Hydrogenation of arenes over catalysts that combine a metal phase and a grafted metal complex: Role of the single-site catalyst

Article abstract of DOI:10.1002/anie.200250667

Cooperating catalysts: The combined single-site/dispersed-metal catalyst [Rh(cod)(sulphos)]-Pd⁰/SiO₂ (cod= cyclooctadiene; sulphos = -O₃S(C₆H₄)CH₂C(CH₂ PPh₂)₃) is more efficient than the dispersed-metal catalyst Pd⁰/SiO₂ for the hydrogenation of arenes to cycloalkanes. The grafted rhodium moieties and the palladium particles in [Rh(cod)(sulphos)]-Pd⁰/SiO₂ cooperate to speed up the hydrogenation of arenes (see scheme).

Full text of DOI:10.1002/anie.200250667

Communications
Single-Site/Dispersed-Metal Catalyst
Hydrogenation of Arenes over Catalysts that
Combine a Metal Phase and a Grafted Metal
Complex: Role of the Single-Site Catalyst**
Claudio Bianchini,* Vladimiro Dal Santo, Andrea Meli,
Simonetta Moneti, Marta Moreno, Werner Oberhauser,
Rinaldo Psaro, Laura Sordelli, and Francesco Vizza
In comparison with the separate components, the combina-
tion on the same support material of dispersed palladium
nanoparticles and grafted molecular rhodium complexes has
provided evidence of improved activity in the hydrogenation
of arenes.[1] It was proposed that the catalytic efficiency is a
consequence of a hydrogen-spillover process that would
enhance specifically the hydrogenation activity of the molec-
ular catalyst.[1] A recent study of the hydrogenation of arenes
with a catalyst obtained by silica sol–gel coentrapment of
metallic palladium and [Rh(cod)(m-Cl)]2 disagrees with the
hydrogen-spillover hypothesis and suggests that the action of
both metals is caused by a synergistic effect;[2] however, no
explanation of the nature of this effect has been forthcoming.
Herein, we provide a rationale for this synergistic effect, and
show that the grafted rhodium complex speeds up the
hydrogenation of arenes to cyclic dienes.
Three different heterogeneous catalysts were tested in the
hydrogenation of some arenes in n-pentane at relatively low
temperature (40–608C) and 30 bar H2. A highly dispersed
palladium metallic phase and the RhI complex [Rh(cod)(sul-
phos)] (cod = cyclooctadiene; sulphos = ÀO3S(C6H4)CH2-
C(CH2PPh2)3) were separately supported on high-surface-
area porous silica . A calcination/reduction procedure of
silica-supported PdCl2 was employed to prepare Pd0/SiO2, 1,
with a metal content of about 10 wt%, while the RhI complex
was grafted by using a known procedure involving hydrogen
bonds between silanols of the support and sulfonate groups
from the sulphos ligand.[3] The single-site catalyst, [Rh(cod)-
(sulphos)]/SiO2, 2 (Figure 1A) with a metal content of about
0.5 wt%, was previously employed to catalyze the hydro-
genation and hydroformylation of olefins in either gas–solid
or liquid–solid phase.[3a] Upon hydrogenation, the cod ligand
[*] Dr. C. Bianchini, Dr. A. Meli, Dr. S. Moneti, Dr. M. Moreno,
Dr. W. Oberhauser, Dr. F. Vizza
Istituto di Chimica dei Composti Organometallici (ICCOM) – CNR
Via J. Nardi 39, 50132 Firenze (Italy)
Fax: (+39)055–2478366
E-mail: bianchin@fi.cnr.it
Dr. V. Dal Santo, Dr. R. Psaro, Dr. L. Sordelli
Istituto di Scienze e Tecnologie Molecolari (ISTM) – CNR
Via C. Golgi 19, 20133 Milano (Italy)
[**] This work was partially supported by the RTN contract HPRN-CT-
2002-00196, and a bilateral project CNR/FONACIT. M.M. acknowl-
edges a contribution from the European Commission for a
postdoctoral grant (Marie Curie Contract no. HPMF-CT-2000–
00568).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or fromthe author.
2636
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200250667
Angew. Chem. Int. Ed. 2003, 42, 2636 – 2639

Angewandte
Chemie
Catalyst 3 was also the best for the hydrogenation of
cyclohexa-1,3-diene (Table 2). The presence of the
grafted RhI complex apparently minimizes the amount
of benzene produced by the dehydrogenation of cyclo-
hexa-1,3-diene over the Pd particles (see entries 1 and
3).[5] The ability of supported Pd0 particles to dispropor-
tionate cyclohexa-1,3-diene to benzene and cyclohexene
is well-known, and the hydrogenation of benzene to
cyclohexa-1,3-diene was proved to be the rate limiting
step along the reduction to cyclohexane.[5]
Figure 1. Sketches of the precursors employed in this study.
in 2 is removed as cyclooctane and replaced in the metal
coordination sphere by hydride ligands.[3a] Studies of model
compounds showed that the most likely stucture of RhIII
tetrahydride is nonclassical (Figure 1B).[4] The third catalyst,
[Rh(cod)(sulphos)]–Pd0/SiO2 ,3 , was obtained by straightfor-
wardly tethering the RhI precursor [Rh(cod)(sulphos)] onto 1
by using the same procedure adopted for 2 (Figure 1C).[3]
The high resolution transmission electron-microscopy
(HRTEM) analysis of 1 showed a uniform particle dispersion
with a narrow distribution centered at 1.65 nm, which is in
good agreement with the value estimated from the EXAFS
data. The surface area was found to decrease from 344 to
304 m2 gÀ1 with a corresponding decrease in the pore volume
from 1.19 to 1.02 cm3 gÀ1. Diffuse reflectance infrared fourier
transform spectroscopy (DRIFTS) and extended X-ray
absorption fine-structure (EXAFS) measurements showed
[Rh(cod)(sulphos)] to be tethered through hydrogen bonds
on preformed 1 with no change in the chemical environment
of the rhodium center. The surface area and pore volume for 3
were 284 m2 gÀ1 and 0.95 cm3 gÀ1, respectively. HRTEM data
were consistent with a uniform palladium-particle dispersion
with a mean size of 1.79 nm. Table 1 reports data obtained for
the hydrogenation of benzene, toluene, styrene, and ethyl-
benzene. Aunique sample of 3 containing a Pd/Rh molar ratio
of 17.6:1 was used in all reactions. Irrespective of the arene,
the combined single-site/dispersed-metal catalyst 3 was from
four to six times more active than 1, while 2 proved to be
totally inactive. In contrast, the grafted RhI complex was as
active as 1 for the hydrogenation of the styrene double bond
(entries 7, 8).
Table 2: Hydrogenation of cyclohexa-1,3-diene with 1, 2 or 3.[a]
Entry Catalyst Substrate/
M ratio
% Conversion[b] (TOF, M)[c]
benzene cyclohexene cyclohexane
1
2
3
1
2
3
525
9200
525/9200
27 (284)
0
6 (63, Pd)
5 (52)
11 (2024)
6 (63, Pd) 88 (924 Pd)
68 (714)
2 (368)
[a] Experimental conditions: 1 (9.86 wt% Pd), 0.044 mmol Pd; 2 (0.56
wt% Rh), 0.0025 mmol Rh; 3 (0.56 wt% Rh, 9.86 wt% Pd), 0.044 mmol
Pd, 0.0025 mmol Rh; 30 bar H2, 30 m Ln-pentane, 258C, 30 min,
1500 rpm. [b] Average values over at least three runs. [c] Mol product
(mol Mh)À1 (M=Pd, Rh; TOF=turnover frequency).
Control experiments of the dehydrogenation of cyclo-
hexa-1,3-diene under N2 proved that 1 is an effective
disproportionation catalyst (cyclohexane 2%, cyclohexene
46%, benzene 48% under the conditions of Table 2), while 2
is unable to perform such a reaction.
Finally, 1 proved much better than 2 for the hydro-
genation of cyclohexene to cyclohexane (Table 3 entries 1, 3),
and did not promote the dehydrogenation of cyclohexene as
shown by a reaction under N2 (Table 3, entry 2).
A number of independent experiments were performed to
gain further insight into the role played by the tethered RhI
catalyst in the hydrogenation of benzene. The hydrogenation
was carried out in the presence of a mechanical mixture of 1
and 2 under the experimental conditions of Table 1, entry 3.
No appreciable synergistic effect was observed, the conver-
sion to cyclohexane being almost identical to that obtained
with 1 alone.
Table 1: Hydrogenation of arenes with 1, 2 or 3.[a]
Entry
Catalyst
T [8C]
Substrate
Substrate/M ratio
% Conversion[b] (TOF, M)[c] product
1
2
3
4
5
6
7
8
1
2
3
1[d]
2[e]
3[f]
1
2
3
1
2
40
40
40
40
40
40
60
60
60
60
60
60
benzene
benzene
benzene
toluene
toluene
toluene
styrene
styrene
styrene
525
9200
525/9200
426
7520
426/7520
400
8750
400/8750
400
8750
4 (11) cyclohexane
0
15 (39, Pd) cyclohexane
4 (8) methylcyclohexane
0
16 (32, Pd) methylcyclohexane
97 (194) ethylbenzene; 3 (6) ethylcyclohexane
98 (4287) ethylbenzene
81 (162, Pd) ethylbenzene; 19 (38 Pd) ethylcyclohexane
3 (6) ethylcyclohexane
9
10
11
12
ethylbenzene
ethylbenzene
ethylbenzene
0
3
400/8750
20 (40, Pd) ethylcyclohexane
[a] Experimental conditions: 1 (9.86 wt% Pd), 0.044 mmol Pd; 2 (0.56 wt% Rh), 0.0025 mmol Rh; 3 (0.56 wt% Rh, 9.86 wt% Pd), 0.044 mmol Pd,
0.0025 mmol Rh; 30 bar H2; 30 m Ln-pentane, 2 h, 1500 rpm. [b] Average values over at least three runs. [c] Mol product (mol Mh)À1 (M=Pd, Rh).
[d] 0.088 mmol Pd. [e] 0.005 mmol Rh. [f] 0.088 mmol Pd, 0.005 mmol Rh.
Angew. Chem. Int. Ed. 2003, 42, 2636 – 2639
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2637

Communications
Table 3: Hydrogenation of cyclohexene to cyclohexane with 1, 2 or 3.[a]
The possibility that rhodium metal, eventually formed
upon partial decomposition of the grafted complex, might
contribute to enhance the hydrogenation activity of the
palladium catalyst was ruled out on the basis of five recycling
experiments that gave comparable conversions. Washing the
recycled catalyst with MeOH to remove the grafted rhodium
complex gave a much less active catalyst with a conversion to
cyclohexane similar to that observed with 1 alone. Mecha-
nistic details for cyclohexa-1,3-diene hydrogenation to cyclo-
hexene were provided for [Ir(h4-C6H8)(triphos)]+.[9]
Entry
Catalyst
Substrate/M ratio
% Conversion[b] (TOF, M)[c]
1
1
1
2
3
525
525
9200
525/9200
99.8 (1048)
0
0.5 (92)
2[d]
3
4
100 (1050 Pd)
[a] Experimental conditions: 1 (9.86 wt% Pd), 0.044 mmol Pd; 2 (0.56
wt% Rh), 0.0025 mmol Rh; 3 (0.56 wt% Rh, 9.86 wt% Pd), 0.044 mmol
Pd, 0.0025 mmol Rh; 30 bar H2; 30 m Ln-pentane, 258C, 30 min,
1500 rpm. [b] Average values over at least three runs. [c] Mol product
(mol Mh)À1 (M=Pd, Rh). [d] Under 1 bar N2.
The incorporation of the experimental evidence reported
above leads us to conclude that 1) an intimate contact
between the Pd nanoparticles and the Rh single sites in the
mixed catalyst 3 is mandatory to observe improved activity in
the hydrogenation of arenes to cycloalkanes; 2) the inhibition
of diene disproportionation by the grafted rhodium complex,
though important, cannot account alone for the increased
conversions to cycloalkanes. These conversions are at least
four times higher than those obtained with 1, and even under
the precondition that all cyclic dienes undergo disproportio-
nation over palladium the total production of cycloalkane
would not have been more than two times higher than that
obtained with palladium alone. It is therefore reasonable to
conclude that the grafted rhodium moieties cooperate with
the palladium nanoparticles in promoting the hydrogenation
of arenes to cyclic 1,3-dienes.
The known RhI complex [Rh(CO)2(sulphos)], unable to
hydrogenate double bonds,[3a] was anchored over 1, and the
resulting combined molecular–nanoparticle catalyst was used
to hydrogenate benzene under the conditions of Table 1. The
conversion to cyclohexane (4%) was similar to that observed
in the presence of 1 alone (Table 1, entry 1).
Both 2 and the molecular precursor [Rh(cod)(sulphos)]
did not hydrogenate arenes, irrespective of the phase
variation, even at very high H2 pressure, which rules out the
promoting effect of an eventual hydrogen spillover promoted
by palladium particles, and the action of leached RhI catalyst,
respectively.
The hydrogenation of cyclohexa-1,3-diene in the presence
of 2 under the conditions of Table 2, entry 2 was quenched
after 30 min by depressurising the reactor, which was then
replenished with N2 (Figure 2). The solid compound was
collected by filtration, dried in vacuo and treated with
MeOD/[D7]DMF (1:1, v/v) under N2 to break the hydrogen
bonds between the support and the molecular complex.[3]
31P{1H} and 1H NMR spectra of the resulting solution
showed the formation of [Rh(h4-C6H8)(sulphos)].[6] Crystals
of [Rh(h4-C6H8)(sulphos)]·2MeOH were obtained by react-
ing [Rh(cod)(sulphos)] with cyclohexa-1,3-diene in MeOH
and authenticated by a single-crystal X-ray analysis.[7] The
structure consists of discrete molecules of [Rh(h4-C6H8)(sul-
phos)] with MeOH filling holes in the lattice (ORTEP
drawing, Figure 2). Three fac phosphorus atoms and two
double bonds from cyclohexa-1,3-diene coordinate the rho-
dium center in a distorted square-pyramidal geometry. The
bond distances and angles are typical of metal complexes with
either sulphos or MeC(CH2PPh2)3 (triphos) and h4-dienes.[3a,8]
A simplified mechanism for the overall hydrogenation of
benzene to cyclohexane is proposed in Figure 3, in which the
rate-limiting hydrogenation of benzene to cyclohexa-1,3-
diene is assisted by both palladium and rhodium centers. In
view of the catalytic results reported in Table 2 and Table 3,
Figure 3. Proposed mechanism for the hydrogenation of benzene by 3.
cyclohexa-1,3-diene should be more rapidly reduced to
cyclohexene at rhodium, while cyclohexene should be pre-
dominantly reduced at palladium. The intimate mechanism of
the Pd–Rh interaction in the first hydrogenation step is still
rather obscure, yet its discovery is of paramount importance
as it may open the door to the rational design of a new
generation of tailored heterogeneous catalysts for diverse
applications through appropriate combinations of supported
metals and grafted molecular complexes.
Experimental Section
Details of the synthesis and characterization of the heterogeneous
catalysts 1, 2, and 3 and of the complex [Rh(h4-C6H8)(sulphos)] are
described in the Supporting Information.
Hydrogenation procedure: A stainless steel Parr reactor (100 mL),
equipped with a temperature and pressure controller, and a paddle
stirrer was charged with the appropriate amount of catalyst, the
Figure 2. Proposed mechanism for the formation of [Rh(h4-C6H8)(sul-
phos)].
2638
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2003, 42, 2636 – 2639

Angewandte
Chemie
unsaturated substrate, n-pentane (30 mL), and H2 (30 bar). The
ensemble was heated to the appropriate temperature and then stirred
(1500 rpm) for the desired time, after which the vessel was cooled to
ambient temperature and depressurised. The liquid contents were
analysed by GC and GC–MS. Above 1500 rpm, the rates were
independent of the agitation speed at all the temperatures studied,
thus indicating the absence of mass transfer resistance.
Received: December 2, 2002
Revised: March 13, 2003 [Z50667]
Keywords: arenes · heterogeneous catalysis · hydrogenation ·
.
palladium· rhodium
[1] a) H. Gao, R. J. Angelici, J. Am. Chem. Soc. 1997, 119, 6937; b) H.
Yang, H. Gao, R. J. Angelici, Organometallics 2000, 19, 622, and
references therein.
[2] R. Abu-Reziq, D. Avnir, I. Miloslavski, H. Schumann, J. Blum, J.
Mol. Catal. A 2002, 185, 179.
[3] a) C. Bianchini, D. G. Burnaby, J. Evans, P. Frediani, A. Meli, W.
Oberhauser, R. Psaro, L. Sordelli, F. Vizza, J. Am. Chem. Soc.
1999, 121, 5961; b) C. Bianchini, P. Barbaro, V. Dal Santo, R.
Gobetto, A. Meli, W. Oberhauser, R. Psaro, F. Vizza, Adv. Synth.
Catal. 2001, 343, 41; c) C. Bianchini, V. Dal Santo, A. Meli, W.
Oberhauser, R. Psaro, F. Vizza, Organometallics 2000, 19, 2433.
[4] V. I. Bakhmutov, C. Bianchini, M. Peruzzini, F. Vizza, E.
Vorontsov, Inorg. Chem. 2000, 39, 1655.
[5] a) A. Bunjes, I. Eilks, M. Pahlke, B. Ralle, J. Chem. Educ. 1997,
74, 1323; b) L. Cerveny, S. Rehurkova, Collect. Czech. Chem.
Commun. 1987, 52, 2909; c) L. Cerveny, P. Kluson, I. Paseka,
React. Kinet. Catal. Lett. 1991, 43, 533.
[6] NMR Data ([D7]DMF, 208C,
J
in Hz): 31P{1H} NMR
(81.01 MHz): d = 11.0 ppm (d, 1J(P,Rh) = 122.6); 1H NMR
(200.13 MHz): d = 7.88 (d, 3J(H,H) = 8.1, 2H; m-H-SO3C6H4),
7.71 (d, 3J(H,H) = 8.1, 2H; o-H-SO3C6H4), 7.30 (m, 6H; p-H-
PPh2), 7.2–7.1 (m, 24H; o- and m-H-PPh2), 6.61 (m, 2H; CH-
C6H8), 3.52 (s, 2H; CH2C), 3.45 (m, 2H; CH-C6H8), 2.91 (br s, 6H;
CH2P), 1.35 (m, 2H; Hb(CH2)-C6H8), 1.23 ppm (m, 2H;
Ha(CH2)-C6H8).
[7] Crystal data: yellow crystals (0.3 ” 0.25 ” 0.25 mm), monoclinic,
space group P21/n, a = 13.513(5), b = 18.308(5), c = 19.539(5) Š,
b = 94.350(5)8, V= 4820(3) Š3, Z = 4, 1calcd = 1.415 gcmÀ3, m =
0.546 mmÀ1, MoKa (l = 0.71609 Š), T= 293(2) K, 6674 independ-
ent reflections. Refinement on all data converged at R1 = 0.0678,
wR2 = 0.1113, GOF = 0.9912. CCDC-198527 contains the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge via www.ccdc.cam.ac.uk/conts/retrieving.
html (or from the Cambridge Crystallographic Data Centre, 12
Union Road, Cambridge CB21EZ, UK; fax: (+ 44)1223-336-033;
or deposit@ccdc.cam.ac.uk).
[8] a) C. Bianchini, P. Frediani, V. Herrera, M. V. Jimꢀnez, A. Meli, L.
Rincꢁn, R. Sꢂnchez-Delgado, F. Vizza, J. Am. Chem. Soc. 1995,
117, 4333; b) C. Bianchini, K. G. Caulton, C. Chardon, M.-L.
Double, O. Eisenstein, S. A. Jackson, T. J. Johnson, A. Meli, M.
Peruzzini, W. E. Streib, F. Vizza, Organometallics 1994, 13, 2010.
[9] C. Bianchini, K. G. Caulton, K. Folting, A. Meli, M. Peruzzini, A.
Polo, F. Vizza, J. Am. Chem. Soc. 1992, 114, 7290.
Angew. Chem. Int. Ed. 2003, 42, 2636 – 2639
www.angewandte.org
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2639