Entrapment of rhodium complexes in inorganic or hybrid matrices via the sol-gel method

Article abstract of DOI:10.1039/b207522a

Rhodium complexes have been entrapped inside the porous systems of inorganic or hybrid matrices via the sol-gel method. The resulting materials were tested as catalysts in hydroformylation reactions. The microporous materials could be recycled without any rhodium leaching, being active even in the absence of a solvent. High turnover numbers were obtained in the hydroformylation of either 1-hexene or 1-decene. However, the characteristics of the matrix could not be related to the addition (or not) of a hybrid organic-inorganic co-condensation agent, but appeared to depend on the nature of the rhodium complex.

Full text of DOI:10.1039/b207522a

View Article Online / Journal Homepage / Table of Contents for this issue
Entrapment of rhodium complexes in inorganic or hybrid matrices via
the sol–gel method
´
Jose Daniel Ribeiro de Campos and Regina Buffon*
´
Instituto de Quımica, Universidade Estadual de Campinas, P.O. Box 6154, 13083-970,
Campinas-SP, Brazil. E-mail: rbuffon@iqm.unicamp.br; Fax: +55 19 3788 3023;
Tel: +55 19 3788 3090
Received (in Montpellier, France) 31st July 2002, Accepted 7th October 2002
First published as an Advance Article on the web 2nd January 2003
Rhodium complexes have been entrapped inside the porous systems of inorganic or hybrid matrices via the
sol–gel method. The resulting materials were tested as catalysts in hydroformylation reactions. The
microporous materials could be recycled without any rhodium leaching, being active even in the absence of a
solvent. High turnover numbers were obtained in the hydroformylation of either 1-hexene or 1-decene.
However, the characteristics of the matrix could not be related to the addition (or not) of a hybrid
organic–inorganic co-condensation agent, but appeared to depend on the nature of the rhodium complex.
The development of well-defined catalyst systems that combine
good catalytic activity with facile and quantitative catalyst/
products separation is still a challenge. An approach to facili-
tate catalyst recovery is the attachment of homogeneous cata-
lysts to inorganic or hybrid organic–inorganic matrices.
Inorganic matrices such as silica are advantageous catalyst
supports due to their physical strength and chemical inertness.
The entrapment of homogeneous catalysts inside the porous
system of silica matrices prepared by the sol–gel method
appears to be a promising strategy for catalyst recovery.1
The addition of hybrid organic-inorganic co-condensation
agents [(RO)3Si–R0–Si(OR)3] to the classical sol–gel precursors
[Si(OR)4] leads to three-dimensional networks with an organic
fragment as an integral component. Such hybrid inorganic-
organic matrices have been studied by many groups,2–4 with
polysiloxane-bound transition metal complexes being exten-
sively studied by Lindner et al.5 Although the entrapment of
hydroformylation catalysts, bearing or not a hydrolysable
ligand, inside the porous system of inorganic matrices has been
reported,6,7 there are no studies concerning the characteristics
of the matrix leading to a leach-proof system. In related works
we have noticed that, at least in the absence of a hydrolysable
ligand [i.e., a ligand bearing an –Si(OR)3 group], in order
to avoid leaching the matrix should present a microporous
system.8 In this work we investigated the effects of the nature
of the precursor complex, of the phosphine and also of the
presence of a hybrid co-condensation agent on the properties
of the matrix as well as on the recycling of the resulting hydro-
formylation catalysts.
Catalyst preparation. In a typical preparation, 0.01 mmol of
the rhodium complex and 0.05 mmol of the phosphine were
dissolved in 6 ml of THF (degassed through several cool-thaw
cycles) in a 50 ml Schlenk flask. After 15 min stirring, 2.0 ml
(11.10 mmol) of deionized water, 2.0 ml (13.56 mmol) of
TMOS (tetramethylorthosilicate), 1.0 ml (3.39 mmol) of 1,4-
bis(triethoxysilyl)benzene (co-condensation agent), 1.0 ml
(24.69 mmol) of methanol and 4 drops of a 3% solution of
(AcO)2Sn(Bu)2 in polydimethylsiloxane (Dow Corning) were
added to the yellow solution. The inorganic matrices were pre-
pared in the same way, but without the addition of the co-
condensation agent. The resulting solution was stirred for
15 min and then allowed to stand for at least overnight until
gelation. The gel was then dried under vacuum, washed with
CH2Cl2 in a Soxhlet and dried again under vacuum. The
resulting yellow-orange materials were stored under air at
room temperature.
Catalytic experiments
All catalytic experiments were performed in a 100 ml stainless
steel Parr reactor. The reaction temperature was kept at 80 ꢁC
and the solution was stirred at 300 rpm. In a typical experi-
ment, 2–5 mmol of the rhodium complex (ꢂ0.250 g of the cat-
alyst), 0.28 g (2–5 mmol) 1-hexene ([Rh]/[olefin] ꢂ 1/1000); 0.1
g cyclooctane (internal standard) and 30 ml of THF (solvent)
were employed. The reactor was first purged with H2 , and then
pressurized at 50 bar (CO/H2 ¼ 1/1). For recycling experi-
ments, the catalyst was separated by filtration, washed in a
Soxhlet with CH2Cl2 , dried under vacuum and used in a
new run. GC analyses were performed in an HP5890 series
II gas chromatograph, equipped with an HP5 capillary column
(50 m ꢃ 0.2 mm) and a flame ionization detector. Products
were quantified using calibration curves obtained with stan-
dard solutions.
Experimental
Materials
RhCl3ꢀ3H2O (Strem), [RhCl(COD)]2 (Strem), [Rh(CO)2(acac)]
(Strem), tetramethylorthosilicate (TMOS, Aldrich), dichloro-
methane (Merck), 1-hexene (Aldrich), styrene (Aldrich) and
cyclooctane (Fluka) were used as received. Toluene (Merck)
and tetrahydrofuran (Merck) were distilled under argon over
sodium/benzophenone. [Rh(OMe)(COD)]2 , [RhCl(CO)2]2 ,
[Rh(acac)(COD)] and 1,4-bis(triethoxysilyl)benzene were pre-
pared according to published methods.9–12
Catalyst characterization
Nitrogen adsorption isotherms were determined at ꢄ196 ꢁC
with a Micromeritics ASAP 2010 automated porosimeter. All
calculations were performed using the associated Micromeri-
tics software. Samples were degassed at 80 ꢁC for a minimum
446
New J. Chem., 2003, 27, 446–451
DOI: 10.1039/b207522a
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
of 24 h prior to measurements. Sample size varied from 200 to
300 mg.
Rhodium concentrations were determined through ICP-
AES analyses (Perkin–Elmer Optima 3000 DV).
([Rh(OMe)(COD)]2 + dppe) was active in at least 8 runs, the
activity increasing in each run. Conversely, the catalytic activ-
ity of system 3a ([RhCl(CO)2]2 + dppe) decreased in each of 5
runs. Except in two cases with high rhodium leaching, isomer-
ization of the substrate, mainly to trans-2-hexene, was always
observed. The ratio between linear and branched aldehydes
varied from 1.6 to 2.6 for the leachless systems (1a, 3a, 3b
and 4b), the dppe-based catalysts leading to higher n/i values.
Hybrid matrices based on [RhCl(CO)2]2 + dppe or dppf (sys-
tems 3b and 4b, respectively) turned out to be the best catalysts
among those studied (higher turnover numbers and hydro-
formylation/isomerization ratios).
These two systems were also tested in reactions without a
solvent. As it can be seen in Table 2, high turnover numbers
were obtained in all cases, even when using 1-decene as a sub-
strate (Table 2, 2nd row) without any rhodium leaching. Table 2
also shows that the entrapped catalysts are less active than
their homogeneous counterparts, which was indeed expected
owing to the diffusion problems intrinsic to a porous material.
Nevertheless, for similar conversions, using THF as a solvent
the entrapped catalysts were more selective (higher n/i values)
than their homogeneous counterparts. For comparison, some
homogeneous experiments were also carried out without addi-
tion of phosphines: these complexes were far less selective than
the entrapped catalysts, suggesting that the complexes depicted
in the Scheme are indeed the precursors of the active species.
System 3b was further tested in the hydroformylation of
styrene. Using [Rh]/[olefin] ¼ 1/1000, a turnover number of
770 was observed after 24 h, with n/i ¼ 0.5.
Solution 31P NMR spectra were recorded on a Gemini 300 P
instrument at 121.5 MHz, in CDCl3 . 31P CP-MAS NMR ana-
lyses were carried out on a Varian INOVA 500 spectrometer,
at 202 MHz, using H3PO4 as a reference. The contact time
was 1 ms with a 4 s delay between each scan. Around 16 000
scans were accumulated. 29Si CP-MAS NMR analyses were
performed on the same spectrometer. The contact time was 2
ms with a 1.1 s delay between each scan. Around 8000 scans
were accumulated. Double-bearing zirconia rotors were
employed.
Scanning electron micrographs were obtained using a Jeol–
JMS T300 equipment. TEM images were obtained on a Zeiss
CEM-902 apparatus equipped with a CCD-Proscan camera
and a high speed/slow scan system controller. The samples
were suspended in iso-propanol and dispersed on carbon-
coated copper grids.
Results
In order to determine the effects of the nature of the rhodium
complex on the structure of the matrix, four different precur-
sors, viz. [RhCl(CO)2]2 , [Rh(OMe)(COD)]2 , [Rh(CO)2(acac)]
and [Rh(COD)(acac)], as well as three chelating phosphine
ligands, viz. dppe, DPEphos and dppf, characterized by dif-
ferent bite angles (78, 102 and 99ꢁ, respectively13), were
employed. The expected resulting complexes, according to
the 31P NMR spectra of the corresponding solutions, are
shown in the Scheme. We tried to prepare both inorganic, sys-
tems a, and hybrid matrices [in this case by using (EtO)3Si–Ph–
Si(OEt)3 as a co-condensation agent], systems b, with each
rhodium complex/phosphine pair. However, in the case of
[RhCl(CO)2]2/dppf we failed to prepare the inorganic matrix
owing to precipitation of the rhodium complex. All systems
were tested in the hydroformylation of 1-hexene and most
were also characterized by N2 adsorption-desorption iso-
therms. The corresponding results are summarized in Table 1.
Some systems were also characterized by 31P and 29Si CP-MAS
NMR and/or electron microscopy.
Nitrogen adsorption/desorption experiments
The nitrogen adsorption-desorption isotherms of catalysts 1a,
3b and 4b are of type I (IUPAC classification14), characteristic
of microporous systems. In only one case was a small hyster-
esis in their desorption branches observed (system 3b, Fig. 1),
suggesting that the pores are mainly smooth and cylindrical,
with a low contribution of mesopores.15 The volumes adsorbed
at the lowest relative pressure indicate, in all cases, a large
volume of extremely small pores. The Horvath–Kawazoe dif-
ferential pore volume plots for catalysts 1a (as a fresh sample
and after 8 runs) and 3b are shown in Fig. 2 (catalyst 4b pre-
sented a homogeneous pore distribution). BET surface areas,
pore volumes and average pore size determined from the iso-
therms are presented in Table 1.
Catalytic experiments
Catalysts 2b, 5a,b and 6a,b are characterized by type IV iso-
therms (Fig. 1), with catalyst 6a (as well as 5a) presenting a
type H2 hysteresis loop (IUPAC classification14). The shape
of this isotherm indicates the presence of ink-bottle or narrow-
mouth shaped pores.15 The hysteresis loop observed in the
isotherm of catalyst 2b is of type H4, indicating that its meso-
porous system is not well defined. The volumes adsorbed at the
lowest relative pressure represent ca. 50% and 30% of the total
pore volume for catalysts 2b and 6a (as well as 5a), respec-
tively, indicating in both cases a significant volume of extre-
mely small pores. Systems 5b and 6b present larger pores (up
to 12 nm), with the volume adsorbed at the lowest relative
pressure representing less than 10% of the total pore volume
for these catalysts. The BJH adsorption pore size distributions
for some of these systems are shown in Fig. 2. We failed to
obtain isotherms for systems 1b and 3a.
All systems were active in the hydroformylation of 1-hexene
but in several cases (systems 1b, 2a,b, 5a,b, and 6a,b) a
high level of Rh leaching was observed (Table 1). System 1a
29Si and 31P CP-MAS NMR
In order to try to correlate the degree of condensation with the
porosity of the hybrid or inorganic matrices, some catalysts
were also analyzed by 29Si CP-MAS NMR. These catalysts
were chosen based both on their different performances in cat-
alysis and on the nature of their porous systems. Fig. 3 shows
the spectra obtained for catalysts 2b, 4b and 5a. For the hybrid
Scheme 1 Rhodium species as characterized by 31P NMR.
New J. Chem., 2003, 27, 446–451
447

View Article Online
Table 1 Composition, physical characteristics and catalytic activity of the inorganic (a) and hybrid (b) systems.a
BET
surface
Pore
volume/ pore
diameter/nm Run % Convb
Average
Rhodium
System complex
Rh wt % Phosphine area/m2 gꢄ1 cm3 gꢄ1
n/ic
H/Id
TONe
Rh leachingf
1a
[Rh(OMe)(COD)]2 0.25
dppe
480
0.21g
0.70g
1
2
4
5
8
1
1
1
1
2
3
5
1
5
1
2
6
1
1
1
1
34
32
40
57
65
59
38
55i
65
46
39
10
46
67
74
70
71
27j
91
51
90
2.6
2.5
2.5
2.5
2.0
2.6
2.6
2.3
2.0
2.4
2.4
2.4
2.5
2.1
1.6
1.9
1.7
2.5
0.8
0.7
0.7
1.7
1.7
1.6
1.5
1.8
1.8
1.1
1.3
2.0
1.7
1.7
1.9
1.7
1.9
3.0
2.4
2.6
–
340
320
400
570
650
590
380
550
650
460
390
100
810k
990k
740
700
710
270
910
510
900
Colorless solution
420
0.18g
0.72g
1b
2a
2b
3a
[Rh(OMe)(COD)]2 0.15
[Rh(OMe)(COD)]2 0.24
[Rh(OMe)(COD)]2 0.15
dppe
High
DPEphos
DPEphos
dppe
n.d.
420
n.d.
0.27h
n.d.
4.36h
0.7%
20%
[RhCl(CO)2]2
0.33
Colorless solution
3b
4b
[RhCl(CO)2]2
[RhCl(CO)2]2
0.23
0.21
dppe
dppf
370
410
0.17g
0.10g
0.85g
0.66g
Colorless solution
Colorless solution
5a
5b
6a
6b
[Rh(CO)2(acac)]
[Rh(CO)2(acac)]
[Rh(COD)(acac)]
[Rh(COD)(acac)]
0.12
0.07
0.03
0.02
dppf
dppf
670
390
750
360
0.53h
1.04h
0.60h
1.00h
2.95h
12.1h
2.68h
12.5h
High
High
High
High
19.2
4.1
–
DPEphos
DPEphos
a
b
c
Reaction conditions: [Rh]/[1-hexene] ꢂ 1/1000; 50 bar ([CO]/[H2] ¼ 1/1; 80 ꢁC; 24 h. Conversion to aldehydes. n/i ¼ ratio between linear and branched alde-
d
e
f
hydes. H/I ¼ ratio between conversion to hydroformylation and isomerization products. TON ¼ mol of aldehydes per mol of rhodium. Colorless solution:
g
h
i
j
k
ꢅ0.7% Rh. Horvath–Kawazoe method. BJH method. 3 h. 5 h. [Rh]/[1-hexene] ꢂ 1/1800.
matrices (2b, 4b), signals corresponding to T1 [–Si(OR)2R0,
for all systems showed only resonances ascribed to the free
2
3
0
0
=
=
=
d ꢂ ꢄ62), T [ Si(OR)R , d ꢂ ꢄ71] and T ( Si–R , d ꢂ ꢄ79)
ligand and/or to the corresponding phosphine oxide. Owing
to the low rhodium concentration in all cases, and also to
the excess of phosphine employed, the nonobservation of a sig-
nal ascribed to a rhodium-coordinated phosphine does not
imply that such a species was not formed.
sites, owing to the presence of the co-condensation agent, were
observed. Since all spectra were recorded using exactly the
same conditions, the relative intensities of the signals of the
Q2 [ Si(OH)2 , d ꢂ ꢄ91], Q [ Si–OH, d ꢂ ꢄ101] and Q
3
4
=
=
=
=
=
=
( Si–O–Si , d ꢂ ꢄ101) sites, Table 3, can be used to qualita-
=
tively compare the relative degree of TMOS condensation
between hybrid and inorganic matrices. Higher Q4/Q3 and
Q4/Q2 values were obtained with hybrid matrices (Table 3),
showing that the condensation degree of TMOS is higher when
the co-condensation agent is present. (The absolute degree of
condensation cannot be directly determined since the spectra
were collected using the cross-polarization technique.) In the
hybrid matrices, the relative degree of condensation of 1,4-bis-
(triethoxysilyl)benzene, which is based on the T3/T2 ratios,
does not seem to be related to the porosity of the matrix:
T3/T2 ratios for a mesopoporous and a microporous matrices
do not vary significantly (Table 3).
Electron microscopy
Fig. 5 shows transmission electron micrographs obtained from
microporous, 3b (top left), and mesoporous, 5a (top right),
materials. The TEM image from system 3a (bottom, left) sug-
gests a very dense matrix. The corresponding SEM image (bot-
tom, right) shows holes with diameters varying from 0.4 to
1.3 mm. Taking into account that this material was able to en-
trap a relatively large amount of rhodium and did not present
any leaching in the catalytic experiments, these micrographs
suggest a nonporous structure with packed microporous
domains. TEM experiments also showed that clustered rho-
dium particles were not formed in the cases studied.
Samples 1a,b, 2b, 3a, 4b and 5a were also analyzed by 31P
CP-MAS NMR. Nevertheless, except for the system 3a, which
was analyzed as a fresh sample and also after the catalytic
experiments (i.e., after 5 runs, seeFig. 4), the spectra obtained
Discussion
The co-condensation agent was expected to reduce the degree
of 3D-cross-linking, decreasing the rigidity and improving the
swelling properties of the resulting matrix. It was also expected
to affect the porous system and thus the diffusion of the
reagents inside the matrix.
Comparison of the systems 1a and 1b with 3a and 3b shows
that there is no clear correlation between the presence of the
co-condensation agent and the catalytic performance: when
the precursor is [RhCl(CO)2]2 (systems 3a and 3b), the hybrid
catalyst shows no leaching and keeps its catalytic activity for at
least 5 runs. If the precursor is [Rh(OMe)(COD)]2 (systems 1a
and 1b), the presence of the co-condensation agent leads to a
high rhodium leaching and catalysis seems to take place in
solution. On the other hand, in the absence of the co-conden-
sation agent (system 1a), the catalytic activity increases in each
run, suggesting that the formation of the active species (a rho-
dium hydride) is rather slow, its concentration increasing in
Table 2 Catalytic performance of some hybrid systems without sol-
vent as compared to homogeneous systems in THF.a
System
% Conv.b
n/i
H/I
TONc
3b [RhCl(CO)2]2 + dppe
3b [RhCl(CO)2]2 + dppe
4b [RhCl(CO)2]2 + dppf
35
72
42
64
56
67
34
46
46
1.7
1.4
2.2
2.4
1.7
1.5
2.1
0.80
0.67
11.9
2.9
1750
3600d
2300
9.2
21.3
1.9
3180e
2800f g
3350f h
1700f g
2300f g
2300f g
[Rh(OMe)(COD)]2 + dppe
2.5
1.3
[RhCl(CO)2]2 + dppe
[Rh(OMe)(COD)]2
[RhCl(CO)2]2
0.96
0.88
a
b
c
[Rh]/[1-hexene] ¼ 1/5000; 24 h. Conversion to aldehydes. mol
d
e
of aldehydes per mol of Rh. [Rh]/[1-decene] ¼ 1/5000. 48 h.
f
g
In THF. 2.5 h. 4 h.
h
448
New J. Chem., 2003, 27, 446–451

View Article Online
Fig. 1 Nitrogen adsorption/desorption isotherms for several catalysts.
each run. Considering only systems 1 and 3, it can also be seen
that the best catalysts, 1a and 3b (without and with co-conden-
sation agent, respectively), are characterized by a microporous
system, with a low mesoporous contribution.
system is significantly higher in the absence of the co-conden-
sation agent. The higher degree of TMOS condensation
obtained by addition of the co-condensation agent could
account for this observation: when microporous matrices were
obtained (catalysts 1a, 3b and 4b), the hybrid systems showed
smaller BET surface areas and pore volumes than the in-
organic ones. On the other hand, the relative degree of conden-
sation of the co-condensation agent seems to depend on the
nature of the entrapped rhodium complex and cannot be
directly correlated to the porosity of the matrix.
Systems 1 to 4 were prepared using (AcO)2SnBu2 as a con-
densation catalyst, allowing gelation to take place overnight.
In the other cases it was not employed and gelation took at
least 3 days. Systems 5 and 6 were prepared without the tin cat-
alyst since it was expected that cationic rhodium complexes
would be able to catalyze the condensation reaction. Rhodium
The continuous deactivation of system 3a could be due to
degradation of the rhodium complex since metal leaching
was not detected; a color change, from yellow to orange, was
indeed observed after the first run. Nevertheless, 31P CP-
MAS NMR spectra of a fresh sample of system 3a and after
5 runs show only one broad signal at d ꢂ 38, tentatively
ascribed to [RhCl(dppe)]2 (d 35) and/or to the corresponding
dppe-oxide (d 33). Thus, the catalyst deactivation is probably
due to pore blockage, as suggested by the dense structure
revealed by the transmission electron image.
Comparing the systems 1 through 4, it can also be seen that
the amount of rhodium complex entrapped inside the porous
Fig. 2 Pore size distribution for several catalysts.
New J. Chem., 2003, 27, 446–451
449

View Article Online
Fig. 4 31P CP-MAS NMR spectra for 3a: (a) fresh sample; (b) after
5 runs.
Fig. 3 29Si CP-MAS NMR spectra for catalysts 2b, 4b and 5a.
complexes containing the acacꢄ ligand react with biphosphines
leading to cationic species of the type [RhL2(biphosphine)]+-
[acac]ꢄ (L ¼ neutral ligand).7 The data in Table 1 show that
this strategy has led to the entrapment of still lower amounts
of rhodium. It can also be observed that when the acac-
containing precursors were employed the resulting matrices
presented a mesoporous system, which does not seem to be
able to entrap adequately the rhodium complexes. Moreover,
for systems 5 and 6 the influence of the co-condensation agent
was clear: it led to a significant decrease in the volume
of micropores and to a large distribution of mesopores.
Therefore, the rhodium complexes were washed away during
Table 3 Relative intensities of 29Si CP NMR signals for some cata-
lysts
Catalyst
Q4/Q2
Q4/Q3
T3/T2
Porous system
5a
1a
2b
1b
4b
1.39
1.77
2.46
2.63
3.15
0.43
0.36
0.48
0.52
0.58
Mesoporous
Microporous
Mesoporous
0.40
0.50
0.36
Microporous
Fig. 5 TEM images (bar: 10 nm) from systems 3b (top left), 5a (top right), 3a (bottom left); SEM image from system 3a (bottom right; bar: 2 mm).
450
New J. Chem., 2003, 27, 446–451

View Article Online
the Soxhlet washing, and the very small amount (ꢂ0.02 wt %)
of remaining complexes leached in the catalytic experiments.
All systems prepared with DPEphos, starting either with
[Rh(OMe)(COD)]2 or [Rh(COD)(acac)], showed a high rho-
dium leaching and also a mesoporous system (for the three
analyzed cases). The higher bite angle of this ligand along with
its large flexibility (86–120ꢁ)13 might favor larger pores in order
to better accommodate a more voluminous complex.
From all data available, it appears that a microporous sys-
tem is necessary to avoid leaching. This conclusion is in agree-
ment with results reported in the case of entrapped rhenium or
molybdenum epoxidation catalysts, at least when no hydroly-
sable ligand was employed.8 However, the ambiguous results
obtained in this work concerning the effects of the co-conden-
sation agent are in contrast with previous works: for rhenium
and molybdenum epoxidation and ruthenium hydrogenation
catalysts, at least with the same TMOS (or TEOS)/1,4-bis-
(triethoxysilyl)benzene molar ratio (4/1), the co-condensation
agent ensured a microporous system and higher surface
areas.8,16
nature of the final rhodium complexes. Thus, it seems that
cationic rhodium species favor the formation of mesoporous
structures; therefore, acac-containing precursors are to be
avoided when no hydrolysable ligand is employed.
Acknowledgements
Financial support from FAPESP, as well as a fellowship to
J.D.R.C., are gratefully acknowledged.
References
1
J. Blum, D. Avnir and H. Schumann, Chemtech, 1999, 29(Feb.),
32.
U. Schubert, New J. Chem., 1994, 18, 1049.
K. J. Shea and D. A. Loy, Acc. Chem. Res., 2001, 34, 707.
(a) R. J. P. Corriu and D. Leclercq, Angew. Chem., Int. Ed Engl.,
1996, 35, 1420; (b) P. Chevalier, R. J. P. Corriu, P. Delord, J. J. E.
Moreau and M. W. Chi Man, New J. Chem., 1998, 22, 423.
E. Lindner, T. Schneller, F. Auer and H. A. Mayer, Angew.
Chem., Int. Ed., 1999, 38, 2154.
2
3
4
5
6
(a) S. Wieland and P. Panster, Catal. Org. React., 1994, 62, 383;
(b) J. Blum, A. Rosenfeld, N. Polak, O. Israelson, H. Schumann
and D. Avnir, J. Mol. Catal., 1996, 107, 217.
Conclusions
7
8
9
A. J. Sandee, L. A. van der Veen, J. N. H. Reek, P. C. J. Kamer,
M. Lutz, A. L. Spek and P. W. N. M. van Leeuwen, Angew.
Chem., Int. Ed., 1999, 38, 3231.
(a) K. Dallmann and R. Buffon, Catal. Commun., 2000, 1, 9; (b) S.
Teixeira, K. Dallmann, U. Schuchardt and R. Buffon, J. Mol.
Catal. A: Chem., 2002, 182–183, 167.
Rhodium complexes derived from [RhCl(CO)2]2 or [Rh(O-
Me)(COD)]2 and dppe or dppf could be prepared in situ and
entrapped inside the porous systems of inorganic or hybrid
matrices via the sol–gel method. Only the microporous systems
could be re-used without rhodium leaching. The hybrid sys-
tems 3b and 4b turned out to be the best catalysts studied in
this work, being highly active and stable for at least 7 days.
They are prepared from commercially available phosphines
and easily synthesized rhodium precursors. Higher olefins such
as 1-decene and styrene could also be hydroformylated. High
turnover numbers were obtained in the absence of a solvent.
The selectivity to linear aldehydes was, in some cases, higher
than that observed in homogeneous solution.
´
R. Uson, L. A. Oro and J. Cabeza, Inorg. Synth., 1985, 23, 126.
10 J. A. McCleverty and G. Wilkinson, Inorg. Synth., 1966, 8, 211.
11 R. Cramer, J. Am. Chem. Soc., 1964, 86, 217.
12 D. A. Loy, G. M. Jamison, B. M. Baugher, S. A. Myers, R. A.
Assink and K. J. Shea, Chem. Mater., 1996, 8, 656.
13 P. C. J. Kamer, J. N. H. Reek and P. W. N. M. van Leeuwen,
Chemtech, 1998, 28(Sept.), 27.
14 K. S. W. Sing, D. H. Everett, R. A. Haul, L. Moscou, R. A.
Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem.,
1985, 57, 603.
15 C. J. Brinker and G. W. Scherer, Sol–Gel Science: The Physics and
Chemistry of Sol–Gel Processing, Academic Press, San Diego,
1990, pp. 522–525.
16 K. Dallmann and R. Buffon, J. Mol. Catal. A: Chem., 2002, 185,
187.
Although the hybrid matrices led to the best catalysts, the
presence of the hybrid co-condensation agent did not ensure
the formation of a microporous matrix. Its presence seemed
to increase the degree of condensation of TMOS. The porosity
of the matrix, however, appeared to be more dependent on the
New J. Chem., 2003, 27, 446–451
451