Synthesis and catalytic properties of neutral and cationic rhodium- And iridium-containing carbosilane dendrimers

Article abstract of DOI:10.1039/b211960a

The reaction of phosphanyl-terminated carbosilane dendrimers containing only one phosphorus ligand per arm with [MCl(cod)]₂ (M = Rh, Ir; cod = cycloocta-1,5-diene) resulted in the grafting of MCl(cod) moieties on the surface of the dendrimer. However, dendrimers displaying two phosphorus ligands per arm gave very unstable species which were not isolated. On the other hand, the last group of dendrimers reacted with [MCl(cod)]₂, in the presence of silver salts, to afford cationic dendrimers in which a metal fragment is attached simultaneously to both phosphorus atoms of the same arm. The substitution of cod with 1,1′-bis(diphenylphosphino)ferrocene (dppf) in one example of the latter group yielded a bimetallic Rh/Fe layer-containing dendrimer. The new metallodendrimers were tested as catalysts in the hydrogenation of 1-hexene. The Royal Society of Chemistry 2003.

Full text of DOI:10.1039/b211960a

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and catalytic properties of neutral and cationic
rhodium- and iridium-containing carbosilane dendrimers
Inma Angurell, Guillermo Muller, Mercè Rocamora, Oriol Rossell* and Miquel Seco
Departament de Química Inorgànica, Universitat de Barcelona, Martí i Franquès, 1-11,
E-08028 Barcelona, Spain
Received 5th December 2002, Accepted 29th January 2003
First published as an Advance Article on the web 11th February 2003
The reaction of phosphanyl-terminated carbosilane dendrimers containing only one phosphorus ligand per arm with
[MCl(cod)]2 (M = Rh, Ir; cod = cycloocta-1,5-diene) resulted in the grafting of MCl(cod) moieties on the surface of
the dendrimer. However, dendrimers displaying two phosphorus ligands per arm gave very unstable species which
were not isolated. On the other hand, the last group of dendrimers reacted with [MCl(cod)]2, in the presence of silver
salts, to afford cationic dendrimers in which a metal fragment is attached simultaneously to both phosphorus atoms
of the same arm. The substitution of cod with 1,1Ј-bis(diphenylphosphino)ferrocene (dppf ) in one example of the
latter group yielded a bimetallic Rh/Fe layer-containing dendrimer. The new metallodendrimers were tested as
catalysts in the hydrogenation of 1-hexene.
Introduction
Neutral metal complexes dendrimers
Metallodendrimers constitute a very interesting class of macro-
molecules because of their useful properties in areas such as
catalysis, electrochemistry and photophysics, among others.1
Although metals have served as branching centers and building
block connectors, a number of them have also been incorpor-
ated on the surface after dendritic construction. This is the case,
for example, of dendrimers containing Au, Pd, Pt, Co, Rh, Fe,
W, Ru, Ti, Zr or Nb.2 In the vast majority of the cases, the
species were obtained by direct complexation of alkyne-,3
nitrogen-,4 or phosphine-terminated dendrimers.5 Very recently,
we described the synthesis of the very stable (both kinetically
and thermodynamically) diphenylphosphino-terminated carb-
osilane dendrimers which, after reaction with metal com-
plexes, such as AuCl(tht) (tht = tetrahydrothiophene) or
[MCl2(cod)] (M = Pd, Pt), gave dendrimers decorated on the
surface by transition metals. Interestingly, gold-complexed
dendrimers were found to be suitable species for the synthesis
of the first transition metal clusters grafted on the periphery
of the dendrimers.6 In addition, we found that palladium
dendrimers retain the catalytic properties of the mononuclear
species in the hydrovinylation of styrene.7 Here, we report an
extension of our studies concerning the synthesis of neutral
and cationic rhodium- and iridium-containing dendrimers and
their catalytic behavior in the hydrogenation of 1-hexene.
The complexation ability of the diphenylphosphino groups of
dendrimers shown in Chart 1 was verified by allowing them to
react with [MCl(cod)]2 (M = Rh, Ir) in thf. Majoral and co-
workers described the grafting of rhodium (but not iridium)
units to some phosphino-terminated dendrimers by an identical
method.5f The complete disappearance of the signal at about
Ϫ(22–25) ppm in the 31P NMR spectra due to the diphenyl-
phosphane dendrimers indicated that the reaction occurred
quantitatively in all cases. However, according to the nature of
the dendrimer employed, two kinds of metallodendrimers are
formed: those showing one only MCl(cod) unit grafted per
branch (1a,b and 3a,b) and those displaying two metal frag-
ments bonded per arm (2a,b and 4a (Chart 2)). Dendrimers of
the first type were easily isolated and spectroscopically charac-
terized. The reaction was monitored by 31P NMR spectroscopy,
which showed that complexation induces the expected deshield-
ing effect with the occurrence of a doublet at 25 ppm (J(PRh) =
148 Hz) and a singlet at about 15 ppm for the rhodium
and iridium complexes, respectively. In contrast, the metallo-
dendrimers 2a,b and 4a could not be isolated because they
rapidly decomposed in solution even at low temperatures. Thus,
31P NMR spectroscopy showed, after mixing the reagents, the
emergence of a doublet at 39 ppm (J(PRh) = 196 Hz) along
with another minor doublet (J(PRh) = 150 Hz) at about
17 ppm. Interestingly, the first (probably 2a or 4a) disappeared
rapidly to give broad signals at about 28–30 ppm, which is typi-
cal for oxidized phosphorus species, while the second remained
unaltered. According to the 31P NMR parameters we assigned
the last doublet to cationic species containing only one metal
center per branch (2 : 1 P : Rh molar ratio) (see below). Most
probably, the high steric congestion on the periphery of these
species is responsible for their instability. The orange com-
pounds 1a,b and 3a,b are pure within the limits of the NMR
spectroscopic detection. They are soluble in most common
organic solvents and were fully characterized by elemental
analyses, 1H, 13C, 31P and 29Si NMR and FAB or ES mass
spectrometry. For example, only one signal in the 31P NMR
spectrum as well as the two signals for 1a,b and the four for 3a,b
in the 29Si NMR spectra, confirmed the structures proposed.
Moreover, the presence of the cyclooctadiene ligand and the
methyl, phenyl, methylene and ethylene protons in the expected
Results and discussion
The carbosilane dendrimers used for this study are shown in
Chart 1. The synthesis and characterization of 1, 2 and 3 have
been reported elsewhere.8 Dendrimer 4 was obtained similarly
by the repetitive stepwise synthetic route of hydrosilylation of
the vinylic end-groups of the dendrimer with chlorosilanes, fol-
lowed by alkenylation with vinylmagnesium chloride. Unlike 3,
the third hydrosilylation step involved the use of HSiMeCl2.
Finally, the reaction with the tetramethylenediamine com-
plex of (diphenylphosphino)methyllithium in thf permitted the
incorporation of terminal CH2PPh2 groups. Dendrimer 4 was
obtained as a colorless oil in moderate yields, accompanied
with minor quantities of PPh2CH3 which were removed by
column chromatography on silica gel. The 31P{1H} NMR
spectrum showed one signal at Ϫ21.4 ppm in agreement with
the values for 1–3. The 29Si{1H} NMR spectrum revealed four
different types of silicon atoms, as expected. The signal corre-
sponding to the external silicon atoms appeared as a triplet
1
integrated ratio is corroborated from the H and 13C NMR
spectra. Although molecular parent ion peaks were not
shown in the FAB or ES mass spectrum, several fragments
were detected. For example, the FAB mass spectrum of 1a con-
tained two ion peaks at m/z 2120.8 (calc. 2120.5) and 1042.3
1
(J(PSi) = 14.0 Hz). Both the H and 13C NMR spectra were
consistent with the structure proposed.
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
1194

View Article Online
on the surface of the dendrimers given that each branch
should support two metal units MCl(cod). In order to minim-
ize this effect we envisaged the synthesis of the correspond-
ing metallodendrimers containing only one metal per branch;
for this, it was necessary to abstract the chlorine ligands
of the dinuclear [MCl(cod)]2 by using silver salts. Thus, we
reacted 2 or 4 with [MCl(cod)]2 in the presence of AgCF3SO3 or
AgBF4 in thf at room temperature. The process was monitored
by 31P NMR spectroscopy that showed that the reaction pro-
ceeded readily to completion. All the NMR data were in
agreement with the formation of the cationic dendrimers
[2a]4ϩ(CF3SO3)4, [2b]4ϩ(CF3SO3)4, [4a]8ϩ(CF3SO3)8 and [4b]8ϩ
-
(BF4)8 (Chart 3). 31P NMR spectra exhibited doublets (J(PRh)
= 145.3 and 145.4 Hz) at 18.9 and 19.3 ppm for [2a]4ϩ and
[4a]8ϩ, respectively, and singlets at 18.9 and 19.3 ppm for [2b]4ϩ
and [4b]8ϩ, respectively. It is worth noting that the M(cod)
units can be grafted on the surface of the dendrimers through
two P atoms belonging to the same or different branches. On
the other hand, the possible interaction through two differ-
ent molecules of dendrimer was ruled out because of the
very low solubility expected for the resulting polymers. Since in
all cases the 31P NMR spectra showed only one signal, it
can be concluded that only a single stereoisomer is, in fact,
formed. By comparison with our recent studies on platinum
and palladium-containing dendrimers, we tentatively assigned
compounds [2a,b]4ϩ and [4a,b]8ϩ to the stereoisomer that dis-
plays the metal centers bonded to the phosphine ligands of the
same branch.
The 29Si{1H} NMR spectrum showed the two or the four
different silicon environments expected for [2a,b]4ϩ or [4a,b]8ϩ
,
1
respectively, and their H NMR spectra were consistent with
the structures proposed, showing characteristic data of com-
plexed cyclooctadiene. However, the low solubility of [4a]8ϩ and
[4b]8ϩ precluded obtaining their 13C NMR spectra. In order to
overcome this, we planned to substitute the cyclooctadiene
ligand for carbonyl groups, which were expected to increase the
solubility of the final species. This process is known to occur for
monometallic rhodium or iridium complexes.10 Thus, CO was
bubbled through a thf solution of [4a]8ϩ for 2 h and 31P NMR
and IR spectroscopies showed that the substitution process
occurred easily to yield [4aЈ]8ϩ . The ν(CO) IR pattern in the
1966–2064 cmϪ1 region confirmed the nature of the resulting
compound. Unfortunately, as found for [4a]8ϩ, the poor solubil-
ity in the common organic solvents precluded registering its 13
NMR spectrum.
C
Reaction with 1,1Ј-bis(diphenylphosphino)ferrocene (dppf)
One of the most interesting goals in metallodendrimer chem-
istry is the formation of molecular multiple layer structures,
which can be useful for the construction of nanoscale molecu-
lar architectures.11 Since the dendrimers reported here display
metal centers containing cyclooctadiene ligands, we reacted
them with dppf, in an attempt to displace the molecule of cod
and to form a double metal layer dendritic system. The reaction
did not proceed with the cationic dendrimers, apparently
because of the positive charge on metal atoms. However, the
neutral dendrimer 1a gave, after reaction with dppf in thf, the
new double metallic layer dendritic system 1a-dppf (Chart 4)
which is a yellow air-sensitive solid. It was characterized by
NMR spectroscopy. The 31P NMR spectrum confirmed its
formation, showing two double doublet signals at 24.3 ppm
(J(PRh) = 141 Hz; J(PP) = 36 Hz), and at 22.5 ppm (J(PRh) =
151 Hz; J(PP) = 39 Hz) for the dppf ligand and a signal for the
phosphorus nucleus of the dendrimer at 44.4 ppm (J(PRh) =
192 Hz; J(PP) = 36 Hz; J(PP) = 39 Hz). The synthesis of 1a-dppf
is not clean because another unidentified rhodium-containing
species, displaying one signal at 47.0 ppm (J(PRh) = 200 Hz),
was always present. However, this minor product could be
easily separated by washing the solid 1a-dppf with methanol.
Chart 1
(calc. 1042.5) corresponding to [MϪCl]ϩ and [MϪ2Cl]2ϩ
,
respectively, in accordance with the proposed formula. Other
fragments for 1a and for 1b and 3a derivatives are listed in the
Experimental section. Unfortunately, neither FAB and ES mass
spectrometry nor MALDI-TOF techniques gave fragments for
3b, which could not be identified. This fact has precedents in the
literature.9
Cationic metallodendrimers
The instability of the rhodium or iridium complexed den-
drimers 2a,b and 4a,b can be attributed to the high congestion
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
1195

View Article Online
Chart 2
Eventually, traces of dppf were also eliminated by washing the
solid with diethyl ether.
hydrogenation processes. The results showed no significant
decrease of the enantioselectivity compared with mononuclear
analogues.14–16 Very recently, asymmetric hydrogenation has
been tested using a group of cationic rhododendrimers con-
taining up to 32 metal centres, the largest chiral phosphine
dendrimer reported up to now.17 In this case, a decrease in both
activity and selectivity of the dendrimer catalysts was observed
when going to the higher generations. The length of the spacer
between the phosphorus atoms and the carbon–silicon linkage
could be an important factor in the control of the selectivity of
the reaction as reported in the catalysed hydroformylation of
alkenes.18
Catalytic studies
In recent years, the attachment of homogeneous catalysts to
soluble dendrimer supports has received considerable attention
owing to its potential advantages related to the fixation and
recovery or recycling of the catalyst.12 The dendrimers have a
well-defined structure with the possibility of attaching a known
large number of active sites on the periphery, retaining basically
the properties of the simple molecular counterpart.
Therefore, dendrimers loaded mainly with Ni, Pd, Ru and
Rh complexes have been tested as supports for homogeneous
catalysts in a wide number of reactions. The results have been
generally similar, but with a slightly lower range of selec-
tivities to those obtained with the homogeneous mononuclear
species, which indicates that the catalytic centres act as
independent sites. Several reports concerning the catalytic
hydrogenation of olefins in the presence of metallodendrimers
have been published,13 including the use of dendrimers contain-
ing 8 or 16 terminal chiral diphosphine ligands in asymmetric
In this work, the Rh() bound metallodendrimers were
employed as efficient catalysts for olefin hydrogenation under
standard conditions (25 ЊC, 10 bar H2, 1 h, acetone) and using
1-hexene as substrate in a 1 : 500 metal-to-substrate ratio. For
good reproducibility of the catalytic results the catalyst had to
be freshly prepared under an inert atmosphere.
As shown in Table 1, the catalytic activity (TOF) of neutral
Rh() dendrimers 1a and 3a was found to be higher than both
monomeric [RhCl(cod)(PPh3)] and dimeric [RhCl(cod)]2 com-
plexes (entries 6, 7, 10, 11, 1 and 2). There was a slight decrease
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
1196

View Article Online
Chart 4
Fig. 1 Mol% of hydrogenation reaction products of 1-hexene vs. time
using [2a]4ϩ as precursor.
hydrogenation ability of the catalyst towards hexenes: 1-hexene
> cis-2-hexene > trans-2-hexene.
Neutral dentritic species showed higher catalytic activities if
the hydrogenation mixture is exposed to the air for 30 min after
1 h reaction and reintroduced into the autoclave for another
hour in the same conditions. The ionic [2a]4ϩ dendrimer was less
sensitive to the air exposure. This can be related to the observ-
ation reported by Marciniec et al.21 for the catalytic hydrosilyl-
ation reaction of 1-hexene, the presence of oxidants probably
shortens the induction period of the active catalyst.
The iridium dendrimers were also tested as hydrogenation
catalyst for 1-hexene in the same conditions. Compound 1b
showed similar turnover frequency to that observed for Rh()
metallodendrimer 1a but unreproducible results have been
found. This is related to the low solubility observed for the Ir()
dendrimers under the reaction conditions.
Chart 3
in turnover frequency upon growth of the Rh() metal-
lodendrimer (entries 6, 7, 10, 11). Kakkar et al. have reported
similar results.13
The cationic Rh() metallodendrimer [2a]4ϩ is less active than
the corresponding ionic monomeric Rh() species [Rh(cod)-
(dppp)][CF3SO3] (dppp = 1,3-bis(diphenylphosphino)propane),
but shows a higher activity than the neutral dendritic species of
the same generation 1a (entries 6, 7, 8, 9). Fig. 1 shows mol%
values of the hydrogenation reaction products of 1-hexene vs.
time using [2a]4ϩ as precursor. The same behaviour is observed
for the monomeric neutral species [RhCl(cod)(PPh3)] or [RhCl-
(cod)]2, in comparison with the ionic [Rh(cod)(dppp)][CF3SO3],
in agreement with the results already reported by Schrock and
Osborn19 and Crabtree et al.20
Experimental
When hydrogenation reactions were run in the same condi-
tions but for 2 h the turnover frequencies are similar except for
compound 3a which shows a decrease presumably due to a
partial decomposition of the third-generation Rh() den-
drimer (entries 10, 11). Moreover, a decrease of the cis- and
trans-2-hexene products was observed in agreement with a slow
hydrogenation of the secondary olefin formed in the isomeriz-
ation process. The results also reflect the expected relative
All reactions were carried out under an atmosphere of dry
nitrogen using standard Schlenk techniques. Solvents were dis-
tilled from sodium/benzophenone ketyl (thf and Et2O), CaCl2
and stored over molecular sieves (acetone) or dried with CaCl2
and distilled from CaH2 (CH2Cl2) under N2 prior to use.
Elemental analysis (C, H) were performed at the Servicio de
Microanálisis del Centro de Investigación y Desarrollo del
Consejo Superior de Investigaciones Científicas (CSIC). 1H
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
1197

View Article Online
Table 1 Hydrogenation reaction of 1-hexene
Run
Precursor [Rh]
t/h
Conversion (%)
a (%)
b ϩ c (%)
b/c
TOF
1
2
3
4
5
6
7
8
9
[RhCl(cod)(PPh3)]
[RhCl(cod)]2
1
2
1
2
1
1
2
1
2
1
2
6
17
8
74.5
68
75.5
85
98.5
83
86.5
75
25.5
32
24.5
15
1.5
17
13.5
25
0.5
17.5
16.5
1.3
1.4
0.8
1.05
6.7
0.9
1.05
1.2
0.6
0.9
1.0
30
42
38
70
472
72
28
99
14
29
47
99
12
17
[Rh(cod)(dppf )][CF3SO3]
1a
68
[2a]4ϩ[CF3SO3]4
220
250
57
99.5
82.5
83.5
10
11
3a
45
Reaction performed in acetone solution at 25 ЊC. Hydrogen pressure 10 bar. Ratio Rh/1-hexene 1/500. TOF = mola molRhϪ1 hϪ1
.
(250 MHz), 13C (62.86 MHz), 29Si (49,66 MHz) and 31P (101.26
Mz) NMR spectra were recorded at 25 ЊC on a Bruker 250
spectrometer anf 19F (282.23 MHz) NMR spectra on a Varian
Unity 300 spectrometer. Chemical shifts are reported in ppm
2155.95); δH (CDCl3): 0.35 (m, 40H, CH3, CH2Si), 1.74
(d, 8H, J(HP) = 14.9 Hz, CH2P), 1.8–2.01 (m, 16H, cod
CH ), 2.38 (s br, 16H, cod CH ), 3.01 (s br, 8H, cod CH᎐CH),
᎐
2
2
5.42 (s br, 8H, cod CH᎐CH) and 7.24–7.64 (m, 40H, C H );
᎐
6
5
1
3
relative to external standards (SiMe4 for H, 13C and 29Si, 85%
δC (CDCl3): Ϫ0.6 (d, J(CP) = 2.9 Hz, CH3), 2.9 (s, C1H2),
H3PO4 for 31P, CF3COOH for 19F) and coupling constants are
given in Hz. Infrared spectra were recorded with FT-IR 520
Nicolet or Impact 400 Nicolet spectrometers in the 4000–
400 cmϪ1 range as KBr pellets. MS (FAB and ES) spectra were
recorded on a Fisons VGQuatro spectrometer using as a matrix
NBA (3-nitrobenzylalcohol) for FAB spectra. MALDI-TOF
spectra were recorded on a Voyager DE-RP (Perspective Bio-
systems) time-of-flight (TOF) spectrometer using as a matrix
DBH (2,5-dihydrobenzoic acid).
10.0 (d, J(CP) = 3.8 Hz, C2H2), 12.8 (d, J(CP) = 13.1 Hz,
3
CH2P), 28.7 (s, cod CH2), 32.9 (d, J(CRh) = 2.4 Hz, cod
CH ), 69.8 (d, J(CRh) = 14.0 Hz, cod CH᎐CH), 103.7
᎐
2
2
(dd, J(CRh) = 12.5 Hz, J(CP) = 7.0 Hz, cod CH᎐CH), 128.0
᎐
(d, 3J(CP) = 9.5 Hz, m-C6H5), 129.7 (s, p-C6H5), 133.9 (d,
2J(CP) = 10.9 Hz, o-C6H5) and 135.2 (d, J(CP) = 39.7 Hz,
1
ipso-C6H5); δSi (CDCl3): 9.41 (Si0), 3.30 (Si1); δP (CDCl3):
25.2 (d, J(PRh) = 147.8 Hz, PPh2). MS (ESϩ): m/z = 2120.8
[MϪCl]ϩ, 1874.4 [MϪClϪRhCl(cod)]ϩ, 1042.3 [MϪ2Cl]2ϩ
,
The starting materials [RhCl(cod)]2,22 [IrCl(cod)]2,23 and
phosphine-terminated dendrimers 1, 2 and 3 were prepared fol-
lowing published procedures.6,7 Other reagents were purchased
from commercial suppliers.
919.2 [MϪ2ClϪRhCl(cod)]2ϩ
,
795.8 [MϪ2ClϪ 2RhCl-
(cod)]2ϩ
,
683.2 [MϪ3Cl]3ϩ
.
MS (FABϩ): m/z
=
2121.3
[MϪCl]ϩ, 1800.1 [MϪClϪ3cod]ϩ, 1627.9 [MϪClϪ2RhCl-
(cod)]ϩ.
Reaction of 1a with 1,1Ј-bis(diphenylphosphino)ferrocene
(1a-dppf). A solution of dppf (0.09 g, 0.16 mmol) in 5 mL of thf
was added slowly to a solution of 1a (0.09 g, 0.04 mmol) in
10 mL of thf at Ϫ50 ЊC. The reaction was monitored by 31P
NMR. The solvent was evaporated to dryness and the solid 1a-
dppf was washed with methanol and diethyl ether twice and
dried under vacuum (105 mg, 66%); δP (thf ): 44.4 (ddd, J(CRh)
= 192 Hz; J(PP) = 36 Hz; J(PP) = 39 Hz, PPh2), 24.3 (dd,
J(PRh) = 141 Hz, 2J(PP) = 36 Hz, dppf) and 22.5 (dd, 1J(PRh) =
152 Hz, 2J(PP) = 39 Hz, dppf ); δSi (thf ): Ϫ0.16 (s, Si1);
δH (CDCl3): Ϫ0.6–0.3 (m, CH2, CH3), 1.8 (m br, CH2P), 3.4, 4.0,
4.3, 4.7 (s, Cp) and 7.04–8.1 (m, C6H5).
Syntheses
Si0(C1H2C2H2Si1(C1H3)2C3H2C4H2Si2(C2H3)(C5H2C6H2Si3-
(C3H3)(CH2PPh2)2)2)4 (4). A solution of Si(CH2CH2SiMe2CH2-
CH2SiMe(CH2CH2SiMeCl2)2)4 (2.29 g, 1.28 mmol) in 10 mL of
thf was added dropwise to a solution of LiCH2PPh2ؒTMEDA
(7.2 g, 0.022 mol) in 20 mL of thf at Ϫ10 ЊC. After stirring for
4 h part of the solvent was removed under vacuum and hexane
was slowly added (40 mL). The suspension was filtered through
Celite and the solvent was removed. The crude product was
purified by chromatography on a silica gel column eluting with
3 : 10 thf–hexane, under N2, to give 4 as a colourless oil (50%
overall yield); δH (CDCl3): Ϫ0.27 (s, C2H3), Ϫ0.23 (s, C1H3),
Ϫ0.03 (s, C3H3) (60H), 0.18–0.4 (m, 64H, CH2Si), 1.64 (s br,
32H, CH2P) and 7.26–7.38 (m, 160H, C6H5); δC (CDCl3): Ϫ6.5
(s, C2H3), Ϫ4.1 (s, C1H3), Ϫ2.8 (s, C3H3), 1.2, 3.0, 4.5 (s br), 7.1
(s br), 8.2 (CH2Si), 12.6 (d, J(PC) = 28 Hz, CH2P), 128.4 (s br,
C6H5), 131.0 (d, J(PC) = 42.8 Hz, ipso-C6H5), 132.6 (d, J(PC) =
18.6 Hz, o-C6H5) and 140.9 (s br, C6H5); δSi (CDCl3): 9.4 (Si0),
8.3 (Si2), 5.9 (Si1) and 4.4 (t, 2J(PSi) = 14.0 Hz, Si3); δP (CDCl3):
Ϫ21.4 (s, PPh2).
2
2
Si0(C1H2C2H2Si1(CH3)2CH2PPh2IrCl(cod))4 (1b). Experi-
mental conditions and workup were identical to those for the
preparation of 1a. 1b (501 mg, 92%) (Found: C, 46.90; H, 5.52.
C100H136Cl4Ir4P4Si5 requires C, 47.78; H, 5.45%; M 2513.19);
δH (CDCl3): 0.2–0.3 (m, 40H, CH3, CH2Si), 1.68 (s br, 8H, cod
CH2), 1.84 (dϩm, 16H, J(HP) = 13.8 Hz, CH2P, cod CH2),
2.14–2.16 (m, 16H, cod CH ), 2.68 (s br, 8H, cod CH᎐CH),
᎐
2
5.00 (s br, 8H, cod CH᎐CH) and 7.2–7.8 (m, 40H, C H );
᎐
6
5
δC (CDCl3): Ϫ0.8 (s, CH3), 2.8 (s, C1H2), 9.7 (d, 3J(CP) = 3.5 Hz,
C2H2), 12.1 (d, J(CP) = 19.4 Hz, CH2P), 29.3 (s, cod CH2),
32.3 (s, cod CH ), 52.9 (s, cod CH᎐CH), 91.9 (d, 2J(CP) =
Si0(C1H2C2H2Si1(CH3)2CH2PPh2RhCl(cod))4 (1a). To a solu-
tion of 1 (0.22 g, 0.19 mmol) in CH2Cl2 (10 mL) was added
[RhCl(cod)]2 (0.19 g, 0.38 mmol) and the mixture was stirred
at room temperature for 1 h. The solvent was evaporated to
dryness and the residue was washed with diethyl ether. The
product was then dried under vacuum. Complex 1a was
obtained as a yellow solid (360 mg, 88%) (Found: C, 54.64;
H, 5.72. C100H136Cl4P4Rh4Si5 requires C, 55.71; H, 6.36% M
᎐
2
3
13.5 Hz, cod CH᎐CH), 127.9 (d, J(CP) = 9.7 Hz, m-C H ),
᎐
6
5
129.8 (s, p-C6H5), 133.3 (d, 2J(CP) = 10.6 Hz, o-C6H5) and
134.4 (d, J(CP) = 47.8 Hz, ipso-C6H5); δSi (CDCl3): 9.5 (Si0) and
3.2 (Si1); δP (CDCl3): δ 15.6 (s, PPh2). MS (ESϩ) (CH2Cl2):
m/z = 2477.0 [MϪCl]ϩ, 2143.5 [MϪClϪIrCl(cod)]ϩ, 1053.4
[MϪ2ClϪIrCl(cod)]2ϩ. MS (FABϩ) (CH2Cl2): m/z = 2477.0
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
1198

View Article Online
[MϪCl]ϩ, 2143.3 [MϪClϪIrCl(cod)]ϩ, 1807.6 [MϪClϪ
washed twice with Et2O and dried (443 mg, 80%) (Found: C,
2IrCl(cod)]ϩ.
54.72; H, 5.36. C152H172F12O12P8Rh4S4Si5 requires C, 54.54; H
5.18%; M 3347.10); δH (CDCl3): Ϫ0.71, Ϫ0.41 (m, 28H, CH3,
CH2Si), 1.57 (s br, 16H, CH2P), 2.25 (m, 32H, cod CH2), 4.36 (s,
Si(CH2CH2Si(CH3)(CH2PPh2RhCl(cod))2)4 (2a). Experi-
mental conditions and workup were identical to those for the
preparation of 1a. The synthesis was assayed in CH2Cl2 and in
thf but the product could not be isolated pure in the solid state;
δP (thf ): 39.3 (d, J(PRh) = 196.3 Hz, PPh2) and 17.3 (d, J(PRh)
= 151.1 Hz, PPh2) with integrals in the ratio 1 : 0.04.
16H, cod CH᎐CH) and 7.25–7.71 (m, 80H, C H ); δ (CDCl ):
᎐
6
5
C
3
Ϫ2.53 (CH3), 0.87 (C1H2), 7.83 (C2H2), 10.8 (s br, CH2P), 30.4
(cod CH2), 100.0 (d, 2J(CP) = 25 Hz, cod CH᎐CH), 121.0
᎐
(q, J(CF) = 320 Hz, CF3SO3) and 129.4–133.3 (m, C6H5).
δSi (CDCl3): 0.81 (Si1); δP (CDCl3): 18.90 (d, J(PRh) = 145.3
Hz); δF (CDCl3): Ϫ80.2 (s, CF3SO3).
Si(CH2CH2Si(CH3)(CH2PPh2IrCl(cod))2)4 (2b). Experi-
mental conditions and workup were identical to those for the
preparation of 2a; δP (thf ): Ϫ8.35 (s br, PPh2) and 8.2 (s, PPh2)
with integrals in the ratio 1 : 0.06.
[Si0(C1H2C2H2Si1(CH3)(CH2PPh2)2Ir(cod))4][CF3SO3]4
([2b]4؉(CF3SO3)4). Experimental conditions and workup were
identical to those for the preparation of [2a]4ϩ(CF3SO3)4 (174
mg, 88%) (Found: C, 49.01; H, 4.72. C152H172F12Ir4O12P8S4Si5
requires C, 49.29; H, 4.68%; M 3704.34); δH (CDCl3): Ϫ0.71,
Ϫ0.1 (m, 28H, CH3, CH2Si), 1.8–2.16 (m, 48H, CH2P, cod
Si0(C1H2C2H2Si1(C1H3)2C3H2C4H2Si2(C2H3)(C5H2C6H2Si3-
(C3H3)2CH2PPh2RhCl(cod))2)4 (3a). Experimental conditions
and workup were identical to those for the preparation of
1a but using thf as solvent (310 mg, 81%) (Found: C, 56.11;
H, 7.42. C228H340Cl8P8Rh8Si17 requires C, 55.74; H, 6.97%;
M 4913.34); δH (CDCl3): Ϫ0.12 (s, C2H3), Ϫ0.09 (s, C1H3)
(36H), 0.3–0.5 (m, 112H, CH2Si, C3H3), 1.73 (d, 16H, 2J(HP) =
14.7 Hz, CH2P), 1.83 (m, 16H, cod CH2), 1.97 (m, 16H, cod
CH ), 3.95 (s, 16H, cod CH᎐CH) and 7.30–7.64 (m, 80H,
᎐
2
C6H5). δC (CDCl3): Ϫ2.5 (s, CH3), 0.92 (s, C1H2), 7.9 (s, C2H2),
10.7 (s br, CH2P), 31.0 (s, cod CH2), 88.0 (d, 2J(CP) = 30 Hz, cod
CH᎐CH), 120.7 (q, J(CF) = 319 Hz, CF SO ) and 129.1–133.5
᎐
3
3
(m, C6H5). δSi (CDCl3): 0.31 (Si1). δP (CDCl3): 18.90 (d, J(PRh)
= 145.3 Hz, PPh2). δF (CDCl3): Ϫ80.0 (s, CF3SO3). MS (ESϩ):
m/z = 777.8 [M]4ϩ, 2202.3 [MϪ3Ir(cod)]ϩ. MS (ESϪ): m/z =
149.4 [CF3SO3]Ϫ.
CH ), 2.31 (m, 32H, cod CH ), 3.01 (s, 16H, cod CH᎐CH), 5.42
᎐
2
2
(s, 16H, cod CH᎐CH), 7.31 (m, 48H, C H ) and 7.60 (m, 32H,
᎐
6
5
C6H5). δC (CDCl3): Ϫ6.3 (s, C2H3), Ϫ4.19 (s, C1H3), Ϫ0.48
(s, C3H3), 2.8, 3.8, 4.5, 4.8, 6.9 (s, CH2Si), 10.1 (s br, C6H2Si),
13.0 (d, J(CP) = 10.3 Hz, CH2P), 28.9 (s, cod CH2), 33.1 (s,
[Si0(C1H2C2H2Si1(C1H3)2C3H2C4H2Si2(C2H3)(C5H2C6H2Si3-
(C3H3)(CH2PPh2)2Rh(cod))2)4][CF3SO3]8
([4a]8؉(CF3SO3)8).
cod CH ), 70.0 (d, J(CRh) = 14.0 Hz, cod CH᎐CH), 104.0
᎐
2
Experimental conditions and workup were identical to those
for the preparation of [2a]4ϩ(CF3SO3)4 (281 mg, 73%) (Found:
C, 54.33; H, 5.66. C332H412F24O24P16Rh8S8Si17 requires C, 54.66;
H, 5.69%; M 7295.30); δH (CDCl3): Ϫ0.69, Ϫ0.47, Ϫ0.19 (m,
60H, C2H3, C1H3, C3H3), 0.06 (m, 64H, CH2Si), 1.57 (s br, 32H,
2
(dd, J(CRh) = 18.2 Hz, J(CP) = 6.7 Hz, cod CH᎐CH), 128.2
᎐
(d, 3J(CP) = 9.7 Hz, m-C6H5), 129.9 (s, p-C6H5), 133.3 (d,
2J(CP) = 10.9 Hz, o-C6H5) and 135.4 (d, J(CP) = 39.4 Hz,
ipso-C6H5); δSi (CDCl3): 8.7 (Si0), 7.3 (Si2), 4.96 (Si1) and 3.3
(Si3). δP (CDCl3): 24.2 (d, J(PRh) = 149.1 Hz). MS (ESϩ):
m/z = 1531.8 [MϪ3ClϪ2cod]3ϩ. MS (MALDIϩ): m/z = 610.99
CH P), 2.23 (m, 64H, cod CH ), 4.37 (s br, 32H, cod CH᎐CH)
᎐
2
2
and 7.5 (m, 160H, C6H5). δSi (CDCl3): 8.5 (Si2), 5.9 (Si1) and
0.79 (Si3); δP (CDCl3): 19.3 (d, J(PRh) = 145.4 Hz, PPh2);
δF (CDCl3): Ϫ79.9 (s, CF3SO3).
[MϪ8ClϪ7codϪ2Rh]6ϩ
.
Si0(C1H2C2H2Si1(C1H3)2C3H2C4H2Si2(C2H3)(C5H2C6H2Si3-
(C3H3)2CH2PPh2IrCl(cod))2)4 (3b). Experimental conditions
and workup were identical to those for the preparation of 3a
(220 mg, 94%) (Found: C, 49.02; H, 6.39. C228H340Cl8Ir8P8Si17
requires C, 48.66; H, 6.09%; M 5627.78); δH (CDCl3): Ϫ0.16
(s, C2H3), Ϫ0.08 (s, C1H3) (36H), 0.22 (s, C3H3), 0.3–0.45 (m,
112 H, CH2Si), 1.54 (m, 16H, cod CH2), 1.8 (m, 32H, cod CH2,
Reaction of [4a]8؉(CF3SO3)8 with CO. In a suspension of
[4a]8ϩ(CF3SO3)8 in 15 mL of thf is bubbled CO for 2 hours. The
yellow solid [4aЈ]8؉(CF3SO3)8 is filtered off and dried under
vacuum. IR: νmax/cmϪ1(CO) 1966s, 1999m and 2064w (KBr).
[Si0(C1H2C2H2Si1(C1H3)2C3H2C4H2Si2(C2H3)(C5H2C6H2Si3-
(C3H3)(CH2PPh2)2Ir(cod))2)4] [BF4]8 ([4b]8؉(BF4)8). Experi-
mental conditions and workup were identical to those for the
preparation of [2a]4ϩ(CF3SO3)4 (123 mg, 71%) (Found: C,
51.44; H, 5.50%; C324H412B8F32Ir8P16Si17 requires C, 51.80; H,
5.53%; M 7512.04); δH (CDCl3): Ϫ0.65–0.1 (m, 124H, CH2Si,
C2H3, C1H3, C3H3), 1.8–2.1 (m, 96H, CH2P, cod CH2), 3.98
CH P), 2.19 (m, 32H, cod CH ), 2.65 (s, 16H, cod CH᎐CH),
᎐
2
2
5.00 (s, 16H, cod CH᎐CH) and 7.34–7.59 (m, 80H, C H );
᎐
6
5
δC (CDCl3): Ϫ6.3 (s, C2H3), Ϫ4.1 (s, C1H3), Ϫ0.5 (s, C3H3), 1.2,
2.8, 4.8, 6.9 (s, CH2Si), 10.0 (s br, C6H2Si), 12.3 (d, J(CP) = 19
Hz, CH2P), 29.6 (s, cod CH2), 33.6 (s, cod CH2), 53.1 (s, cod
CH᎐CH), 92.2 (d, 2J(CP) = 14.1 Hz, cod CH᎐CH), 128.1
᎐
᎐
(d, 3J(CP) = 9.5 Hz, m-C6H5), 130.1 (s, p-C6H5), 133.5 (d,
2J(CP) = 10.5 Hz, o-C6H5) and 134.5 (d, J(CP) = 47.8 Hz, ipso-
C6H5); δSi (CDCl3): 8.75 (Si0), 7.3 (Si2), 4.9 (Si1) and 2.4 (Si3);
δP (CDCl3): 14.8 (s, PPh2).
(s br, 32H, cod CH᎐CH) and 7.26–7.7 (m, 160H, C H );
᎐
6
5
δP (CDCl3): 7.6 (s, PPh2); δSi (CDCl3): 8.5 (Si2), 5.9 (Si1) and 0.9
(Si3); δF (CDCl3): Ϫ148.4 (s, BF4).
Catalytic reactions
Si0(C1H2C2H2Si1(C1H3)2C3H2C4H2Si2(C2H3)(C5H2C6H2Si3-
(C3H3)(CH2PPh2RhCl(cod))2)4 (4a). Experimental conditions
and workup were identical to those for the preparation of 2a;
δP (thf ): 39.5 (d, J(PRh) = 196.5 Hz, PPh2) and 17.2 (d,
J(PRh) = 151.0 Hz, PPh2) with integrals in the ratio 1 : 0.1.
Hydrogenation reactions were performed in a Berghoff auto-
clave in a 50 mL glass vessel adapted to a Teflon liner. The
internal temperature was monitored by means of a thermo-
couple. Gas chromatography was run on a HP 5890 Series II
chromatograph equipped with an Ultra 2 column 50 m ×
0.2 mm × 0.33 µm, for quantitative determination of hexane,
1-hexene, and cis, trans-2-hexene. A solution of 10 mL of
acetone and 1-hexene (1.39 mol) was introduced in a thermo-
stated autoclave, which had previously been purged with
successive applications of vacuum and argon. After this a
freshly prepared solution of the catalyst is introduced in the
autoclave under N2 atmosphere (in a ratio of Rh/1-hexene =
1/500) and hydrogen was admitted until a pressure of 10 bar
was reached. The catalyst solution is prepared under N2
[Si0(C1H2C2H2Si1(CH3)(CH2PPh2)2Rh(cod))4][CF3SO3]4
([2a]4؉(CF3SO3)4). [RhCl(cod)]2 (0.16 g, 0.33 mmol) was dis-
solved in 10 mL of thf and AgCF3SO3 was added (0.17 g,
0.67 mmol). The mixture was stirred at room temperature for
1 h in the dark and the AgCl formed was removed by filtration
through Celite. The solution was added to a solution of
dendrimer 2 (0.315 g, 0.165 mmol) in 10 mL of thf and stirred
for 1 h. The resulting orange solution was concentrated to small
volume and diethyl ether was added. The solid was filtered off,
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
1199

View Article Online
atmosphere, dissolving 0.0150 g (6.96 × 10Ϫ3 mmol) of 1a,
0.0230 g (6.96 × 10Ϫ3 mmol) of [2a]4ϩ(CF3SO3)4 or 0.0170 g
(3.48 × 10Ϫ3 mmol) of 3a in 5 mL of freshly distilled and
deoxygenated acetone. After the desired time the autoclave was
slowly depressurised and the conversion determined by GC
analysis.
(d ) P. Lange, A. Schier and H. Schmidbaur, Inorg. Chem., 1996, 35,
637; (e) M. Bardaji, M. Kustos, A. M. Caminade, J. P. Majoral
and B. Chaudret, Organometallics, 1997, 16, 403; ( f ) M. Slany,
M. Bardaji, A. M. Caminade, B. Chaudret and J. P. Majoral,
Inorg. Chem., 1997, 36, 19391939; (g) M. B. Peori and A. K. Kakkar,
Organometallics, 2002, 21, 3860.
6 M. Benito, O. Rossell, M. Seco and G. Segalés, Organometallics,
1999, 18, 5191.
7 M. Benito, O. Rossell, M. Seco, G. Muller, J. I. Ordinas,
M. Font-Bardia and X. Solans, Eur. J. Inorg. Chem., 2002, 2477.
8 M. Benito, O. Rossell, M. Seco and G. Segalés, Inorg. Chim. Acta,
1999, 291, 247.
9 C. Blais, C. O. Turrin, A. M. Caminade and J. P. Majoral,
Anal. Chem., 2000, 72, 5097.
Acknowledgements
Financial support for this work was generously given by the
DGICYT (Projects BQU2000-0644 and PB97-0407-C05-04)
and the CIRIT (Project 2001SGR 00054). I. A. is indebted to
the Ministerio de Ciencia y Tecnologia for a scholarship.
10 A. P. Martínez, P. García, F. Lahoz and L. Oro, Inorg. Chem.
Commun., 2002, 5, 245.
11 N. Feeder, J. Geng, P. G. Goh, B. F. G. Johnson, C. M. Martin,
D. S. Shephard and W. Zhou, Angew. Chem., Int. Ed., 2000, 39,
1661.
12 (a) G. E. Oosterom, J. N. H. Reek, P. C. J Kamer and P. W. N. M. van
Leeuwen, Angew. Chem., Int. Ed., 2001, 40, 1829; (b) D. Astruc and
F. Chardac, Chem. Rev., 2001, 101, 2991; (c) L. J. Twyman, A. S. H.
King and I. K. Martin, Chem. Soc. Rev., 2002, 31, 69; (d ) R. van
Heerbeek, P. C. J. Kamer, P. W. N. M. van Leeuwen and J. N. H.
Reek, Chem. Rev., 2002, 102, 3717.
References
1 (a) For reviews, see, for example: D. A. Tomalia and H. D. Durst,
Top. Curr. Chem., 1993, 165, 193; (b) G. R. Newkome, E. He and
C. N. Moorefield, Chem. Rev., 1999, 99, 1689; (c) A. W. Bosman,
H. M. Janssen and E. W. Meijer, Chem. Rev., 1999, 99, 1665;
(d ) J. P. Majoral and A. M. Caminade, Chem. Rev., 1999, 99, 845;
(e) M. Fischer and F. Vögtle, Angew. Chem., Int. Ed., 1999, 38, 884;
( f ) R. van Heerbeek; P. C. J. Kamer, P. W. N. M. van Leeuwen and
J. N. H. Reek, Chem. Rev., 2002, 102, 3717.
13 M. Petrucci-Samija, V. Guillemette, M. Dasgupta and A. Kakkar,
J. Am. Chem. Soc., 1999, 121, 1968.
2 (a) See, for example: F. J. Stoddart and T. Welton, Polyhedron, 1999,
18, 3575; (b) A. M. Caminade, R. Laurent, B. Chaudret and J. P.
Majoral, Coord. Chem. Rev., 1998, 178–180, 793; (c) M. Dasgupta,
M. B. Peori and A. K. Kakkar, Coord. Chem. Rev., 2002, 233–234,
223.
3 (a) G. R. Newkome and C. N. Moorefield, Macromol. Symp., 1994,
77, 63; (b) H. W. Marx, F. Moulines, T. Wagner and D. Astruc,
Angew. Chem., Int. Ed., 1996, 35, 1701.
4 (a) M. F. Ottaviani, S. Bossmann, N. J. Turro and D. A. Tomalia,
J. Am. Chem. Soc., 1994, 116, 661; (b) J. W. J. Knapen, A. W. Van der
Made, J. C. De Wilde, P. W. N. M. Van Leeuven, P. Wijkens,
D. M. Grove and G. Van Koten, Nature, 1994, 372, 659.
14 C. Köllner and A. Togni, Can. J. Chem., 2001, 79, 1762.
15 R. Schneider, C. Köllner, I. Weber and A. Togni, Chem. Commun.,
1999, 2415.
16 C. Köllner, B. Pugin and A. Togni, J. Am. Chem. Soc., 1998, 120,
10274.
17 G. D. Engel and L. H. Gade, Chem. Eur. J., 2002, 8, 4319.
18 L. Ropartz, K. J. Haxton, D. F. Foster, R. E. Morris, A. M. Z.
Slawin and D. J. Cole-Hamilton, J. Chem. Soc., Dalton Trans., 2002,
4323–4334.
19 R. R. Schrock and J. A. Osborn, J. Am. Chem. Soc., 1976, 98, 2134.
20 R. H. Crabtree, A. Gautier, G. Giordano and T. Kahn,
J. Organomet. Chem., 1977, 141, 113.
5 (a) A. M. Herring, B. D. Steffey, A. Miedaner, S. A. Wander and
D. L. Dubois, Inorg. Chem., 1995, 34, 1100; (b) M. Slany,
M. Bardaji, M. J. Casanove, A. M. Caminade, J. P. Majoral and
B. Chaudret, J. Am. Chem. Soc., 1995, 117, 9764; (c) P. Lange,
A. Schier and H. Schmidbaur, Inorg. Chim. Acta, 1995, 235, 263;
21 B. Marciniec, W. Duczmal, W. Urbaniak and E. Sliwinska,
J. Organomet. Chem., 1990, 385, 319.
22 J. L. Herde, J. C. Lambert and C. V. Senoff, Inorg. Synth., 1974, 15,
18.
23 G. Giordano and R. H. Crabtree, Inorg. Synth., 1979, 19, 218.
D a l t o n T r a n s . , 2 0 0 3 , 1 1 9 4 – 1 2 0 0
1200