New Rh derivatives of s-indacene active in dehydrogenative silylation of styrene

Article abstract of DOI:10.1016/j.jorganchem.2006.03.009

The mono- and bimetallic complexes [(2,6-diethyl-4,8-dimethyl-s-indacenide){Rh(COD)}] (1), anti-[(2,6-diethyl-4,8-dimethyl-s-indacenediide){Rh(COD)}₂] (2a), syn-[(2,6-diethyl-4,8-dimethyl-s-indacenediide){Rh(COD)}₂] (2b) were synthes

Full text of DOI:10.1016/j.jorganchem.2006.03.009

Journal of Organometallic Chemistry 691 (2006) 3011–3017
www.elsevier.com/locate/jorganchem
New Rh derivatives of s-indacene active in dehydrogenative
silylation of styrene
Edgardo Esponda , Christopher Adams , Francisco Burgos , Ivonne Chavez a,*,
a
a
a
a
b,
b
b
*
Juan M. Manriquez , Fabien Delpech , Annie Castel , Heinz Gornitzka ,
b
b
Monique Rivi e` re-Baudet , Pierre Rivi e` re
a
Departamento de Qu ı´ mica Inorg a´ nica, Facultad de Qu ı´ mica, Pontificia Universidad Cat o´ lica de Chile, Casilla 306 Correo 22, Santiago, Chile
Laboratoire d’H e´ t e´ rochimie Fondamentale et Appliqu e´ e, UMR 5069, Universit e´ Paul Sabatier, 118 route de Narbonne, 31062 Toulouse, Cedex 9, France
b
Received 23 January 2006; received in revised form 6 March 2006; accepted 6 March 2006
Available online 10 March 2006
Abstract
The mono- and bimetallic complexes [(2,6-diethyl-4,8-dimethyl-s-indacenide){Rh(COD)}] (1), anti-[(2,6-diethyl-4,8-dimethyl-s-inda-
cenediide){Rh(COD)} ] (2a), syn-[(2,6-diethyl-4,8-dimethyl-s-indacenediide){Rh(COD)} ] (2b) were synthesized and characterized spec-
2
2
1
3
103
troscopically and in the case of complex 2b, by means of X-ray diffraction. The C and Rh NMR studies suggest that the bonding
3
5
mode of the indacenediide ligand can be described as intermediate between g - and g -coordination. This result was confirmed by the
crystal structure of 2b as evidenced by the slippage of the rhodium atom towards the periphery of the ligand. Cyclic voltammetry studies
revealed a strong intermetallic communication through the fused ring ligand. This property was further illustrated by higher activity and
selectivity of binuclear complexes 2 for the catalytic dehydrogenative silylation of styrene.
ꢀ
2006 Elsevier B.V. All rights reserved.
Keywords: Rhodium complex; s-indacene; Silylation
1
. Introduction
Among the multiple candidates, fulvalene-derived ligands
and fused delocalized polycyclic arenes have been shown
to be notably efficient for allowing electronic interaction
between two metals [4]. For instance, homo- and heterobi-
metallic complexes incorporating pentalene and s-indacene
as bridging ligand, possess the highest electronic interaction
between two metals, as in systems such as [Mn(CO) ] (l-
Electronic communication between metals in polynu-
clear complexes has focused considerable attention [1]. This
interest has been stimulated by the dramatic modifications
of the physical and chemical properties of complexes com-
pared to those of mononuclear species [2]. This process,
which is termed ‘‘cooperative effect’’, has found significant
implications for the design of materials exhibiting nonlinear
optical activity or unusual electrochemical properties [3]. In
this context, it is evident, that these properties are pro-
foundly affected by the link which mediates electron interac-
tion between metals. Depending on the flexibility and on the
delocalization along this ligand, the degree of metal-
to-metal communication may vary to a large extend.
3
2
5
3
*
g ,g -C H ) (C H = pentalenediide) or [Cp M-Spacer-
8
6
8
6
0
0
*
M Cp ] (Spacer = pentalenediide, s-indacenediide; M,
M = Fe, Ru, Ni, Co) [5]. Thus, their potential of applica-
tion for electronic devices is wide and attracting. However,
the relevance of cooperative effect goes well beyond the field
of material chemistry since electron transfer is a fundamen-
tal process which is involved in the chemical reactivity of
molecules [6]. The presence of two metals in close proximity
potentially may lead to cooperative interactions which can
facilitate transformations that cannot occur for the mono-
metallic species. Additionally, an expanded range of oxida-
tion states is accessible to dinuclear complexes as a result of
*
Corresponding authors. Tel.: +33 561558412; fax: +33 561558204.
E-mail addresses: ichavez@puc.cl (I. Chavez), delpech@chimie.ups-
tlse.fr (F. Delpech).
0
022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.03.009

3012
E. Esponda et al. / Journal of Organometallic Chemistry 691 (2006) 3011–3017
stabilizing metal–metal interactions. These characteristics
offer significant opportunities in the field of metal-mediated
reactions for the design of new and efficient catalytic
systems.
[10]. Cooperative effect was proposed to be involved, how-
ever, dihydropentalenes are known to oligomerize readily
and are reacted in situ at low temperature to form pentale-
nyl complexes [11]. Thus, the employment of alternative
ligand is highly desirable.
Following our interest in the synthesis and the study of
the physical and chemical properties of binuclear complexes
bridged by fused ring systems, we have recently developed a
new effective and selective route to polyalkylated s-indac-
enes. These compounds are obtained in high yields and
are easy to handle and to convert to the corresponding
mono- or dianions [12]. In this context, s-indacene presents
an attractive alternative as a bridging ligand. This work
describes the synthesis and characterization of the mono-
metallic [(2,6-diethyl-4,8-dimethyl-s-indacenide)Rh(COD)],
and of the bimetallic anti-[(2,6-diethyl-4,8-dimethyl-s-inda-
Extensive works have been focused on complexes incor-
porating Cp or Cp-derived ligands. Indeed, cyclopentadie-
nyl ligand is ubiquitous in organometallic chemistry since it
can bind to a large number of metals with different oxida-
tion states. However, in contrast to the well-documented
chemistry of fulvalene and fulvalene-like bridged binuclear
complexes [4,7], chemical reactivity of dimetallic fused
rings systems has been much less developed. These rare
investigations were focused on ligand mobility and substi-
tutional lability, steps which are fundamental for the devel-
opment of potential catalyst. Indeed, comparison of the
reactivity of the ancillary NBD or COD ligands (NBD =
norbornadiene; COD = 1,5-cyclooctadiene) with carbon
monoxide has been used as a criterion for evaluating influ-
ence of ‘‘indenyl effect’’ in several indenyl and indacene
anions derived mono- and bimetallic complexes.
cenediide){Rh(COD)} ] and syn-[(2,6-diethyl-4,8-dimethyl-
2
s-indacenediide){Rh(COD)} ] complexes. As part of our
2
effort to add to our understanding of the influence of
metal–metal interaction, the spectroscopic, electrochemical
and catalytic properties of mono- and bimetallic have been
investigated in details. The particular physical and chemical
characteristics of the binuclear compounds are strongly
influenced by the presence of the second metal centers and
will be rationalized in terms of cooperative effect.
However, to the best of our knowledge, extension of the
scope of this stoichiometric reaction to a catalytic process
has been successfully realized only in two cases.
6
3
[
(CO) Cr(l-g ,g -Ind)Rh(COD)] (Ind = heptamethylinde-
3
nyl) catalyzes the trimerization of various alkynes, the pres-
ence of (CO) Cr being responsible of an increase of
3
3
efficiency (factor of 10 ). This behaviour was mainly attrib-
2. Results and discussion
uted to an enhanced coordinative insaturation at rhodium,
an effect dubbed ‘‘extra indenyl effect’’ [8]. In the second
example, we successfully employed the pentalene complex
2.1. Synthesis
5
3
*
[
Cp Ru(l-g ,g -C H )Rh(COD)] (C H = pentalenediide)
The Rh(I) homobimetallic complexes of the substituted
s-indacenediide were prepared by two paths as shown in
Scheme 1. In the two steps process (path I), the addition
of one equivalent of n-BuLi to a THF solution of the
2,6-diethyl-4,8-dimethyl-s-indacene ligand at ꢀ80 ꢁC
8
6
8
6
for the highly effective and selective dehydrogenative silyla-
tion of styrene with triethylsilane [9]. Its selectivity and
activity were found to be similar to those of the most effi-
cient rhodium catalysts reported [Rh(COD) ]BF /PPh
2
4
3
CH3
7
8
1
3
CH3
7
a
8a
CH3
CH3
6
2
path Ia
CH3
CH3
CH3
Rh
5
4a
3a
4
CH3
path II
path Ib
1
CH3
CH3
Rh
CH3
CH3
+
CH3
CH3
Rh
Rh
CH3
Rh
CH3
2a
2b
Scheme 1.

E. Esponda et al. / Journal of Organometallic Chemistry 691 (2006) 3011–3017
3013
affords the transient monolithiated compound which in
are listed in Table 1. The homobinuclear complex syn-
[(2,6-diethyl-4,8-dimethyl-s-indacenediide){Rh(COD)}2],
presented an apparent symmetrical structure, slightly
curved on the tips, with a molecular angle distortion of
6.89ꢁ. Consequently, the s-indacenediide loses its aromatic-
ity as evidenced by the noticeable deviation from planarity
of the five membered rings.
4
turn reacts with [{Rh(l-Cl)(g -COD)} ]. The mono-rho-
2
dium compound 1 was isolated as a yellow solid in an
almost quantitative yield. No trace of homobimetallic com-
plex was observed. Subsequent metallation of 1 by n-BuLi
4
and reaction with one equivalent of [{Rh(l-Cl)(g -
COD)} ] gives the homobimetallic complexes 2 as orange
2
solids in good yield (65%). Compounds 2 can be also pre-
pared in a one-step bis-metallation reaction (path II) by
addition of two equivalents of n-BuLi to the starting s-
indacene followed by reaction with rhodium chloride dimer
in a comparable yield. In both cases, a mixture of syn and
anti isomers was obtained in the same ratio (syn/anti, 2:1)
The puckering of the five membered ring of s-indacene
has already been noted for homobinuclear s-indacene
*
* n+
[Cp M(s-indacenediide)MCp ]
(M = Fe and n = 0, 1;
M = Co and n = 0, 2; M = Ni and n = 0, 1) [5a]. As a mat-
ter of fact, the complex 2b displays Rh–C (of the ring junc-
˚
tion) (Rh–C: 2.407(3)–2.479(3) A) bonds significantly
1
determined by H NMR. A lot of hypotheses have been
longer than those involving the periphery carbon atoms
˚
already proposed to explain the different syn/anti ratio in
related homo- and heterobinuclear complexes. Among
them, we can cite: the structures and the relative stabilities
of both the lithiated intermediates and the metal dimers,
the steric hindrance, the solvent effect (THF), the influence
of the ancillary ligand of rhodium (COD vs. CO). Ceccon
et al. [13] who studied in more details this aspect, con-
cluded that the role of ligand seems preponderant: a large
preference for the syn isomer was observed with ancillary
ligands as COD or NBD and for the anti form with CO.
In our case where two different pathways were used, imply-
ing various reactional intermediates but the same deproto-
nation/metallation procedures, it seems reasonable to
envisage that the same hypothesis (dominant influence of
the COD ligand) could explain the identical ratio obtained
for the two paths. Meanwhile, this chelating olefin affords a
smaller excess of the syn isomer than that obtained for s-
indacenediide spacer (90%) [13]. This presumably results
from the slightly higher hindrance of the alkylated s-inda-
cene compared to that of the unsubstituted analogue. Frac-
tional precipitation using benzene as solvent allows us to
separate these isomers, the syn form 2b precipitates
whereas the anti form 2a remains in the benzenic solution.
(Rh–C: 2.169(3)–2.226(3) A). The extend of the differences
between these distances is larger than those found for ana-
logues in the literature. Indeed, this effect is emphasized
when comparing D(M–C) [14–16] values of 2b (D(Rh1–
˚
˚
C) = 0.25 A and D(Rh2–C) = 0.24 A) with those for
[
[
˚
(hydro-s-indacenide)Rh(COD)] (D(Rh–C) = 0.15 A) [17],
(5-hydro-2,6-dimethyl-s-indacenide)Rh(COD)] (D(Rh–
˚
C) = 0.16 A) [18], [(6-hydro-2,7-dimethyl-as-indacenide)
˚
Rh(COD)], (D(Rh–C) = 0.14 A) [18], or syn-[(2,7-dimethyl-
as-indacenediide){Rh(COD)} ] (D(Rh1–C) = 0.07 A and
D(Rh2–C) = 0.19 A) [13], anti-[(2,7-dimethyl-as-indacen-
˚
2
˚
˚
ediide){Rh(COD)} ] (D(Rh1–C) = 0.00 A and D(Rh2–
2
˚
C) = 0.01 A) [13].
The COD ligands assume the usual conformation: the
two olefin double bonds of each COD lie in a plane almost
parallel to the indacenediide ligand and are oriented in a
direction almost perpendicular to the ring junction bonds
(C4a–C7a and C3a–C8a). This orientation allows the
absence of short intramolecular distances between COD
˚
shortest non-bonded C. . .C: 3.94 A). In contrast, in the
(
related syn-[(2,7-dimethyl-as-indacenediide){Rh(COD)}2],
˚
the proximity of the olefinic ligand (C. . .C: 3.47 A) results
in the change of conformation of one COD and in the
emergence of intramolecular p-hydrogen bonds between
indacenediide and COD ligand [13]. In 2b, the COD
2
.2. X-ray and NMR structural analysis
X-ray quality single crystals of 2b were grown from benze-
Table 1
nic solution: the solid state structure of 2b is depicted in Fig. 1
and confirms the synfacial coordination of the Rh(COD)
moieties on the polycyclic ligand. Relevant bond distances
˚
Selected bond distances (A) for complex 2b
Rh(1)–C(8a)
Rh(1)–C(3)
Rh(1)–C(1)
Rh(2)–C(7a)
Rh(2)–C(6)
C(3a)–C(8a)
C(1)–C(8a)
C(1)–C(2)
2.419(3)
2.226(3)
2.169(3)
2.441(3)
2.223(3)
1.448(3)
1.463(4)
1.426(4)
1.468(4)
1.423(4)
Rh(1)–C(3a)
Rh(1)–C(2)
Rh(2)–C(4a)
Rh(2)–C(7)
Rh(2)–C(5)
C(4a)–C(7a)
C(3)–C(3a)
C(2)–C(3)
2.479(3)
2.218(3)
2.407(3)
2.202(3)
2.178(3)
1.448(3)
1.460(4)
1.420(4)
1.427(4)
1.448(4)
C(4a)–C(5)
C(6)–C(7)
C(5)–C(6)
C(7)–C(7a)
Rh(1)–C(15)
Rh(1)–C(19)
Rh(2)–C(23)
Rh(2)–C(27)
C(15)–C(16)
C(23)–C(24)
2.123(3)
2.122(3)
2.110(3)
2.127(3)
1.407(4)
1.393(4)
Rh(1)–C(16)
Rh(1)–C(20)
Rh(2)–C(24)
Rh(2)–C(28)
C(19)–C(20)
C(27)–C(28)
2.114(3)
2.150(3)
2.133(3)
2.130(3)
1.404(4)
1.388(5)
Fig. 1. Crystal structure of 2b at the 50% probability level for the thermal
ellipsoids.

3014
E. Esponda et al. / Journal of Organometallic Chemistry 691 (2006) 3011–3017
ligands are far enough to allow typical COD conformation.
Thus, steric repulsions do not seem to be the sole factor
responsible for the larger shift of Rh(COD) fragments
toward the periphery of the indacenediide ligand.
Rh NMR has also been shown to be an effective tool
for the determination of cyclopentadienyl analogues hap-
ticity in organometallic complexes [16]. Chemical shifts of
2.3. Cyclic voltammetry studies
The resulting cyclic voltammograms of all products
present cathodic and anodic processes that are summarized
in Table 3. This tool has been previously used as an indica-
tor for the interaction between two metallic centers bonded
by a bridging ligand, even if quantification remains difficult
[21]. Complex 1 can be used as a reference compound, in
order to evidence the effect of the second metal listed as
DEp.
1
03
1
–2 have been determined by using HMBC between the
1
03
olefinic protons of the COD ligand and the
Rh atom
1
03
(
JH–Rh = 2 Hz) and are reported in Table 2. The
Rh
NMR spectra of complexes 1–2 display signal with chemi-
cal shifts similar to those found for the related indenyl- and
alkylated (indenyl)Rh(COD) complexes (ꢀ278 to
All complexes, presented a one electron oxidation of
the Rh center. Complex 1 owned a very positive poten-
tial, indicative of the low electronic density the mononu-
clear complex may have had in order to stabilize higher
oxidation states. Comparing complex 1 with its binuclear
analogues these presented over 950 mV shift into less
positive potentials. This large shift was due to electronic
factors given by a ligand which was then more negatively
charged. The values of I /I listed in Table 3 are higher
than unity for complexes 1 and 2b, suggesting the pres-
ence of a coupled irreversible chemical reaction while
2a has quasi-reversible electrochemistry. This result could
be indicative of a cooperative effect between both metal-
lic centers, absent in isomer 2b, possibly related to a spa-
tial interaction between both Rh atoms, yielding a
decomposition of the oxidized specie, consistent with its
electrochemical irreversibility.
ꢀ
519 ppm) [19]. The bonding mode has been described as
3
5
intermediate between g - and g -coordination (i.e.,
2
3
g + g ). The situation in s-indacene complexes is compa-
rable since chemical shifts of 1–2 are in-between that of
3
3
true g -bonding mode in [(g -cyclooctenyl)Rh(COD)]
5
(
ꢀ9 ppm) and that of [(g -Cp)Rh(COD)] which exhibits a
a
c
5
quasi-perfect g -hapticity (ꢀ777 ppm).
As previously shown in related as-indacenediide com-
1
03
plexes,
Rh NMR chemical shift can be used as a good
indicator to evaluate the relative Rh centers slippage.
1
03
d( Rh) of syn complex 2b is shifted about 70 ppm down-
field compared to that of anti analogue 2a. This difference
3
may reflect a slightly more pronounced g -character in the
syn complex, presumably to possible higher steric repulsion
of Rh(COD) fragments in 2a which, then, slips away.
1
3
C NMR spectra are also of diagnostic importance in
2.4. Catalysis
the determination of the hapticity of Cp-derived ligand.
Indeed, in the case of indenyl, it can be deduced by com-
paring the NMR chemical shifts of the five-membered
ring carbon atoms in the complex with those of indenyl
Spectroscopic and CV studies have underlined that the
presence of a second metal center dramatically affects the
physical and chemical properties of Rh(COD) entities.
Similar phenomena associated to the electronic communi-
cation between two metal centers have been shown to have
implications in stoichiometric reaction such substitution
reactions in the related bimetallic complexes syn- and
1
3
3
sodium (Dd C). For g -indenyl ligand, periphery carbon
atoms are shielded while ring-junction ones are not. The
5
closer to a g -bonding mode, the more shielded the ring-
1
3
junction carbons signal (Dd C = ꢀ20 to ꢀ40 ppm in
5
true g -coordination) [20]. Same considerations can also
anti-[(as-indacenediide){Rh(COD)} ]: the substitution by
2
be done in s-indacene complexes. Significant shielding
carbonyl ligands of the first COD is much faster than that
of the second one [13]. However, definitive proofs for the
involvement of cooperative effect in catalytic processes
remain very rare [4].
(
ꢀ20 to ꢀ35 ppm in 1 and 2) clearly shows the presence
of interaction between ring-junction carbon atoms and
2
3
the rhodium center and confirms g + g coordination
mode for the indacene ligand. However, no detailed
study correlating chemical shift differences and hapticity
are currently available, and thus this prevents further
analysis of these values.
Table 3
Cyclic voltammetry results
a
b
a
c
d
Complex
E
DE
I /I
a c
DE
2
1
1.42
0.46
0.37
0.14
0.17
0.09
1.6
1.0
1.8
0.00
0.96
1.05
2
2
a
b
a
Table 2
103
Experimental chemical shifts, d( Rh) of the studied mono- and bimetallic
rhodium complexes 1, 2a and 2b
Conditions: solvent CH
2
Cl
2
4 4
, 0.1 M [NBu ][BF ] as supporting elec-
trolyte, using Pt as working electrode, a Pt wire as an auxiliary electrode,
and SCE as the reference electrode. All complexes had a working con-
Complex
d
ꢀ4
centration of 5 · 10 M. All potential values listed in volts.
(
ppm)
b
Potential value for the respective anodic peak.
Difference between respective anodic and cathodic peak.
Difference between anodic peak of reference complex 1 and that of
[
(2,6-Diethyl-4,8-dimethyl-s-indacenide){Rh(COD)}] (1)
anti-[(2,6-Diethyl-4,8-dimethyl-s-indacenediide){Rh(COD)}
syn-[(2,6-Diethyl-4,8-dimethyl-s-indacenediide){Rh(COD)}
ꢀ486
c
2
] (2a) ꢀ334
d
2
] (2b) ꢀ261
listed compound.

E. Esponda et al. / Journal of Organometallic Chemistry 691 (2006) 3011–3017
3015
We have recently reported that the related heterobime-
leads to similar product distribution, i.e., 85–88% of 3
and 12–15% of hydrosilylation products 4 and 5. In con-
trast, the selectivity of the reaction in the presence of
mononuclear complex 1a drops to 75%. However, these
differences remain weak and hazardous to rationalize.
Thus, the difference of activity provides direct evidence
for the involvement of cooperative effect between both
metal centers during the catalysis. Dehydrogenative silyla-
tion and hydrosilylation products result from two compet-
itive catalytic cycles and electronic communication may
have stabilized intermediaries, which may have not taken
place with mononuclear complexes. Detailed mechanism
remain to be clarified and in particular, the precise role
of the electronic communication during the reaction. Fur-
ther exploration in this direction is currently in progress.
*
5
3
tallic pentalenediide complex Cp Ru(l-g ,g -pentalenedi-
ide)Rh(COD) was one of the most active rhodium
catalysts reported to date for dehydrogenative silylation
[
9]. This prompted us to test the activity of 2 for this reac-
tion. In addition, the availability of mono- and bimetallic
complexes 1–2 offers a unique opportunity to study poten-
tial involvement of a cooperative effect. The reaction of tri-
ethylsilane (1.5 mmol) with styrene (4.5 mmol) in the
presence of catalytic amount of complexes 1–2 (0.030 mmol
of Rh) at 80 ꢁC gave a mixture of (E)-1-phenyl-2-(triethyl-
silyl)ethene (3, dehydrogenative silylation product),
-phenyl-2-(triethylsilyl)ethane (4) 1-phenyl-1-(triethylsi-
lyl)-ethane (5) (4 and 5: hydrosilylation products) (Eq.
1)). Along with these products, ethylbenzene was obtained
1
(
in an equal amount to that of 3, and this result clearly
showed that styrene acted also as hydrogen acceptor. The
reaction conditions and the product distribution of the dif-
ferent catalytic tests are summarized in Table 4
3. Conclusions
We have synthesized and fully characterized new bimetal-
lic rhodium complexes incorporating the fused delocalized
ligand (2,6-diethyl-4,8-dimethyl-s-indacenediide). X-ray
diffraction analysis of the syn isomer clearly shows that, in
Ph
Ph
Ph
Ph
Ph
+
HSiEt
+
+
+
3
SiEt
3
SiEt
3
Et Si
3
3
4
5
3
the solid state, the rhodium moieties were closer to a g
ð1Þ
Comparison between catalytic results of runs with
bonding than those previously in related complexes. In addi-
tion, the preparation of the mononuclear compound
allowed a relevant comparison of the influence of electronic
communication on physical and chemical properties. Cyclic
voltammetry studies together with catalytic properties of 2
for the dehydrogenative silylation of styrene insinuate a
cooperative effect between both metallic centers. In order
to verify this, EPR experiments of these complexes and their
oxidized derivatives are being carried out for a future article.
mono- and binuclear complexes show that in the latter
case, the second metallic center created a remarkable effect,
the most pronounced consequence being on the activity of
the catalyst. Indeed, concentration of Rh was kept constant
for all catalytic tests and, this allowed direct assessment of
relative activity. Mononuclear complexes presented a reac-
tion time of 450 min (see Table 4), while complete silane
consumption is observed within 35 and 180 min, for 2a
and 2b, respectively. The strong electronic interaction
between both metals channeled through the bridging ligand
can be invoked to account for the higher activities of dinu-
clear complexes. The differences of efficiency between 2a
and 2b, can be rationalized in terms of steric hindrance
due to the proximity of the two Rh(COD) entities in the
syn isomer. Similarly, different behaviours between syn
and anti isomers of bimetallic compounds have been
already reported and identified by different chemical reac-
tivity or physicochemical properties [4]. The selectivity
was also found to depend on the nature (mono- vs. bime-
tallic) of the catalyst. Indeed, homobinuclear compounds
4. Experimental
4.1. General considerations
All manipulations were carried out under pure dinitro-
gen atmosphere using a vacuum atmosphere drybox
equipped with a Model HE 493 Dri-Train purifier or a vac-
uum line using standard Schlenk-tube techniques. Reagent
grade solvents were distilled under dinitrogen from sodium
benzophenone ketyl (tetrahydrofuran, toluene, petroleum
4
ether). [{Rh(l-Cl)(g -COD)} ] was prepared by the litera-
2
ture procedures [22]. 2,6-Diethyl-4,8-dimethyl-s-indacene
was prepared as previously described [12]. Styrene and tri-
ethylsilane were purchased from Aldrich and were degassed
before use. n-Butyllithium (1.6 M in hexane) was purchased
Table 4
a
Effect of complexes on dehydrogenative silylation of styrene
1
13
1
from Aldrich. H and
C{ H} NMR spectra were
b
c
Entry Complex HSiEt
3
Time (min) Product distribution 3:4:5
103
recorded on Bruker Avance 400 MHz. All
Rh NMR
1
2
3
1
100
100
100
450
35
180
75:14:11
88:12:0
85:11:4
spectra were performed on a Bruker AMX-400 spectrome-
2a
2b
103
ter. The d( Rh) values were calculated by determining the
absolute frequency of the cross-peak (HMBC experiments)
relating it to the reference frequency of 12.64 MHz
a
A mixture of styrene (4.5 mmol), HSiEt
3
(1.5 mmol), 1 (0.030 mmol),
2
a and 2b (0.015 mmol) and toluene (5 mL) was stirred under argon at
(N = 3.16 MHz at 100 MHz). GC and GC–MS spectra
8
0b ꢁC.
(EI, 70 eV) were recorded on HP 5890 series II and HP
Silane consumed %.
c
Determined by GC.
5889 A spectrometers. Elemental analyses (C and H) were

3016
E. Esponda et al. / Journal of Organometallic Chemistry 691 (2006) 3011–3017
made with a Fisons EA 1108 microanalyzer. Cyclic voltam-
metry experiments were performed in an airtight three-elec-
trode cell connected to an argon line. The reference
electrode was an SCE. The counter electrode was a plati-
num wire, and the working electrode was a platine disc
of ca. 3 mm in diameter. The currents and potentials were
recorded on a Pentium II 350 MHz processor, with a BAS
CV-50 Voltammetric Analyzer Potentiometer. Each exper-
iment employed CH Cl refluxed over phosphorous pent-
the insoluble lithium chloride is removed by filtration.
The solution was concentrated giving an orange solid made
up of an anti/syn isomer mixture in a 1:2 ratio, respectively
(Yield: 0.21 g, 65%).
Direct synthesis (Path II): A solution of n-butyllithium
(5.88 mmol) in hexane (1.6 M) was added to a solution of
2,6-diethyl-4,8-dimethyl-s-indacene (0.70 g, 2.94 mmol) in
THF (25 mL) at ꢀ80 ꢁC. The reaction mixture was allowed
to reach room temperature and stirred for 1.5 h, followed
2
2
oxide and [NBu ]BF , was dried prior use.
by stirring at 50 ꢁC for another 2 h. The solution was
4
4
4
cooled to ꢀ80 ꢁC and [{Rh(l-Cl)(g -COD)} ] (1.44 g,
2
4
(
.2. [(2,6-Diethyl-4,8-dimethyl-s-indacenide){Rh(COD)}]
1)
2.94 mmol) in THF (30 mL) was slowly added via syringe.
The mixture was allowed to reach room temperature and
stirred for 2 h. The solvents were evaporated under reduced
pressure. To the remaining solid, 15 mL of toluene is added
and the insoluble lithium chloride is removed by filtration.
This solution was concentrated giving an orange solid con-
sisting of an anti/syn isomer mixture in a 1:2 ratio (Yield:
1.20 g, 62%). Fractional precipitation using benzene as sol-
vent allows us to separate these isomers: the syn form 2b
precipitates (Yield: 0.72 g, 37% – related to s-indacene) giv-
ing suitable crystals for X-ray study whereas the anti form
2a remains in the benzenic solution (Yield: 0.43 g, 22% –
related to s-indacene).
A solution of n-butyllithium (2.52 mmol) in hexane
(
1.6 M) was added to a solution of 2,6-diethyl-4,8-
dimethyl-s-indacene (0.60 g, 2.52 mmol) in THF (25 mL)
at ꢀ80 ꢁC. The solution was allowed to reach room tem-
perature and stirred for 1.5 h. Then, the resulting mixture
4
was cooled to ꢀ80 ꢁC and a solution of [{Rh(l-Cl)(g -
COD)} ] (0.62 g, 1.26 mmol) in THF (20 mL) was added
2
via syringe. The temperature was raised to room tempera-
ture, and the mixture was stirred for 2 h. The solvents were
removed under reduced pressure. To the remaining solid,
1
0 mL of pentane was added and the insoluble lithium
Anti-[(2,6-diethyl-4,8-dimethyl-s-indacenediide){Rh-
1
chloride was removed by filtration. The solution was con-
(COD)} ] (2a): H NMR (400 MHz, C D ): d 1.31 (t,
2
6
6
3
centrated giving a yellow solid of 1 (Yield: 1.07 g, 95%)
JHH = 7.5 Hz, 6H, C2,6–CH –CH ), 1.68 and 1.78 (m,
2 3
1
3
H NMR (400 MHz, C D ): d 1.08 (t, JHH = 7.5 Hz,
16H, COD–CH ), 2.29 (s, 6H, C –CH ), 2.52 (q.d,
2 4,8 3
6
6
3
3
3
3
H, C –CH –CH ), 1.38 (t, J = 7.5 Hz, 3H, C –CH –
JHH = 7.5 Hz, J
= 1.4 Hz, 4H, C2,6–CH –CH ), 4.16
HRh 2 3
6
2
3
HH
2
2
1
3
1
CH ), 1.73 and 1.88 (m, 8H, COD–CH ), 2.21 (s, 3H,
CH –C ), 2.28 (q, JHH = 7.5 Hz, 2H, C –CH –CH ),
2
(br.s, 8H, COD–CH), 5.00 (s, 4H, C1,3,5,7–H). C{ H}
(100 MHz, C D ), 15.21 (C –CH –CH ), 15.59 (C4,8–
3
2
3
3
8
2
2
3
6
6
2,6
2
3
3
.34 (s, 3H, CH –C ), 2.59 (q, J = 7.5 Hz, 2H, C6–
CH ), 23.55 (C –CH –CH ), 32.12 (COD–CH ), 68.56
3 2,6 2 3 2
3
4
HH
1
1
CH –CH ), 2.99 (br s, 2H, C –H ), 4.10 (m, 4H,
(d, JCRh = 13 Hz, CH–COD), 73.75 (d, JCRh = 5 Hz,
2
3
5
2
2
4
1
JHRh = 2 Hz, COD–CH), 4.98 and 5.00 (d, J = 1.6 Hz,
C1,3,5,7), 110.94 (C4,8), 114.28 (d,
C3a,4a,7a,8a), 117.50 (d, JCRh = 6 Hz, C2, 6). Anal.
JCRh = 1.6 Hz,
HH
1
3
1
1
2
H, C1,3–H), 6.67 (m, 1H, C –H). C{ H} (100 MHz,
7
C D ), 13.92 (C –CH –CH ), 15.32 (CH –C ), 15.51
Found: C, 61.95; H, 6.74. C H Rh : Calc.: C, 62.01;
6
6
6
2
3
3
4
34 44
2
(
CH –C ), 16.00 (C –CH –CH ), 23.09 (C –CH –CH ),
H, 6.69. Syn-[(2,6-diethyl-4,8-dimethyl-s-indacenediide)
3
8
2
2
3
2
2
3
1
2
3
7
1
5.00 (C –CH –CH ), 31.12 and 31.67 (COD–CH ),
{Rh(COD)} ], isomer, (2b): H NMR (400 MHz, C D ):
6
2
3
2
2
6
6
1
3
9.61 (C ), 66.38 and 66.63 (d, J
= 2 Hz, CH–COD),
d 1.31 (t, J = 7.5 Hz, 6H, C2,6–CH –CH ), 1.69 and
HH 2 3
5
CRh
1
5.33 and 75.62 (d, JCRh = 4.3 Hz, C1,3), 110.86 and
2.10 (m, 16H, COD–CH ), 2.20 (s, 6H, C –CH ), 2.56
2 4,8 3
1
1
3
3
13.10 (d, J
= 2.5 Hz, C3a,8a), 114.93 (d, J
= 5 Hz,
CRh
(q.d, JHH = 7.5 Hz, JHRh = 1.5 Hz. 4H, C2,6–CH2–
CRh
C ), 115.37 (C ), 119.56 (C ), 124.84 (C ), 136.23 (C ),
CH ), 4.33 (br.s, 8H, COD–CH), 4.93 (br.s, 4H, C1,3,5,7–
H). C{ H} (100 MHz, C D ), 15.22 (C –CH –CH ),
6 6 2,6 2 3
2
4a
7a
7
8
3
1
3
1
1
39.57 (C ), 149.12 (C ). Anal. Found: C, 69.59; H, 7.40.
4
6
C H Rh: Calc.: C, 69.64; H, 7.37.
15.08 (C4,8–CH ), 23.78 (C –CH –CH ), 32.19 (COD–
3 2,6 2 3
2
6
33
1
CH ), 68.33 (d, JCRh = 13 Hz, CH–COD), 73.69 (d,
2
1
1
4
{
.3. Anti-[(2,6-diethyl-4,8-dimethyl-s-indacenediide)-
Rh(COD)} ] (2a) and syn (2b)
JCRh = 4.5 Hz, C1,3,5,7), 112.13 (C ), 115.92 (d, J
1.2 Hz, C3a,4a,7a,8a), 117.18 (d, J
=
CRh
4, 8
1
= 6 Hz, C2,6). Mass
CRh
2
+
spectra: m/z [M, %]: 658 [M , 100], 447 [M ꢀ Rh(COD),
Stepwise synthesis (Path Ib): A solution of n-butylli-
44]. Anal. Found: C, 61.98; H, 6.73. C H Rh : Calc.:
3
4
44
2
thium (0.49 mmol) in hexane (1.6 M) was added to a solu-
tion of 1 (0.22 g, 0.49 mmol) in THF (15 mL) at ꢀ80 ꢁC.
The temperature was raised to room temperature and the
C, 62.01; H, 6.69.
4.4. Crystallography
mixture was stirred for 1.5 h. The solution was cooled to
4
ꢀ
80 ꢁC and a solution of [{Rh(l-Cl)(g -COD)} ] (0.12 g,
Crystal data for (2b): C H Rh , M = 658.51, ortho-
2
34 44
2
˚
˚
0
2
.25 mmol) in THF (10 mL) was added. After stirring for
h, the solvents were removed under reduced pressure.
To the remaining solid, 15 mL of toluene is added and
rhombic,
Pbca,
a = 13.809(5) A,
b = 13.909(5) A,
3
˚
˚
c = 29.277(10) A, V = 5623(3) A , Z = 8, T = 193(2) K.
30,597 reflections (5717 independent, Rint = 0.0365) were

E. Esponda et al. / Journal of Organometallic Chemistry 691 (2006) 3011–3017
3017
ꢀ
3
˚
[2] (a) M.D. Ward, Chem. Soc. Rev. 24 (1995) 121;
b) J.P. Sauvage, J.P. Collin, J.C. Chambron, S. Guillerez, C.
Coudret, V. Balzani, F. Bargelletti, L. De Cola, L. Flamigni, Chem.
Rev. 94 (1994) 993.
[3] (a) S. Barlow, D. O’Hare, Chem. Rev. 97 (1997) 637;
(b) R.G. Denning, J. Mater. Chem. 5 (1995) 365.
collected. Largest electron density residue: 0.514 e A , R1
(
(
for I > 2r(I)) = 0.0259 and wR = 0.0629 (all data)
2
P
P
P
2
o
2
c
2
with R = iF | ꢀ |F i/ |F | and wR
2
¼ ð wðF ꢀ F Þ =
1
2
o
o
c
o
P
2
0:5
wðF Þ Þ . The data for 2b were collected at low temper-
atures using an oil-coated shock-cooled crystal on a Bru-
[
4] A. Ceccon, S. Santi, L. Orian, A. Bisello, Coord. Chem. Rev. 248
2004) 683.
ker-AXS CCD 1000 diffractometer with Mo Ka radiation
k = 0.71073 A). The structure was solved by direct meth-
ods [23] and all non-hydrogen atoms were refined aniso-
tropically using the least-squares method on F [24].
(
˚
(
[
5] (a) J.M. Manriquez, M.D. Ward, W.M. Reiff, J.C. Calabrese, N.L.
Jones, P.J. Carroll, E.E. Bunel, J.S. Miller, J. Am. Chem. Soc. 117
(1995) 6182;
2
(
1
b) S.C. Jones, T. Hascall, S. Barlow, D. O’Hare, J. Am. Chem. Soc.
24 (2002) 11610.
4.5. General procedure for the dehydrogenative silylation of
styrene
[
6] P. Diversi, S. Iacoponi, G. Ingrosso, F. Laschi, A. Lucherini, C.
Pinzino, G. Ucello-Barreta, P. Zanello, Organometallics 14 (1995)
3
275.
A two-necked flask equipped with a magnetic bar was
charged with the complex (0.030 mmol for 1 and
.015 mmol for 2). The reactor was evacuated and filled
with argon. Styrene (0.52 mL, 4.5 mmol) and toluene
5 mL) were added, and the mixture was stirred for
min. Then Et SiH (0.24 mL, 1.5 mmol) was added. The
[7] C.G. de Azevedo, K.P.C. Vollhardt, Synlett (2002) 1019.
[8] A. Ceccon, A. Gambaro, S. Santi, A. Venzo, J. Mol. Catal. 69 (1991)
L1.
0
[9] F. Burgos, I. Chavez, J.M. Manriquez, M. Valderrama, E. Lago, E.
Molins, F. Delpech, A. Castel, P. Rivi e` re, Organometallics 20 (2001)
(
5
1
287.
3
[10] R. Takeuchi, H. Yasue, Organometallics 15 (1996) 2098.
reaction flask was immersed in an 80 ꢁC thermo-stabilized
[11] (a) T.J. Katz, M. Rosenberger, R.K. O’Hara, J. Am. Chem. Soc. 86
(
(
1964) 249;
bath. The progress of the reaction was monitored by GC
b) S.C. Jones, P. Roussel, T. Hascall, D. O’Hare, Organometallics
(
disappearance of Et SiH). The products were identified
3
2
5 (2006) 221.
1
by H NMR, by GC–MS, and by comparison with the lit-
erature data [10].
[
12] M.R. Dahrouch, P. Jara, L. Mendez, Y. Portilla, D. Abril, G.
Alfonso, I. Ch a´ vez, J.M. Manriquez, M. Rivi e` re-Baudet, P. Riviere,
A. Castel, J. Rouzaud, H. Gornitzka, Organometallics 20 (2001)
5
591, and references therein.
Acknowledgements
[
13] A. Ceccon, A. Bisello, L. Crociani, A. Gambarro, P. Ganis, F.
Manoli, S. Santi, A. Venzo, J. Organomet. Chem. 600 (2000)
The authors thank projects FONDECYT Nos. 1020525
and 1020314, and project ECOS-CONICYT C01E06, and
F. Burgos for Support Scholarship in Ph.D. Thesis.
9
4.
[14] The slip distortion was calculated according to: D(M–C) = 1/2[(M–
C3a) + (M–C8a)] ꢀ 1/2[(M–C1) + (M–C3)] (carbon atom labels are
indicated in Fig. 1).
[
15] L. Orian, P. Ganis, S. Santi, A. Ceccon, J. Organomet. Chem. 690
2005) 482.
Appendix A. Supplementary material
(
[16] L. Orian, A. Bisello, S. Santi, A. Ceccon, G. Saielli, Chem. Eur. J. 10
(2004) 4029.
CCDC 295720 (2b) contains the supplementary crystal-
lographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/conts/retriev-
ing.html (or from the CCDC, 12 Union Road, Cambridge
CB2 1EZ, UK; fax: +44 1223 336033; e-mail: depos-
it@ccdc.cam.ac.uk). Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.jorganchem.2006.03.009.
[17] A. Ceccon, P. Ganis, F. Manoli, A. Venzo, J. Organomet. Chem. 601
2000) 267.
(
[18] A. Bisello, A. Ceccon, A. Gambaro, P. Ganis, F. Manoli, S. Santi,
A. Venzo, J. Organomet. Chem. 593–594 (2000) 315.
[19] A. Ceccon, C.J. Elsevier, J.M. Ernsting, A. Gambaro, S. Santi, A.
Venzo, Inorg. Chim. Acta 204 (1993) 15.
[20] R.T. Baker, T.H. Tulip, Organometallics 5 (1986) 839.
[21] (a) S. Santi, A. Ceccon, F. Carli, L. Crociani, A. Bisello, M. Tiso, A.
Venzo, Organometallics 21 (2002) 2679;
(
b) F. Barri e` re, N. Camire, W.E. Geiger, U.T. Mueller-Westerhoff,
References
R. Sanders, J. Am. Chem. Soc. 124 (2002) 7262.
22] J. Chatt, L. Venanzi, J. Chem. Soc. (1957) 4735.
[
[
1] (a) J.A. McCleverty, M.D. Ward, Acc. Chem. Res. 31 (1998) 842;
b) P. Belser, S. Bernhard, C. Blum, A. Beyeler, L. De Cola, V.
Balzani, Coord. Chem. Rev. 190–192 (1999) 155.
[23] G.M. Sheldrick, Acta Crystallogr. A46 (1990) 467.
[24] G.M. Sheldrick, SHELXL-97, Program for Crystal Structure Refine-
ment, University of G o¨ ttingen, G o¨ ttingen, 1997.
(