Heterogeneous heck catalysis with palladium-grafted molecular sieves

Article abstract of DOI:10.1021/ja971637u

The synthesis and characterization of palladium-grafted mesoporous MCM- 41 material, designated Pd-TMS11, are described. The material is investigated for carbon-carbon coupling reactions (Heck catalysis) with activated and nonactivated aryl substrates. Fo

Full text of DOI:10.1021/ja971637u

J. Am. Chem. Soc. 1998, 120, 12289-12296
12289
Heterogeneous Heck Catalysis with Palladium-Grafted Molecular
Sieves
Christian P. Mehnert, David W. Weaver, and Jackie Y. Ying*
Contribution from the Department of Chemical Engineering, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
ReceiVed May 20, 1997. ReVised Manuscript ReceiVed September 30, 1998
Abstract: The synthesis and characterization of palladium-grafted mesoporous MCM-41 material, designated
Pd-TMS11, are described. The material is investigated for carbon-carbon coupling reactions (Heck catalysis)
with activated and nonactivated aryl substrates. For the preparation of the new catalyst, a volatile organometallic
precursor is reacted in the gas phase with the surface of the porous framework, generating a highly dispersed
metal deposition. The ultrahigh surface area, the large pore opening, and the highly dispersed catalyst species
in Pd-TMS11 material create one of the most active heterogeneous catalysts for Heck reactions.
Introduction
interest for catalysis involving bulky substrates, which have been
a challenge for zeolitic materials due to their pore opening
restrictions (4-12 Å). We have focused our attention on
carbon-carbon coupling reactions (Heck catalysis)5 whereby
palladium is used for the olefination of aryl halides with vinyl
substrates, forming bulky coupling products. Heck catalysis is
one of the most versatile tools in modern synthetic chemistry
and has great potential for future industrial applications.
Although recent advances in homogeneous6 and heterogeneous7
Heck catalysis have attracted considerable attention, these
catalysts currently suffer from several drawbacks, such as low
turnover numbers (TON) and limited lifetime.8
Recently discovered hexagonally packed mesoporous mo-
lecular sieves1 (designated MCM-41) are of great interest to
catalysis because their large and uniform pore sizes (20-100
Å) allow sterically hindered molecules facile diffusion to their
internal active sites. These nanostructured materials have a well-
defined cylindrical pore structure and ultrahigh surface areas
of up to 1200 m2/g, which make them excellent supports for a
new generation of heterogeneous catalysts.2 This research
focuses on the synthesis and application of modified mesoporous
materials3,4 that have an active species attached to the framework
via host-guest interactions, creating discrete and uniform sites
on the inner walls of the porous systems. Due to their large
pore sizes (>20 Å), these molecular sieves are of particular
(5) Reviews: (a) Herrmann, W. A. In Applied Homogeneous Catalysis
with Organometallic Compounds; Cornils, B., Herrmann, W. A., Eds.;
VCH: Weinheim, Germany, 1996; Vol. 2, pp 712-732. (b) Cabri, W.;
Candiani, I. Acc. Chem. Res. 1995, 28, 2-7. (c) Daves, G. D., Jr.; Hallberg,
A. Chem. ReV. 1989, 89, 1433-1445. (d) de Meijer, A.; Meyer, F. E. Angew.
Chem., Int. Ed. Engl. 1994, 33, 2379-2411. (e) Grushin, V. V.; Alper, H.
Chem. ReV. 1994, 94, 1047-1062.
* To whom correspondence should be addressed.
(1) (a) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.;
Beck, J. S. Nature 1992, 359, 710-712. (b) Beck, J. S.; Vartuli, J. C.;
Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C.
T-W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; Higgens, J. B.;
Schlenker, J. L. J. Am. Chem. Soc. 1992, 114, 10834-10843.
(2) Reviews: (a) Ying, J. Y.; Mehnert, C. P.; Wong, M. S. Angew. Chem.
1998, in press. (b) Sayari, A. Chem. Mater. 1996, 8, 1840-1852. (c)
Behrens, P. AdV. Mater. 1993, 5, 127-132.
(6) (a) Beller, M.; Fischer, H.; Herrmann, W. A.; O¨ fele, K.; Brossmer,
C. Angew. Chem., Int. Ed. Engl. 1995, 34, 1848-1849. (b) Beller, M.;
Riermeier, T. H. Tetrahedron Lett. 1996, 37, 6535-6538. (c) Beller, M.;
Riermeier, T. H.; Haber, S.; Kleiner, H.-J.; Herrmann, W. A. Chem. Ber.
1996, 129, 1259-1264. (d) Ben-David, Y.; Portnoy, M.; Milstein, D.
Organometallics 1992, 11, 1995-1996. (e) Herrmann, W. A.; Elison, M.;
Fischer, J.; Ko¨cher, C.; Artus, G. R. J. Angew. Chem., Int. Ed. Engl. 1995,
34, 2371-2374. (f) Jeffery, T. Tetrahedron Lett. 1994, 35, 3051-3054.
(g) Loiseleur, O.; Meier, P.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1996,
35, 200-202. (h) Portnoy, M.; Ben-David, Y.; Milstein, D. Organometallics
1994, 13, 3465-3479. (i) Herrmann, W. A.; Brossmer, C.; O¨ fele, K.;
Reisinger, C.-P.; Priermeier, T.; Beller, M.; Fischer, H. Angew. Chem., Int.
Ed. Engl. 1995, 34, 1844-1848. (j) Herrmann, W. A.; Brossmer, C.; O¨ fele,
K.; Beller, M.; Fischer, H. Mol. Catal. A 1995, 103, 133-146. (k) Herrmann,
W. A.; Brossmer, C.; Reisinger, C.-P.; Riermeier, T. H.; O¨ fele, K.; Beller,
M. Chem. Eur. J. 1997, 3, 1357-1364. (l) Reetz, M. T.; Lohmer, G.;
Schwickardi, R. Angew. Chem. 1998, 110, 492-495. (m) Reetz, M. T.;
Lohmer, G.; Schwickardi, R. Angew. Chem., Int. Ed. Engl. 1997, 36, 1526-
1529. (n) Ohff, M.; Ohff, A.; van der Boom, M. E.; Milstein, D. J. Am.
Chem. Soc. 1997, 119, 11687-11688.
(7) (a) Reetz, M. T.; Lohmer, G. Chem. Commun. 1996, 1921-1922.
(b) Augustine, R. L.; O’Leary, S. T. J. Mol. Catal. A 1992, 72, 229-242.
(c) Augustine, R. L.; O’Leary, S. T. J. Mol. Catal. A 1995, 95, 277-285.
(d) Beller, M.; Ku¨hlein, K. Synlett 1995, 441-442. (e) Beller, M.; Fischer,
H.; Ku¨hlein, K.; Reisinger, C.-P.; Herrmann, W. A. J. Organomet. Chem.
1996, 520, 257-259.
(8) (a) Baumeister, P.; Seifert, G.; Steiner, H. (Ciba-Geigy), EP Patent
No. A 584 043, 1992. (b) Beller, M.; Tafesh, A.; Herrmann, W. A. DE
Patent No. 19 503 119, 1996. (c) DeVries, R. A.; Mendoza, A. Organo-
metallics 1994, 13, 2405-2411.
(3) Mehnert, C. P.; Ying, J. Y. Chem. Commun. 1997, 2215-2216.
(4) (a) Ryoo, R.; Jun, S.; Kim, J. M.; Kim, M. J. Chem. Commun. 1997,
2225-2226. (b) Yonemitsu, M.; Tanaka, Y.; Iwamoto, M. Chem. Mater.
1997, 9, 2679-2681. (c) Maschmeyer, T.; Oldroyd, R. D.; Sankar, R. D.;
Thomas, J. M.; Shannon, I. J.; Klepetko, J. A.; Masters, A. F.; Beatie, J.
K.; Catlow, C. R. A. Angew. Chem., Int. Ed. Engl. 1997, 109, 11713-
1716. (d) Diaz, J. F.; Balkus, K. J., Jr.; Bedioui, F.; Kurshev, V.; Kevan,
L. Chem. Mater. 1997, 9, 61-67. (e) Liu, C.-J.; Li, S. G.; Pang, W.-Q.;
Che, C.-M. Chem. Commun. 1997, 65-66. (f) Maschmeyer, T.; Rey, F.;
Sankar, G.; Thomas, J. M. Nature 1995, 378, 159-162. (g) Morey, M.;
Davidson, A.; Eckert, H.; Stucky, G. Chem. Mater. 1996, 8, 486-492. (h)
Rao, Y. V. S.; De Vos, D. E.; Bein, T.; Jacobs, P. A. Chem. Commun.
1997, 355-356. (i) Sutra, P.; Brunel, D. Chem. Commun. 1996, 2485-
2486. (j) Wu, C.-G.; Bein, T. Chem. Mater. 1994, 6, 1109-1112. (k) Zhang,
W.; Wang, J.; Tanev, P. T.; Pinnavaia, T. J. Chem. Commun. 1996, 979-
980. (l) Abdel-Fattah, T. M.; Pinnavaia, T. J. Chem. Commun. 1996, 665-
666. (m) Abe, T.; Tachibana, Y.; Uematsu, T.; Iwamoto, M. Chem.
Commun. 1995, 1617-1618. (n) Butterworth, A. J.; Clark, J. H.; Walton,
P. H.; Barlow, S. J. Chem. Commun. 1996, 1857-1858. (o) Zhang, Z.;
Suo, J.; Zhang, X.; Li, S. Chem. Commun. 1998, 241-242. (p) Van Rhijin,
W. M.; De Vos, D. E.; Sels, B. F.; Bossaert, W. D.; Jacobs, P. A. Chem.
Commun. 1998, 317-318. (q) Rao, Y. V. S.; De Vos, D. E.; Jacobs, P. A.
Angew. Chem., Int. Ed. Engl. 1997, 36, 2661-2663. (r) Fowler, C. E.;
Burkett, S. L.; Mann, S. Chem. Commun. 1997, 1769-1770. (s) Macquarrie,
D. J.; Jackson, D. B. Chem. Commun. 1997, 1781-1782.
10.1021/ja971637u CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/12/1998

12290 J. Am. Chem. Soc., Vol. 120, No. 47, 1998
Mehnert et al.
Figure 1. Transmission electron micrograph of palladium-containing Nb-MCM-41 material prepared via impregnation with PdCl2.
Table 1. Palladium-Containing Catalysts Prepared via
Impregnation or Vapor Grafting Techniques
Herein we report the synthesis and characterization of
palladium-grafted mesoporous materials, designated Pd-TMS11,
metal
metal
and their evaluation as heterogeneous catalysts for Heck
reactions of activated and nonactivated aryl substrates.
Pd-containing material
(method, precursor)
Pd
(wt %)
dispersion surface area
(%)
(m2/g Pd)
Results and Discussions
Pd/SiO2 (impregnation, PdCl2)
Pd/Al2O3 (impregnation, PdCl2)a
Pd/Nb-MCM-41 (impregnation,
PdCl2)
Pd/Nb-MCM-41 (impregnation,
[Pd(η-C5H5)(η3-C3H5)])
Pd-TMS11s (vapor grafting,
[Pd(η-C5H5)(η3-C3H5)])b
Pd-TMS11 (vapor grafting,
[Pd(η-C5H5)(η3-C3H5)])
2.0
9.1
1.0
4
15
9
18
60
40
In our effort to design a heterogeneous Heck catalyst, we
synthesized the mesoporous Pd-TMS11 material, which has
highly dispersed palladium and the ability to accommodate
sterically demanding substrates. Due to the easy accessibility
of its large pores, mesoporous Nb-MCM-41 material was
employed as the support framework for the deposition of the
catalytically active species. For the preparation of palladium-
loaded mesoporous materials, we initially applied the impregna-
tion method. After treating porous Nb-MCM-41 material with
an aqueous solution of PdCl2, the resulting material was reduced
under a stream of hydrogen, giving a gray palladium-containing
powder. The isolated material retained its hexagonally packed
porous structure (XRD: (100) 40.4 Å, (110) 22.8 Å, (200) 19.8
Å, (210) 14.9 Å) with a high surface area (948 m2/g) and a
palladium loading of 1.0 wt %. Although the well-defined pore
structure of the Nb-MCM-41 material was intact, the palladium
dispersion on the surface was only 9%. Investigation of the
impregnated material with TEM (Figure 1) and EDAX mapping
showed the formation of palladium clusters on the porous
support materials, which accounted for the low palladium metal
dispersion. In an impregnation experiment with the organo-
metallic complex9 [Pd(η-C5H5)(η3-C3H5)] as the palladium
source, a metal dispersion of 17% was established and was the
highest value obtained via the impregnation technique. A
comparison of metal dispersions for various palladium-contain-
ing materials is shown in Table 1. Further impregnation attempts
were undertaken to prepare highly dispersed palladium inside
the porous framework of Nb-MCM-41, but the majority of the
samples showed substantial cluster formation.
4.4
1.6
17
25
32
78
110
140
22.3
a From Sigma-Aldrich. b Silanized Nb-MCM-41 material was used
as the support.
onto the walls of the porous framework, followed by reduction
with H2. The success of grafting the inside of a porous material
with a highly dispersed metal film strongly depends on the
properties of the volatile complex and the reaction conditions.
To maintain the uniform structure, high surface area, and
accessibility of the mesoporous support, it is important to
minimize cluster growth inside and outside the molecular sieve.
Vapor grafting provides a uniform distribution of volatile
complex, and ensures the discrete deposition of metal fragments
throughout the porous structure of the support without cluster
formation, giving metal-grafted molecular sieves.
The Nb-MCM-41 support material was prepared by using a
slightly modified procedure from that reported by Mobil
researchers. This gave materials with BET surface areas of 950-
1050 m2/g and pore diameters of 24-28 Å. The palladium
grafting of the degassed Nb-MCM-41 material was carried out
via sublimation of the volatile organometallic complex [Pd(η-
C5H5)(η3-C3H5)] under reduced pressure through the porous
(10) (a) Zinn, A. A.; Brandt, L.; Kaesz, H. D.; Hicks, R. F. In Chemical
Vapor Deposition of Platinum, Palladium and Nickel; Kodas, T. T.,
Hampden-Smith, M. J., Eds.; VCH: Weinheim, Germany, 1994; pp 329-
355. (b) Gates, B. C. In Metal Cluster Catalysts Dispersed on Solid Supports;
Adams, R. D., Cotton, F. A., Eds.; Wiley-VCH: New York, 1998; pp 509-
538. (c) Kirillov, V. L.; Zaikovskii, V. I.; Ryndin, Yu. A. React. Kinet.
Catal. Lett. 1998, 64, 169-175.
Therefore, a different approach was adopted that involved
introduction of palladium species from the vapor phase.10
Palladium-grafted mesoporous material, designated Pd-TMS11,
was made by vapor deposition of a volatile palladium complex
(9) Shaw, B. L. J. Chem. Soc. 1960, 247.

Heck Catalysis with Pd-Grafted Molecular SieVes
J. Am. Chem. Soc., Vol. 120, No. 47, 1998 12291
Figure 2. Apparatus for the preparation of vapor-grafted Pd-TMS11.
Figure 3. Powder XRD patterns of (a) MCM-41 starting material, (b)
Pd-TMS11 catalyst after H2 reduction, and (c) Pd-TMS11 catalyst
recovered after a TON of 3000.
Scheme 1
Scheme 2
material (Scheme 1). The volatile palladium complex was loaded
in a small round-bottom flask and attached to a short-path frit
(Figure 2). After the degassed Nb-MCM-41 material was loaded
onto the porous frit and secured with a layer of glass wool, a
condensation bridge with a cold trap was attached and evacuated
to 10-2 Torr. The round-bottom flask with palladium precursor
and the frit with Nb-MCM-41 were placed in an oil bath that
was heated to 85 °C. It is important that the oil bath fully covers
the frit with the Nb-MCM-41 material during the vapor grafting
process, as otherwise the volatile palladium complex condenses
in excess on the porous support. In a successful grafting reaction,
the white-colored Nb-MCM-41 material turned uniformly
grayish-black, and excess volatile palladium complex could be
recovered as red crystals from the cold trap. After the palladium
complex had been deposited onto the walls of the porous
framework, the material was reduced under a stream of hydrogen
at 350 °C for 3 h, yielding an air-stable, black powder. Following
the above procedure, materials with typical palladium loadings
of 20-25 wt % were obtained. The level of palladium loading
depends on the condition used in the Nb-MCM-41 pretreatment,
which controls the number of reactive oxygen groups that are
available for reaction with the palladium complex. For example,
Nb-MCM-41 that had been heated and degassed under reduced
pressure (10-2 Torr, 600 °C) gave palladium loadings between
20 and 25 wt %, whereas material degassed under milder
condition (10-2 Torr, 150 °C) led to palladium loadings between
15 and 20 wt %. In both cases, the degassed Nb-MCM-41 was
exposed to an excess of volatile palladium precursor. When the
amount of the palladium complex was reduced, the resulting
sample would not be uniformly grafted with palladium. This
was readily noticeable, since the palladium precursor would be
consumed by the lower level of the Nb-MCM-41 bed, while
the top part of the bed remained untreated.
Following the successful synthesis of uniformly grafted
Pd-TMS11s material with a reduced palladium content, we
scaled up the vapor grafting technique. To ensure uniform
exposure of large amounts of support material to the volatile
palladium complex during the vapor grafting process, a fluidized
bed reactor was designed and constructed. The silanized Nb-
MCM-41 material and the organometallic precursor were loaded
under an inert atmosphere, before a stream of argon saturated
with the volatile palladium precursor was applied to generate
the fluidized bed. After the column was heated to 125 °C for 2
h, the white Nb-MCM-41 material turned grayish-black and an
excess of the palladium complex was deposited at the top of
the column. The unreacted palladium complex was recycled,
and the grafted porous material was reduced under a stream of
hydrogen, producing the Pd-TMS11s catalyst on a multigram
scale.
The Pd-TMS11 catalyst retained its hexagonally packed
porous structure as shown by the XRD pattern (Figure 3b).
Although the diffraction patterns of Pd-TMS11 and the MCM-
41 starting material (Figure 3a) have similar peaks, the peak
intensity for Pd-TMS11 was lower, possibly due to radiation
diffusion caused by the grafted palladium metal. Palladium metal
has major diffraction peaks at 2θ ) 40.1° (111) and 46.7° (200),
which were not found in the XRD pattern of Pd-TMS11 (Pd
content: 19.1 wt %), indicating that palladium was highly
dispersed in the latter. The BET surface area of a typical Pd-
TMS11 sample with a palladium content of 19.1 wt % was 750
m2/g by N2 adsorption (Figure 4). The BJH (Brunauer-Joyner-
Halenda) pore size distribution of Pd-TMS11 had a narrow peak
centered at 26.1 Å, which was, as expected, slightly smaller
than that determined for the MCM-41 starting material (27.4
Å) due to the presence of grafted palladium metal in Pd-TMS11.
TEM examination indicated the hexagonally packed pore
structure of Pd-TMS11 (Figure 5), with no noticeable palladium
clusters. In addition, energy dispersive analysis by EDAX on
Pd-TMS11 particles indicated that palladium was highly dis-
persed over the pore surface of the mesoporous support (Figure
6).
In an effort to significantly reduce the palladium content of
the material while maintaining a uniform Pd distribution,
silanized Nb-MCM-41 (Scheme 2) was used as the support
material. The fractional capping of the reactive hydroxy groups
with an inert silane fragment reduced the susceptibility of the
support surface to electrophilic attack, providing a convenient
method for synthesizing Pd-grafted materials (designated Pd-
TMS11s) with much lower levels of palladium loading (values
as low as 1.6 wt % were obtained). Although the palladium
loading was reduced in Pd-TMS11s, the metal dispersion did
not exceed 30% and was in a similar range to that for Pd-TMS11
with loadings of 15-25 wt %.
Elemental analysis of Pd-TMS11 showed palladium metal
contents between 1 and 25 wt %, depending on the pretreatment
of the MCM-41 starting material and the amount of volatile
complex [Pd(η-C5H5)(η3-C3H5)] used in the vapor grafting

12292 J. Am. Chem. Soc., Vol. 120, No. 47, 1998
Mehnert et al.
The yields for the activated aryl halides with respect to reaction
time and amount of catalyst showed that n-butyl acrylate is a
more reactive substrate than styrene. Investigation of the kinetics
of the reaction of n-butyl acrylate with 4-bromoacetophenone
(Figure 7) showed a typical initiation period of 5 min required
to heat the reaction mixture to the appropriate temperature.
Addition of catalyst to a preheated reaction mixture resulted in
instantaneous reactions. Full conversion and a yield of 99% were
obtained after only 1 h for the reaction of n-butyl acrylate and
4-bromoacetophenone with as little as 0.02 mol % catalyst.
Reaction between bromobenzene and n-butyl acrylate at 170
°C gave a TON of 624, which is higher than that achieved by
many of the most active homogeneous Heck catalysts.6 To
ensure that no reactive palladium was dissolved in the reaction
mixture, we isolated the catalyst after 20% conversion and
monitored the resulting filtrate under identical reaction condition
for another 2 h; no further conversion was detected on removal
of Pd-TMS11. Addition of elemental mercury to a progressing
reaction resulted in the immediate poisoning of the catalyst. Both
studies indicated a reaction mechanism based on a heterogeneous
catalysis cycle for carbon-carbon bond formation with
Pd-TMS11.
Figure 4. N2 adsorption-desorption isotherm of Pd-TMS11 with a
Pd content of 19.1 wt %.
process (Table 1). CO chemisorption on Pd-TMS11 showed
metallic surface areas between 110 and 140 m2/g (Pd metal),
with a metal dispersion as high as 32%. In comparison, materials
prepared by using impregnation methods showed dispersions
between 4% and 17% and metallic surface areas of 18-78 m2/g
(Pd metal). Although a metal dispersion of 17% was obtained
with the complex [Pd(η-C5H5)(η3-C3H5)] in solution impregna-
tion, introduction of the same complex via the gas phase
generated metal dispersions as high as 32%, clearly demonstrat-
ing the advantages of the vapor grafting approach.
The Pd-TMS11 materials catalyzed the Heck carbon-carbon
coupling reactions (Scheme 3). The catalytic activity of these
materials was investigated with activated and nonactivated aryl
halides, with styrene or n-butyl acrylate as the vinylic substrate.
The reactions were conveniently carried out in batch reactors
in air with temperatures between 120 and 170 °C (Table 2).
For further comparison, we investigated commercially avail-
able palladium catalysts (supported on silica, alumina, and
carbon), as well as palladium-impregnated Nb-MCM-41. Com-
pared to Pd-TMS11, these systems showed significantly lower
reactivity (up to 70% conversion for the activated aryl substrates)
with respect to their total TON, which might be attributed to
their low palladium dispersion (e.g. metal dispersion was only
4% for palladium on silica), and with the exception of the
impregnated Nb-MCM-41 sample, to limited diffusion in the
micropores of the support materials. Investigation of nonacti-
vated aryl substrates gave only negligible conversion over the
impregnated palladium materials. However, surprisingly high
conversions were achieved for the nonactivated aryl substrates
with Pd-TMS11 (67% for bromobenzene with n-butyl acrylate),
Figure 5. Transmission electron micrograph of vapor-grafted Pd-TMS11.

Heck Catalysis with Pd-Grafted Molecular SieVes
J. Am. Chem. Soc., Vol. 120, No. 47, 1998 12293
Figure 6. EDAX study of Pd-TMS11. Top row: annular dark field (ADF) image (left), oxygen map (middle), and silicon map (right). Bottom
row: palladium map (left).
Table 2. Heck Olefinationaof Aryl Halides Over Pd-TMS11 Catalysts
vinyl
substrate
aryl halide
substrateb
amount of
catalyst (mol %)
% conversionc
(% yield)d
T (°C)
time
48 h
20 min
8 h
E:Z
TONe
styrene
C6H5Br
4-BrC6H4NO2
4-BrC6H4NO2
4-BrC6H4C(O)CH3
0.1
1.0
0.1
0.1
170
120
120
120
39 (82)
99 (98)
98 (99)
99 (98)
99:1
95:5
96:4
95:5
200
50
500
1000
45 min
n-butyl acrylate
C6H5Cl
C6H5Br
4-BrC6H4NO2
4-BrC6H4C(O)CH3
4-BrC6H4C(O)CH3
0.1
0.1
0.1
0.1
0.02
170
170
120
120
120
32 h
48 h
90 min
25 min
60 min
16 (40)
67 (92)
100 (99)
100 (99)
100 (99)
99:1
99:1
99:1
99:1
99:1
64
624
1000
1000
5000
a All reactions were carried out in air. b 1.1 equiv of base [N(CH2CH3)3] with respect to the aryl halide substrate was added to the reaction
mixture. c Conversion of reactant was determined by gas chromatography. d Yield ) (mol of coupling product (E+Z))/(mol of reactant converted).
e TON ) (mol of product)/(mol of catalyst).
Scheme 3
although systems with even higher activity are needed for
efficient industrial applications.
Advantages of the Pd-TMS11 material were that the catalytic
reactions could be conveniently carried out in air, and the
separation of this heterogeneous catalyst could be achieved
easily by filtration. The recovered Pd-TMS11 catalyst showed
some agglomeration of palladium and partial structural damage
of the Nb-MCM-41 support material. Detailed studies were
carried out on Pd-TMS11 material that had been recovered after
three catalysis cycles in the reaction between n-butyl acrylate
and 1-bromo-4-nitrobenzene, with a total TON of 3000. The
XRD pattern (Figure 3c) of the used catalyst showed reduced
intensity for the diffraction peaks at low 2θ angles for the porous
framework of the support. Between 2θ ) 40° and 50°, two XRD
peaks for palladium were noted, which indicated metal cluster
formation. The widths of these diffraction peaks were too broad
to allow a precise crystallite size determination. The BET surface
area and average pore diameter of the used catalyst were 780
m2/g and 27 Å, respectively, compared to 750 m2/g and 26 Å

12294 J. Am. Chem. Soc., Vol. 120, No. 47, 1998
Mehnert et al.
opening provided easy access for large substrate molecules. The
introduction of palladium from the vapor phase resulted in highly
dispersed metal sites on the pore walls of the support material,
providing excellent activity for Heck carbon-carbon coupling
reactions in air.
Experimental Section
Materials. Sodium silicate solution (27% SiO2, 14% NaOH),
n-heptane, chlorotrimethylsilane, N,N-dimethylacetamide, dodecane,
styrene, n-butyl acrylate, triethylamine, 4-bromoacetophenone, chlo-
robenzene, bromobenzene, 1-bromo-4-nitrobenzene, niobium ethoxide,
palladium (10 wt %) on alumina, palladium (10 wt %) on activated
carbon, silica gel (70-230 mesh, BET surface area ∼500 m2/g)
(Aldrich); hexadecyltrimethylammonium bromide (Alfa Chemicals); and
allylpalladium chloride dimer (Strem Chemicals) were used as received.
Instrumentation. The powder X-ray diffraction (XRD) data were
recorded on a Siemens D5000 θ-θ diffractometer with nickel-filtered
Cu KR radiation. Transmission electron micrographs were taken on a
JEOL 200CX transmission electron microscope equipped with a
lanthanum hexaboride (LaB6) gun operating at an accelerating voltage
of 200 kV and with an objective aperture of 60 mm. Samples for
transmission electron microscopy (TEM) were ground, sonicated in
2-propanol, and supported on a carbon-coated copper grid. Nitrogen
adsorption isotherms were collected at 77 K on a Micromeritics ASAP
2010 Gas Sorption and Porosimetry System. Samples were typically
prepared for measurement by degassing at 150 °C under vacuum until
a final pressure of 10-2 Torr was reached. BET (Brunauer-Emmett-
Teller) surface areas were determined over a relative pressure range of
0.005 to 0.20. Mesopore size distributions were calculated with the
BJH (Brunauer-Joyner-Halenda) method off the adsorption branch
of the isotherms. Diffuse reflectance infrared Fourier transform (DRIFT)
spectra were collected with a Harrick HVC-DRA2 cell on a Bio-Rad
FTS-60A spectrometer. 13C and 29Si solid-state magic angle spinning
nuclear magnetic resonance (MAS NMR) spectra were obtained by
Spectral Data Services, Inc. on a SDS270 instrument operating at a
frequency of 270.618 MHz with an Oxford standard bore (54 mm)
magnet and a home-built Nicolet-based spectrometer system with Henry
Radio amplifiers for final RF generation.
Figure 7. Conversion as a function of time for Heck olefination of
4-bromoacetophenone (as an example for activated aryl substrate) with
n-butyl acrylate at 120 °C with 0.02 mol % Pd-TMS11 catalyst. For
reaction conditions, see Table 2.
before catalysis. Elemental analysis of the used catalyst showed
a palladium content reduced by ∼5 wt % (for Pd-TMS11
catalysts with 20-25 wt % loading) and an increase in the
carbon content to ∼4 wt %, indicating loss of palladium and
some coking of the catalyst surface. TEM (Figure 8) and EDAX
examination showed the formation of some palladium clusters
and partial structural damage of the Nb-MCM-41 support in
the used catalyst. The palladium clusters had a typical crystallite
size of 150 Å and a tendency to agglomerate over time.
Although the Pd-TMS11 catalyst suffered from deactivation after
long periods of reaction, the palladium metal could be almost
fully recovered. The Pd-TMS11 system is an extremely simple
and efficient Heck catalyst that rivals the best homogeneous
catalysts in use today.
Conclusions
A new heterogeneous catalyst system Pd-TMS11 has been
successfully prepared by Vapor grafting. Mesoporous Nb-MCM-
41 was used as the support since its well-defined, large pore
Figure 8. Transmission electron micrograph of the Pd-TMS11 catalyst recovered after a total TON of 3000 from the reaction between 1-bromo-
4-nitrobenzene and n-butyl acrylate at 120 °C.

Heck Catalysis with Pd-Grafted Molecular SieVes
J. Am. Chem. Soc., Vol. 120, No. 47, 1998 12295
Modified Synthesis of Nb-MCM-41. Hexadecyltrimethylammo-
nium bromide (15.9 g, 44.9 mmol) was dissolved in H2O (1.2 L) and
treated with sodium silicate (27% SiO2, 14% NaOH) (100 g, 44.9 mmol)
in distilled H2O (0.4 L) and niobium ethoxide (1.27 g, 4 mmol) to
give a white precipitate. After the mixture was stirred for 30 min, the
pH value was adjusted to 11.5 with sulfuric acid (30%). The gel was
aged at room temperature for 5 days before it was heated in pressure
tubes to 80 °C for 7 days. The supernatant of the aged gel was decanted
off and the resulting white residue was washed with H2O (1 L) and
EtOH (1 L). The isolated solid was dried in air at room temperature
for 24 h before the material was calcined at 560 °C (ramp ) 1 °C/
min) for 6 h, giving mesoporous Nb-MCM-41 material (XRD: (100)
39.9 Å, (110) 22.5 Å, (200) 19.6 Å, (210) 14.8 Å; BET surface area )
997 m2/g; BJH average adsorption pore size ) 27.4 Å; elemental
analysis (wt %): C < 0.1, H < 0.1, N < 0.01, Nb ) 1.5). The
introduction of niobium dopant increased the long-range pore-packing
order of the mesostructure and led to thicker pore walls (18 Å) in
comparison to undoped MCM-41 material (12 Å).
Silylation of Nb-MCM-41. Nb-MCM-41 (1.8 g) (XRD: (100) 38.4
Å, (110) 22.1 Å, (200) 19.2 Å, (210) 14.6 Å; BET surface area )
1004 m2/g; BJH average adsorption pore size ) 23.9 Å) was degassed
at 150 °C under reduced pressure (10-2 Torr) for 6 h before being
treated with chlorotrimethylsilane (8.56 g, 79 mmol) in n-heptane (100
mL). After the mixture was refluxed for 5 h, the supernatant was filtered
off and the resulting solid was washed with pentane (3 × 100 mL). To
ensure that the silanized support material was free of all solvent and
silane residues, the material was dried under reduced pressure (10-2
Torr) at room temperature for 12 h, giving a white powder (XRD: (100)
38.4 Å, (110) 22.1 Å, (200) 19.2 Å, (210) 14.6 Å; BET surface area )
824 m2/g; BJH average adsorption pore size ) 21.4 Å; DRIFT:
ν(C-H) 2966 (s), 2910 cm-1 (m); 13C-CP/MAS NMR: δ -1.1 ppm
[-Si(CH3)3]; 29Si-CP/MAS NMR: δ -111.0 [Si-O-Si], -106.9 [Si-
O-Si], 14.4 ppm [-Si(CH3)3]; elemental analysis (wt %): C ) 6.4, H
) 1.3, N < 0.05).
Palladium-Grafted Mesoporous Material (Pd-TMS11). (a) Nb-
MCM-41 (0.5 g) (XRD: (100) 39.9 Å, (110) 22.5 Å, (200) 19.6 Å,
(210) 14.8 Å; BET surface area ) 997 m2/g; BJH average adsorption
pore size ) 27.4 Å) was degassed at 600 °C under reduced pressure
(10-2 Torr) for 6 h. The resulting material was loaded into a short-
path frit and contained with glass wool under an inert atmosphere. A
small round-bottom flask was filled with the red complex [Pd(η-C5H5)-
(η3-C3H5)] (0.25 g, 1.2 mmol) and connected to the loaded frit that
was further attached to a condensation bridge with an empty round-
bottom flask. The apparatus (Figure 2) was evacuated, and a constant
pressure (10-2 Torr) was maintained by cooling the empty round-bottom
flask to -196 °C. The small round-bottom flask containing volatile
palladium complex and the loaded frit were heated to 85 °C (ramp )
1 °C/min) with an oil bath. It is important that the level of the oil bath
fully covers the bed of the porous material to ensure uniform grafting.
During the heating process, the white Nb-MCM-41 material turned
black and the excess volatile organometallic complex was condensed
into the cooled round-bottom flask. The resulting solid was reduced
under a stream of hydrogen (50 mL/min) at 350 °C for 3 h, giving a
black powder designated Pd-TMS11 (XRD: (100) 39.6 Å, (110) 22.5
Å, (200) 19.6 Å, (210) 14.8 Å; BET surface area ) 750 m2/g; BJH
average adsorption pore size ) 26.1 Å; CO chemisorption: Pd
dispersion ) 32%, Pd surface area ) 140 m2/g (Pd metal); elemental
analysis (wt %): Pd ) 22.3).
(b) To reduce the loading of palladium on the mesoporous support,
silanized Nb-MCM-41 material (XRD: (100) 38.4 Å, (110) 22.1 Å,
(200) 19.2 Å, (210) 14.6 Å; BET surface area ) 824 m2/g; BJH average
adsorption pore size ) 21.4 Å) was used for the vapor grafting process.
The silanized Nb-MCM-41 material (0.3 g) was treated with [Pd(η-
C5H5)(η3-C3H5)] (0.05 g, 0.23 mmol) following the vapor grafting
preparation described under (a), giving a palladium-grafted Pd-TMS11s
catalyst with a significantly reduced metal content. The XRD pattern
showed a (100) peak at 38.4 Å and distinct (110) and (200) peaks; the
BET method gave 633 m2/g of surface area with a BJH average
adsorption pore size of 21.2 Å. CO chemisorption indicated a Pd
dispersion of 25% and a Pd surface area of 110 m2/g (Pd metal).
Elemental analysis gave 1.6% Pd by weight.
(c) To scale up the preparation of Pd-TMS11s, a fluidized bed reactor
was built that enabled multigram-scale synthesis of the catalyst. In a
typical preparation, silanized Nb-MCM-41 (15 g) and volatile orga-
nometallic precursor [Pd(η-C5H5)(η3-C3H5)] (0.5 g, 2.3 mmol) were
loaded into the fluidized bed reactor, separated by a porous frit, under
an inert atmosphere. After the sealed reactor was exposed to air, a
stream of argon was vented through the reactor, generating the fluidized
bed. The column setup was warmed (ramp ) 1 °C/min) to 125 °C for
2 h with heating tapes. The excess volatile palladium complex was
recycled from the top of the reactor column, while the palladium-grafted
Nb-MCM-41 material was isolated as a grayish-black powder from
the fluidized bed section of the column. In the final step of the
preparation, the material was reduced under a stream of hydrogen at
350 °C for 3 h, giving Pd-TMS11s.
Palladium Deposition onto Nb-MCM-41 Wia Impregnation. (a)
A suspension of Nb-MCM-41 (1.0 g) (XRD: (100) 43.8 Å, (110) 24.5
Å, (200) 21.2 Å, (210) 16.0 Å; BET surface area ) 823 m2/g; BJH
average adsorption pore size ) 21.4 Å) in hexane (200 mL) was treated
with a solution of [Pd(η-C5H5)(η3-C3H5)] (0.05 g, 0.23 mmol) in hexane
(50 mL). The reddish-brown slurry was stirred for 6 h at room
temperature. The suspension was filtered and the solid residue was
washed with hexane (3 × 50 mL). After the impregnated solid was
dried under reduced pressure, the material was reduced under a stream
of hydrogen at 350 °C for 3 h, giving a gray powder (XRD: (100)
41.4 Å, (110) 23.0 Å, (200) 20.0 Å, (210) 14.9 Å; BET surface area )
710 m2/g; BJH average adsorption pore size ) 20.4 Å). Carbon
monoxide chemisorption analysis indicated a metal dispersion of 17%
and a metallic surface area of 78 m2/g (Pd metal). Elemental analysis
gave 4.4% Pd by weight.
(b) A suspension of Nb-MCM-41 (1.0 g) (XRD: (100) 41.0 Å, (110)
23.1 Å, (200) 20.0 Å, (210) 15.0 Å; BET surface area ) 1041 m2/g;
BJH average adsorption pore size ) 24.3 Å) in distilled H2O (200 mL)
was treated with a solution of PdCl2 (100 mg, 0.56 mmol) in distilled
H2O (50 mL). The reddish-brown slurry was stirred for 6 h at room
temperature. The suspension was filtered and the solid residue was
washed with distilled H2O (3 × 250 mL). After the impregnated solid
was dried at 150 °C for 12 h, the material was reduced under a stream
of hydrogen at 350 °C for 3 h, giving a black powder (XRD: (100)
40.4 Å, (110) 22.8 Å, (200) 19.8 Å, (210) 14.9 Å; BET surface area )
948 m2/g; BJH average adsorption pore size ) 24.0 Å). Carbon
monoxide chemisorption analysis indicated a metal dispersion of 9%
and a metallic surface area of 40 m2/g (Pd metal). Elemental analysis
(wt %) gave C ) 0.2, H ) 0.6, N < 0.05, Cl ) 0.2 and Pd ) 1.0.
Palladium Deposition onto Silica Gel Wia Impregnation. A
suspension of silica gel (1.0 g) in distilled H2O (200 mL) was treated
with a solution of PdCl2 (100 mg, 0.56 mmol) in distilled H2O (50
mL). The reddish-brown slurry was stirred for 6 h at room temperature.
The suspension was filtered and the solid residue was washed with
distilled H2O (3 × 250 mL). After the impregnated solid was dried at
150 °C for 12 h, the material was reduced under a stream of hydrogen
at 350 °C for 3 h, giving a black powder (BET surface area ) 437
m2/g; carbon monoxide chemisorption analysis showed a metal disper-
sion of 4% and a metallic surface area of 18 m2/g (Pd metal); elemental
analysis (wt %): C < 0.1, H ) 0.5, N < 0.05, Cl ) 0.3, Pd ) 2.0).
General Procedure for Heck Olefination of Aryl Halides. A 250
mL three-necked flask equipped with a thermometer and a reflux
condenser was charged with aryl halide (49 mmol), styrene (6.9 mL,
60 mmol) or butyl acrylate (8.6 mL, 60 mmol), triethylamine (7.3 mL,
52.4 mmol), dodecane (7.2 mL, 31.7 mmol), and N,N-dimethylaceta-
mide (48 mL). The palladium catalyst (Pd-TMS11) was added through
a funnel and the reaction mixture was heated to 120 or 170 °C (mol %
of the catalyst used, reaction time, and temperature are shown in Table
2). To monitor the reaction, 0.2 mL samples were taken in regular
intervals, diluted with N,N-dimethylacetamide, filtered over a bed of
Celite, and analyzed by gas chromatography and mass spectrometry.
After completion of the reaction, the mixtures were quenched with H2O
(400 mL) and the organic products extracted with diethyl ether (3 ×
200 mL). The combined organic phases were dried over MgSO4,
filtered, and concentrated under reduced pressure. Separation of the
e/z-isomers was carried out by chromatography with silica gel with
pentane/diethyl ether (1:1) solvent mixtures. Recrystallization of the

12296 J. Am. Chem. Soc., Vol. 120, No. 47, 1998
Mehnert et al.
resulting material from concentrated diethyl ether solutions at -10 °C
gave analytically pure product. The catalyst was filtered from the
aqueous phase. The solid was washed with ethanol (2 × 200 mL) and
diethyl ether (2 × 200 mL) and dried at 150 °C for several hours to
give a black solid.
added to the reaction mixture after 20% conversion. The reaction
stopped upon mercury introduction, and no further conversion was
noted. Both experiments indicated that the Heck reactions we inves-
tigated followed a heterogeneous pathway.
Heterogeneity Test. To ensure that the catalysis runs were based
on a heterogeneous pathway, two test experiments were undertaken:
(a) The reaction was carried out in a similar manner as the general
procedure for Heck reactions, with the exception that the catalyst was
filtered off via cannular after approximately 20% conversion. The
resulting solution was heated under the identical reaction condition for
an additional 2 h and analyzed for further conversion. No further
reaction was noted after the catalyst removal. (b) The reaction was
carried out in a similar manner as the general procedure for Heck
reactions, with the exception that a drop of elemental mercury was
Acknowledgment. We thank Dr. T. J. Garrett-Reed and M.
Frongillo (M.I.T. CMSE) for their assistance with transmission
electron microscopy. Financial support from the National
Science Foundation (CTS-9257223), David and Lucile Packard
Foundation, and M.I.T. UROP Office (D.W.W.) is gratefully
acknowledged.
JA971637U