A mechanistic study into the interconversion of rhodium alkyne, alkynyl hydride and vinylidene complexes

Article abstract of DOI:10.1039/b806358c

The conversion of the alkyne complex RhCl(PPrⁱ₃) ₂(η²-PhCCH) into its vinylidene isomer RhCl(PPr ⁱ₃)₂(=C=CPhH) has been monitored via NMR spectroscopy in a range of solvents. In THF and hexane solution the reaction was observed to proceed via the alkynyl hydride complex Rh(H)Cl(PPr ⁱ₃)₂(-CCPh). The kinetic profile of the reaction demonstrates that this species is an intermediate in the formation of the vinylidene complex.

Full text of DOI:10.1039/b806358c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
A mechanistic study into the interconversion of rhodium alkyne, alkynyl
hydride and vinylidene complexes†
Michael J. Cowley, Jason M. Lynam* and John M. Slattery
Received 15th April 2008, Accepted 28th May 2008
First published as an Advance Article on the web 4th July 2008
DOI: 10.1039/b806358c
2
The conversion of the alkyne complex RhCl(PPri3)2(g -
alkyne, alkynyl hydride and vinylidene ligands.10 The precise
mechanism of the conversion of the alkynyl hydride complex to
the corresponding vinylidene has been the subject of recent debate.
Circumstantial experimental11 and some theoretical studies5 have
implied that this transformation occurs via a bimolecular pathway
(involving species such as E, Scheme 1). Although the absence of a
low energy unimolecular route may be an artefact of the symmetry
constraints employed in the calculations by Morokuma.5,6 More
recent isotopic labelling7 and theoretical studies8 appear to
demonstrate that a unimolecular 1,3-shift from the alkynyl hydride
is the preferred pathway.
i
≡
= =
PhC CH) into its vinylidene isomer RhCl(PPr 3)2( C
CPhH) has been monitored via NMR spectroscopy in a
range of solvents. In THF and hexane solution the reaction
was observed to proceed via the alkynyl hydride complex
Rh(H)Cl(PPr 3)2(–C CPh). The kinetic profile of the reaction
demonstrates that this species is an intermediate in the
formation of the vinylidene complex.
i
≡
Although the conversion of terminal alkynes into their vinylidene
tautomers mediated by electron-rich transition metal complexes is
a well-established transformation,1–4 the precise mechanism of this
process is still underdetermined in many cases. Several pathways
have been proposed for this reaction: in nearly all cases the first
step is coordination of the alkyne to the metal to give to give a
complex such as A (Scheme 1) which is then followed by oxidative
addition of the C–H bond of the alkyne to the metal giving
an alkynyl hydride intermediate (B). A 1,3-hydrogen shift then
results in the formation of the vinylidene ligand (C). A more
concerted 1,2-hydrogen shift from the initial alkyne complex has
also been postulated proceeding via a transition state such as D.
The precise pathway is almost certainly system dependant and
detailed mechanistic and theoretical studies have been performed
in order to elucidate these pathways.5–9
Our recent studies focusing on the incorporation of alkynes
containing nucleobases into the coordination sphere of transition
metal complexes12 have required re-evaluation of many reactions
commonly employed to prepare complexes containing vinylidene
ligands due to the poor solubility of these organic substrates. Dur-
ing these studies we have investigated the effect of different solvents
on the alkyne to vinylidene conversion mediated by RhCl(PPri3)2.
These studies have allowed us to perform an investigation into this
reaction and in this manuscript we demonstrate through a kinetic
study that the alkynyl hydride complex is indeed an intermediate
in the formation of the vinylidene complex and furthermore that
the reaction proceeds in a unimolecular fashion.
Reaction of [RhCl(PPri3)2]2, 1, with an equimolar amount
≡
of PhC CH in d8-THF solution resulted in the selective and
i
2
≡
immediate formation of RhCl(PPr 3)2(g -PhC CH), 2a, even at
−78 ◦C. When the reaction was monitored by NMR spectroscopy
over the course of several hours the resonance for 2a in the
31P{ H} NMR spectrum was observed to decrease in intensity to
1
be replaced firstly by a resonance at d 49.31 (1JP–Rh = 98.7 Hz,
2JP–H = 12.4 Hz) assigned to the alkynyl hydride complex 3a
1
(Scheme 2). A resonance in the hydride region of the H NMR
1
2
spectrum at d −27.97 (dt, JRh–H = 42.9 Hz, JP–H = 12.8 Hz),
Scheme 1 Possible mechanistic pathways for the conversion of alkyne
complexes to their vinylidene tautomers.
There has been considerable interest in the alkyne–vinylidene
transformation occurring at the Rh(P)2Cl centre (P = phosphine
ligand). Werner has demonstrated that the rhodium may support
Department of Chemistry, Univeristy of York, Heslington, York, UK
YO10 5DD. E-mail: jml12@york.ac.uk.; Fax: +44(0)1904 432516;
Tel: +44(0)1904 432534
† Electronic supplementary information (ESI) available: Experimental
information, details of the kinetic modelling and DFT calculations. For
ESI see DOI: 10.1039/b806358c
n
≡
Scheme 2 (i) + RC CH. R = Ph, a; R = Bu = b.
4552 | Dalton Trans., 2008, 4552–4554
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
consistent with the previously reported data for 3a,10a was observed
to grow in with this peak.
Over the course of approximately 16 h the resonances for
both 2a and 3a had vanished and were replaced by a se-
ries of peaks which were assigned to the vinylidene complex
extremely small. Importantly, models in which the conversion of
3a → 4a proceeds by a bimolecular pathway gave poor fits to the
experimental data. If the reaction was modelled with competing
uni- and bimolecular pathways for the conversion of 3a → 4a
then identical rate constants and fits to the unimolecular pathway
alone were obtained with a vanishingly small rate constant for any
bimolecular pathway.
In an attempt to gain further insight into this process, the
reaction was repeated in hexane as solvent. A near identical
reaction pathway was obtained in this medium, although the
i
10b
= =
RhCl(PPr 3)2( C C{Ph}H), 4a. The identity of 2a, 3a and
4a were further confirmed by repeating the reaction with
PhC 13CH. In this case, resonances could be observed in the 13
C
≡
{ H} NMR spectrum at d 75.65 (1JRh–C = 14.4 Hz, 2JP–C = 3.0 Hz)
1
1
2
for 2a, at d 102.72 (dt, JRh–C = 47.3 Hz, JP–C 15.3 Hz) for 3a
and at d 296.35 (dt, 1JRh–C = 58.6 Hz, 2JP–C = 15.9 Hz) for 4a. The
intensity of these resonances mirrored those for the corresponding
calculated rates of reaction are slower than in THF solution. In
n
≡
addition, the reaction of BuC CH was also examined and, again
peaks in the 1H and 31P{ H} spectra.
afforded a similar reaction profile with the conversion of 2b, 3b and
4b being successfully modelled. The rate constants are collected in
Table 1. Notably, the rates of interconversion of 2b and 3b are over
an order of magnitude greater than for the corresponding phenyl
derivatives 2a and 3a. In contrast, there is a smaller difference in
the rate constants for the conversion of complexes of type 3 into 4
with k2 3a → 4a being 3.4 times larger than 3b → 4b.
1
A 48 mmol d8-THF solution of 2a was prepared and the
formation of 3a and 4a (and the loss of 2a) was monitored by 1H
and 31P{ H} NMR spectroscopy over a period of several hours.
1
The resulting reaction profile is shown in Fig. 1. Crucially, we were
able to fit this kinetic profile to the reaction profile 2a ꢀ 3a →
4a with first order rate constants of k1 = (4.26 0.14) × 10−4 s−1,
k−1 = (2.16 0.22) × 10−4 s−1 and k2 = (4.59 0.07) × 10−4 s−1.
(See Scheme 1 for the assignment of the rate constants.)
These results provide direct evidence for the intermediacy of
the alkynyl hydride complex 3 in the conversion of 2 to 4 and
demonstrate that the process occurs via a unimolecular pathway,
complementing the labelling studies described by Grotjahn7 and
the theoretical study by De Angelis.8
In a series of studies on related systems with pyridyl-phosphines
and using CD2Cl2 as solvent, Grotjahn has demonstrated that
the conversion of complexes such as 2 to 4 is a unimolecular
process (which is consistent with our kinetic analysis) but, in direct
contrast to our data, intermediates such as 3 are not observed.13
We therefore repeated our experiments in CD2Cl2 as solvent under
identical concentrations and temperature. In this instance the
formation of the alkyne complex 2 from 1 was much slower and,
in direct agreement with the results of Grotjahn, no evidence
for a hydride-containing intermediate was observed during the
formation of 4.
The marked difference in rate of addition of the alkyne to 1 in
THF and CH2Cl2 is intriguing so this process was examined in
more detail. Identical samples of 1 in d8-THF and CD2Cl2 show
markedly different 31P{ H} NMR spectra. In the former case the
1
major component of the spectrum is a broad doublet resonance at
d 56.39 (d, 1JRh–P = 197.4 Hz), whereas in CD2Cl2 solution a sharp
resonance was observed at substantially different chemical shift d
Fig. 1 Reaction profile for the formation of complex 4a from 2a with
theoretical fit.
1
21.52 (d, JRh–P = 96.9 Hz). The addition of CH2Cl2 to d8-THF
Attempts to model the kinetic pathway using other possible
mechanisms were unsuccessful. Notably, attempts to model the
reaction sequence as 2a ꢀ 3a ꢀ 4a returned identical rate
constants for k1, k−1 and k2 as the previous approach: the rate
constant for the reaction 4a → 3a was always predicted to be
1
solutions of 1 resulted in marked changes to the 31P{ H} NMR
spectra. Although we have not been able to identify the species
formed, it seems likely that 1 reacts with CH2Cl2 to give a new
species. It is evident that this process is reversible as reaction with
Table 1 Comparison of kinetic and activation parameters for the conversion of alkyne to vinylidene ligandsa
Substituent Solvent k1/10−4 s−1 k−1/10−4 s−1 k2/10−4 s−1 DG‡23/kJmol−1 DG‡32/kJmol−1 DG23/kJmol−1 DG‡34/kJmol−1 DG34/kJmol−1
Ph
THF
Hexane
THF
n. a.
4.26 0.14 2.16 0.22 4.59 0.07 92.85 0.08
2.10 0.30 2.09 0.54 1.47 0.01 94.62 0.36
53.44 0.39 38.76 0.29 1.23 0.02 86.54 1.64
94.55 0.25
94.62 0.64
87.34 1.68
67
97
88
80
−1.70 0.17 92.67 0.04
0.00 0.74 95.51 0.12
−0.80 2.34 95.94 0.04
—
—
—
−63
−39
−26
−20
Ph
nBu
Model i
Model ii
Model iii
Model iv
—
—
—
—
—
—
—
—
—
—
—
—
98
111
106
97
+31
82
115
105
103
n. a.
+14
+18
+17
n. a.
n. a.
a Experimental DG values for data recorded at 300 K.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 4552–4554 | 4553
©

View Article Online
terminal alkynes still affords 2. Consistent with this argument, the
than hexane. Although this is consistent with the experimental
observations, it is important to note that Werner has shown that
pyridine may act as a Lewis base towards complexes such as 3 and
may affect the conversion to the vinylidene complex.10a,11 Such
an interaction involving THF, which may not be well modelled
by solvation calculations, cannot be discounted and so gas phase
energies are presented here. Solvation-corrected energies can be
found in the ESI.†
Both de Angelis8 and Grotjahn13 have demonstrated that the
nature of the phosphine ligand may have a pronounced effect on
the potential energy surface of this reaction. In addition, both our
experimental and theoretical results suggest that the substituent
on the alkyne, and the solvent employed, may also have a dramatic
effect on the reaction pathway. These results therefore reinforce the
notion that the precise manner of the metal-mediated conversion
of alkynes to vinylidene ligands is extremely system-dependant.
1
31P{ H} spectrum of 1 recorded in hexane solution is essentially
identical to that observed in the THF case as might be expected
on the basis of their similar kinetic profiles.
In order to examine substituent effects on the alkyne to vinyli-
dene transformation in more detail, a series of DFT calculations
exploring the mechanism shown in Scheme 2, similar to those
previously reported,8,13,14 were performed on four model systems
ꢀ
≡
involving different phosphine (PR3) and alkyne (R C CH) sub-
stituents (model i R = H, Rꢀ = H; ii R = Me, Rꢀ = H; iii R =
Me, Rꢀ = Ph; iv R = Me, Rꢀ = Me). Models iii and iv were
chosen as slightly simplified models of the experimental systems
a and b respectively. Geometry optimisations were performed for
model structures 2, 3 and 4 and the transition states that link them
(denoted TS23 and TS34 respectively) at the (RI-)BP86/SV(P) level
and followed by single point calculations at the (RI-)PBE0/TZVP
level to obtain more reliable energies.15,16 The potential energy
surface for model i is shown in Scheme 3 as an example.
Acknowledgements
We are grateful to the EPSRC and the University of York for
funding.
Notes and references
1 M. I. Bruce, Chem. Rev., 1991, 91, 197.
2 H. Werner, J. Organomet. Chem., 1994, 475, 45.
3 Y. Wakatsuki, J. Organomet. Chem., 2004, 689, 4092.
4 M. C. Puerta and P. Valerga, Coord. Chem. Rev., 1999, 193–195,
977.
5 Y. Wakatsuki, N. Koga, H. Werner and K. Morokuma, J. Am. Chem.
Soc., 1997, 119, 360.
6 R. Stegmann and G. Frenking, Organometallics, 1998, 17, 2089.
7 D. B. Grotjahn, X. Zeng and A. L. Coosky, J. Am. Chem. Soc., 2006,
128, 2798.
Scheme 3 Calculated structures and schematic potential energy surface
for model i.
8 F. De Angelis, A. Sgamellotti and N. Re, Organometallics, 2007, 26,
The calculated activation parameters agree relatively well with
those determined experimentally given the expected differences
between gas and solution phase and the different substituents.
Importantly, the trends in substituent effects on DG‡ and DG‡
5285.
9 I. De los Rios, M. Jime´nez Tenorio, M. C. Puerta and P. Valerga, J. Am.
Chem. Soc., 1997, 119, 6529.
10 (a) J. Wolf, H. Werner, O. Serhadli and M. L. Ziegler, Angew. Chem.,
Int. Ed. Engl., 1983, 22, 414; (b) F. J. G. Alonso, A. Ho¨hn, J. Wolf, H.
Otto and H. Werner, Angew. Chem., Int. Ed. Engl., 1985, 24, 406; (c) H.
Werner, F. J. G. Alonso, J. Otto and J. Wolf, Z. Naturforsch., B: Chem.
Sci., 1988, 43, 722; (d) H. Werner and U. Brekau, Z. Naturforsch., B:
Chem. Sci., 1989, 44, 1439; (e) T. Rappert, O. Nu¨rnberg, N. Mahr, J.
Wolf and H. Werner, Organometallics, 1992, 11, 4156; (f) H. Werner,
T. Rappert, M. Baum and Arthur Stark, J. Organomet. Chem., 1993,
459, 319; (g) H. Werner, M. Baum, D. Schneider and B. Windmu¨ller,
Organometallics, 1994, 13, 1089.
11 A. Ho¨hn and J. Werner, J. Organomet. Chem., 1990, 382, 255.
12 (a) H. Hamidov, J. C. Jeffery and J. M. Lynam, Chem. Commun., 2004,
1364; (b) M. J. Cowley, J. M. Lynam and A. C. Whitwood, Dalton
Trans., 2007, 4427.
23
34
(i.e. larger DG‡ for Ph when compared to an alkyl derivative
23
and similar DG‡ for both aryl and alkyl substituents) is reflected
34
in the calculations. These results support the experimental data
and the argument that the non-observance of an alkynyl hydride
n
≡
intermediate in the reaction of BuC CH with 1 in CH2Cl2
is unlikely to be a substituent effect. The calculations also
predict that the introduction of a substituent, be it alkyl or aryl,
destabilises the resulting vinylidene complex relative to the starting
alkyne complex. Similar results have also been obtained by Suresh
and Koga.14
13 D. B. Grotjahn, X. Zeng, A. L. Cooksy, W. S. Kassel, A. G. DiPasquale,
L. N. Zakharov and A. L. Rheingold, Organometallics, 2007, 26, 3385.
14 C. H. Suresh and N. Koga, J. Theor. Comput. Chem., 2005, 4, 59.
The calculations also predict a barrier of between 123 and
154 kJ mol−1 for the conversion of complexes 4 into 3: at 300 K an
activation barrier of 154 kJ mol−1 corresponds to a rate constant
of ca. 2.4 × 10−9 s−1 which is consistent with the experimental
data which demonstrate that the conversion of the alkynyl hydride
complex to the vinylidene is essentially irreversible.
Solvation effects were also considered in the calculations using
the COSMO model.16d This lowered the energies of all the transi-
tion states, with the stabilisation being more pronounced for THF
15 All calculations were performed using TURBOMOLE 5.9.1.16
A
detailed description of the methods used, XYZ coordinates and
vibrational frequencies of the stationary points can be found in the
ESI†.
16 (a) R. Ahlrichs, M. Baer, M. Ha¨ser, H. Horn and C. Koelmel, Chem.
Phys. Lett., 1989, 162, 165; (b) M. von Arnim and R. Ahlrichs, J. Chem.
Phys., 1999, 111, 9183; (c) F. Weigend and M. Ha¨ser, Theor. Chem. Acc.,
1997, 97, 331; (d) A. Klamt and G. Schu¨u¨rmann, J. Chem. Soc., Perkin
Trans. 2, 1993, 799.
4554 | Dalton Trans., 2008, 4552–4554
This journal is
The Royal Society of Chemistry 2008
©