>20 nm) and NaA crystals in
Pd/NaA (Figure S5, Supporting Information). In this sample,
sharp XRD peaks corresponding to metallic Pd were detected
in addition to the peaks for NaA zeolite (Figure S5, Supporting
Information). Selective encapsulation of Pt in NaA was achieved
2
+
16,17
using a stable cationic Pt(NH
)
3 4
precursor
previously, but
PtCl led to
not with anionic H PtCl species. The use of H
2
6
2
6
extremely low metal loading (0.07 wt %, Table 1). This can be
attributed to the anionic nature of the precursor, which results
in the lack of interaction between the metal species and the
zeolite framework. Both cationic and anionic precursors,
however, led to the successful encapsulation of Pt clusters upon
ligation by bifunctional mercaptosilane ligands, because of the
strong and preferential interactions of the mercapto group with
the Pt centers in both precursors. Thus, we conclude that the
method reported here allows successful metal encapsulation with
a much more diverse range of metal precursors and without the
need to synchronize colloidal hydroxide precipitation of precur-
sors with the nucleation of the zeolite framework.
Figure 2. Effect of air treatment temperature on fractional metal dispersions
(
measured by H
2
chemisorption uptakes at 313 K) of Pt (circles) and Rh
(
squares) encapsulated within NaA zeolite (closed symbols) and supported
3
on SiO
g
H
2
(open symbols). Samples were treated in flowing dry air (5 cm
The structure and the chemical state of Pt were determined
by X-ray absorption fine structure (XAFS) spectroscopy at
various stages during the synthesis of Pt/NaA using mercap-
tosilanes as ligands. The resonant feature above the Pt LIII edge
-1 -1
s ) at the specified temperature for 6 h followed by treatment in flowing
3
-1 -1
2
(5 cm g
s ) at 623 K.
These data, taken together, indicate that all of these metals were
encapsulated with very high selectivity within crystalline NaA
R-cages during zeolite crystallization.
(
Pt white line) was stronger for as-synthesized Pt/NaA sample
than for bulk Pt metal or Pt/NaA samples after sequential O
2
30
/
2
H treatments (Figure 3a), indicating the presence of Pt cations
Metal/NaA samples generally showed much higher metal
3
upon encapsulation. Figure 3b shows the k -weighed radial
structure functions obtained from the fine structure in the X-ray
absorption spectrum of these samples (EXAFS). As-synthesized
Pt/NaA gave a feature at 2.31 ( 0.01 Å, corresponding to Pt-S
dispersions (0.65 - 0.99) than the corresponding metal/SiO
2
samples (0.26-0.74); the latter samples were prepared via
impregnation with metal precursors to incipient wetness (Table
1
). Clusters encapsulated within NaA microporous environments
gave much narrower size distributions than those dispersed on
SiO (Figure 1). Similar effects of encapsulation within NaA
R-cages are evident for the other metals (TEM images for Pd,
Ir, Rh, and Ag on SiO are shown in Figure S4, Supporting
31
distances but without detectable features for Pt-Pt coordina-
tion. The Pt-S coordination number (CN) was estimated to be
2
4
.1 ( 0.6, consistent with square planar PtS -ligand complexes.
4
Based on these XANES and EXAFS data, we conclude Pt(II)-
complexes formed via reactions of mercaptosilanes with
2
S
4
Information), indicating that the uniform microporous structure
of NaA zeolite enables the formation of highly dispersed metal
clusters with much greater size uniformity than on mesoporous
supports.
Figure 2 shows the effects of thermal treatment in flowing
dry air on metal dispersion for Pt and Rh clusters encapsulated
2-
PtCl
Pt(II)-S
,2-dithiolate with PtCl
ing ligands act as nucleation points for the aluminolicate LTA
framework. As a result, the extraframework Pt(S-C is
1.1 nm in its diagonal dimension, consistent with its encap-
6
, consistent with the observed formation of square-planar
4
complexes in the reactions of 3,5-ditert-butylbenzene-
2-
32
1
6
. The silane moieties in the protect-
3
7 4
H )
∼
within NaA and dispersed on SiO
Rh clusters in NaA zeolite decreased by less than 10% (0.91 to
.90 for Pt, 0.83 to 0.77 for Rh) upon treatment in flowing dry
air at 673-873 K and subsequent reduction in flowing H at
23 K. In contrast, Pt and Rh clusters supported on SiO showed
2
. The dispersions of Pt and
sulation within R-cages in Na-LTA (1.1 nm diameter). The
flexibility of its C-C and C-S bonds and of the framework
lattice allows these species to reside within nucleating zeolite
crystals during hydrothermal synthesis.
0
2
6
2
2 2
Subsequent thermal treatment in O and H at 623 K led to
a larger decrease in metal dispersion (25-46%) (0.61 to 0.33
for Pt, 0.60 to 0.45 for Rh) compared with those in NaA after
similar treatments, in spite of the greater sintering stability
the predominant presence of Pt-Pt features (2.72 ( 0.01 Å) in
the fine structure and to the disappearance of all Pt-S features,
consistent with the full reduction of cationic precursors and with
the substantial removal of sulfur heteratoms from Pt centers.
The Pt-Pt coordination number measured from the fine structure
was 6.6 ( 1.2; this value is consistent with the presence of
2
expected of the larger clusters prevalent on SiO -supported
samples. The higher stability of smaller clusters in NaA clearly
shows that encapsulation within microporous environments
protects metal clusters against sintering and coalescence during
thermal treatments required in their synthesis and their regenera-
tion during catalytic use.
We note that the use of the same metal precursors without
the use of (3-mercaptopropyl)trimethoxysilane ligands during
hydrothermal NaA synthesis gave very low metal dispersions
or metal loadings (Table 1) because of segregated precipitation
(
(
(
30) Lytle, F. W.; Wei, P. S. P.; Greegor, R. B.; Via, G. H.; Sinfelt, J. H.
J. Chem. Phys. 1979, 70, 4849.
31) Huertas, S.; Hissler, M.; McGarrah, J. E.; Lachicotte, R. J.; Eisenberg,
R. Inorg. Chem. 2001, 40, 1183.
32) Ray, K.; Weyherm u¨ ller, T.; Neese, F.; Wieghardt, K. Inorg. Chem.
2005, 44, 5345.
9
132 J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010

Mercaptosilane-Assisted Synthesis of Metal Clusters
A R T I C L E S
3
Figure 3. (a) X-ray absorption near edge structure (XANES) and (b) Fourier transforms of the k -weighted extended X-ray absorption fine structure (EXAFS)
for 1.1 wt % of Pt/NaA samples at various stages in the mercaptosilane-assisted synthesis process (A, as-synthesized Pt/NaA; B, Pt/NaA treated in O and
at 623 K; C, bulk Pt).
2
H
2
Table 2. Catalytic Properties of Metal-Encapsulated NaA and Metal/SiO
2
Catalysts in Hydrogenation (HD) of Alkenes and Oxidative
a
Dehydrogenation (ODH) of Alkanols
-1
-1
-1 -1
alkene hydrogenation turnover rates (mol·surf. g-atom
s
)
alkanol oxidative dehydrogenation turnover rates (mol·surf. g-atom
s
)
b
b
c
HD
d
HD
b
b
c
ODH
d
ODH
sample
R
ethene
R
isobutene
ꢀ
φ
R
methanol
R
isobutanol
ꢀ
φ
Pt/NaA
0.57
1.3
0.59
1.6
0.27
0.70
0.29
0.47
0.0021
0.61
0.013
0.34
0.0030
0.80
0.0070
0.84
271
2.1
47
4.8
90
0.88
42
0.56
130
9.8
100
75
0.38
0.56
0.11
0.51
0.092
0.19
0.013
0.049
0.043
0.029
0.010
0.35
0.018
0.23
0.011
0.15
0.0015
0.042
38
1.6
24
Pt/SiO
Pd/NaA
2
5.7
2.2
8.1
1.3
8.6
1.2
2.6
6.2
7.2
Pd/SiO
Ir/NaA
2
Ir/SiO
Rh/NaA
Rh/SiO
Ag/NaA
2
2
0.0021
0.035
21
0.83
25
Ag/SiO
2
a
Hydrogenation: 1.5 kPa of alkene, 5 kPa of H
2
, and 93.5 kPa of He at 294 K (except 373 K for Ir). Oxidative dehydrogenation: 4 kPa of alkanol, 9
b
kPa of O
the number of exposed metal atoms derived from chemisorption uptakes. All the rates were determined at the reaction conversion less than 10%.
Rethene/Risobutene, ꢀODH ) Rmethanol/Risobutanol. Relative ratio of ꢀ values for NaA and SiO samples.
2
2
, and 87 kPa of He at 393 K (except 353 K for Pt). Reaction turnover rate defined as molecules of reactant converted per time normalized by
c
ꢀ
HD
d
)
3
3
34
35
spherical clusters with 1.0 ( 0.2 nm diameters in this sample.
These diameters are similar to those of the R-cages in NaA (1.1
nm) and consistent with mean cluster diameters estimated from
chemisorption uptakes and electron microscopy (Table 1). These
data indicate that Pt-S bonds persist during hydrothermal
synthesis and that metal clusters formed via sequential treatments
diameter (ethene: 0.39 nm and isobutene: 0.50 nm ) and for
oxidative dehydrogenation (ODH) of alkanols with different
3
6
kinetic diameter (methanol, 0.37 nm; and isobutanol, 0.55
nm ). The specificity of encapsulation is then demonstrated
by the ratio of turnover rates for small and large reactants on
3
6
2
clusters dispersed within NaA and on SiO samples (Table 2).
The small windows in NaA (0.41 nm) allow essentially
unrestricted access to encapsulated structures by ethene (kinetic
2 2
in O and H are about 1 nm in diameter and do not retain
detectable Pt-S coordination.
3
4
36
diameter: 0.39 nm ) and methanol (0.37 nm ), but not by the
The ultimate arbiter of the successful encapsulation of these
metal clusters is the demonstration of the intended benefits of
constrained environments in their catalytic function, specifically
in the ability of these materials to select reactants and transition
states based on molecular size. We use here the relative turnover
rates for hydrogenation (HD) of alkenes with different kinetic
35
36
larger isobutene (0.50 nm ) or isobutanol (0.55 nm ) reactants.
As a result, the ratio of these turnover rates for small and large
substrates (ꢀ; Table 2) is much larger for clusters encapsulated
(
(
34) Triebe, R. W.; Tezel, F. H.; Khulbe, K. C. Gas. Sep. Purif. 1996, 10,
8
1.
35) Boucheffa, Y.; Thomazeau, C.; Cartraud, P.; Magnoux, P.; Guisnet,
M. Ind. Eng. Chem. Res. 1997, 36, 3198.
(
33) Ramallo-L o´ pez, J. M.; Requejo, F. G.; Craievich, A. F.; Wei, J.;
(36) Zhu, G.; Li, Y.; Zhou, H.; Liu, J.; Yang, W. Mater. Lett. 2008, 62,
Avalos-Borja, M.; Iglesia, E. J. Mol. Catal. A 2005, 228, 299.
4357.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010 9133

A R T I C L E S
Choi et al.
a
a
Table 3. Selectivities in Methanol Oxidative Dehydrogenation
Table 4. Selectivities in Isobutanol Oxidative Dehydrogenation
product selectivity (%)
b
selectivity (%)
catalyst
conv (%)
HCHO
MF
DMM
CO
2
catalyst
conv (%)
i-C H
3 7
CHO
i-C H
3 7
COOH
i-C H COO-i-C H
3 7 4 9
CO
2
Pt/NaA
1.7
2.4
1.1
1.6
1.3
2.2
1.2
1.8
5.2
1.6
91.0
84.0
92.9
90.2
84.9
55.2
94.3
93.5
72.8
73.4
3.4
7.5
2.5
2.4
9.7
30.9
1.4
3.2
22.1
20.0
0.6
1.4
0.4
0.6
2.1
6.2
0.4
0.6
0.5
0.6
4.6
6.8
2.7
5.1
5.4
6.5
4.1
2.5
4.9
6.1
Pt/NaA
0.2
1.4
1.4
1.1
0.6
2.0
0.2
1.1
0.9
0.2
86.5
77.6
79.8
76.7
76.5
88.8
95.2
92.3
82.3
95.1
9.3
0.9
0
3.1
2.7
6.7
5.9
3.7
5.9
0.7
5.1
4.5
2.9
Pt/SiO
Pd/NaA
2
Pt/SiO
Pd/NaA
2
18.3
3.4
1.9
3.7
0.7
2.2
0.4
4.8
0.7
2.5
0.6
4.1
0.5
0.6
0.4
3.7
0.1
Pd/SiO
Ir/NaA
2
Pd/SiO
Ir/NaA
2
Ir/SiO
Rh/NaA
Rh/SiO
Ag/NaA
2
Ir/SiO
Rh/NaA
Rh/SiO
Ag/NaA
2
2
2
Ag/SiO
2
Ag/SiO
2
a
a
4
kPa of methanol, 9 kPa of O
2
, 393 K (except 353 K for Pt).
2
4 kPa of isobutanol, 9 kPa of O , 393 K (except 353 K for Pt).
b
MF: methyl formate. DMM: dimethoxymethane.
within NaA than for clusters dispersed within mesoporous SiO
2
supports (mean pore diameter: 15 nm), which allow unrestricted
access to both large and small reactants. The ratio of these ꢀ
values on NaA and SiO samples, defined as φ, reflects the
2
diffusional constraints in accessing intrazeolite clusters for the
larger reactants. A φ value of unity is indicative of essentially
unconstrained access to metal clusters, as is the case for clusters
2
dispersed within mesoporous SiO supports. Larger φ values
provide strong evidence for preferential containment of active
clusters within spatially constrained micropores that can sieve
reactants and transition states based on their molecular size and
which are inaccessible to the larger reactants.
Alkene hydrogenation led to the exclusive formation of the
corresponding alkane on all samples (Table 2), indicating that
any residual acid sites did not contribute to the measured
chemical reaction rates. Pt/NaA gave much higher turnover rates
for ethene than for isobutene (ꢀHD ) 271). These relative rates
were much larger than on Pt/SiO
2
(ꢀHD ) 2.1), resulting in a
φ
HD value of 130. These data are consistent with the very high
specificity for Pt encapsulation during ligand-assisted hydro-
thermal synthesis. The other metals also gave φHD values much
larger than unity (9.8-100; Table 2), reflecting the faster
diffusion of ethene relative to isobutene through NaA micropores
and the preferential containment of the respective metal clusters
within such micropores. Differences in metal cluster size
Figure 4. Values of φ parameters, reflecting the shape selectivity for various
metal/NaA samples, in (2) alkene hydrogenation (HD), and (O) alkanol
oxidative dehydrogenation (ODH). φ is defined as the ꢀ value (the ratio of
turnover rates for small and large substrates) for M/NaA (M ) Ag, Pt, Ir,
2
Rh, and Pd) divided by that for the same metal dispersed on SiO for a
given reaction; φ values larger than unity indicate the preferential contain-
ment of clusters within zeolite micropores that can sieve reactants based
on size.
2
between NaA and SiO samples are unlikely to account for these
large φHD values because alkene hydrogenation turnover rates
3
7,38
depend only weakly on metal cluster size
and because any
such size effects would be similar for ethene and isobutene
hydrogenation. Thus, we conclude that active metal surfaces in
NaA samples reside preferentially within NaA cages that select
ethene (0.39 nm) over isobutene (0.50 nm) with very high
specificity. Ethene hydrogenation turnover rates are lower on
formaldehyde and isobutylaldehyde decreased with increasing
reactant conversions (varied by changing reactor residence time),
as expected from sequential pathways that convert these primary
3
9,40
reaction products.
As in the case of alkene hydrogenation, alkanol ODH
reactions gave φ values significantly larger than unity on all
samples (φODH ) 2.6-25), but these values were typically
smaller than those for alkene hydrogenation (φHD ) 9.8-130)
NaA than on SiO samples (by a factor of 0.3-0.6), apparently
2
because access to active cluster surfaces is restricted by diffusion
or by limited access to surfaces of clusters tightly contained
within NaA cages, even for small ethene molecules.
(
Table 2 and Figure 4). These trends reflect the different
Alkanol ODH reactions preferentially form selective oxidation
products (formaldehyde from methanol and isobutylaldehyde
from isobutanol) with low selectivity to combustion products
diffusivity ratios for ethene-isobutene and methanol-isobutanol
reactants within NaA windows and perhaps also a strong
dependence of the alkanol ODH reaction on cluster size, which
2
(
CO
products, such as dimethoxymethane (CH
and isobutyl isobutyrate (i-C COO-i-C
on metal-containing NaA and SiO samples. Selectivities to
2
< 7%, Tables 3 and 4). Small amounts of condensation
OCH OCH , DMM)
), were also detected
tends to make larger clusters on SiO more reactive than smaller
3
9
3
2
3
clusters encapsulated within NaA. Differences in relative
diffusivities are more likely to influence φODH values than cluster
size effects, because the latter are expected to influence methanol
H
3 7
4 9
H
2
(
(
37) Schlatter, J. C.; Boudart, M. J. Catal. 1972, 24, 482.
(39) Lichtenberger, J.; Lee, D.; Iglesia, E. Phys. Chem. Chem. Phys. 2007,
9, 4902.
38) Shaikhutdinov, S.; Heemeier, M.; Baumer, M.; Lear, T.; Lennon, D.;
Oldman, R. J.; Jackson, S. D.; Freund, H. J. J. Catal. 2001, 200, 330.
(40) Liu, H.; Iglesia, E. J. Phys. Chem. B 2005, 109, 2155.
9
134 J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010

Mercaptosilane-Assisted Synthesis of Metal Clusters
A R T I C L E S
2 2
Figure 5. Ethene hydrogenation rates for (a) 1.1 wt % of Pt/NaA and 0.79 wt % of PtSiO and (b) 1.0 wt % of Rh/NaA and 1.1 wt % of Rh/SiO in the
presence and absence of thiophene (294 K, 1.5 kPa of ethene, 5 kPa of H , with/without 0.1 kPa of thiophene).
2
and isobutanol ODH turnover rates to similar extents, because
both require kinetically relevant C-H bond activation on
surfaces covered with chemisorbed oxygen, a step disfavored
compounds, while concurrently demonstrating the high encap-
sulation selectivity achieved by the synthetic protocols that we
report here.
4
0
by the larger band gaps of small oxide domains.
The use of mercaptosilane protecting ligands provide a robust
synthetic strategy for the encapsulation of most metals of
catalytic interest (e.g., Pt, Pd, Ir, Rh, and Ag) during the
hydrothermal crystallization of small-pore zeolites, for which
postsynthetic protocols are ineffective, by preventing premature
nucleation of metal hydroxides in alkaline media. Encapsulation
leads to clusters protected against sintering and contact with
toxic impurities within environments that can select reactants
and transition states based on molecular size.
The data in Figure 4 demonstrate that both hydrogenation
and oxidation probe reactions detect the preferential encapsula-
tion of all metals within NaA using the ligand-assisted synthesis
protocols reported here. These data also show the ability of NaA
channels to selectively admit smaller reactants to intracrystalline
active sites on encapsulated clusters. These spatial constraints
also protect metal clusters from larger molecules that can cover
their surfaces with strongly bound poisons (e.g., organosulfur
compounds) during alkene hydrogenation. Ethene hydrogenation
turnover rates were measured on Pt/NaA and Rh/NaA and on
Conclusion
their SiO
added to the ethene-H
micropore apertures (0.41 nm) are expected to prevent access
to encapsulated metal clusters by thiophene (0.46 nm kinetic
diameter). The addition of thiophene (0.1 kPa) to ethene-H
reactants decreased hydrogenation rates by only ∼20% on Pt
and Rh clusters in NaA, while SiO -supported samples showed
no detectable hydrogenation rates immediately after thiophene
introduction. The small reduction in hydrogenation turnover rate
upon thiophene addition for encapsulated clusters appears to
partly reflect diffusional constraints imposed by thiophene
physisorbed on external surfaces of NaA crystals. Indeed,
turnover rates are nearly restored to their initial values upon
removal of thiophene from the inlet stream. The small loss of
catalytic activity even after the removal of thiophene from the
inlet stream can be attributed to the irreversible poisoning of
external Pt clusters by thiophene. In contrast, thiophene removal
did not restore any detectable hydrogenation activity to unpro-
2
-supported counterparts with and without thiophene
feed (Figures 5a and 5b). NaA
We report a general and versatile synthesis protocol for the
encapsulation of metal clusters within zeolite micropores using
mercaptosilane protecting ligands. This strategy avoids the
synchronization of metal hydroxide formation and zeolite
crystallization required for unprotected cationic precursors. The
mercapto (-SH) group of the silane inhibits the formation of
metal hydroxides in the alkaline media required for zeolite
synthesis, while the alkoxysilane moiety is hydrolyzed at such
conditions and condenses with zeolite precursors to form
covalent Si-O-Si or Si-O-Al bonds that induce the crystal-
lization of the inorganic framework around the metal-organosi-
lane complexes. The bifunctional nature of organosilane ligand
enables the selective encapsulation of a broad range of metals
(Pt, Pd, Ir, Rh, and Ag) within NaA, a zeolite with channel
windows too small for postsynthesis encapsulation of metal
precursors. The materials exhibited high catalytic shape-
selectivities in hydrogenation and oxidative dehydrogenation
reactions and remarkable resistance against sulfur poisoning and
thermal sintering as a consequence of metal encapsulation in
the zeolitic micropores. We expect that the strategy can be
extended further to other zeolite and related porous materials
having different pore structures and framework compositions
by the proper design of protecting ligands. It is reasonable to
expect that the present synthetic methods would allow the
2
4
1
2
2
2
tected clusters on mesoporous SiO , indicating that metal
surfaces were irreversibly poisoned by thiophene. These data
show that encapsulated clusters within small-pore zeolites can
be effectively protected against poisoning by organosulfur
(
41) Zenga, Y.; Jub, S. Sep. Purif. Tech. 2009, 67, 71.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010 9135

A R T I C L E S
Choi et al.
4
3
encapsulation of not only the reduced metal clusters but also
various metal oxides and sulfides by the proper gas treatment
in the pressure range of 10-30 kPa. Total gas uptake was
determined by extrapolation of the high-pressure linear portion of
the isotherm to zero pressure. Metal dispersions were calculated
(
2 2
e.g., O and H S) after metal encapsulation.
4
4
by assuming H/Pt
O/Ag )1 stoichiometries.
s
) 1, H/Pd
s
) 1, H/Rh
s
) 1, H/Ir ) 2, and
s
4
3
Experimental Section
s
X-ray absorption spectra were measured at the Stanford Syn-
chrotron Radiation Laboratory using beamline 4-1. Samples (10
mg, 80-120 mesh) were held within a quartz capillary (0.8 mm
Material Synthesis. Chemicals including colloidal silica (30 wt
%
, Aldrich), NaAlO
NaOH (Aldrich, 99.99%), (3-mercaptopropyl)trimethoxysilane (Al-
drich, 95%), H PtCl (Aldrich, 99% trace metal basis), Pd(NH Cl
Aldrich, 99.99%), H IrCl (Aldrich, 99.98%), (NH RhCl (Al-
drich, reagent grade), and AgNO (Aldrich, 99%) were used as
received without further purification.
2 2 3 2
(Riedel-de Ha e¨ n, 53% Al O , 42.5% Na O),
4
5
inner diameter, 0.1 mm wall thickness). The spectra for the Pt/
NaA sample were acquired at ambient temperature after re-reduction
2
6
3
)
4
2
(
2
6
4
)
3
6
3
-1 -1
2
at 623 K under H flow (5 cm g s ). Three individual spectra
3
were averaged. The pre-edge region was fitted to a second order
polynomial curve and postedge background was subtracted using
cubic spline routines after spectra were normalized by the absorption
edge height. The fine structure region was analyzed using Fourier
transform methods and phase shifts and amplitude functions were
obtained from FEFF 6. The molecular size of Pt(S-C H ) complex
3 7 4
was calculated by geometry optimization using molecular mechanics
(Polack-Ribiere algorithm, HyperChem 6.03).
Metal-encapsulated NaA zeolites were prepared by hydrothermal
crystallization in the presence of (3-mercaptopropyl)trimethoxysi-
lane. In a typical synthesis of Pt/NaA, 1.42 g of (3-mercaptopro-
pyl)trimethoxysilane and 9.60 g of NaOH were dissolved in 36.0 g
-1
of deionized H
solution of 0.60 g of H
into the previous solution with vigorous stirring. Colloidal silica
21.3 g) was then added and the mixture was heated to 353 K under
stirring until clear Na-silicate solution was obtained. A solution
of 12.0 g of NaAlO in 36.0 g of H O was added to the silicate
solution and the solution agitated for 2 h at ambient temperature.
The final synthesis gel molar ratios were 1.7 SiO /1.00 Al /3.2
Na O/110 H O/0.019 Pt/0.12 (3-mercaptopropyl)trimethoxysilane.
2
O (∼0.06 µS cm in electrical conductivity). A
2
PtCl in 36 g of H O was added dropwise
6
2
(
Measurements of Reaction Rates and Selectivities. Gases
including He (99.999%, Praxair), H (99.999%, Praxair), ethene
2
2
2
(99.9%, Praxair), and isobutene (99%, Aldrich) were purified by
Agilent O /H O trap to remove trace amounts of H O and O .
2 2 2 2
2
2 3
O
Ethanol (99.99%, Sigama-Aldrich) and isobutanol (99.99%, Sigama-
Aldrich) were used as received. Thiophene (99%, Aldrich) was
purified over degassed molecular sieve 3A and by repeated
freeze-vacuum-thaw cycles by using dry ice/acetone trap
2
2
The resultant gel was then crystallized hydrothermally at 373 K
with stirring for 12 h in polypropylene bottle (Nalgene). The solids
formed were isolated by filtration (Pyrex 36060, 10-15 µm fritted
funnel) and washed with deionized water until the rinse solution
reached a pH of 8. The sample was dried at 393 K for 8 h in ambient
air. These procedures were identical for other metal/NaA samples,
except for the use of different metal precursors. The (3-mercapto-
propyl)trimethoxysilane to metal precursor molar ratio was 6, except
for comparison samples in which the ligand was excluded. The
removal of the organic moieties was carried out by heating the
(
195 K).
Rates and selectivities were measured for hydrogenation of
alkenes and oxidative dehydrogenation of alkanols using a
packed-bed quartz reactor with plug-flow hydrodynamics. Rates
were reported here as turnover rates, defined as the number of
molecules converted per time normalized by the number of
surface metal atoms determined from chemisorption uptakes. All
reaction rates and selectivities were determined at the reactant
conversion less than 10%. The catalyst was diluted 10-fold (by
-1
samples from ambient to 623 at 0.033 K s and held for 2 h in
3
-1 -1
flowing dry air (5 cm g
s ). After being cooled to ambient
3
-1
mass) using SiO (fumed silica, Cab-O-Sil M5), pressed into a
temperature, the samples were reduced in flowing H
2
(5 cm g
s ) at 623 K (ramping rate: 0.0333 (K s ) for 2 h. In the case of
the Ag/NaA sample, the O /H treatments were carried out at 573
2
-
1
-1
pellet, and sieved to retain 60-80 mesh particles. These diluted
catalysts were further mixed with 60-80 mesh quartz granules
(treated in concentrated nitric acid and then calcined at 773 K
for 4 h in flowing dry air) to avoid temperature gradients caused
by the exothermic hydrogenation and oxidative dehydrogenation
reactions. Pre-reduced and passivated samples were treated in
2
2
K in order to minimize metal sintering during the treatment. Before
the exposure to ambient air, all the samples were passivated under
3
-1 -1
flowing 0.5% O
2
/He (Praxair, 5 cm g s ) for 1 h at 298 K.
Pt, Pd, Ir, Rh, and Ag clusters supported on mesoporous SiO
Davisil, grade 646, surface area: 294 m g , mean pore diameter:
5 nm) were prepared by impregnation to incipient wetness with
2
2
-1
3
-1 -1
(
1
flowing H
2
(5 cm g
s
) at 623 K (573 K for Ag) for 1 h
/He flow
before measuring hydrogenation rates and in 20% O
2
aqueous solutions of the same metal precursors as used in the NaA
encapsulation protocols. The samples were dried at 373 K for 6 h
at 393 K for 1 h before oxidation rate measurements. Alkene
hydrogenation was carried out using 1.5 kPa of alkene, 5 kPa
and then treated sequentially in flowing dry air and H
the same procedure used for NaA samples.
2
following
of H , and 93.5 of kPa He as diluent at 294 K (373 K for Ir
2
samples). Alkanol oxidation was carried out using 4 kPa of
Sample Characterization. X-ray diffractograms were measured
using a Siemens diffractometer (model D500). Transmission
electron micrographs were obtained with a Philips 420 TEM at
alkanol, 9 kPa of O
except 353 K for Pt samples). Reactant and product concentra-
tions were analyzed by online gas chromatography (Agilent
890GC) using a methyl silicone capillary column (HP-1; 50 m
0.25 mm, 0.25 µm film thickness) and a Porapak Q packed
2
, and 87 kPa of He as diluent at 393 K
(
1
20 kV. The samples were embedded into an adhesive polymer,
6
×
mechanically thinned and dimpled, and further thinned by Ar ion-
milling (Gatan 691 Precision Ion Polishing System, 3.0 kV). Particle
size distributions were determined by counting at least 200
crystallites. Metal dispersions were determined by chemisorptions
column (80-100 mesh, 1.82 m × 3.18 mm) connected to flame
ionization and thermal conductivity detectors, respectively.
of H
9.999%) for Ag samples using an Autosorb-1 (Quantachrome)
analyzer after re-reduction of samples in flowing H for 2 h at 623
K (ramping: 0.083 K s ) followed by evacuation for 2 h at the
same temperature. H adsorption isotherms for Pt, Ir, and Rh
samples were measured at 313 K in the pressure range of 10-50
kPa, while H adsorption isotherms for Pd samples were measured
at 343 K at 0.4-1.5 kPa to avoid the formation of the ꢁ-hydride
2
(99.999%, Praxair) for Pt, Pd, Ir, and Rh and by O
2
(Praxair,
Acknowledgment. We thank Professor Stacey I. Zones (Chev-
ron; University of California, Berkeley) for technical advice and
extensive discussions about the contents of this manuscript. The
X-ray absorption data were acquired at the Stanford Synchrotron
Research Laboratory, a national user facility operated by Stanford
University on behalf of the US Department of Energy, Office of
9
2
-1
2
2
4
2
phase. For O
for 1 h at 523 K and then evacuated for 2 h at the same
temperature. The O adsorption isotherm was measured at 443 K
2
chemisorption on Ag, samples were re-reduced in
(
42) Aben, P. C. J. Catal. 1968, 10, 224.
43) Hoost, T. E.; Kudla, R. J.; Collins, K. M.; Chattha, M. S. Appl. Catal.,
B 1997, 13, 59.
H
2
(
2
9
136 J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010

Mercaptosilane-Assisted Synthesis of Metal Clusters
A R T I C L E S
Basic Energy Sciences. We acknowledge with thanks the financial
support for this research from the Chevron Energy Technology.
for metal/NaA zeolite and metal/SiO samples. This material is
available free of charge via the Internet at http://pubs.acs.org.
2
JA102778E
Supporting Information Available: Stabilities of various metal
precursors in alkaline solutions, TEM images, and XRD patterns
(
44) McVicker, G. B.; Baker, R. T. K.; Garten, R. L.; Kugler, E. L. J.
Catal. 1980, 65, 207.
(45) Lacheen, H. S.; Cordeiro, P. J.; Iglesia, E. J. Am. Chem. Soc. 2006,
128, 15082.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 26, 2010 9137