An alternative mechanistic paradigm for the β-Z hydrosilylation of terminal alkynes: The role of acetone as a silane shuttle

Article abstract of DOI:10.1002/chem.201303063

The β-Z selectivity in the hydrosilylation of terminal alkynes has been hitherto explained by introduction of isomerisation steps in classical mechanisms. DFT calculations and experimental observations on the system [M(I)₂{κ-C,C,O,O-(bis-NHC)}]BF₄ (M=Ir (3 a), Rh (3 b); bis-NHC=methylenebis(N-2-methoxyethyl)imidazole-2-ylidene) support a new mechanism, alternative to classical postulations, based on an outer-sphere model. Heterolytic splitting of the silane molecule by the metal centre and acetone (solvent) affords a metal hydride and the oxocarbenium ion [R ₃Si - O(CH₃)₂]⁺, which reacts with the corresponding alkyne in solution to give the silylation product [R ₃Si - CHi￡C - R]⁺. Thus, acetone acts as a silane shuttle by transferring the silyl moiety from the silane to the alkyne. Finally, nucleophilic attack of the hydrido ligand over [R₃Si - CHi￡C - R]⁺ affords selectively the β-(Z)- vinylsilane. The β-Z selectivity is explained on the grounds of the steric interaction between the silyl moiety and the ligand system resulting from the geometry of the approach that leads to β-(E)-vinylsilanes. Silanes catch the shuttle: An outer-sphere mechanism that explains the β-Z hydrosilylation of terminal alkynes based on the role of acetone as a silane shuttle is disclosed. Heterolytic splitting of the silane molecule by the metal centre and acetone affords a metal hydride and the oxocarbenium ion [R ₃Si - O(CH₃)₂]⁺, which reacts with the alkyne in solution to give the silylation product [R₃Si - CHi￡C - R]⁺ (see figure).

Full text of DOI:10.1002/chem.201303063

FULL PAPER
DOI: 10.1002/chem.201303063
An Alternative Mechanistic Paradigm for the b-Z Hydrosilylation of
Terminal Alkynes: The Role of Acetone as a Silane Shuttle
Manuel Iglesias,*[a] Pablo J. Sanz Miguel,[a] Victor Polo,[b]
Francisco J. Fernꢀndez-Alvarez,*[a] Jesffls J. PØrez-Torrente,[a] and Luis A. Oro*[a]
Abstract: The b-Z selectivity in the hy-
drosilylation of terminal alkynes has
been hitherto explained by introduc-
tion of isomerisation steps in classical
mechanisms. DFT calculations and ex-
perimental observations on the system
[M(I)2{k-C,C,O,O-(bis-NHC)}]BF4
(M=Ir (3a), Rh (3b); bis-NHC=
methylenebis(N-2-methoxyethyl)imida-
zole-2-ylidene) support a new mecha-
nism, alternative to classical postula-
tions, based on an outer-sphere model.
Heterolytic splitting of the silane mole-
cule by the metal centre and acetone
(solvent) affords a metal hydride and
Thus, acetone acts as a silane shuttle
by transferring the silyl moiety from
the silane to the alkyne. Finally, nucle-
À
the oxocarbenium ion [R3Si O-
A
ophilic attack of the hydrido ligand
(CH3)2]+, which reacts with the corre-
over [R3Si CH=C R] affords selec-
tively the b-(Z)-vinylsilane. The b-Z se-
lectivity is explained on the grounds of
the steric interaction between the silyl
moiety and the ligand system resulting
from the geometry of the approach
that leads to b-(E)-vinylsilanes.
+
À
À
sponding alkyne in solution to give the
+
À
À
silylation product [R3Si CH=C R] .
Keywords: alkynes · carbenes · ho-
mogeneous catalysis · hydrosilyla-
tion · silanes
Introduction
The synthesis of vinylsilanes has attracted considerable at-
tention as a result of their increasing demand as key build-
ing blocks in synthetic organic chemistry.[1,2] Vinylsilanes
have been traditionally prepared by stoichiometric methods
from organometallic reagents.[3] However, transition-metal-
catalysed hydrosilylation of terminal alkynes has become
a highly sought-after route to vinylsilanes due to its efficien-
cy and atom economy.[4]
Three possible isomers can be obtained from hydrosilyla-
tion of terminal alkynes (Figure 1). The most thermodynam-
ically stable and usually major reaction product is the b-(E)-
vinylsilane, whereas b-(Z)- and a-vinylsilanes are more elu-
sive targets.[5] Catalyst design is, therefore, deemed crucial
to overcome the thermodynamic stability of the b-(E)-vinyl-
silane and favour the less stable b-(Z)- and a-vinylsilanes.
Figure 1. Possible products obtained from hydrosilylation of terminal al-
kynes.
A good understanding of the mechanisms involved in the
selectivity of the process is essential to help the develop-
ment of more selective catalysts. To the best of our knowl-
edge, two main mechanisms have been proposed for the hy-
drosilylation of terminal alkynes. The first, proposed by
Chalk and Harrod (Chalk–Harrod mechanism),[6] entails the
À
oxidative addition of the Si H bond of the silane, followed
by coordination of the alkyne and migratory insertion into
À
the metal (M) H bond. Subsequently, reductive elimination
affords the b-(E)-vinylsilane. However, this mechanism does
not explain dehydrogenative silylation products or the for-
mation of b-(Z)-vinylsilanes, usually observed when Ru, Rh
and Ir catalysts are employed. To justify the formation of
these products, a new mechanism was proposed (modified
Chalk–Harrod mechanism).[7] This postulates that the migra-
[a] Dr. M. Iglesias, Dr. P. J. Sanz Miguel, Dr. F. J. Fernꢀndez-Alvarez,
Prof. J. J. Pꢁrez-Torrente, Prof. L. A. Oro
Departmento Quꢂmica Inorgꢀnica
Instituto Sꢂntesis Quꢂmica y Catꢀlisis Homogꢁnea
Universidad de Zaragoza–CSIC
C/Pedro Cerbuna 12, 50009 Zaragoza (Spain)
E-mail: miglesia@unizar.es
À
tory insertion of the alkyne occurs into the M Si bond in-
À
stead of the M H bond, thus rendering an M-alkenyl inter-
paco@unizar.es
oro@unizar.es
mediate with the silyl moiety and the metal complex in rela-
tive cis positions (A; Scheme 1). Then, a metal-assisted iso-
merisation from A to D is invoked to explain the formation
of the b-(Z)-vinylsilane—the release of steric hindrance be-
tween the R3Si moiety and the metal complex being the
driving force of the process. The isomerisation mechanisms
proposed by Ojima and Crabtree independently have been
[b] Dr. V. Polo
Departmento Quꢂmica Fꢂsica
Instituto de Biocomputaciꢃn y Fꢂsica de Sistemas Complejos (BIFI)
Universidad de Zaragoza
C/Pedro Cerbuna 12, 50009 Zaragoza (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201303063.
Chem. Eur. J. 2013, 19, 17559 – 17566
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
17559

Scheme 2. Synthetic pathway to complexes 3a and 3b.
Scheme 1. Isomerisation mechanism. [M]=Transition metal fragment.
The protons corresponding to the imidazole rings appear
as two doublets, with a coupling constant of 2 Hz, at d=7.32
and 7.05 ppm. The resonance assigned to the protons of the
methylene bridge is observed as a singlet at d=6.05 ppm.
These NMR patterns are in agreement with the expected
symmetry (C2v point group) in solution. In the 13C{1H} NMR
spectra the most notable signal is that corresponding to the
carbene carbon, which appears as a doublet centred at d=
ubiquitously employed as models to explain the formation
of b-(Z)-vinylsilanes in the hydrosilylation of terminals al-
kynes.[8] The isomerisation process requires the formation of
a metallacyclopropene (B) in the mechanism proposed by
Crabtree et al.,[9] or a zwitterionic carbene (C) in Ojimaꢅs
mechanism,[10] as intermediates for the A–D isomerisation
of the metal-alkenyl intermediate (Scheme 1). Therefore,
isomerisation of A to the less sterically hindered intermedi-
ate D allows the formation of the less thermodynamically
stable b-(Z)-vinylsilane.
In a previous communication we described a hydrosilyla-
tion catalyst able to afford the b-(Z)-vinylsilane in excellent
yields and selectivities.[11] The stereoselectivity of the process
was tentatively explained as a consequence of the steric con-
trol exerted by the ligand system over the M-alkenyl inter-
mediate, which isomerises according to a classical mecha-
nism. However, on trying to extend this reaction to other
solvents we found that acetone is the only solvent in which
the reaction takes place. This fact, together with the increas-
ing number of outer-sphere mechanisms recently published
on iridium-catalysed reactions,[12] prompted us to study the
role played by acetone in a catalytic cycle based on an
outer-sphere mechanism.
154.3 ppm (JRh C =43 Hz). Moreover, the acetate ligand
À
shows a doublet at d=187.7 ppm (JRh C =1.4 Hz) and a sin-
À
glet at d=24.8 ppm. These two resonances correspond, re-
spectively, to the carboxylic and methyl carbon atoms of the
h2-coordinated acetato ligand. Complexes 2a and 2b were
further characterised by X-ray crystallography as CH2Cl2 ad-
ducts. Figure 2 provides a view of 2a and 2b. Geometries of
Herein, we disclose an outer-sphere mechanism for the
hydrosilylation of terminal alkynes based on experimental
data and theoretical calculations at the DFT level. More-
over, we have extended our previous work[11] to an analo-
gous rhodium catalyst. Its preparation and full characterisa-
tion are described together with a comparative catalytic
study of the iridium and rhodium catalysts.
Results and Discussion
Figure 2. View of complexes [M(h2-CH3COO)(I)
(2a), Rh (2b)) with atom-numbering schemes.
2ACHTUNGTREN(NGNU bis-NHC)] (M=Ir
Synthesis and characterisation of complexes: Reaction of
1 with [MCl
sodium iodide and potassium acetate afforded [M(h2-
CH3COO)(I)2A(bis-NHC)] (M=Ir (2a), Rh (2b); bis-NHC=
methylenebis(N-2-methoxyethyl)imidazole-2-ylidene)) as
ACHTUNGTRENNUNG(cod)]2 (M=Rh, Ir; cod=1,5-cyclooctadiene),
N
both IrIII and RhIII metal centres are octahedral, with iodido
ligands at the apical positions, and two equatorial chelating
acetato and bis-NHC-C,C ligands. The bridging methylene
groups (C31) are situated slightly out of the equatorial
plane of the metals (2a, 1.047(6); 2b, 0.893(3) ꢆ), whereas
the h2-acetato ligands lie almost coplanar. The dihedral
angle between the imidazole rings in 2a (33.8(2)8) is more
pronounced than that in 2b (28.4(1)8). In addition, the side
air-stable orange solids (Scheme 2). The reaction for the for-
mation of complexes 2a and 2b entails a change in the oxi-
dation state of the metal centres from MI to MIII and con-
1
comitant loss of cod. The H NMR spectra of complex 2b in
CDCl3 show a resonance corresponding to the methyl group
of the h2-coordinated acetato ligand at d=2.03 ppm.
17560
www.chemeurj.org
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17559 – 17566

Hydrosilylation of Terminal Alkynes
FULL PAPER
arms extend roughly symmetrically out of the equatorial
planes.
responding to the carbene carbon atoms (C12 and C22).
À
The presence of the BF4 anion was corroborated by
Complexes [M(I)2{k-C,C,O,O-(bis-NHC)}]BF4 (M=Ir
(3a), Rh (3b)) were prepared in virtually quantitative yields
by treatment of the corresponding 2a or 2b species with
one equivalent of HBF4·Et2O in dry dichloromethane at
08C (Scheme 2). Successful removal of the h2-acetato ligand
was indicated by loss of the resonances corresponding to the
19F NMR spectroscopy, in which a peak at d=À151.0 ppm
was observed for 3b.
Catalytic study: At the outset of the catalytic study, we ex-
amined the reaction between phenylacetylene and triethylsi-
lane with 3a as catalyst, to optimise the reaction conditions.
Remarkably, it was found that the reaction would only pro-
ceed in acetone. The use of other solvents, such as toluene,
dichloromethane, chloroform, methanol or acetonitrile, af-
forded only traces of the hydrosilylated product. The reac-
tion, initially attempted at room temperature, was found to
be sluggish even in the presence of high catalyst loadings
(ca. 70% conversion after 5 h with 10 mol% 3a), whereas
temperatures above 608C bring about a loss of selectivity as
the Z/E ratio decreases (from 22:1 at 508C to 1.8:1 at 708C)
and more hydrogenated product is observed (from 8% at
508C to 16% at 708C). Complex 3a effectively catalyses the
hydrosilyation of phenylacetylene with triethylsilane at
508C with 5 mol% catalyst loading and 1.1 equivalents of
the corresponding silane (Scheme 3). However, 3b was
1
acetate in H and 13C NMR spectra. Remarkably, in the ab-
sence of the acetato ligand, coordination of the two oxygen
atoms to the metal centre takes place, as confirmed by X-
ray diffraction studies of 3a and 3b (Figure 3). Coordination
Scheme 3. Catalytic hydrosilylation of terminal alkynes.
found to be considerably less active and selective than its iri-
dium analogue (3a) under the reaction conditions previously
mentioned (Tables 1 and 2; R=Ph).
Figure 3. View of cations [M(I)2{k-C,C,O,O-(bis-NHC)}]+ (M=Ir (3a),
Rh (3b)) with atom-numbering schemes. Both enantiomers are present
in the unit cells.
A detailed catalytic study was carried out under the
above-mentioned reaction conditions employing different si-
lanes and a range of terminal alkynes, which were converted
in excellent yields and selectivities to their corresponding b-
(Z)-vinylsilanes by using 3a as catalyst (Table 1). The best
selectivities for 3a were obtained when the most sterically
hindered silane (Ph2MeSiH) was employed; nonetheless,
good results were also achieved with less encumbered si-
lanes (PhMe2SiH and Et3SiH). Altogether, exceptional re-
sults were obtained with 3a for all the tested alkynes, but
somewhat lower selectivities were found for aliphatic al-
kynes than their aromatic counterparts.
In the case of catalyst 3b, aliphatic alkynes were hydrosi-
lylated in excellent yields and selectivities, especially on em-
ploying Ph2MeSiH as silane. Remarkably, the use of
Ph2MeSiH and PhMe2SiH with 3b as catalyst clearly im-
proves the results obtained with 3a for 1-hexyne (n-butyl;
Table 2) and especially 1-nonyne (n-heptyl; Table 2). How-
ever, at variance with 3a, the use of Ph2MeSiH with 3b as
catalyst leads to very low conversions with aromatic alkynes.
In general, catalyst 3b shows an outstanding difference in
activity upon modification of the silane substituents. As
a matter of fact, good conversions and selectivities were ob-
tained with PhMe2SiH, reaching total conversion after 5 h.
of the O-ethers at the sites occupied in 2a by the h2-acetato
ligands modifies the equatorial coordination sphere of com-
plexes 3a and 3b, with equatorial angles being closer to 90
À
À
À
and 1808. M O bond lengths (Ir1 O43=2.216(6) and Ir1
O53=2.208(6) ꢆ) involving the coordinated side arms are
significantly longer for complexes 3a and 3b than those ob-
served for the h2-acetato ligands in 2a and 2b. Tilt angles
between both imidazole rings (24.8(5) and 29.7(4)8) are in
the range shown by 2a and 2b.
1H NMR spectra of 3b show the resonance corresponding
to the methylene bridge (C31 in Figure 3) as a singlet at d=
6.50 ppm, which suggests retention of the two-fold symmetry
of these complexes in solution. The peaks corresponding to
the -CH2O- and CH3O- groups undergo significant down-
1
field shift in the H NMR spectra (in [D6]acetone) from 2b
to 3b, as a consequence of the coordination to the metal
centre of the oxygen atoms in the wingtip groups. The reso-
nance corresponding to the -CH2O- protons shifts d=3.82–
3.74 ppm in 2b to 4.42–4.37 ppm in 3b. The CH3O- reso-
nance follows the same trend, thus shifting from d=
3.33 ppm in 2b to d=4.06 ppm in 3b. 13C{1H} NMR spectra
show a doublet for 2b at d=146.8 ppm (JRh C =44 Hz) cor-
À
Chem. Eur. J. 2013, 19, 17559 – 17566
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17561

M. Iglesias, F. J. Fernꢀndez-Alvarez, L. A. Oro et al.
Table 1. Hydrosilylation of terminal akynes with 3a.[a] Product selectivi-
ties and conversion in [%].
Alkyne (R)
Ph
Silane
Z
E
RCH=CH2
Conversion
Et3SiH
86
86
86
83
94
84
93
84
94
93
86
98
92
91
92
54
30
59
74
60
75
7
9
8
8
2
8
5
7
0
5
10
0
7
5
2
6
0
7
5
6
9
4
8
2
9
6
2
4
2
1
4
6
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
45
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
4-tBu-Ph
4-MeO-Ph
4-CF3-Ph
4-Me-Ph
n-heptyl
n-butyl
Me2PhSiH
MePh2SiH
Et3SiH
40
41
31
13
18
14
Me2PhSiH[b]
MePh2SiH
Et3SiH
69
38
100
100
100
10
13
22
11
Me2PhSiH
MePh2SiH
Figure 4. Kinetic experiments performed with representative alkynes (re-
action conditions: 508C, 5 mol% catalyst).
[a] Reaction conditions: 5 h/508C/5 mol% catalyst (3a) in acetone.
[b] 29% of the a-vinylsilane is observed.
Figure 4. Remarkably, a significant drop in activity, mainly
in the case of catalyst 3b, is observed when the more steri-
cally hindered Ph2MeSiH is employed instead of PhMe2SiH.
This effect is especially notable for R=4-CF3-Ph (Figure 4).
The reaction rates obtained for aromatic alkynes on using
3a as catalyst with Ph2MeSiH and PhMe2SiH are higher
than those for 3b with the same alkynes, except in the case
of the 4-methoxy derivative, for which 3b with PhMe2SiH
Table 2. Hydrosilylation of terminal akynes with 3b.[a] Product selectivi-
ties and conversion in [%].
Alkyne (R)
Ph
Silane
Z
E
RCH=CH2
Conversion
Et3SiH
63
76
74
88
89
85
91
90
95
57
87
97
80
90
100
94
88
94
95
79
92
32
24
26
11
11
15
7
5
0
0
<1
0
0
1
2
0
1
3
3
3
0
0
3
2
2
38
100
39
32
100
39
53
100
79
70
100
33
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
Me2PhSiH
MePh2SiH
Et3SiH
4-tBu-Ph
4-MeO-Ph
4-CF3-Ph
4-Me-Ph
n-heptyl
n-butyl
shows
a greater reaction rate than that of 3a with
Ph2MeSiH. In the case of aliphatic alkynes, namely 1-
nonyne, the reaction with PhMe2SiH is finished within the
first 30 min with 3b as catalyst. Conversely, the same reac-
tion after 5 h with catalyst 3a does not reach total conver-
sion (Table 1). Therefore, it can be concluded that iridium
catalyst 3a is significantly more active than 3b for the hy-
drosilylation of aromatic alkynes, whereas the trend ob-
served for aliphatic alkynes is the opposite; 3b is remarka-
bly more active than its iridium analogue 3a.
8
5
42
10
0
17
10
0
3
10
4
45
100
44
88
100
100
88
100
100
Mechanistic proposal: In the search for a deeper under-
standing of the catalytic cycle that could explain the key
role of acetone, the investigation was extended to theoreti-
cal calculations at the DFT level. Initial calculations were
aimed at achieving computational support for a classic
mechanism, based on a cycle consistent with the modified
Chalk–Harrod mechanism. Therefore, oxidative addition of
the silane followed by coordination of the alkyne and inser-
5
11
6
0
10
2
Me2PhSiH
MePh2SiH
[a] Reaction conditions: 5 h/508C/5 mol% catalyst (3b) in acetone.
Conversely, in the case of Ph2MeSiH and Et3SiH the conver-
sions after 5 h were frequently below 50%. Notably, those
reactions that did not reach total conversion in 5 h were
completed after extended reaction times maintaining the ini-
tial selectivities (usually 36 h suffices). Therefore, total de-
composition or deactivation of the catalyst can be discarded
as the cause of the low yields obtained on using 3b as cata-
lyst with HSiEt3 or HSiMePh2 after 5 h.
À
tion into the M Si bond is required. However, theoretical
calculations show that MV (M=Ir or Rh) species resulting
À
from oxidative addition of R3Si H to 3a or 3b are not
stable due to the high trans effect of the hydrido ligand
trans to one of the NHC units (Scheme 4a). Hence, no sta-
tionary point on the potential energy surface can be found
for such a structure. A second possibility that entails prior
Kinetic experiments performed with representative al-
kynes (R=4-CF3-Ph, 4-MeO-Ph and n-heptyl) are shown in
17562
www.chemeurj.org
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17559 – 17566

Hydrosilylation of Terminal Alkynes
FULL PAPER
Figure 5. DFT calculated relative energy [DE in kcalmolÀ1] profile for
the heterolytic splitting of SiMe3.
Scheme 4. Possible initial steps of the catalytic cycle.
coordination of the alkyne to give a vinylidene (Scheme 4b)
can be excluded as no reaction was observed between al-
kynes and complexes 3a or 3b in the absence of silane
under catalytic conditions—not even at temperatures as
high as 808C. Consequently, once all possibilities of classical
mechanisms have been ruled out, a solvent-assisted outer-
sphere mechanism seems to be the most likely alternative.
What could the connection between acetone and hydrosi-
lylation be? As an answer to this question, we have pro-
posed a new mechanism for hydrosilylation of terminal al-
kynes based on an acetone-assisted heterolytic splitting of
silane (Scheme 4c), in which the reaction solvent (acetone)
transfers the R3Si+ moiety from the silane to the terminal
alkyne by means of an oxocarbenium ion. In this line, recent
reports by Brookhart et al. on hydrosilylation of ketones
show evidence of the formation of an oxocarbenium ion,
generated by reaction of a ketone with R3Si+ (resulting
from the abstraction of a hydride from R3SiH by the metal
centre).[13]
ed metal complexes bear an acetone molecule. This may
force the h1-H(Si) coordination of the silane, thus resulting
in an increase of electrophilicity at the silicon atom. Subse-
quently, the oxygen lone pair of another acetone molecule
can interact with the back side of the silane molecule form-
ing species C-3. The formation of the oxocarbenium ion and
the hydride takes place via the transition structure C-4-TS
(Figure 6), affording an energy barrier of 16.0 kcalmolÀ1
The molecular geometries were optimised at the B3LYP
level using the def2-SVP basis set for all atoms and the asso-
ciated core pseudopotentials for the iridium and iodine. En-
ergies were corrected using the def2-TZVP basis set. Due to
the decisive role of acetone, up to two molecules were con-
sidered in the calculations of the reaction mechanism.
Single-point calculations using a continuum model on the
gas-phase optimised structures were carried out to simulate
solvent effects with a dielectric constant for acetone of
20.493. In this computational study, all reported energies are
relative to the reactants (3a and the corresponding isolated
molecules) and include the solvent correction.
This mechanism starts with the dissociation of the two
ether functions in the wingtip groups, followed by coordina-
tion to the metal centre of an acetone molecule and the
silane in an end-on fashion to yield intermediate C-2, which
is 11.5 kcalmolÀ1 less stable than the reactants (see
Figure 5). It is worth mentioning that the crystal structure of
a complex with a triethylsilane coordinated end-on through
the hydrogen atom to an iridium centre, comparable to in-
termediate C-2, has been recently reported.[14] The authors,
based on DFT studies, suggest a greater electrophilicity at
the Si atom in an h1-H(Si) species than in its analogue h2-
H(Si). Remarkably, our theoretical calculations show that
acetone coordinates to the metal better than the ether at
the wingtip groups in solution; consequently, all the calculat-
Figure 6. DFT optimised structure of C-4-TS, leading to the iridium hy-
drido complex and oxocarbenium ion. Wingtip groups and hydrogen
atoms have been omitted for clarity.
from the reactants. This transition structure can be described
as an SN2 silyl transfer to the oxygen and concomitant for-
À
mation of the Ir H bond. Although the formation of C-5 is
an endothermic process (+5.0 kcalmolÀ1), it can be achieved
at ambient temperature.
À
After formation of the M H bond and the oxocarbenium
ion, three possible reaction pathways have been considered
(Scheme 5). First, acetone hydrosilylation is obtained by nu-
cleophilic attack of the hydride over the oxocarbenium ion.
Figure 7 shows the calculated energy profile, which reveals
a very low activation barrier (2.0 kcalmolÀ1) for the hydride
transfer from the metal to the carbon (I-2-TS). Although
this process is exothermic (À3.5 kcalmolÀ1), the energy bar-
rier for the reverse reaction is only 16.0 kcalmolÀ1. There-
fore, in the absence of alkyne or other reactants, hydrosilyla-
tion of acetone is a kinetically and thermodynamically fav-
oured reaction. As a matter of fact, this process is observed
experimentally in the absence of alkyne. However, as a re-
versible process, the existence of a competitive pathway (re-
action of the oxocarbenium ion with alkynes) may yield
a different product.
Chem. Eur. J. 2013, 19, 17559 – 17566
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17563

M. Iglesias, F. J. Fernꢀndez-Alvarez, L. A. Oro et al.
Figure 8. DFT calculated relative energy [DE in kcalmolÀ1] profile for
the silyl transfer from the oxocarbenium ion to the alkyne.
features the Me3Si moiety pointing outside the metal com-
plex (III-5-TS, Figure 9), which leads to the formation of the
b-(Z)-vinylsilane with an activation barrier of 4.3 kcalmolÀ1.
However, in the second approach, the Me3Si (trans posi-
tioned with respect to the phenyl group) points toward the
metal centre (III-5-TS’, see Figure 9), the activation barrier
being 17.0 kcalmolÀ1 and leading to the b-(E)-vinylsilane.
Both processes are very exothermic, the b-(E)-vinylsilane
being the thermodynamically favoured product (Figure 10).
The origin of the b-Z selectivity can be explained on the
grounds of the greater steric congestion in III-5-TS’ relative
to III-5-TS, a consequence of the silyl moiety interacting
Scheme 5. Possible non-classical reaction pathways.
Figure 7. DFT calculated relative energy [DE in kcalmolÀ1] profile for
acetone hydrosilylation.
Two possible reaction mechanisms have been considered
for the reaction of the hydrido complex and oxocarbenium
ion with an alkyne. In mechanism II (Scheme 5), coordina-
tion of the alkyne to the metal followed by insertion into
À
the M H bond is proposed; however, this mechanism may
yield only the b-(E)- or a- but not b-(Z)-vinylsilane. Alter-
natives through a metallacycle intermediate and further cis–
trans isomerisation have been proposed[9] but the calculated
relative energy for the metallacycle is too high (19.5 kcal
molÀ1), and therefore this pathway can be discarded.
Another alternative mechanism is the reaction of the oxo-
carbenium ion with the alkyne to yield a b-silyl carbocation
(Scheme 5, mechanism III). The existence of such a species
has been reported previously. The stability of this carbocat-
ion can be attributed to the large hyperconjugative nature
of the b-silyl effect.[15] Therefore, acetone acts as a silane
shuttle transferring the R3Si+ moiety from the silane to
a molecule of alkyne by means of an oxocarbenium ion. The
energy profile for the silyl transfer from the oxocarbenium
ion to the alkyne to form III-3 is shown in Figure 8, and fea-
tures an activation barrier of 17.0 kcalmolÀ1.
The carbocation resulting from the silylation of the
alkyne reacts with the hydrido complex without previous co-
ordination due to steric interaction between them. There-
fore, the vinylsilane is formed by nucleophilic attack of the
hydride over the carbocation. Two geometries are conceiva-
ble for the approach of the carbocation. The first possibility
Figure 9. DFT optimised structure of III-5-TS (top) and III-5-TS’
(bottom) leading to the b-(Z)-vinylsilane and b-(E)-vinylsilane, respec-
tively. Wingtip groups and hydrogen atoms have been omitted for clarity.
17564
www.chemeurj.org
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17559 – 17566

Hydrosilylation of Terminal Alkynes
FULL PAPER
actions, which could direct the selectivity of the hydrosilyla-
tion of alkynes towards the desired vinylsilane, are currently
being investigated by our group.
Experimental Section
General information: All manipulations were performed using standard
Schlenk techniques under an argon atmosphere, except when otherwise
noted. All complexes after their formation were treated under aerobic
conditions. Solvents were obtained dried from a solvent purification
system from Innovative Technology Inc. Salt 1 and complexes 2a and 3a
were prepared according to a synthetic procedure recently reported by
us.[11] All other reagents were used as received. NMR spectra were re-
corded on Bruker Avance 300 MHz, Bruker ARX 300 or Bruker Avance
400 MHz spectrometers. The chemical shifts are given as dimensionless
d values and are frequency referenced relative to residual solvent peaks
Figure 10. DFT calculated relative energy [DE in kcalmolÀ1] profile for
the nucleophilic attack of the hydride to the carbocation to yield b-(Z)-
and b-(E)-vinylsilanes.
1
for H and 13C. Coupling constants J are given in Hertz as positive values
with the two iodido ligands in the apical positions and the
wingtip groups in the equatorial plane, as supported by the
different energy barrier between both processes.
regardless of their real individual signs. The multiplicity of the signals is
indicated as s, d or m for singlet, doublet or multiplet, respectively. Mass
spectra and high-resolution mass spectra were obtained on an Esquire
3000+ instrument with ion-trap detector interfaced on an Agilent 1100
HPLC analyser, in electrospray mode unless otherwise reported. Elemen-
tal C/H/N analyses were carried out in a PerkinElmer 2400 CHNS/O an-
alyser.
Conclusion
Methylenebis(N-2-methoxyethyl)imidazole-2-ylidene)acetato-
ACHUTNGRENUN(G diiodo)rhodiumACHTUNGTNER(NUGN III) (2b): A mixture of [RhCAHTUNGTRNEN(GNU m-Cl)ACTHGNUTREN(UNGN cod)]2 (0.123 g,
We have developed a straightforward method for the syn-
0.25 mmol), 1 (0.260 g, 0.5 mmol), NaI (0.70 g, 5.0 mmol), KOAc (0.65 g,
6.5 mmol) and CH3CN (15 mL) was heated at reflux for 3 days, after
which the solvent was evaporated under reduced pressure. The remaining
residue was dissolved in CH2Cl2 and any insoluble impurities were isolat-
ed by filtration. The solution was concentrated to approximately 1 mL
and the product precipitated with diethyl ether. The solid was washed
with diethyl ether (3ꢇ10 mL) and dried in vacuo to give an orange solid
thesis of cationic (bis-NHC)-C,C,O,O iridium
ACHTUNGTRENN(UNG III) and
rhodium(III) complexes (3a and 3b). These complexes
ACHTUNGTRENNUNG
showed good activities in the hydrosilylation of terminal al-
kynes and excellent selectivities towards b-(Z)-vinylsilanes
only when acetone was used as solvent. Remarkably, iridium
complex 3a performs better in the hydrosilylation of aro-
matic alkynes than rhodium complex 3b. Conversely, 3b
shows better selectivities and activities than 3a in the hydro-
silylation of aliphatic alkynes.
The reaction mechanism was investigated in a combined
computational and experimental study. The possibility of
a classical mechanism was discarded as 1) the oxidative ad-
(0.15 g,
88%).
1H NMR
(acetone,
400 MHz):
d=7.48
(d,
JHÀH =2.1 Hz, 2H; CHim), 7.42 (d, JHÀH =2.1 Hz, 2H; CHim), 6.27 (s, 2H;
NCH2N), 4.71–4.58 (m, 4H; CH2N), 3.82–3.74 (m, 4H; CH2O), 3.33 (s,
6H; CH3O), 1.82 ppm (s, 3H; CH3AcO); 1H NMR (CDCl3, 400 MHz): d=
7.32 (d, JHÀH =2 Hz, 2H; CHim), 7.05 (d, JHÀH =2 Hz, 2H; CHim), 6.05 (s,
2H; NCH2N), 4.71–4.65 (m, 4H; NCH2), 3.85–3.81 (m, 4H; CH2O), 3.36
(s, 6H; CH3O), 2.03 ppm (s, 3H; CH3AcO); 13C{1H} NMR plus attached
proton test (CDCl3, 101 MHz): d=187.7 (d, JRh C =1.4 Hz; CCOOÀ),
À
À
dition of the Si H bond to the rhodium or iridium centre
154.3 (d, JRh C =43 Hz; NCimN), 125.4 (CHim), 119.6 (CHim), 72.4 (OCH2),
À
63.3 (NCH2N), 58.9 (CH3O), 51.1 (CH2N), 24.8 ppm (CH3COOÀ); ele-
mental analysis calcd (%) for C15H24I2RhN4O4 (680.89): C 26.45, H 3.55,
N 8.23; found: C 26.76, H 3.56, N 8.49.
was found to be unfeasible by calculations at the DFT level,
and 2) it was found experimentally that alkynes do not react
with 3a or 3b.
Methylenebis(N-2-methoxyethyl-kO,kO’)imidazole-2-ylidene)-
An outer-sphere mechanism for the hydrosilylation of ter-
minal alkynes was proposed based on Si–O interactions be-
tween the silane and the solvent (acetone). The mechanism
proposal is based on three main steps: 1) the heterolytic
splitting of the silane molecule by the metal centre and the
acetone molecule, 2) the reaction of the resulting oxocarbe-
À
ACHUTNGERN(NUG diiodo)rhodiumACHTUGNTREN(NGUN III) tetrafluoroborate (3b): HBF4 (75 mL, 0.55 mmol)
was added dropwise under argon to a solution of 2b (0.35 g, 0.51 mmol)
in dry CH2Cl2 (30 mL) at 08C, then the mixture was stirred for 1 h and
allowed to warm to room temperature, thus affording an orange suspen-
sion. Subsequently, the solvent was evaporated under reduced pressure
to a volume of approximately 1 mL and precipitation was achieved with
Et2O (10 mL). The resulting residue was washed with diethyl ether (3ꢇ
10 mL) to afford an orange solid (0.30 g, 83%). 1H NMR ([D6]acetone,
nium ion ([R3Si OACHTUNGTRENNUNG
(CH3)2]+) with the corresponding alkyne
+
400 MHz):
d=7.72
(d,
JHÀH =2 Hz,
2H;
CHim),
7.70
À
À
to give the silylation product ([R3Si CH=C R] ), and
(d, JHÀH=2 Hz, 2H; CHim), 6.50 (s, 2H; NCH2N), 4.80–4.74 (m, 4H;
CH2N), 4.42–4.37 (m, 4H; OCH2), 4.06 ppm (s, 6H; CH3O);
À
3) nucleophilic attack of the hydrido ligand over [R3Si CH=
+
À
C R] . The b-Z selectivity of the reaction is explained as
13C{1H} NMR ([D6]acetone, 101 MHz): d=146.8 (d, JRh C =44 Hz;
À
a consequence of the higher steric interaction resulting from
the geometry of the approach that leads to b-(E)-vinylsi-
lanes.
NCimN), 124.6 (CHim), 122.1 (CHim), 76.7 (OCH2), 67.6 (CH3O), 62.5
(NCH2N), 49.6 ppm (CH2N); 19F NMR ([D6]acetone, 282 MHz): d=
À151.0 ppm; elemental analysis calcd (%) for C13H20BF4I2RhN4O2
(707.88): C 22.06, H 2.85, N 7.92; found: C 22.10, H 2.77, N 8.16.
This mechanism proposal represents the first example of
an outer-sphere mechanism for the hydrosilylation of termi-
nal alkynes. The development of new catalytic systems oper-
ating through outer-sphere mechanisms based on Si–O inter-
NMR studies of the hydrosilylation of alkynes: The corresponding cata-
lyst precursor, 3a or 3b (6.0 mg, 7.5ꢇ10À3 mmol), was dissolved in
[D6]acetone (0.5 m) in an NMR tube. Subsequently, the silane
(0.16 mmol), the corresponding alkyne (0.15 mmol) and mesitylene
Chem. Eur. J. 2013, 19, 17559 – 17566
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17565

M. Iglesias, F. J. Fernꢀndez-Alvarez, L. A. Oro et al.
(21.0 mL, 0.15 mmol) as internal standard were added. The NMR tube
was then closed under argon and placed in an oil bath for 5 h at 508C.
Yields and selectivities were calculated by 1H NMR spectroscopy. The
[2] G. W. Gribble, J. J. Li, in Palladium in Heterocyclic Chemistry:
a Guide for the Synthetic Chemist (Eds.: G. W. Gribble, J. J. Li),
Elsevier, Oxford, 2006, p. 12.
[3] For examples, see: a) J. D. Sunderhaus, H. Lam, G. B. Dudley, Org.
Lett. 2003, 5, 4571–4573; b) K. Murakami, H. Yorimitsu, K.
Oshima, J. Org. Chem. 2009, 74, 1415–1417.
[4] For a reviews see: a) B. Marciniec, C. Pietraszuk, I. Kownacki, M.
Zaidlewicz, in Comprehensive Organic Functional Group Transfor-
mations (Eds.: A. R. Katrizky, J. K. Taylor), Elsevier, Oxford, 2005,
different isomers were unambiguously identified by means of the 3JHÀ
H
coupling constants of the vinylic protons.
X-ray crystallography: Intensity data of 2a, 2b, 3a and 3b were collected
on a Bruker Kappa APEX2 diffractometer (MoKa 0.71069 ꢆ, graphite
monochromator). Data reduction and cell refinements were carried out
using the APEX2 software package.[16] All the structures were solved by
direct methods and refined by full-matrix least-squares based on F2 using
the SHELXL-97 and WinGX programs.[17] Positions of all non-hydrogen
atoms were deduced from difference Fourier maps and refined anisotrop-
ically. Distance and/or thermal restraints were applied to CH2Cl2 (2a),
BF4 (3a) and N13, N22, O53, F12, F14 (3b). Hydrogen atoms were posi-
tioned in geometrically calculated positions and refined with isotropic
displacement parameters according to the riding model.
´
p. 941; b) P. Pawluc, W. Prukala, B. Marciniec, Eur. J. Org. Chem.
2010, 219–229; c) D. S. W. Lim, E. A Anderson, Synthesis 2012,
983–1010.
[5] a) A. K. Roy, Adv. Organomet. Chem. 2007, 55, 1–59; b) B. Marci-
´
niec, K. H. Maciejewski, C. Pietraszuk, P. Pawluc, in Hydrosilyla-
tion: A Comprehensive Review on Recent Advances (Ed.: B. Marci-
niec), Springer, Heidelberg, 2008.
[6] A. J. Chalk, J. F. Harrod, J. Am. Chem. Soc. 1965, 87, 16–21.
[7] a) J. F. Harrod, A. J. Chalk, in Organic Synthesis via Metal Carbon-
yls, Vol. 2 (Eds.: I. Wender, P. Pino), Wiley, New York, 1977, p. 673;
b) T. D. Tilley, in The Chemistry of Organic Silicon Compounds
(Eds.: S. Patai, Z. Rappoport), Wiley, New York, 1989, p. 1415; c) I.
Ojima, in The Chemistry of Organic Silicon Compounds (Eds.: S.
Patai, Z. Rappoport), Wiley, New York, 1989, p. 1479; d) M. A.
Schroeder, M. S. Wrighton, J. Organomet. Chem. 1977, 128, 345–
358; e) C. L. Reichel, M. S. Wrighton, Inorg. Chem. 1980, 19, 3858–
3860; f) C. L. Randolph, M. S. Wrighton, J. Am. Chem. Soc. 1986,
108, 3366–3374; g) A. Millan, E. Towns, P. M. J. Maitlis, Chem. Soc.
Chem. Commun. 1981, 673–674; h) A. Milan, M.-J. Fernandez, P.
Bentz, P. M. Maitlis, J. Mol. Catal. 1984, 26, 89–104; i) I. Ojima, M.
Yatabe, T. Fuchikami, J. Organomet. Chem. 1984, 260, 335–346.
[8] For examples see: a) J. W. Faller, D. G. D’Alliessi, Organometallics
2002, 21, 1743–1746; b) R. H. Crabtree, New J. Chem. 2003, 27,
771–772; c) M. V. Jimꢁnez, J. J. Pꢁrez-Torrente, M. I. Bartolomꢁ, V.
Gierz, F. J. Lahoz, L. A. Oro, Organometallics 2008, 27, 224–234.
[9] R. S. Tanke, R. H. Crabtree, J. Am. Chem. Soc. 1990, 112, 7984–
7989.
Crystal
[C16H25Cl2I2IrN4O4]; orthorhombic; space group Pbca; a=16.0598(12),
b=13.9481(10), c=22.4432(17) ꢆ; Z=8; Mr =854.3 gmolÀ1
V=
data
for
[Ir(I)
(bis-NHC)(h2-CH3COO)]·CH2Cl2
(2a):
;
5027.4(6) ꢆ3; Dcalcd =2.257 gcmÀ3; l
ACTHNUTRGEN(UNG MoKa)=0.71073 ꢆ; T=100 K; m=
8.008 mmÀ1; 53124 reflections collected, 6809 observed (Rint =0.0435);
2
R1(Fo)=0.0273 [I>2s(I)], wR2
N
Crystal data for [Rh(I)2A
G
(2b):
¯
[C16H25Cl2I2N4O4Rh]; triclinic; space group P1; a=8.6375(7), b=
12.0197(10), c=13.3547(11) ꢆ; a=97.0160(10), b=108.7340(10), g=
105.6760(10)8; Z=2; Mr =765.01 gmolÀ1
ACHTUNGTRENNUNG
2.065 gcmÀ3; l(MoKa)=0.71073 ꢆ; T=100 K; m=3.448 mmÀ1; 13529 re-
flections collected, 6234 observed (Rint =0.0207); R1(Fo)=0.0217 [I>
2
2s(I)], wR2ACHTUNGTRENNUNG
Crystal
G
(3a):
¯
[C13H20BF4I2IrN4O2]; triclinic; space group P1; a=8.1369(6), b=
12.3989(10), c=13.1088(10) ꢆ; a=62.7780(10), b=76.3250(10), g=
75.5850(10)8; Z=2; Mr =797.14 gmolÀ1
2.348 gcmÀ3; l
flections collected, 5718 observed (Rint =0.0220); R1(Fo)=0.0482 [I>
2
2s(I)], wR2
ACHTUNGTRENNUNG(Fo )=0.1276 (all data); GOF=1.031.
[10] I. Ojima, N. Clos, R. J. Donovan, P. Ingallina, Organometallics 1990,
9, 3127–3133.
[11] M. Iglesias, M. Pꢁrez-Nicolꢀs, P. J. Sanz Miguel, V. Polo, F. J. Fernꢀn-
dez-Alvarez, J. J. Pꢁrez-Torrente, L. A. Oro, Chem. Commun. 2012,
48, 9480–9482.
Crystal data
for
[Rh(I)2{h4:k-O,k-O-(bis-NHC)}]
(BF4)
(3b):
¯
[C13H20BF4I2N4O2Rh]; triclinic; space group P1; a=8.6767(13), b=
10.3696(15), c=12.4048(18) ꢆ; a=84.859(2), b=70.430(2), g=
77.625(2)8; Z=2; Mr =707.85 gmolÀ1
;
V=1027.0(3) ꢆ3; Dcalcd
=
2.289 gcmÀ3; l
ACHTUNGTRENNUNG
(MoKa)=0.71073 ꢆ; T=100 K; m=3.889 mmÀ1; 11996 re-
[12] a) M. Martꢂn, E. Sola, S. Tejero, J. A. Lꢃpez, L. A. Oro, Chem. Eur.
J. 2006, 12, 4057–4068; b) M. Martꢂn, E. Sola, S. Tejero, J. L.
Andrꢁs, L. A. Oro, Chem. Eur. J. 2006, 12, 4043–4056; c) G. E. Do-
bereiner, A. Nova, N. D. Schley, N. Hazari, S. J. Miller, O. Eisen-
stein, R. H. Crabtree, J. Am. Chem. Soc. 2011, 133, 7547–7562; d) I.
Mena, M. A. Casado, V. Polo, P. Garcꢂa-OrduÇa, F. J. Lahoz, L. A.
Oro, Angew. Chem. 2012, 124, 8384–8388; Angew. Chem. Int. Ed.
2012, 51, 8259–8263; e) R. Lalrempuia, M. Iglesias, V. Polo, P. J.
Sanz Miguel, F. J. Fernꢀndez-Alvarez, J. J. Pꢁrez-Torrente, L. A.
Oro, Angew. Chem. 2012, 124, 12996–12999; Angew. Chem. Int. Ed.
2012, 51, 12824–12827; f) P. C. Roosen, V. A. Kallepalli, B. Chatto-
padhyay, D. A. Singleton, R. E. Malezca Jr, M. R. Smith III, J. Am.
Chem. Soc. 2012, 134, 11350–11353; g) W. W. N. O, A. J. Lough,
R. H. Morris, Organometallics 2012, 31, 2152–2165; h) N. D. Schley,
S. Halbert, C. Raynaud, O. Eisenstein, R. H. Crabtree, Inorg. Chem.
2012, 51, 12313–12323.
[13] S. Park, M. Brookhart, Organometallics 2010, 29, 6057–6064.
[14] J. Yang, P. S. White, C. K. Schauer, M. Brookhart, Angew. Chem.
2008, 120, 4209–4211; Angew. Chem. Int. Ed. 2008, 47, 4141–4143.
[15] V. Gabelica, A. J. Kresge, J. Am. Chem. Soc. 1996, 118, 3838–3841.
[16] APEX2 Bruker AXS Inc., Madison, Wisconsin, USA, 2011.
[17] a) G. M. Sheldrick, SHELXS-97 and SHELXL-97, University of
Gçttingen (Germany), 1997; b) L. J. Farrugia, WinGX, University of
Glasgow (UK), 1998.
flections collected, 4702 observed (Rint =0.0255); R1(Fo)=0.0597 [I>
2
2s(I)], wR2ACHTUNGTRENNUNG(Fo )=0.1434 (all data); GOF=1.084.
CCDC-953145 (2a), CCDC-953146 (2b), CCDC-953147 (3a) and
CCDC-953148 (3b) contain the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
Acknowledgements
This work was supported by the Spanish Ministry of Economy and Com-
petitiveness (MINECO/FEDER) (CONSOLIDER INGENIO-2010,
CTQ2011-27593 projects, and “Ramꢃn y Cajal” (P.J.S.M.) and “Juan de
la Cierva” (M.I.) programmes) and the DGA/FSE (E07). V.P. gratefully
acknowledges the resources from the supercomputer “Terminus”, techni-
cal expertise and assistance provided by the Institute for Biocomputation
and Physics of Complex Systems (BIFI), Universidad de Zaragoza.
[1] a) I. Ojima, in The Chemistry of Organic Silicon Compounds (Eds.:
S. Patai, Z. Rappoport), Wiley, New York, 1989, p. 1479; b) I.
Ojima, Z. Li, J. Zhu, in The Chemistry of Organic Silicon Com-
pounds (Eds: Z. Rappoport, Y. Apeloig), Wiley, New York, 1998,
p. 1687.
Received: August 2, 2013
Published online: November 14, 2013
17566
www.chemeurj.org
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17559 – 17566