Supported organometallic complexes. Part 27: Novel sol-gel processed rhodium(I) complexes: Synthesis, characterization, and catalytic reactions in interphases

Article abstract of DOI:10.1016/S0020-1693(01)00675-2

A novel T-silyl functionalized cationic (COD)(dppp)rhodium(I) complex was sol-gel processed with various amounts of the co-condensing agents MeSi(OMe)₂(CH₂)₆(OMe)₂SiMe and MeSi(OMe)₂(CH₂)₃ (C₆H₄)(CH₂)₃(OMe)₂SiMe to give novel stationary phases for 'Chemistry in Interphases'. The polysiloxane matrices and the integrity of the rhodium(I) complex centers were investigated by means of multinuclear solid-state NMR (¹³C, ²⁹Si, ³¹P) and EXAFS spectroscopies. Dynamic NMR measurements show an increasing mobility of the matrix and the reactive centers with a higher amount of the co-condensing component. The accessibility of the anchored rhodium(I) centers was scrutinized by the metal catalyzed hydrogenation of 1-hexene. All applied xerogels show remarkable activities and selectivities. An enhancement of the activities is achieved when polar solvents are used. SEM micrographs reveal the morphology of the hybrid materials and energy dispersive X-ray spectroscopy (EDX) suggests that the distribution of the elements is in satisfying agreement with the applied composition.

Full text of DOI:10.1016/S0020-1693(01)00675-2

www.elsevier.com/locate/ica
Inorganica Chimica Acta 327 (2002) 54–65
Supported organometallic complexes. Part 27: novel sol–gel
processed rhodium(I) complexes: synthesis, characterization, and
catalytic reactions in interphasesꢀ
Ekkehard Lindner a,*, Stefan Brugger a, Stefan Steinbrecher b, Erich Plies b,
Michael Seiler c, Helmut Bertagnolli c, Peter Wegner a, Hermann A. Mayer a
a Institut fu¨r Anorganische Chemie der Uni6ersita¨t Tu¨bingen, Auf der Morgenstelle 18, D-72076 Tu¨bingen, Germany
b Institut fu¨r Angewandte Physik der Uni6ersita¨t Tu¨bingen, Auf der Morgenstelle 10, D-72076 Tu¨bingen, Germany
c Institut fu¨r Physikalische Chemie der Uni6ersita¨t Stuttgart, Pfaffenwaldring 55, D-70569 Stuttgart, Germany
Received 9 April 2001; accepted 27 July 2001
Dedicated to Professor Kees Vrieze
Abstract
A novel T-silyl functionalized cationic (COD)(dppp)rhodium(I) complex was sol–gel processed with various amounts of the
co-condensing agents MeSi(OMe)2(CH2)6(OMe)2SiMe and MeSi(OMe)2(CH2)3(C6H4)(CH2)3(OMe)2SiMe to give novel stationary
phases for ‘Chemistry in Interphases’. The polysiloxane matrices and the integrity of the rhodium(I) complex centers were
investigated by means of multinuclear solid-state NMR (13C, 29Si, 31P) and EXAFS spectroscopies. Dynamic NMR measurements
show an increasing mobility of the matrix and the reactive centers with a higher amount of the co-condensing component. The
accessibility of the anchored rhodium(I) centers was scrutinized by the metal catalyzed hydrogenation of 1-hexene. All applied
xerogels show remarkable activities and selectivities. An enhancement of the activities is achieved when polar solvents are used.
SEM micrographs reveal the morphology of the hybrid materials and energy dispersive X-ray spectroscopy (EDX) suggests that
the distribution of the elements is in satisfying agreement with the applied composition. © 2002 Published by Elsevier Science B.V.
Keywords: Rhodium(I) complexes; Sol–gel process; Xerogels; Solid-state NMR spectroscopy; EXAFS spectroscopy; EDX measurements;
Catalytic reactions in interphases
troduced recently [11]. An interphase is established if a
stationary phase and a mobile component penetrate
1. Introduction
each other on a molecular scale without forming a
In order to achieve the separability of homogenous
homogeneous phase. If such interphases are provided
catalysts from the product, many efforts have been
with a swellable polymer, they are able to imitate
made to heterogenize transition metal complexes [2–6].
homogenous conditions, because the active centers (e.g.
Anchoring of reactive centers to polymers [7] or inor-
a catalytically active metal complex) become highly
ganic materials [8,9] entails several momentous disad-
mobile simulating the properties of a solution and
vantages such as high degree of metal leaching or
hence they are accessible for substrates.
inhomogeneity of the reactive centers [10]. As an effi-
cient alternative for the reduction or even elimination
Typical interphases are generated by simultaneous
co-condensation of T-functionalized ligands or transi-
of these handicaps, ‘Chemistry in Interphases’ was in-
tion-metal complexes with various alkoxysilanes [11,12].
This procedure affords a possibility to modify materials
ꢀ For Part 26 see Ref. [1].
* Corresponding author. Tel.: +49-7071-29 72039/22; fax: +49-
7071-29 5306.
by controlling the density and the distance of the
reactive centers [6,13]. Frequently applied co-condens-
E-mail address: ekkehard.lindner@uni-tuebingen.de (E. Lindner).
ing agents are alkoxysilanes such as Si(OEt)4 (Q0),
0020-1693/02/$ - see front matter © 2002 Published by Elsevier Science B.V.
PII: S0020-1693(01)00675-2

E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
55
MeSi(OMe)3 (T0), and Me2Si(OMe)2 (D0) [14–18]. By
the development of the D-bifunctionalized organosi-
lanes MeSi(OMe)2(CH2)z(MeO)2SiMe (z=6, 8, 14;
D0ꢀCzꢀD0) [19] and MeSi(OMe)2(CH)z(C6H4)(CH2)z-
(MeO)2SiMe (z=3, 4; Ph(1,4-CzD0)2) [20] the highly
crosslinked and mobile of Q- and D-groups, respec-
tively have been combined successfully. If these D-bi-
functionalized organosilanes are co-condensed with
metal complexes or ligands, then highly mobile inter-
phases are formed and the reactive centers are accessi-
ble even for large molecules [21–24].
In this work we wish to report on the synthesis and
characterization of novel cationic rhodium(I) complexes
anchored to polysiloxane-based inorganic–organic hy-
brid materials. The rhodium centers in these complexes
are coordinated to a modified 1,3-bis(diphenylphos-
phinylpropane) ligand (dppp) [1(T0)] with a spacer of
six methylene units. The monomeric T-silyl functional-
ized complex 2(T0) was sol–gel processed with different
ratios of the D-bifunctionalized co-condensation
agents, D0ꢀC6ꢀD0 and Ph(1,4-C3D0)2. Structural investi-
gations of these novel stationary phases were carried
out by multinuclear solid-state NMR spectroscopy.
With these methods, information on the dynamic pro-
perties of the new materials is available [25–27]. An
exemplary examination of the coordination sphere of
the rhodium(I) center was performed by means of
extended X-ray absorption fine structure (EXAFS)
spectroscopy, as this method operates in the case of
amorphous materials like interphases [19,21]. The mor-
phology of the materials was investigated by scanning
electron microscopy (SEM) and the elemental distribu-
tion is revealed by EDX spectroscopy. Finally, the
accessibility of these polysiloxane-bound complexes was
investigated by the catalyzed hydrogenation of 1-
hexene.
spectra (FD) were acquired on a Finnigan MAT 711A
instrument (8 kV, 333 K) and reported as mass/charge
(m/z).
CP/MAS solid-state NMR spectra were recorded on
Bruker DSX 200 and Bruker ASX 300 (temperature
dependent measurements) multinuclear spectrometers
equipped with wide bore magnets (field strength 4.7 and
7.05 T). Magic angle spinning was applied at 10 kHz (4
mm ZrO2 rotors) and 3–4 kHz (7 mm ZrO2 rotors),
respectively. Frequencies and standards: 31P: 81.961
MHz (4.7 T), 121.442 MHz (7.05 T) [85% H3PO4,
NH4H2PO4 (l 0.8) as second standard]; 13C: 50.228
MHz (4.7 T), 75.432 MHz (7.05 T) [TMS, carbonyl
resonance of glycine (l 176.05) as second standard];
29Si: 39.73 MHz (4.7 T), 59.595 MHz (7.05 T) (Q8M8 as
second standard). All samples were packed under exclu-
sion of molecular oxygen. The cross polarization con-
stants TXH were determined by varying the contact
time, Tc (14–16 experiments). The proton relaxation
times in the rotating frame T1rH were measured by
direct proton spin lock-t-CP experiments [30] via 29Si
and 31P. The relaxation parameters were obtained using
the Bruker software SIMFIT and WINFIT following the
procedure described in Ref. [19].
EXAFS measurements were performed at the Rh
K-edge (23220 eV) at the beamline X1.1 of the Ham-
burger Synchrotron–Strahlungslabor (HASYLAB) in
DESY, Hamburg with
a Si(311) double crystal
monochromator under ambient conditions (5.46 GeV,
beam current 93 mA). Data were collected in transmis-
sion mode with ion chambers. Energy calibration was
monitored with a 20 mm thick rhodium metal foil at the
K-edge (23 220 eV). All measurements were performed
under an inert gas atmosphere. The samples were pre-
pared out of
a
mixture of the xerogels and
polyethylene. Data were analyzed with a program pack-
age specially developed for the requirements of amor-
phous samples [31]. The program AUTOBK of the
University of Washington [32] was used for background
removal, and the program EXCURV92 [33] was used for
the evaluation of the XAFS function. The resulting
EXAFS function was weighted with k3. Data analysis
in k space was performed according to the curved-wave
multiple-scattering formalism of the program EX-
CURV92. The mean free path of the scattered electrons
was calculated from the imaginary part of the potential
(VPI was set to −4.00), the amplitude reduction factor
AFAC was fixed at 0.8, and an overall energy shift DE0
was introduced to fit the data. In the fitting procedure
the intermolecular coordination numbers were varied
and the intramolecular coordination numbers were
fixed according to the known values of the ligands
around the Rh atom.
2. Experimental
2.1. Materials and instrumentation
Elemental analysis was carried out on a Vario EL
Analyzer (Elementar Analytische Systeme Hanau). IR
data were obtained on a Bruker IFS 48 FT-IR spec-
trometer. Solution nuclear magnetic resonance spectra
were recorded on a Bruker DRX 250 spectrometer
(field strength 5.87 T) at 296 K. Frequencies are as
1
follows: H NMR: 250.13 MHz; 31P{1H} NMR: 101.25
MHz (referenced to 85% H3PO4); 13C{1H} NMR: 62.90
MHz. 103Rh NMR resonances were measured using a
2D (31P, 103Rh){1H} experiment [28]. Chemical shifts in
1H and 13C{1H} NMR spectra were measured relative
to the partially deuterated solvent peaks which are
reported relative to TMS. The 103Rh chemical shift
values are referred to ](Rh)=3.16 MHz [29]. Mass
The surface areas were determined by analyzing the
N2 adsorption isotherms according to the BET method
using a Micromeritics Gemini II.

56
E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
Scanning electron micrographs and energy dispersive
kPa); integrator Hewlett–Packard 3390 A]. Leaching
investigations were carried out with a Varian Spectr
AA 20 Plus atomic absorption spectrometer (wave
length: 343.5 nm; lamp current: 5 mA; fuel: acetylene;
support: air).
X-ray analysis (EDX) were performed on a Philips XL
30 scanning electron microscope (SEM) equipped with
a DX-4 X-ray detection system by EDAX. This consists
of an energy dispersive Si(Li)-detector with an active
area of 10 mm2 and the eDX software package. The
primary beam energy was set to 20 keV for all measure-
ments. Micrographs were recorded detecting secondary
electrons generated by a probe current of 168 pA,
whereas 623 pA probe current were applied for carrying
out elemental analysis by EDX. Quantification of X-ray
emission spectra was achieved employing the ZAF cor-
rection procedure to convert X-ray intensities to ele-
mental contents [34]. The sample powder was placed in
a specimen stub covered with a conductive adhesive tab
and subsequently provided with a sputtered 20 nm gold
layer to avoid specimen charging. Owing to its high
fluorescence yield, gold emission lines are observed with
all spectra. However, the gold layer is too thin to
absorb X-rays efficiently and thus does not affect quan-
tification. Spectra were recorded with spot illumination
of the sample for 240 live seconds, yielding count rates
of about 2000 s−1. Spectra were acquired several times
at different sample positions to ensure reproducibility.
The various measurements were found to differ by
about 3%, which is within the limits of error, especially
for light element samples with pronounced topography.
Therefore, special care was taken to find specimen areas
exhibiting flat surfaces to ensure the validity of the
correction models. All measurements were performed at
ambient temperature.
2.3. Preparation of p4-1,5-cyclooctadiene[2-diphenyl-
phosphinylmethyl-1-diphenyl-phosphinyl-8-trimethoxy-
silyloctanerhodium(I)]hexafluoroantimonate(V) [2(T0)]
To a solution of AgSbF6 (164 mg, 0.48 mmol) in 3 ml
of THF a solution of [m-ClRh(COD)]2 (118 mg, 0.24
mmol) in 3 ml of THF was added under exclusion of
light. After stirring for 1 h a solution of 1(T0) (295 mg,
0.48 mmol) in 3 ml of THF was added. The reaction
mixture was stirred for 2 h and the precipitated AgCl
was removed by centrifugation. The solvent was then
removed under reduced pressure. The residue was dis-
solved in 3 ml of THF and 10 ml of n-pentane was
added to this solution. After stirring for 10 min, filtra-
tion (P3) and drying for 1 h 452 mg (88%) of 2(T0) was
obtained as a microcristalline, orange–yellow solid. FD
MS; m/z: 827.1 (M+−SbF6). Anal. Calc. for
C44H58F6O3P2RhSbSi (MW 1062.1): C, 49.69; H, 5.50.
1
Found: C, 49.32; H, 4.87%. H NMR (CDCl3, l ppm):
7.56–7.19 (m, 20H, H-phenyl), 4.86, 4.24 [br, 4H, H-
olefin (COD)], 3.47 (s, 9H, OCH3), 2.60–1.05 (m, 23H,
CH2, CH), 0.53–0.47 (m. 2H, CH2ꢀCH2ꢀSi). 13C{1H}
NMR (CDCl3, l ppm): 134.6–126.3 (m, C-phenyl),
102.2, 95.6 [br, C-olefin (COD)], 48.7 (OCH3), 38.4 (m,
CH2P), 31.0–20.7 (CH2, CH), 7.2 (CH2ꢀCH2ꢀSi).
1
31P{1H} NMR (CDCl3, l ppm): 15.5 (d, JRhP=141.4
All reactions and manipulations were carried out
under argon with the usual Schlenk techniques.
Methanol was dried with magnesium and distilled. All
other solvents were distilled from sodium benzophe-
none ketyl or calcium hydride. H2O and (n-
Bu)2Sn(OAc)2 were distilled under inert gas prior to
use. All solvents and reagents were stored under argon.
[m-ClRh(COD)]2 [35], the T-functionalized diphosphine
1(T0) [36] and the co-condensation agents D0ꢀC6ꢀD0
[19] and Ph(1,4-C3D0)2 [20] were synthesized as de-
scribed earlier.
Hz). 103Rh{1H} NMR (CDCl3, l ppm): −350. IR
(KBr, cm−1): w(SiꢀOMe) 1096, w(SbF6−) 669.
2.4. General procedure of sol–gel processing
To a solution of 2(T0) in 10 ml of THF the corres-
ponding amount of the co-condensing agent D0ꢀC6ꢀD0
or Ph(1,4-C3D0)2, 500 mg (28 mmol) of H2O and the
catalyst (n-Bu)2Sn(OAc)2 (30 mg, 0.087 mmol) was
added. The reaction mixture was stirred at room tem-
perature for 24 h until a gel was formed. The solvent
was then removed under reduced pressure and the
crude product was dried for 2 h. After washing three
times with toluene (5 ml), diethyl ether (5 ml) and
n-pentane (10 ml) and drying in vacuum overnight the
xerogels Xa–i were obtained as yellow powders.
2.2. Catalysis
The hydrogenation experiments were carried out in a
100 ml stainless steel autoclave equipped with a me-
chanical stirring bar. The autoclave was flushed with
argon prior to the introduction of the reaction mixture
(1–10 mmol catalyst with respect to rhodium, 6–60
mmol of 1-hexene and 25 ml of the solvent). The
suspension was set under hydrogen pressure and stirred
prior to heating to the desired temperature. The quanti-
tative analyses were performed on a GC 6000 Vega
Series 2 (Carlo Erba Instruments) with an FID and a
capillary column CP Sil 88 [17 m, carrier gas helium (50
Xa: initial weight of 2(T0) 513 mg (0.48 mmol). Yield:
467 mg. Anal. Calc. for C82H98F12O3P4Rh2Sb2Si2 (ideal-
ized stoichiometry): C, 49.52; H, 4.97. Corrected stoi-
chiometry [19]: C, 49.47; H, 5.06. Found: C, 42.87; H,
3.07% [37]1. 31P CP/MAS NMR (l ppm): 15.3. 13C
1 The low carbon value is due to the incomplete combustion of
silicon-containing compounds, which may be responsible for the
formation of SiC.

E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
57
CP/MAS NMR (l ppm): 144.4–121.1 (br, C-phenyl),
103.8, 98.0 [br, C-olefin (COD)], 40.2, 31.1, 28.8, 23.7,
12.0 (br, CH2, CH). 29Si CP/MAS NMR (l ppm):
−57.7 (T2), −64.8 (T3).
and spacer), −0.2 (SiꢀCH3). 29Si CP/MAS NMR (l
ppm): −9.6 (D1), −20.5 (D2), −55.5 (T2), −65.5
(T3).
Xg: initial weight of 2(T0) 315 mg (0.38 mmol) and of
Ph(1,4-C3D0)2 704 mg (1.9 mmol). Yield: 764 mg. Anal.
Calc. for C222H318F12O23P4Rh2Sb2Si22 (idealized stoi-
chiometry): C, 55.85; H, 6.71. Corrected stoichiometry
[19]: C, 54.49; H, 6.69. Found: C, 52.40; H, 6.08% [37].
31P CP/MAS NMR (l ppm): 14.0. 13C CP/MAS NMR
(l ppm): 144.4–122.6 (br, C-phenyl), 102.5–95.4 [br,
C-olefin (COD)], 49.8 (OCH3), 44.1–7.3 (br, CH2, CH
of backbone and spacer), −0.1 (SiꢀCH3). 29Si CP/
MAS NMR (l ppm): −12.5 (D1), −21.7 (D2), −54.7
(T2), −67.0 (T3).
Xb: initial weight of 2(T0) 500 mg (0.47 mmol) and of
D0ꢀC6ꢀD0 345 mg (1.18 mmol). Yield: 730 mg. Anal.
Calc. for C122H188F12O13P4Rh2Sb2Si12 (idealized stoi-
chiometry): C, 48.83; H, 6.31. Corrected stoichiometry
[19]: C, 47.87; H, 6.61. Found: C, 42.21; H, 5.13% [37].
31P CP/MAS NMR (l ppm): 13.9. 13C CP/MAS NMR
(l ppm): 145.4–120.1 (br, C-phenyl), 103.7, 97.9 [br,
C-olefin (COD)], 42.7–10.5 (br, CH2, CH of backbone
and spacer), −0.2 (SiꢀCH3). 29Si CP/MAS NMR (l
ppm): −9.8 (D1), −20.8 (D2), −57.1 (T2), −65.4
(T3).
Xh: initial weight of 2(T0) 185 mg (0.17 mmol) and of
Ph(1,4-C3D0)2 644 mg (1.70 mmol). Yield: 496 mg.
Anal. Calc. for C362H538F12O43P4Rh2Sb2Si42 (idealized
stoichiometry): C, 57.52; H, 7.17. Corrected stoichiome-
try [19]: C, 56.89; H, 7.00. Found: C, 57.90; H, 7.02%.
31P CP/MAS NMR (l ppm): 10.9. 13C CP/MAS NMR
(l ppm): 149.3–115.8 (br, C-phenyl), 108.4–97.3 [br,
C-olefin (COD)], 48.4–8.2 (br, CH2, CH of backbone
and spacer), −0.1 (SiꢀCH3). 29Si CP/MAS NMR (l
ppm): −12.8 (D1), −22.5 (D2), −67.7 (T3).
Xc: initial weight of 2(T0) 315 mg (0.38 mmol) and of
D0ꢀC6ꢀD0 559 mg (1.90 mmol). Yield: 705 mg. Anal.
Calc. for C162H278F12O23P4Rh2Sb2Si22 (idealized stoi-
chiometry): C, 48.49; H, 6.98. Corrected stoichiometry
[19]: C, 48.54; H, 6.88. Found: C, 45.44; H, 6.70% [37].
31P CP/MAS NMR (l ppm): 14.1. 13C CP/MAS NMR
(l ppm): 145.0–120.3 (br, C-phenyl), 104.2, 98.1 [br,
C-olefin (COD)], 49.5 (OCH3), 44.1–7.3 (br, CH2, CH
of backbone and spacer), −0.4 (SiꢀCH3). 29Si CP/
MAS NMR (l ppm): −11.5 (D1), −22.2 (D2), −59.0
(T2), −67.9 (T3).
Xi: initial weight of 2(T0) 205 mg (0.19 mmol) and of
Ph(1,4-C3D0)2 1430 mg (3.86 mmol). Yield: 1420 mg.
Anal. Calc. for C642H978F12O83P4Rh2Sb2Si82 (idealized
stoichiometry): C, 58.73; H, 7.51. Corrected stoichiome-
try [19]: C, 59.09; H, 7.61. Found: C, 58.74; H, 6.07%.
31P CP/MAS NMR (l ppm): 14.3. 13C CP/MAS NMR
(l ppm): 148.3–117.9 (br, C-phenyl), 108.3–92.4 [br,
C-olefin (COD)], 46.5–8.9 (br, CH2, CH of backbone
and spacer), −0.1 (SiꢀCH3). 29Si CP/MAS NMR (l
ppm): −11.8 (D1), −21.1 (D2), −54.7 (T2), −65.8
(T3).
Xd: initial weight of 2(T0) 185 mg (0.17 mmol) and of
D0ꢀC6ꢀD0 510 mg (1.74 mmol). Yield: 466 mg. Anal.
Calc. for C242H458F12O43P4Rh2Sb2Si42 (idealized stoi-
chiometry): C, 48.15; H, 7.65. Corrected stoichiometry
[19]: C, 46.93; H, 7.57. Found: C, 46.04; H, 7.38% [37].
31P CP/MAS NMR (l ppm): 14.1. 13C CP/MAS NMR
(l ppm): 140.2–116.1 (br, C-phenyl), 106.8–90.7 [br,
C-olefin (COD)], 49.8 (OCH3), 44.7–7.9 (br, CH2, CH
of backbone and spacer), −0.2 (SiꢀCH3). 29Si CP/
MAS NMR (l ppm): −12.4 (D1), −22.2 (D2), −59.0
(T2), −68.0 (T3).
Xe: initial weight of 2(T0) 205 mg (0.19 mmol) and of
D0ꢀC6ꢀD0 1130 mg (3.86 mmol). Yield: 1080 mg. Anal.
Calc. for C402H818F12O83P4Rh2Sb2Si82 (idealized stoi-
chiometry): C, 47.88; H, 8.18. Corrected stoichiometry
[19]: C, 46.08; H, 8.36. Found: C, 46.63; H, 7.56%. 31P
CP/MAS NMR (l ppm): 14.2. 13C CP/MAS NMR (l
ppm): 145.3–120.8 (br, C-phenyl), 110.5–91.4 [br, C-
olefin (COD)], 49.8 (OCH3), 44.4–8.2 (br, CH2, CH of
backbone and spacer), −0.2 (SiꢀCH3). 29Si CP/MAS
NMR (l ppm): −11.5 (D1), −20.6 (D2), −56.6 (T2),
−64.2 (T3).
3. Results and discussion
3.1. Synthesis of the monomeric rhodium(I) complex
[2(T0)]
The cationic rhodium(I) complex 2(T0) was obtained
in good yields as a orange–yellow microcrystalline
powder by the reaction of the diphos ligand 1(T0) with
a 0.5 equiv. of [m-ClRh(COD)]2 and 1 equiv. of AgSbF6
(Scheme 1). Complex 2(T0) was characterized by H,
1
13C{1H}, 31P{1H}, and 103Rh NMR spectroscopies and
mass spectrometry.
Xf: initial weight of 2(T0) 500 mg (0.47 mmol) and of
Ph(1,4-C3D0)2 437 mg (1.18 mmol). Yield: 665 mg.
Anal. Calc. for C152H208F12O13P4Rh2Sb2Si12 (idealized
stoichiometry): C, 53.99; H, 6.20. Corrected stoichiome-
try [19]: C, 53.04; H, 6.05. Found: C, 53.35; H, 9.23%.
31P CP/MAS NMR (l ppm): 14.3. 13C CP/MAS NMR
(l ppm): 148.6–115.1 (br, C-phenyl), 103.4–98.5 [br,
C-olefin (COD)], 47.1–9.5 (br, CH2, CH of backbone
3.2. Sol–gel processing of 2(T0)
The xerogel Xa was prepared by sol–gel processing
of the monomeric T-functionalized rhodium(I) complex
2(T0) without any co-condensing agent, whereas in the
case of Xb–i 2(T0) was condensed into a matrix with

58
E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
Scheme 1.
Table 1
Labeling of the compounds
Complex
Co-condensation agent
Ideal T/D ratio
Compound
Xerogel
2(T0)
2(T0)
2(T0)
2(T0)
2(T0)
2(T0)
2(T0)
2(T0)
2(T0)
2a(Tn)
Xa
Xb
Xc
Xd
Xe
Xf
Xg
Xh
Xi
D0ꢀC6ꢀD0
5
10
20
40
5
10
20
40
2b(Tn)(DiꢀC6ꢀDi)2.5
2c(Tn)(DiꢀC6ꢀDi)5
2d(Tn)(DiꢀC6ꢀDi)10
2e(Tn)(DiꢀC6ꢀDi)20
2f(Tn)(Ph(1,4-C3Di)2)2.5
2g(Tn)(Ph(1,4-C3Di)2)5
2h(Tn)(Ph(1,4-C3Di)2)10
2i(Tn)(Ph(1,4-C3Di)2)20
D0ꢀC6ꢀD0
D0ꢀC6ꢀD0
D0ꢀC6ꢀD0
Ph(1,4-C3D0)2
Ph(1,4-C3D0)2
Ph(1,4-C3D0)2
Ph(1,4-C3D0)2
Scheme 2.

E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
59
tee reproducible materials uniform reaction conditions
were maintained.
3.3. Solid-state NMR spectroscopic in6estigations
Owing to the amorphous nature of the polymers
Xa–i, solid-state NMR spectroscopy plays an impor-
tant role in investigating the structure and the dynamic
behavior of xerogels [25,26,38]. 29Si NMR spectroscopy
enables the characterization of the carrier matrix, the
degree of condensation and the stoichiometric composi-
tion of the materials. 13C and 31P NMR spectroscopies
allow an insight into the hydrocarbon backbone and
the reactive center, respectively (Fig. 1).
3.3.1. 29Si CP/MAS NMR spectroscopy
As a result of an incomplete condensation the 29Si
CP/MAS NMR spectra of the above-mentioned xero-
gels reveal signals of various substructures with cor-
responding Di- and Tn-functions. Typical chemical
shifts are l= −12.5 (D1), −21.7 (D2), −57.1 (T2),
and −67.0 (T3). They remain unchanged with respect
to the stoichiometric ratio between the co-condensation
agent and the functionalized rhodium(I) complex. All
silicon atoms in the polysiloxane matrix are in direct
proximity of protons, thus silyl species are detectable
via cross polarization [39,40]. The degree of condensa-
tion of D- and T-groups and the real T/D-ratios were
determined by contact time variation experiments
(Table 2) [19,41]. However, due to the low concentra-
tion of the T-functions in the xerogels Xd, e, h, i
T-groups can hardly be detected. Therefore contact
time variation experiments to determine the realistic
composition of these xerogels are not feasible within a
reasonable time.
With the exception of Xg and Xh, in the above-men-
tioned xerogels the degrees of condensation for the D-
and T-functions range between 79 and 95%, which is in
agreement with the earlier investigations [19,42]. In the
case of Xg the T-functions show a very low degree of
condensation, whereas in the case of Xh T- and D-
groups are highly condensed. The experimentally deter-
mined compositions do not differ significantly from the
applied stoichiometries, but it seems that in each sol–
gel process a small amount of the D- or T-groups were
washed out during the solvent processing.
Fig. 1. Solid-state NMR spectra of Xb: 29Si CP/MAS (top); 13C
CP/MAS (middle); 31P CP/MAS (bottom).
3.3.2. 31P and 13C CP/MAS NMR spectroscopies
In the 31P CP/MAS NMR spectra of the polysilox-
ane-bound rhodium(I) complexes Xa–i only one broad
signal is observed with chemical shifts between 10.9 and
15.3 ppm. This result reminds of the monomeric com-
plex measured in solution. No other phosphine species
is detectable in the freshly synthesized polymer, but
exposure to air for some hours leads to the formation
of phosphine oxide resonating at 32 ppm.
variable amounts of D0ꢀC6ꢀD0 or Ph(1,4-C3D0)2 (Table
1, Scheme 2). The properties of such xerogels strongly
depend on the reaction conditions during the sol–gel
process such as concentration of the monomers, type of
solvent, temperature, and kind of catalyst. All sol–gel
processes were carried out in THF with an excess of
water and (n-Bu)2Sn(OAc)2 as the catalyst. To guaran-

60
E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
The 13C CP/MAS NMR spectra of Xa–e reveal a
X-ray diffraction. However, the properties of EXAFS
spectroscopy allow the determination of the local struc-
ture around the excited Rh atoms, independent of the
samples’ physical state. An analysis of the EXAFS
provides information on the bond distance, the coordi-
nation number, the ‘Debye–Waller’ factor, and the
nature of the scattering atoms surrounding an excited
atom [43,44]. The k3 weighted EXAFS functions (Fig.
2) of Xa, Xb, and Xf can be described by two different
atom shells.
For the above-mentioned complexes the first inten-
sive peak in the corresponding Fourier transforms is
mainly due to carbon and phosphorus atoms. In the
case of Xa the assumption of four equivalent carbon
broad signal in the aromatic region, which stems from
the phenyl carbon atoms of the functionalized dppp
ligand. At around 100 ppm, two broad signals occur
originating from the olefinic C-atoms of the COD
ligand. The resonances of the different methylene car-
bon atoms between 40 and 10 ppm of Xa cannot be
resolved. In the case of Xb–e the aliphatic region is
dominated by four major signals resonating at around
33, 23, 18, and 0 ppm, which are assigned to the
DiꢀC6ꢀDi moiety. In the case of the hybrid polymers
Xf–i all main signals (average shifts: 139.4, 128.9, 39.1,
25.1, 16.8, and −0.3 ppm) with the exception of the
peaks originating from the COD ligand at around 100
ppm stem from the Ph(1,4-C3Di)2 component. The
weak or missing signal for the SiꢀOMe function indi-
cates a very high degree of hydrolysis.
,
atoms with an RhꢀC bond length of 2.14 A leads to a
good agreement between the experimental and the cal-
culated functions. Additionally, two equivalent phos-
phorus atoms were established with an average RhꢀP
,
distance of 2.28 A leading to a significant improvement
3.4. EXAFS structure determination of the rhodium(I)
complexes Xa, Xb, and Xf
in the fit (Table 3). In the case of Xb the assumption of
four equivalent carbon atoms with an RhꢀC distance of
,
The amorphous character of the xerogels makes it
impossible to get structural data with conventional
2.21 A and two equivalent phosphorus atoms with an
,
RhꢀP bond length of 2.29 A leads to a good agreement
Table 2
Relative I0, TSiH, and T1rH data of the silyl species in the xerogel
Xerogel
Relative I0 data of D and T species a Degree of
condensation (%)
Real T/D moiety
TSiH (ms) b
T1rH (ms) c
D1
D2
T2
T3
D
T
D1
D2
T2
T3
Xa
Xb
Xc
Xd
Xe
Xf
Xg
Xh
Xi
24.0
9.8
100
12.5
94
81
84
91
88
82
73
100
90
0.81
1.49
1.34
1.60
15.31
12.65
8.11
8.02
2.74
11.84
5.76
3.31
1.79
41.1
100
100
85
87
79
88
89
77
98
95
1:6.3
1:8.9
1.44
1.34
1.66
1.61
0.83
1.42
1.42
1.97
1.98
0.98
1.41
1.00
0.89
36.3
7.3
8.0
1.11
1.25
72.4 d
31.3 d
26.9
100 d 2.9 d
7.7 d
1:16.4 d
1:33.6 d
1:3.4
e
e
100 d 1.4 d
2.5 d
e
e
100
17.7
19.7
1.04
1.19
0.55
80.1
100
13.5
8.0
1:7.9
1.24
0.92
f
e
f
4.9 d
10.5 d
100 d
6.0 d
1.5 d
1:17.4 d
1:52.6 d
1.32
100 d 0.6 d
0.64
e
e
a I0 values calculated according to the literature methods.
b Determined by contact time variation.
c Determined via 29Si with the experiment according to Ref. [30].
d Determined via deconvolution of 29Si CP/MAS NMR spectra (Tc=5 ms).
e Intensity to low for a precise determination of TSiH
.
f Species not detectable.
Fig. 2. Calculated (solid line) and experimental (dotted line) k3(k) function of Xa (left) and their Fourier transforms (right) (Rh K-edge) in the
−1
,
k-range 3.85–13.00 A
.

E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
61
Table 3
EXAFS-determined structural data of the Xa, Xb, and Xf
a
N
Xa
Xb
Xf
b
c
b
c
b
c
,
,
,
,
,
,
| (A)
r (A)
| (A)
r (A)
| (A)
r (A)
RhꢀC d
RhꢀP d
4
2
2.1490.02
2.2890.02
0.08190.01
0.10990.03
2.2190.02
2.2990.02
0.05990.02
0.08190.01
2.2390.02
2.3190.02
0.06790.02
0.06390.01
a Coordination number N.
b Interatomic distance r.
c Debye–Waller factor |.
d Absorber-backscatterer.
between the experimental and the calculated functions.
3.5.2. Mobility of the reacti6e center
Temperature dependent measurements of the T1rH
values (via 31P) are summarized in Table 4. Owing to the
decreasing T1rH values in the applied temperature range,
all xerogels with the exception of Xe and Xi reside on
the slow motion regime of the correlation time curve
[25]. The trend of the T1rH values of compounds Xe and
Xi show a similar behavior. First, they decrease, and
then they pass through a minimum to increase again
with raising temperature. Therefore, these two materials
are located in the medium motion regime of the correla-
tion time curve and show the highest mobility of all the
xerogels mentioned [25]. These results indicate that the
mobility of the reactive centers as well as the mobility of
the matrix (vide supra) increases with a higher amount
of the co-condensing components. Owing to the electro-
static repulsion of the cationic rhodium(I) centers the
space of motion is reduced. If these positively charged
complexes are diluted across the carrier matrix, their
mobility is enhanced, because of the enlargement of the
motion radius.
In the xerogel Xf the found RhꢀC and RhꢀP bond
,
lengths are 2.23 and 2.31 A, respectively.
The structural data are in a good agreement with the
proposed structures of the Xa, Xb, and Xf. Further, in
comparison with the structural data of Xa, the RhꢀC
distances of Xb and Xf are slightly enlarged but still in
agreement with the determined bond lengths of similar
(COD)diphosphinerhodium(I) complexes [45,46].
3.5. Studies on the dynamic beha6ior of the xerogels by
solid-state NMR spectroscopy
To optimize the stationary component to be em-
ployed as catalyst in interphases, it is necessary to get
detailed information on the dynamic properties of the
materials and reactive centers. Owing to the sensitivity
toward motion, NMR parameters like cross-polariza-
tion constants and relaxation times give an access to the
dynamic properties of materials. Motions in the kilo-
hertz region are responsible for the spin-lattice relax-
ation time of the protons in the rotating frame (T1rH),
which was determined by a direct spin lock-t-CP experi-
ment [30] via 29Si and 31P.
3.6. Accessibility studies of the anchored rhodium(I)
complexes by catalytic reactions
Diphosphinerhodium(I) complexes are applied as
catalysts for the hydrogenation of various olefins
3.5.1. Mobility of the matrix
In solid state the xerogels do not differ significantly in
the cross polarization constant TSiH. The relaxation time
parameter (T1rH) data extracted from the 29Si CP/MAS
NMR measurements of the xerogels depend on the
amount of DiꢀC6ꢀDi and Ph(1,4-C3Di)2 moieties, respec-
tively (Table 2). Shorter T1rH values were recorded for
a higher amount of the co-condensing agents, thus
indicating an increase in the mobility of the matrix,
because, according to recent studies the T1rH values of
the xerogels are considered to reside on the slow motion
regime of the correlation time curve [19,23,24]. Compar-
ing the mobilities of the two different co-condensates, it
seems that there is no essential difference between the
materials with the phenyl ring in the backbone and their
counterparts with the alkyl chain as the main building
block.
Table 4
Temperature dependence of the T1rH values
Xerogel
T1rH (ms) a
295 K
305 K
315 K
325 K
335 K
b
b
b
b
b
b
Xa
Xb
Xc
Xd
Xe
Xf
Xg
Xh
Xi
9.01
11.98
5.09
10.20
4.15
13.81
7.55
6.82
10.68
7.15
11.30
1.25
13.17
10.14
5.88
9.94
3.81
9.36
0.91
9.22
6.21
5.30
2.62
9.96
2.75
10.46
3.08
9.79
2.87
3.29
2.92
8.85
3.13
8.06
b
2.49
b
6.80
0.89
a Determination via 31P according to Ref. [30].
b Not determined.

62
E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
Table 5
Hydrogenation of 1-hexene in toluene a
Xerogel
Conversion (%)
Hexane (%)
cis, trans 2-Hexene (%)
TON b
TOF c
Xa
Xb
Xc
80.8
40.9
71.6
61.4
86.1
77.1
38.6
13.9
22.9
2811
2136
3326
703
534
832
Xd d
Xe
45.7
100
40.4
91.0
84.7
73.3
9.0
13.3
26.7
2480
5035
1828
620
1259
457
Xf
Xg
Xh d
Xi
100
91.9
8.1
5575
1394
a Reaction conditions: H2 pressure 10 bar; Rh:1-hexene=1:6000; temperature 313 K; reaction time 4 h.
b Turnover number (molsub molc−at1).
c Turnover frequency (molsub molc−at1 h−1).
d Not determined.
Table 6
Influence of the solvent on conversion and selectivity a
Catalyst
Solvent
Conversion (%)
Hexane (%)
cis, trans 2-Hexene (%)
TON b
TOF c
Xc
Toluene
1,4-dioxane
methanol
71.6
75.5
85.5
77.1
75.5
78.7
22.9
24.5
21.3
3326
3368
4072
832
842
1018
Xg
Toluene
1,4-dioxane
methanol
40.4
36.7
86.6
72.3
71.1
76.0
26.7
28.9
24.0
1826
1571
3968
457
393
992
a Reaction conditions: H2 pressure 10 bar; Rh:1-hexene=1:6000; temperature 313 K; reaction time 4 h.
b Turnover number (molsub molc−at1).
c Turnover frequency (molsub molc−at1 h−1).
organic hybrid materials do not exhibit good swelling
abilities in polar solvents like methanol [20].
[45,47–51]. Therefore, the accessibility of the polysilox-
ane-bound rhodium(I) complexes Xa–c, Xe–g, Xi was
investigated by hydrogenation of 1-hexene (Table 5,
Fig. 3). All applied complexes exhibit rather good
turnover frequencies in toluene and therefore the metal
centers are readily accessible for hydrogen and 1-hex-
ene. However, due to solvation and swelling effects of
the anchored rhodium(I) centers and of the matrix,
respectively, it seems that there is no predictable corre-
lation between the catalytic activity, selectivity and type
of co-condensing moiety.
To investigate the stability of the catalysts, the xe-
rogels Xc and Xg were separated from the reaction
mixture and then used again in a catalytic reaction with
new solvent and 1-hexene (Table 7). In each case an
enhancement of the conversion from run 1 to run 3 was
observed, which indicates that the formation of the
An enhancement of the conversion rate was achieved
with an increase of the solvent polarity (Table 6). In the
case of Xc, only a slight increase of the activity is
observable if dioxane or methanol is used as solvent.
The selectivities are comparable in each solvent. In the
case of Xg the conversion rates and selectivities are
comparable in toluene and dioxane, respectively. How-
ever, if methanol is used then the selectivity is barely
influenced by the solvent but the conversion rate is
twice as high compared with the activities in toluene or
dioxane. This is another example that the metal cata-
lyzed hydrogenation is favored in polar solvents, espe-
cially in alcohols; although these applied inorganic-
Fig. 3. Catalytic activity of the applied anchored rhodium complexes.

E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
63
Table 7
Conversions and selectivities of consecutive runs with a recycling of the catalyst a
Catalyst
Run no.
Conversion (%)
Hexane (%)
cis, trans 2-Hexene (%)
TON b
TOFc
Xc
1
2
3
14.2
18.5
31.1
77.5
65.4
83.0
22.5
34.6
17.0
664
730
1557
332
365
778
Xg
1
2
3
10.5
27.4
34.8
67.6
85.4
67.8
32.4
14.6
32.2
427
1407
1419
213
703
709
a Reaction conditions: toluene as solvent; H2 pressure 10 bar; Rh:1-hexene=1:6000; temperature 313 K; reaction time 2 h.
b Turnover number (molsub molc−at1).
c Turnover frequency (molsub molc−at1 h−1).
catalytic active species takes place after a relative long
possible with chemical elemental analysis. In Table 8
the quantification results of the EDX measurements are
summarized and compared with the reference data
from NMR measurements.
period. In the case of Xc the selectivity decreases from
run 1 to run 2 to increase again in run 3, whereas in the
case of Xg the highest selectivity was achieved in run 2.
Leaching investigations on the catalysis solution
point out, that no rhodium was found to be detached
from the polysiloxane support.
4. Conclusion
3.7. Surface area measurements, SEM images and
EDX measurements
Different xerogels Xa–i were synthesized, in which
the T-functionalized (COD)(dppp)rhodium(I) complex
[2(T0)] was incorporated into a polysiloxane matrix.
Owing to the high organic moieties in these station-
ary phases, the surface area measurements according to
the BET method reveal very low values (B3 m2 g−1),
which is in agreement with former results [19,20,52,53].
SEM images of Xb and Xf are depicted in Fig. 4.
Unlike Xf the xerogel Xb consists of large particles with
sharp fracture edges, whereas Xf looks cauliflower-like
with mainly small-sized particles.
Because of difficulties in classical elemental analysis
of Si-containing materials [37] EDX-measurements
were carried out. Fig. 5 displays typical EDX spectra of
Xa and Xi including peak assignment. Qualitative
analysis confirms the presence of carbon, oxygen, sili-
con, fluorine, phosphorus, antimony, and rhodium. For
quantification purposes, however, the measured compo-
sition has to be compared to theoretical values calcu-
lated excluding hydrogen, since hydrogen is a single
electron atom and thus does not emit characteristic
X-rays.
Owing to uncertainties in fundamental parameters at
low X-ray energies, spectrometer calibration and the
limited detector efficiency in the low energy region,
correct quantification of light elements is a principal
problem in EDX. Partial line overlap between the P K-
and the Au M-line introduces further uncertainties to
quantification. Errors are significant for the quantifica-
tion of minor constituents as the bremsstrahlung back-
ground causes signal-to-background ratios below 3 in
some cases. However, all elements except for hydrogen
can be detected and quantified directly, which is not
Fig. 4. SEM images of Xb (top) and Xf (bottom).

64
E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
Table 8
EDX data of Xa–i
Xerogel
Reference data a
Composition (%) c
EDX (ZAF correction) b
Composition (%) c
C
O
Si
3.0
13.2
15.8
19.3
24.5
9.5
P
F
Sb
Rh
C
O
Si
P
F
Sb
Rh
Xa
Xb
Xc
Xd
Xe
Xf
Xg
Xh
Xi
52.1
51.3
52.2
50.8
50.3
56.5
58.4
61.7
64.0
2.5
9.7
9.5
14.0
17.7
6.5
9.6
9.1
11.3
6.6
4.0
3.5
2.5
1.2
4.3
3.1
2.1
0.8
12.1
7.4
6.4
4.5
2.1
7.9
5.8
3.9
1.5
12.8
7.8
6.8
4.8
2.3
8.3
6.1
4.2
1.6
10.9
6.6
5.8
4.1
1.9
7.0
5.2
3.5
1.3
55.0
55.8
51.2
49.0
52.5
55.3
61.5
61.6
62.6
5.7
4.0
6.0
4.7
3.9
2.9
1.8
3.9
3.2
1.7
1.8
4.6
6.4
4.7
1.8
2.3
6.0
4.2
2.9
2.7
13.4
7.8
5.9
4.3
3.0
7.1
4.7
3.5
3.8
11.3
5.6
4.4
1.9
1.6
5.1
3.4
1.5
2.2
11.5
11.1
11.6
14.5
12.6
9.5
8.2
18.8
28.5
24.3
10.0
13.5
17.5
17.5
11.8
15.5
19.5
11.3
9.4
a Derived from NMR data according to Ref. [19].
b Calculated according to Ref. [34].
c Calculated excluding hydrogen.
hydrogenation of 1-hexene. All applied materials
showed rather good turnover frequencies indicating a
good accessibility of the anchored transition metal com-
plexes. Their activity can be increased by using polar
solvents. After three runs no reduction of the activity
and no metal leaching were observed.
Acknowledgements
The support of this research by the Deutsche
Forschungsgemeinschaft (Graduiertenkolleg ‘Chemie in
Interphasen’, Grant No 441/2 Bonn, Bad Godesberg)
and by the Fonds der Chemischen Industrie, Frankfurt/
Main, is gratefully acknowledged. We thank T. Salesch,
Institut fu¨r Anorganische Chemie II, University of
Tu¨bingen, for BET measurements. We thank Dr. W.
Wischert, Institut fu¨r Anorganische Chemie II, Univer-
sity of Tu¨bingen, for the AAS measurements. Finally,
we thank U. Kehrer, Institut fu¨r Anorganische Chemie
II, University of Tu¨bingen, for measurement of the
103Rh NMR spectra. We are grateful to Degussa AG,
Germany, for a generous gift of RhCl3.
Fig. 5. Typical EDX spectra of Xa (bottom) and Xi (top). The
spectrum of Xi is omitted for clarity. Assignment of the characteristic
X-ray peaks is indicated. Due to the necessity of a conductive
coating, the Au M line is also observed. Partial line overlap of the P
K line with the Au M line introduces uncertainties to quantification.
The backbone of the matrices was modified by employ-
ment of the different co-condensing agents MeSi-
(OMe)2(CH2)6(OMe)2SiMe (D0ꢀC6ꢀD0) and MeSi-
(OMe)2(CH2)3(C6H4)(CH2)3(OMe)2SiMe
[Ph(1,4-C3-
D0)2] in various ratios. Multinuclear CP/MAS NMR
spectroscopy and EXAFS studies proved the integrity
of the complexes. According to further solid-state
NMR experiments, the realistic compositions do not
differ significantly from the applied T/D ratios. Studies
on the mobility of the matrix and the reactive centers
revealed, that an enhancement of the mobility is
achieved with a larger amount of the co-condensing
components and that both kinds of materials show a
similar mobility. The examination of the accessibility of
the incorporated rhodium(I) centers within these differ-
ent types of polysiloxane matrices was carried out by
References
[1] E. Lindner, F. Hoehn, T. Salesch, S. Singh, K. Mu¨ller, H.A.
Mayer, submitted for publication.
[2] P. Panster, S. Wieland, in: B. Cornils, W.A. Herrmann (Eds.),
Applied Homogeneous Catalysis with Organometallic Com-
pounds, VCH, Weinheim, 1996.
[3] F.R. Hartley, P.N. Vezey, Adv. Organomet. Chem. 15 (1977)
189.
[4] J. Blum, D. Avnir, H. Schumann, Chemtech 29 (1999) 32.
[5] M. Ann, D.N. Perera, R.J. Angelici, J. Mol. Catal. A 149 (1999)
99.
[6] U. Schubert, N. J. Chem. 18 (1994) 1049.

E. Lindner et al. / Inorganica Chimica Acta 327 (2002) 54–65
65
[7] F.R. Hartley, Supported Metal Complexes, Reidel, Dordrecht,
1985.
[31] T.S. Ertel, H. Bertagnolli, S. Hu¨ckmann, U. Kolb, D. Peter,
Appl. Spectrosc. 46 (1992) 690.
[8] D.W. Sindorf, G.E. Maciel, J. Am. Chem. Soc. 105 (1983) 3767.
[9] J. Blu¨mel, Inorg. Chem. 33 (1994) 5050.
[32] M. Newville, P. Livins, Y. Yakobi, J.J. Rehr, E.A. Stern, Phys.
Rev. B 47 (1993) 14126.
[10] I.S. Khatib, R.V. Parish, J. Organomet. Chem. 369 (1989) 9.
[11] E. Lindner, F. Auer, T. Schneller, H.A. Mayer, Angew. Chem.
Int. Ed. Engl. 38 (1999) 2154.
[12] E. Lindner, A. Ja¨ger, T. Schneller, H.A. Mayer, Chem. Mater. 9
(1997) 81.
[13] C.J. Brinker, G.W. Scherer, Sol Gel Science, Academic Press,
London, 1990.
[14] B. Breitscheidel, J. Zieder, U. Schubert, Chem. Mater. 3 (1991)
559.
[15] R.J.P. Corriu, C. Hoarau, A. Mehdi, C. Reye, Chem. Commun.
(2000) 71.
[16] E. Lindner, M. Kemmler, H.A. Mayer, P. Wegner, J. Am.
Chem. Soc. 116 (1994) 348.
[17] J. Bu¨chele, H.A. Mayer, Chem. Commun. (1999) 2165.
[18] E. Lindner, R. Schreiber, M. Kemmler, T. Schneller, H.A.
Mayer, Chem. Mater. 7 (1995) 951.
[19] E. Lindner, T. Schneller, H.A. Mayer, H. Bertagnolli, T.S. Ertel,
W. Ho¨rner, Chem. Mater. 9 (1997) 1524.
[20] E. Lindner, T. Salesch, F. Hoehn, H.A. Mayer, Z. Anorg. Allg.
Chem. 625 (1999) 2133.
[21] E. Lindner, A. Baumann, P. Wegner, H.A. Mayer, U. Reino¨hl,
A. Weber, T.S. Ertel, H. Bertagnolli, J. Mater. Chem. 10 (2000)
1655.
[22] E. Lindner, T. Schneller, F. Auer, P. Wegner, H.A. Mayer,
Chem. Eur. J. 3 (1997) 1833.
[23] E. Lindner, S. Brugger, S. Steinbrecher, E. Plies, H.A. Mayer, Z.
Anorg. Allg. Chem. 1627 (2001) 1731.
[24] E. Lindner, T. Salesch, S. Brugger, F. Hoehn, P. Wegner, H.A.
Mayer, J. Organomet. Chem., in press.
[25] C.A. Fyfe, Solid State NMR for Chemists, CRC Press, Gulph,
ON, 1984.
[26] G. Engelhardt, D. Michel, High-Resolution NMR of Silicates
and Zeolithes, Wiley, New York, 1987.
[33] S.J. Gurman, N. Binsted, I. Ross, J. Phys. C 19 (1986) 1845.
[34] J.I. Goldstein, D.E. Newbury, P. Echlin, D.C. Joy, A.D. Romig,
C.E. Lyman, C. Fiori, E. Lifshin, Scanning Electron Microscopy
and X-Ray Microanalysis, 2nd ed., Plenum Press, New York,
1992.
[35] G. Giordano, R.H. Crabtree, Inorg. Synth. 28 (1990) 88.
[36] E. Lindner, A. Enderle, A. Baumann, J. Organomet. Chem. 558
(1998) 235.
[37] R. Kalfat, F. Babonneau, N. Gharbi, H. Zarrouk, J. Mater.
Chem. 6 (1996) 1673.
[38] K. Schmidt-Rohr, H.W. Spiess, Multidimensional Solid State
NMR and Polymers, Academic Press, London, 1994.
[39] D.W. Sindorf, G.E: Maciel, J. Am. Chem. Soc. 105 (1983) 3767.
[40] E. Bayer, K. Albert, J. Reiners, M. Nieder, D. Mu¨ller, J.
Chromatogr. 33 (1983) 197.
[41] R.K. Harris, Analyst 110 (1985) 649.
[42] E. Lindner, W. Wielandt, A. Baumann, H.A. Mayer, U.
Reino¨hl, A. Weber, T.S. Ertel, H. Bertagnolli, Chem. Mater. 11
(1999) 1833.
[43] E.A. Stern, Phys. Rev. B 10 (1974) 3027.
[44] F.W. Lytle, D.E. Sayers, E.A. Stern, Phys. Rev. B 11 (1975)
4825.
[45] J. Karas, G. Huttner, K. Heinze, P. Rutsch, L. Zsolnai, Eur. J.
Inorg. Chem. 3 (1999) 405.
[46] A. Jacobi, G. Huttner, U. Winterhalter, J. Organomet. Chem.
571 (1998) 231.
[47] Y.-Y. Yan, T.V. RajanBabu, Org. Lett. 26 (2000) 4137.
[48] P.A. MacNeil, N.K. Roberts, B. Bosnich, J. Am. Chem. Soc. 103
(1981) 2273.
[49] G. Descotes, D. Lafont, D. Sinou, J.M. Brown, P.A. Chaloner,
D. Parker, N. J. Chem. 3 (1981) 167.
[50] U. Nagel, A. Bublewitz, Chem. Ber. 125 (1992) 1061.
[51] O. Pamies, G. Net, A. Ruiz, C. Claver, Eur. J. Inorg. Chem. 9
(2000) 2011.
[52] E. Lindner, S. Brugger, S. Steinbrecher, E. Plies, H.A. Mayer, J.
Mater. Chem. 11 (2001) 1393.
[27] H. Eckert, Prog. Nucl. Magn. Reson. Spectrosc. 24 (1992) 159.
[28] R. Benn, C. Brevard, J. Am. Chem. Soc. 108 (1986) 5622.
[29] B.E. Mann, in: P.S. Pregosin (Ed.), Transition Metal Nuclear
Magnetic Resonance, Elsevier, Amsterdam, 1991.
[30] R.S. Aujla, R.K. Harris, K.J. Packer, M. Parameswaran, B.J.
Say, A. Bunn, M.E.A. Cudby, Polym. Bull. 8 (1982) 253.
[53] E. Lindner, A. Ja¨ger, F. Auer, W. Wielandt, P. Wegner, J. Mol.
Catal. A 129 (1998) 91.