Binuclear activation of cumulenes: Roles of the adjacent metals and the cumulene binding mode in the activation process

Article abstract of DOI:10.1021/om0580102

The roles of the adjacent metals and the binding mode in the activation process of cumulenes were investigated. A binuclear complex was made to react with allene and methylallene to ultimately yield the vinylcarbene products. Nuclear magnetic resonance spectroscopy was used to monitor the reactions. It was proposed that the formation of the vinyl carbene products results from migration of the methyl ligand to the central cumulene carbon.

Full text of DOI:10.1021/om0580102

Organometallics 2005, 24, 3711-3724
3711
Binuclear Activation of Cumulenes: Roles of the
Adjacent Metals and the Cumulene Binding Mode in the
Activation Process
Dusan Ristic-Petrovic, D. Jason Anderson, Jeffrey R. Torkelson,
Michael J. Ferguson,‡ Robert McDonald,‡ and Martin Cowie*
Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2
Received February 18, 2005
The binuclear complex [Ir2(CH3)(CO)(µ-CO)(dppm)2][CF3SO3] (1; dppm ) µ-Ph2PCH2PPh2)
reacts with allene and methylallene to ultimately yield the vinylcarbene products [Ir2H-
(CO)2(µ-η1:η3-HCC(CH3)C(H)R)(dppm)2][CF3SO3] (R ) H (6), CH3 (7)). Monitoring the
reactions by NMR spectroscopy (1H, 13C, 31P) between -78 °C and ambient temperature
allows the observation of several intermediates in each of these transformations in which
the allene moves from an η2 binding site on one metal, through an η1:η1-bridging geometry
in which the cumulene is coordinated through the “H2CdC” moiety, to an η1:η3-bridging
geometry in which the central carbon of the cumulene is σ-bound to one metal, adjacent to
the methyl ligand, while the three cumulene carbons are η3-bound to the adjacent metal.
We propose that formation of the respective vinyl carbene products results from migration
of the methyl ligand to the central cumulene carbon followed by activation of a cumulene
C-H bond. 1,1-Dimethylallene reacts with 1 at -78 °C to yield a methylene hydride product
containing an η2-bound cumulene on one metal, much as observed for the first products in
the allene and methylallene reactions. Upon warming, this intermediate isomerizes to the
final product containing a methyl ligand on one metal and an η2-bound cumulene on the
other. No cumulene-bridged products are observed with this disubstituted allene. 1,1-
Difluoroallene also yields a methylene hydride product at -78 °C, which is analogous to the
first species observed in all cases noted above. In this case, warming results in movement
of the cumulene to an η1:η1-bridging position in which this group binds to the metals via the
“H2CdC” moiety. Unlike the transformations observed with allene and methyl allene,
difluoroallene undergoes no additional transformations as the temperature is raised. A
rationalization of these transformations is presented together with a perspective on how
the cumulene ligand moves over the dimetallic framework leading to the final products.
Introduction
adjacent metal centers in the former. So although the
chemistry occurring in each environment (surface and
single metal) is fundamentally no different, it neverthe-
less seems intuitively clear that the presence of adjacent
metals should give rise to some degree of reactivity
differences compared to those occurring at single metals.
Certainly, the existence of bridged-ligand binding modes
is only possible in complexes containing adjacent metals,
and one might assume that such bonding should give
rise to differences in reactivity (either subtle or dra-
matic) compared to that of terminally bound groups.2
Although a large number of polymetallic complexes
have been studied,3 surprisingly little is understood
about the roles of the adjacent metals in the activation
Chemistry occurring on metal surfaces is notoriously
difficult to study,1a,b and it is especially difficult to obtain
detailed information about the natures of surface-bound
organic fragments and their roles in the facile trans-
formations occurring on catalyst surfaces. Fortunately,
information about the possible natures of these surface-
bound fragments can be obtained through analogies
with well-defined organometallic complexes, based on
the concept, championed by the late E. L. Muetterties
and others,1 that chemical transformations occurring on
metal surfaces and in metal complexes are fundamen-
tally the same. The most obvious difference between
metal surfaces and mononuclear metal complexes (apart
from the well-defined geometry of ancillary ligands
surrounding the metal in the latter) is the presence of
(2) (a) Shriver, D. F.; Sailor, M. J. Acc. Chem. Res. 1988, 21, 374.
(b) Xiao, F.-S.; Fukuoka, A.; Ichikawa, M. J. Catal. 1992, 138, 206. (c)
Araki, M.; Ponec, V. J. Catal. 1976, 44, 439. (d) Vites, J.; Housecroft,
C. E.; Eigenbrot, C.; Buhl, M.; Long, G. J.; Fehlner, T. P. J. Am. Chem.
Soc. 1985, 107, 8147. (e) Horvitz, C. P.; Shriver, D. G. J. Am. Chem.
Soc. 1985, 107, 8147.
(3) See for example: (a) Braunstein, P. In Perspectives in Coordina-
tion Chemistry; Williams, A. F.; Floriani, C.; Merbach, A. E., Eds.;
Verlag Helvetica Chim. Acta: Basel, 1992; p 67. (b) Shriver, D. F.;
Kaesz, H. D.; Adams, R. D. The Chemistry of Metal Cluster Complexes;
VCH: Weinheim, 1990. (c) Stone, F. G. A. Angew. Chem., Int. Ed. Engl.
1984, 23, 89.
* Corresponding author. E-mail: martin.cowie@ualberta.ca.
‡ X-ray Crystallography Laboratory.
(1) (a) Zaera, F. Chem. Rev. 1995, 95, 2651. (b) Muetterties, E. L.
Pure Appl. Chem. 1982, 54, 83. (b) Muetterties, E. L.; Rhodin, T. N.;
Band, E.; Brucker, C. F.; Pretzer, W. R. Chem. Rev. 1979, 79, 91. (c)
Muetterties, E. L. Chem. Soc. Rev. 1982, 11, 283. (d) Marks, T. J. Acc.
Chem. Res. 1992, 25, 57. (e) Schmid, G. Chem. Rev. 1992, 92, 1709. (f)
Schmid, G., Ed. Clusters and Colloids; VCH: Weinheim, 1994. (g)
Johnson, B. F. G. J. Mol. Catal. 1994, 86, 51.
10.1021/om0580102 CCC: $30.25 © 2005 American Chemical Society
Publication on Web 06/22/2005

3712 Organometallics, Vol. 24, No. 15, 2005
Ristic-Petrovic et al.
Scheme 1
methylene groups have been reported, and we earlier
reported the reactions of 1 with a series of alkynes8,9
and olefins.4d,10
In this paper we report the reactivity of 1 with a series
of cumulenes in order to determine if the C-H activa-
tion process reported in Scheme 1 has a role in subse-
quent C-C bond formation. A preliminary report of
some of this chemistry has appeared.8
of substrate molecules. Our approach to studying such
processes has involved a series of binuclear, diphos-
phine-bridged complexes in which an abundance of
NMR-active nuclei (31P, 19F, 13C, 1H) allows an in-depth
investigation of solution intermediates, often allowing
the determination of substrate and ancillary ligand
coordination modes throughout much of the activation
process.4 Other groups have taken a similar approach
using related diphosphine-bridged systems.5
One example of unusual reactivity initiated by the
presence of an adjacent metal is the facile C-H activa-
tion of a methyl group at one metal upon ligand addition
at the adjacent metal, as was observed in reactions of
[Ir2(CH3)(CO)2(dppm)2][CF3SO3] (1; dppm ) µ-Ph2PCH2-
PPh2) diagrammed in Scheme 1.4d,6 Although R-hydro-
gen elimination involving a methyl ligand at an early-
metal center is common,7 such C-H activation at a
single-site, late-metal complex is not. In the example
given above, the presence of an adjacent metal induces
reactivity not routinely observed with single metals. As
part of a strategy to exploit the above C-H bond
activation in subsequent carbon-carbon bond forma-
tion, we have investigated the addition of unsaturated
hydrocarbons to compound 1, in attempts to induce
coupling between these unsaturated substrates and the
resulting bridging methylene group.4d,8-10 Numerous
examples involving insertion of olefins,11 alkynes,12 and
cumulenes13 into the metal-carbon bonds of bridging
Experimental Section
General Comments. All solvents were dried (using ap-
propriate drying agents), distilled before use, and stored under
dinitrogen. Deuterated solvents used for NMR experiments
were freeze-pump-thaw degassed (three cycles) and stored
under nitrogen or argon over molecular sieves. Reactions were
carried out under argon using standard Schlenk techniques,
and compounds that were used as solids were purified by
recrystallization. Prepurified argon and nitrogen were pur-
chased from Linde, carbon-13 enriched CO (99%) was supplied
by Isotec Inc., and methylallene was supplied by Fluka. All
gases were used as received, and all other reagents were
purchased from Aldrich and were used as received, except as
noted. The compound [Ir2(CH3)(CO)(µ-CO)(dppm)2][CF3SO3] (1)
was prepared as previously reported.6 The perdeuteromethyl
analogue 1-d3 was also prepared by the literature method,
except that methyl-d3 trifluoromethanesulfonate was reacted
with [Ir2(CO)3(dppm)2], and the trimethylamine N-oxide used
to remove one of the carbonyls was purified to avoid incorpora-
tion of protium into the methyl group of the product, by first
recrystallizing commercial anhydrous trimethylamine N-oxide
from D2O and then subliming it in vacuo.
Proton NMR spectra were recorded on Varian Unity 400,
500, or 600 spectrometers or on a Bruker AM400 spectrometer.
Carbon-13 NMR spectra were recorded on Varian Unity 400
or Bruker AM300 spectrometers. Phosphorus-31 and fluorine-
19 NMR spectra were recorded on Varian Unity 400 or Bruker
AM400 spectrometers. Two-dimensional NMR experiments
(COSY, ROESY, TOCSY, and 13C-1H HMQC) were obtained
on Varian Unity 400 or 500 spectrometers. NMR spectral data
for all compounds are given in Table 1, whereas IR data, when
available, are given with the details of preparation.
(4) See for example: (a) George, D. S. A.; Hilts, R. W.; McDonald,
R.; Cowie, M. Organometallics 1999, 18, 5330. (b) George, D. S. A.;
Hilts, R. W.; McDonald, R.; Cowie, M. Inorg. Chim. Acta 2000, 300-
302, 353. (c) Trepanier, S. J.; McDonald, R.; Cowie, M. Organometallics
2003, 22, 2638. (d) Ristic-Petrovic, D.; Anderson, D. J.; Torkelson, J.
R.; McDonald, R.; Cowie, M. Organometallics 2003, 22, 4647. (e)
Rowsell, B. D.; McDonald, R.; Cowie, M. Organometallics 2004, 23,
3873.
(5) See for example: (a) Johnson, K. A.; Vashan, M. D.; Moassen,
B.; Warmka, B. K.; Gladfelter, W. L. Organometallics 1995, 14, 461.
(b) Oldham, S. M.; Houlis, J. F.; Sleigh, C. J.; Duckett, S. B.; Eisenberg,
R. Organometallics 2000, 19, 2985. (c) Gao, Y.; Hennings, M. C.;
Puddephatt, R. J. Organometallics 2001, 20, 1882. (d) Veige, A. S.;
Gray, T. G.; Nocera, D. G. Inorg. Chem. 2005, 44, 17. (e) Rashidi, M.;
Kamali, K.; Jennings, M. C.; Puddephatt, R. J. J. Organomet. Chem.
2005, 690, 1600. (f) Alvarez, M. A.; Anaya, Y.; Garc´ıa, M. E.; Ruiz, M.
A.; Vaissermann, J. Organometallics 2005, 24, 2452.
(6) Torkelson, J. R.; Antwi-Nsiah, F. H.; McDonald, R.; Cowie, M.;
Pruis, J. G.; Jalkanen, K. J.; DeKock, R. L. J. Am. Chem. Soc. 1999,
121, 3666.
(7) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA. 1987; Chapter 3.
(8) Torkelson, J. R.; McDonald, R.; Cowie, M. J. Am. Chem. Soc.
1998, 120, 4047.
(9) Torkelson, J. R.; McDonald, R.; Cowie, M. Organometallics 1999,
18, 4134.
(10) Ristic-Petrovic, D.; Torkelson, J. R.; Hilts, R. W.; McDonald,
R.; Cowie, M. Organometallics 2000, 19, 4432.
(11) (a) Howard, T. R.; Lee, J. B.; Grubbs, R. H. J. Am. Chem. Soc.
1980, 102, 6876. (b) Lee, J. B.; Gajola, G. J.; Schaefer, W. P.; Howard,
T. R.; Ikariya, T.; Straus, D. A.; Grubbs, R. H.; J. Am. Chem. Soc. 1981,
103, 7358. (c) Motyl, K. M.; Norton, J. R.; Schauer, C. K.; Anderson,
O. P. J. Am. Chem. Soc. 1982, 104, 7325. (d) Sumner, C. E., Jr.; Riley,
P. E.; David, R. E.; Pettit, R. J. Am. Chem. Soc. 1980, 102, 1752. (e)
Theopold, K. H.; Bergman, R. G. J. Am. Chem. Soc. 1981, 103, 2489.
(f) Kao, S. C.; Thiel, C. H.; Pettit, R. Organometallics 1983, 2, 914.
1,1-Difluoroallene was prepared using a modification of a
published procedure,14 in which a solution of 20 µL of 2-bromo-
3,3,3-trifluoro-1-propene in 2 mL of Et2O in a round-bottom
flask was cooled to -90 °C in a heptane/liquid N2 bath. To
this was added dropwise 100 µL of a 1.6 M solution of
n-butyllithium in hexanes, followed by stirring of the solution
at this temperature for 10 min. The -90 °C bath was replaced
(12) (a) Dyke, A. F.; Knox, S. A. R.; Naish, P. J.; Taylor, G. E. J.
Chem. Soc., Chem. Commun. 1980, 803. (b) Gracey, B. P.; Knox, S. A.
R.; Macpherson, K. A.; Orpen, A. G.; Stobart, S. R. J. Chem. Soc.,
Dalton Trans. 1985, 1935. (c) Sumner, C. E.; Collier, J. A.; Pettit, R.
Organometallics 1982, 1, 1350. (d) Akita, M.; Hua, R.; Nakanishi, S.;
Tanaka, M.; Moro-oka, Y. Organometallics 1987, 16, 5572. (e) Adams,
P. Q.; Davis, D. L.; Dyke, A. F.; Knox, S. A. R.; Mead, K. A.; Woodward,
P. J. Chem. Soc., Chem. Commun. 1983, 222. (f) Colborn, R. E.; Dyke,
A. F.; Knox, S. A. R.; Macpherson, K. A.; Orpen, A. G. J. Organomet.
Chem. 1982, 239, C15. (g) Colborn, r. E.; Davies, D. L.; Dyke, A. F.;
Knox, S. A. R.; Mead, K. A.; Orpen, A. G.; Guerchais, J. E.; Roue, J. J.
Chem. Soc., Dalton Trans. 1989, 1799. (h) Dennett, J. N. L. Ph.D.
Thesis, Chapter 4, University of Bristol, Bristol, UK, 2000. (i) Kaneko,
Y.; Suzuki, T.; Isobe, K.; Maitlis, P. M. J. Organomet. Chem. 1998,
554, 155. (j) Navarre, D.; Parlier, A.; Rudler, H.; Daran, J. C. J.
Organomet. Chem. 1987, 322, 103. (k) Levisalles, J.; Rose-Munch, F.;
Rudler, H.; Daran, J. C.; Dranzee, Y.; Jeannin, Y. J. Chem. Soc., Chem.
Commun. 1981, 152.
(13) (a) Fildes, M. J.; Knox, S. A. R.; Orpen, A. G.; Turner, M. L.;
Yates, M. I. J. Chem. Soc., Chem. Commun. 1989, 1680. (b) Chetcuti,
M. J.; Fanwick, P. E.; Grant, B. E. Organometallics 1991, 10, 3003.
(c) Chokshi, A.; Rowsell, B. D.; Trepanier, S. J.; Ferguson, M. J.; Cowie,
M. Organometallics 2004, 23, 4759.
(14) Drakesmith, F. G.; Stewart, O. J.; Tarrant, P. J. Org. Chem.
1967, 33, 280.

Binuclear Activation of Cumulenes
Organometallics, Vol. 24, No. 15, 2005 3713

3714 Organometallics, Vol. 24, No. 15, 2005
Ristic-Petrovic et al.
by a chlorobenzene/liquid N2 bath (ca. -40 °C), resulting in
the evolution of 1,1-difluoroallene gas.
those of the authentic sample prepared by the reaction of
compound 1 with 2-butyne.9 Isomer 7b was identified by the
close similarity of its spectral parameters with 7a.
Preparation of Compounds. (a) [Ir2(CH3)(CO)2(µ-η1:η3-
H2CdCdCH2)(dppm)2][CF3SO3] (2e). A solution of com-
pound 1 (50 mg, 0.036 mmol) in 5 mL of CH2Cl2 was cooled to
-78 °C, and allene was passed through the solution for 1 min,
causing an immediate color change from red to light yellow.
The solution was stirred for 30 min at -78 °C and then
warmed to room temperature and stirred for 1 h. The solvent
was reduced to ca. 2 mL, and a yellow powder was precipitated
with addition of 10 mL of Et2O. The isolated solid was washed
with Et2O (2 × 10 mL) and dried under vacuum. Yield: 73%.
Anal. Calcd for Ir2SP4F3O5C57H51: C, 48.43; H, 3.64. Found:
C, 48.23; H, 3.50. HRMS m/z calcd for Ir2P4O2C56H51: 1265.2098.
Found: 1265.2108. IR: ν(CO) ) 1998(s).
Low-Temperature Reactions. In a typical experiment,
in which allene, methylallene, and 1,1-dimethylallene were
reacted with compound 1, 1-3 equiv of the gaseous olefin was
added slowly by means of a gastight syringe to a solution of
30 mg (0.02 mmol) of 1 in 0.7 mL of CD2Cl2 in an NMR tube
cooled to -78 °C in a solid CO2/acetone bath. One-dimensional
NMR spectra (1H, 31P) were recorded from -80 °C to ambient
temperature at 10 °C intervals. As discussed in the Results
section, two-dimensional NMR studies (1H-COSY, TOCSY,
ROESY, and 13C-1H HMQC) were carried out for any new
compounds observed at the appropriate temperatures. The
results of these studies are summarized in Table 1. The details
of these low-temperature reactions are given below for each
cumulene studied.
(g) Allene. Addition of allene to a red solution of 1 at -78
°C resulted in a color change to yellow. Monitoring the 1H and
31P NMR spectra at -80 °C showed the presence of two major
species, 2a and 2b, along with minor amounts of unidentified
products and approximately 5% of a third isomer, 2c. Warming
the sample resulted in an increase in the amount of 2c at the
expense of 2a and 2b, which disappeared together. By -30
°C 2c was the only significant product observed, although
small amounts of new species 2d and 2e were becoming
evident. Upon further warming, 2c disappeared while 2d and
2e increased, such that by 0 °C all three were present in
comparable proportions. By 10 °C only 2d and 2e remained,
and maintaining this temperature for 1-2 h resulted in the
disappearance of 2d, leaving only 2e. Spectral parameters for
2a and 2b were obtained at -80 °C, those for 2c were obtained
at -30 °C, and those for 2d were obtained at 10 °C.
(h) Methylallene. Addition of methylallene at -78 °C
followed by immediate monitoring of the 31P NMR spectrum
at -80 °C showed the presence of 3a and 3b in a 1:1 ratio
together with trace amounts of 3c. Varying amounts of 3c
could be observed depending upon our level of success at
maintaining the temperature at -78 °C during cumulene
addition. Warming resulted in a disappearance of 3a and 3b
while the concentration of 3c increased. By -40 °C 3c was
the only species present. Raising the temperature resulted in
the appearance of 3d and 3e. By 0 °C 3c has disappeared,
leaving only 3d and 3e, and by 20 °C 3d had disappeared,
leaving only 3e. The NMR spectral parameters for 3a and 3b
were obtained at -80 °C, those for 3c were obtained at -40
°C, those for 3d at -10 °C, and those for 3e at ambient
temperature.
(i) 1,1-Dimethylallene. Addition of 1,1-dimethylallene at
-78 °C and immediate monitoring by 31P NMR at -80 °C
showed 4a as the only product. Warming the sample resulted
in the appearance of 4d at approximately -20 °C, and by 20
°C this was the only product. No additional species were
observed.
(b) [Ir2(CH3)(CO)2(µ-η1:η3-H2CdCdC(H)CH3)(dppm)2]-
[CF3SO3] (3e). A solution of compound 1 (50 mg, 0.036 mmol)
in 5 mL of CH2Cl2 was cooled to 0 °C, and methylallene was
then passed through the solution for 1 min, causing an
immediate color change from red to light yellow. The solution
was stirred for 30 min at 0 °C and then warmed to room
temperature and stirred for 1 h, during which time the solution
changed to bright yellow. The solvent was reduced to ca. 2 mL,
and a bright yellow powder was precipitated with addition of
10 mL of Et2O. The isolated solid was washed with Et2O (2 ×
10 mL) and dried under vacuum. Yield: 63%. NMR charac-
terization of 3e showed it to contain approximately 10%
unidentified impurities from which the desired product could
not be separated. HRMS m/z calcd for Ir2P4O2C57H53: 1279.2255.
Found: 1279.2254. IR: ν(CO) ) 1994(s).
(c) [Ir2(CH3)(CO)2(η2-H2CdCdC(CH3)2)(dppm)2][CF3SO3]
(4d). Dimethylallene (3.2 µL, 0.033 mmol) was added to a
solution of compound 1 (45 mg, 0.033 mmol) in 5 mL of CH2-
Cl2 at -78 °C, and the solution was stirred for 10 min, during
which time the solution changed to bright yellow. The solution
was warmed to room temperature, the solvent was reduced to
ca. 2 mL, and a bright yellow powder was precipitated with
addition of 10 mL of Et2O. The isolated solid was washed with
Et2O (2 × 10 mL) and dried under vacuum. Yield: 55%. HRMS
m/z calcd for Ir2P4O2C58H55: 1293.2411. Found: 1293.2415.
IR: ν(CO) ) 1793(m), 1963(br).
(d) [Ir2(CH3)(CO)2(µ-η1:η1-H2CdCdCF2)(dppm)2][CF3SO3]
(5c). The 1,1-difluoroallene gas that was evolved in the
reaction described earlier was passed into a solution of 50 mg
(0.036 mmol) of compound 1 in 10 mL of CH2Cl2 at -78 °C,
resulting in the solution color changing to magenta. The
cooling bath was removed, causing the solution to slowly turn
orange as it warmed. The solution volume was reduced to ca.
2 mL, and addition of 10 mL of Et2O resulted in the precipita-
tion of a brick-red solid, which was dried in vacuo overnight,
yielding 39 mg (83%) of the compound. HRMS m/z calcd for
Ir2P4F2O2C56H49: 1301.1910. Found: 1301.1918. IR: ν(CO) )
1992(m), 1949(w).
(e) [Ir2(H)(CO)2(µ-η1:η3-HCC(Me)dCH2)(dppm)2][CF3SO3]
(6). Compound 2e (20 mg, 0.014 mmol) was dissolved in 10
mL of CH2Cl2 and stirred for 7 days. The solvent volume was
reduced to ca. 2 mL, and a yellow powder was obtained on
addition of 10 mL of Et2O. The solid was isolated, washed with
Et2O (2 × 10 mL), and dried under vacuum. NMR spectra
showed that yields of 6 were variable, with the highest
observed being 80%, among a mixture of other related un-
characterized products. Compound 6 was characterized by
comparison of its 1H and 31P NMR spectra with those of closely
related compounds that had been previously characterized.9
(f) [Ir2(H)(CO)2(µ-η1:η3-HCC(Me)dCHMe)(dppm)2][CF3-
SO3] (7a/7b). Leaving a solution of 3e for approximately 5
days resulted in 50% conversion to a mixture of isomers 7a
and 7b in an approximate 2:1 ratio, together with minor
amounts of 1 and unidentified decomposition products. These
isomers were never obtained as pure species. Isomer 7a was
identified by comparison of its NMR spectral parameters with
(j) 1,1-Difluoroallene. In a typical experiment 0.5-2.0
equiv of 1,1-difluoroallene, generated as described above at
-40 °C, was passed into the solution of 30 mg (0.02 mmol) of
1 in 0.7 mL of CD2Cl2 in an NMR tube at -78 °C. Diethyl
ether-d10 was used as solvent in the generation of the 1,1-
difluoroallene so that the ether carried over in the difluoro-
1
allene stream did not interfere with the H NMR spectrum.
One-dimensional NMR spectra (1H, 13C, 19F, 31P) were recorded
from -80 °C to ambient temperature at 10 °C intervals. As
discussed in the Results section, two-dimensional NMR studies
(1H-COSY and ROESY) were carried out at particular tem-
peratures. The results of these studies are summarized in
Table 1. At -80 °C, the NMR studies showed the presence of
only one product (5a), the characterization of which is dis-
cussed later. At temperatures above -20 °C compound 5a was
replaced by 5c. If 1,1-difluoroallene was present in excess,
varying amounts of other unidentified products were obtained,

Binuclear Activation of Cumulenes
Organometallics, Vol. 24, No. 15, 2005 3715
so reactions were best carried out with a slight excess of 1,
leading to some unreacted 1 in the final solution.
Gaussian integration. See Table 2 for a summary of X-ray data
collection information.
Structure Solution and Refinement. Both structures
were solved using the direct methods program SHELXS-86,16
and refinement was completed using SHELXL-93.17 Hydrogen
atoms were assigned positions based on the geometries of the
attached carbon atoms and were assigned thermal parameters
20% greater than those of the attached carbons. For compound
2e there are two unique molecules per asymmetric unit,
although the parameters for both agree very well. In compound
5c, one of the CH2Cl2 molecules of crystallization was disor-
dered over two positions of half occupancy in the unit cell.
Further details of structure refinement and final residual
indices are found in Table 2.
Low-Temperature Ligand Exchange. In a typical ex-
periment, 1 mL of gaseous cumulene, delivered via a gastight
syringe, was reacted with 30 mg of compound 1 in 0.7 mL of
CD2Cl2 in an NMR tube that had been cooled to -78 °C. The
reaction was taken to completion at this temperature, ensuring
that no compound 1 remained, as established by 31P{1H} and
1H NMR spectroscopy. At this point, 1 mL of a different
cumulene was added by syringe while maintaining the tem-
perature at -78 °C. Upon returning the sample to the
spectrometer at -80 °C, monitoring was continued while
raising the temperature in 20 °C intervals.
(k) Methylallene Addition to Compounds 2a and 2b.
To a mixture of 2a and 2b generated at -78 °C as noted above
was added 1 mL of methylallene. Monitoring the 31P{1H} NMR
spectrum at this temperature showed the immediate appear-
ance of small amounts of 3a/3b together with 2a/2b as the
dominant species. Warming and subsequent mixing led to
mixtures of 2c/2d/2e and 3c/3d/3e in which the 3:2 ratio slowly
increased with time. While the proportions of isomers varied
throughout the different temperatures, as explained earlier,
the final proportions of compounds 2 and 3 were comparable.
Leaving the sample at ambient temperature for several hours
resulted in the formation of only 2e and 3e in comparable
proportion.
(l) Allene Addition to Compounds 3a and 3b. To a
mixture of 3a and 3b at -78 °C was added 1 mL of allene,
leading to the immediate appearance of 2a and 2b together
with 3a and 3b. Warming gave rise to mixtures of the isomers
of compounds 2 and 3 as described in part (k).
(m) Methylallene Addition to Compound 2c. The iso-
meric mix of 2a and 2b was warmed to -20 °C until the major
species present was 2c. Addition of methylallene at this
temperature resulted in the immediate appearance of 3c, and
warming led to mixtures of 2d, 3d, 2e, and 3e, which after
several hours had converted to approximately equal amounts
of 2e and 3e.
(n) Addition of Allene to Compound 4a. To a solution
of 4a, generated at -78 °C as described above, was added 1
mL of allene. No additional species were observed until the
sample was warmed to -60 °C, at which point 2c began to
appear. As the temperature was raised, the proportion of 4a
slowly diminished while that of 2c increased. At -10 °C 2d
and 2e began to appear together with 4a, and raising the
temperature to ambient yielded mainly 2e with small amounts
(ca. 10%) of 4d.
(o) Allene Addition to Compound 5a. Addition of 1 mL
of allene to a solution of 5a at -78 °C, generated as described
above, showed no spectroscopic evidence for formation of any
product other than 5a. Upon warming to ambient temperature,
only 5a and 5c were observed at the appropriate temperatures.
X-ray Data Collection. X-ray quality crystals of 2e were
grown by slow diffusion of diethyl ether into a concentrated
1:1 CH2Cl2/toluene solution of the compound. Crystals of 5c
were grown from ether/CH2Cl2. Data for 2e were collected at
-60 °C on a Siemens P4/RA SMART 1K CCD diffractometer,
while for 5c data were collected on a Bruker PLATFORM/
SMART 1000 CCD diffractometer at -80 °C.15 Unit cell pa-
rameters were obtained from a least-squares refinement of
8192 and 6256 reflections for the respective compounds. For
both compounds, the cell parameters and the diffraction
symmetry established the space groups as P1 or P1h, the latter
of which was established as correct by the successful refine-
ment of both structures. Absorption corrections were applied
to 2e by the SADABS method, supplied through the Bruker
software, while for 5c the crystal faces were indexed and
measured, with absorption corrections carried out using
Results and Compound Characterization
(a) Cumulene Adducts. The binuclear methyl com-
plex [Ir2(CH3)(CO)(µ-CO)(dppm)2][CF3SO3] (1) reacts
readily with allene, methylallene, 1,1-dimethylallene,
and 1,1-difluoroallene at -78 °C, but upon warming, the
initial products undergo a number of subsequent rear-
rangements. Each of the intermediates in these trans-
formations has been characterized at the appropriate
temperature by NMR techniques, while species that
could be isolated have also been characterized by
additional methods, as is described in what follows.
(i) Allene. At -78 °C the reaction of 1 with allene
yields two products (2a and 2b) in approximately equal
proportions as outlined in Scheme 2. The 31P{1H} NMR
spectrum of this mixture shows three resonances at δ
-9.0, -6.7, and -1.6, in an approximate 1:1:2 intensity
ratio. Whereas the two high-field signals each appear
as one-half of an AA′BB′ pattern from two separate
species, the low-field signal of double intensity is
complex and corresponds to two coincidentally overlap-
ping signals from the other halves of the above patterns.
1
The H NMR spectrum at this stage of the reaction is
complicated owing to the overlapping, broad resonances
of both species. Nevertheless, two broad signals are
identified at δ -12.25 and 4.99, corresponding to hy-
dride and methylene groups, respectively. Although
selective 31P decoupling experiments fail to give useful
resolution of the overlapping methylene signals, these
experiments clearly resolve two different, but closely
spaced hydride resonances (δ -12.20, -12.30) for 2a and
2b. The 13C{1H} NMR spectrum of a 13CO-enriched
sample shows four resonances of comparable intensity
at δ 190.3, 189.3, 176.1, and 174.8, consistent with the
presence of two dicarbonyl complexes. We were unable
to identify which pairs of carbonyl resonances belonged
together, although we assume that the spectrum for
each compound is very similar, giving rise to one high-
and one low-field resonance in each case; certainly, the
two high-field signals are very close, as are the two low-
field ones.
The presence of methylene and hydride fragments in
these products is reminiscent of our previous work in
which ligand addition to one metal in 1 resulted in C-H
bond cleavage of the methyl ligand by the adjacent
metal, as diagrammed in Scheme 1.4d,6 We therefore
propose that addition of allene has resulted in a similar
C-H activation process yielding two isomers, [Ir2(H)-
(16) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
(17) Sheldrick, G. M. SHELXL-93, Program for structure determi-
nation; University of Go¨ttingen: Germany, 1993.
(15) Programs for diffractometer operation, data reduction, and
absorption correction were those supplied by Bruker or Siemens.

3716 Organometallics, Vol. 24, No. 15, 2005
Table 2. Crystallographic Data for Compounds 2e and 5c
Ristic-Petrovic et al.
[Ir2(CH3)(CO)2(µ-η1:η3-H2CdCdCH2)-
(dppm)2][CF3SO3] (2e)‚1/3CH2Cl2
[Ir2(CH3(CO)(µ-H2CdCdCF2)-
(dppm)2][CF3SO3](5c)‚2CH2Cl2
A. Crystal Data
formula
fw
cryst dimens (mm)
color
C
57.33H51.67Cl0.67F3Ir2O5P4S
C59H53Cl4F5Ir2O5P4S
1619.15
1441.65
0.24 × 0.14 × 0.03
0.26 × 0.24 × 0.10
yellow
orange
cryst syst
space group
unit cell paramsa,b
a (Å)
triclinic
triclinic
P1h (No. 2)
P1h (No. 2)
10.4603(1)
14.2406(1)
38.7838(4)
91.887(1)
97.732(1)
90.281(1)
5721.39(9)
4
12.327(1)
17.297(2)
17.821(2)
63.059(2)
76.554(2)
70.189(2)
3172.8(6)
2
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
F
cacd (g cm-3
)
1.674
4.882
1.695
4.552
µ (mm-1
)
B. Data Collection and Refinement Conditions
diffractometer
radiation (λ [Å])
temp (°C)
scan type
data collection 2θ limit (deg)
total data collected
Siemens P4/RA/SMART 1K CCDc
Bruker PLATFORM/SMART 1000 CCDd
graphite-monochromated Mo KR (0.71073)
-60
-80
φ rotations (0.3°) and ω scans (0.3°) (10 s exposures)
55.00
63 233 (-13 e h e 13, -18 e k e 18,
-50 e l e 50)
ω scans (0.2°) (25 s exposures)
52.90
17 665 (-15 e h e 15, -19 e k e 21,
-22 e l e 21)
no. of indep reflns
23 466 (Rint ) 0.1958)
10 928 (Fo2 g 2σ(Fo2))
0.6468-0.4616
12 736 (Rint ) 0.0416)
8785 (Fo2 g 2σ(Fo2))
0.6589-0.3840
no. of obsd refns (NO)
range of transm factors
no. of data/restraints/params
extinction coeff (ø)e
23466 [Fo2 g -3σ(Fo2)]/0/1312
0.00027(3)
12 736[Fo2 g -3σ(Fo2)]/0/748
largest diff peak and hole (e Å-3
final R indicesf
)
1.521 and -1.149
4.735 and -2.150
R1 (Fo2 g 2σ(Fo2))
0.0867
0.1853
1.034
0.0549
0.1447
1.026
wR2 (Fo2 g -3σ(Fo2))
goodness of fit (S)g
a For 2e, obtained from least-squares refinement of 8192 centered reflections. b For 5b, obtained from least-squares refinement of 6256
centered reflections. c Programs for diffractometer operation, data collection, data reduction, and absorption correction were those supplied
by Siemens. d Programs for diffractometer operation, data collection, data reduction, and absorption corrections were those supplied by
Bruker. e Fc*
)
kFc[1
+
x{0.001Fc λ3/sin(2θ)}]-1/4 where
k
is the overall scale factor. f R1
)
∑||Fo|
-
|Fc||/∑|Fo|; wR2
)
2
2
2
2
2
[∑w(Fo - Fc )2/∑w(Fo4)]1/2
.
g S ) [∑w(Fo - Fc )2/(n - p)]1/2 (n ) number of data; p ) number of parameters varied; w )
2
[σ2(Fo2) + (aoP)2 + a1P]-1 where P ) [Max(Fo2, 0) + 2Fc ]/3). For 2e ao ) 0.0456, a1 ) 22.8603; for 5b ao ) 0.0766, a1 ) 1.4138.
(η2-C3H4)(CO)2(µ-CH2)(dppm)2][CF3SO3] (2a, 2b), which
differ only in the orientation of the η2-allene fragment
with respect to the bridging methylene group. This
interpretation is consistent with the very close spectral
similarities in these two products, which additionally
bear a close resemblance to the ethylene adduct [Ir2H-
(η2-C2H4)(CO)2(µ-CH2)(dppm)2][CF3SO3], reported else-
where.4d Owing to the overlap of most 1H NMR signals
of these isomers, appropriate experiments to determine
which isomer was which could not be carried out.
Upon warming this sample, an additional species (2c),
which at -78 °C is barely discernible in the 31P{1H}
NMR spectrum, grows in intensity at the expense of 2a
and 2b, and by -30 °C this product accounts for 90% of
the integrated intensity. The 31P{1H} NMR pattern for
2c at this temperature is that of an AA′BB′ spin system
with resonances at δ 25.3 and -5.5. This new product
displays no hydride resonance in the 1H NMR spectrum,
and the only methylene signals observed are those
corresponding to the dppm ligands (δ 3.62, 3.44) and
the allene ligand (vide infra). Instead, this spectrum
displays a triplet signal at δ 0.18, corresponding to an
iridium-bound methyl group. The allene protons appear
in a 1:1:2 intensity ratio, respectively, as two singlets,
at δ 6.06 and 4.01 and a triplet displaying coupling to
two adjacent 31P nuclei at δ 2.46. To ascertain the
bonding mode of the allene ligand, a 1H-ROESY experi-
ment was carried out at -30 °C. NOE contacts were
observed between the high-field allene protons and one
set of dppm methylene protons (δ 3.62), while one of
the protons (δ 4.01) at the other end of the allene ligand
showed a contact with the iridium-bound methyl group.
The remaining allene proton (δ 6.06) showed a strong
NOE with its geminal partner, but with no other proton.
These results suggest the bridging arrangement for the
allene ligand as shown in Scheme 2, in which one allene
proton on the uncoordinated allene double bond is in
close proximity to the iridium-bound methyl ligand,
while the other is aimed away. The high-field signal that
displays 31P coupling clearly corresponds to the allene
methylene group that is bound to one metal. These
protons lie above and below the Ir2-allene plane and
could come into close contact with one proton on each
of the dppm methylene groups. This bridged-allene
binding mode has been proposed in the thermally
unstable Os2(CO)8(µ-η1:η1-H2CdCdCH2),18 and such
bridged binding modes are common for alkynes9,19 and
also known for olefins.4d,20 It is also the coordination
(18) Spetseris, N.; Norton, J. R.; Hunter, D. R. Organometallics
1995, 14, 603.

Binuclear Activation of Cumulenes
Organometallics, Vol. 24, No. 15, 2005 3717
Scheme 2
Figure 1. Perspective view of one of the two crystallo-
graphically independent [Ir2(CH3)(CO)2(µ-η1:η3-H2CdCd
CH2)(dppm)2]+ cations of compound 2e. Thermal ellipsoids
are shown at the 20% probability level except for hydro-
gens, which are shown artificially small. Only the ipso
carbons of the dppm phenyl rings are shown.
which is discussed in detail later in the text, that 2d
has one terminal and one bridging carbonyl, which
interchange rapidly on the NMR time scale. An alter-
nate structure in which the positions of the methyl
group and the terminal carbonyl are interchanged can
be ruled out on the basis that the low-field carbonyl
resonances are indicative of carbonyls that lie in the
vicinity of an adjacent metal rather than being remote
from this metal.21
Crystals of compound 2e were obtained, and an X-ray
investigation established the unusual structure of this
product, as diagrammed in Figure 1. This fifth isomer,
[Ir2(CH3)(CO)2(µ-η1,η3-C3H4)(dppm)2][CF3SO3] (2e), has
the allene ligand η1 bound via the central carbon to one
metal, while being η3 bound to the second metal. Two
crystallographically distinct molecules are present in the
asymmetric unit cell, and these are identified as mol-
ecules A and B in Table 3. The geometry at Ir(1) is
distorted octahedral, in which the two phosphines are
mutually trans, the carbonyl is essentially opposite the
central carbon of the allene moiety, and the methyl
ligand is almost opposite the metal-metal bond. At Ir(2)
the geometry can also be viewed as distorted octahedral,
however with a cis arrangement of phosphines in which
each phosphorus is opposite the ends of the η3-allyl unit.
At this metal, the carbonyl is opposite the metal-metal
bond. The bonding parameters associated with the
bridging allene ligand are essentially as expected. The
Ir(1)-C(4) distance (2.09(2), 2.10(2)Å for the two mol-
ecules) is somewhat shorter than the Ir-CH3 distance
(2.19(2), 2.20(2)Å), a difference that is consistent with
the hybridization differences of the bound carbons (sp2
vs sp3, respectively), and the distances from Ir(2) to the
three allylic carbons (2.12-2.21 Å) are, as expected,
longer than the Ir(1)-C(4)) σ bond. Within the bridging
allene group, the distances (1.41-1.45 Å) and the C(3)-
C(4)-C(5) angle (108(2)°, 106(2)°) are typical of η3-bound
allyl moieties.22,23 Although not common, similar bridg-
ing allene coordination has previously been observed.24
In solution the NMR spectral results are consistent
with the above geometry being maintained. The 13C-
mode observed, apart from the orientation of the cu-
mulene ligand, in the 1,1-difluoroallene adduct [Ir2-
(CH3)(CO)2(µ-η1:η1-H2CdCdCF2)(dppm)2][CF3SO3] (5c)
(vide infra). In the 13C NMR spectrum of 2c the expected
two carbonyl resonances are observed at δ 192.4 and
186.2, consistent with terminal carbonyl binding.
At approximately -10 °C two additional products, 2d
and 2e, begin to appear in the 31P{1H} NMR spectrum,
and their appearance corresponds to the disappearance
of the resonance for 2c. Compound 2d appears as two
resonances at δ 19.5 and 2.5, with patterns typical of
an AA′BB′ spin system, while 2e appears as a single
complex multiplet at δ -15.7. At 10 °C compound 2c
has disappeared, and maintaining this temperature
results in the disappearance of 2d over a 2 h period,
1
leaving only 2e. The H NMR spectrum of 2d shows a
methyl resonance as a triplet at δ 1.23 with coupling to
the adjacent pair of 31P nuclei, and three allene reso-
nances at δ 5.61, 4.81, and 4.24 in a 1:1:2 ratio. Two
closely spaced carbonyl resonances are observed in the
13C NMR spectrum at δ 206.6 and 206.2. Although these
13C chemical shifts suggest the presence of two semibridg-
ing carbonyls having somewhat different environments,
we propose, based on analogies with the ethylene adduct
[Ir2(CH3)(η2-C2H4)(CO)2(dppm)2][CF3SO3]4d and with the
dimethylallene adduct (4d), the characterization of
(19) See for example: (a) Gagne´, M. R.; Takats, J. Organometallics
1988, 7, 561. (b) Dickson, R. S. Polyhedron 1991, 10, 1995. (c) Adams,
R. D.; Huang, M. Organometallics 1995, 14, 2887. (d) George, D. S.
A., McDonald, R.; Cowie, M. Can. J. Chem. 1996, 74, 2289.
(20) See for example: (a) Grevels, F.-W.; Klotzbu¨cher, W. E.; Siels,
F.; Schaffner, K.; Takats, J. J. Am. Chem. Soc. 1990, 112, 1995. (b)
Bender, B. R.; Ramage, D. L.; Norton, J. R.; Wiser, D. C.; Rappe´, A. K.
J. Am. Chem. Soc. 1997, 119, 5628.
(21) George, D. S. A.; McDonald, R.; Cowie, M. Organometallics
1998, 17, 2553.
(22) Hay, C. M. Horton, A. D.; Mays, M. J.; Raithby, P. R. Polyhedron
1988, 7, 987.
(23) Orpen, A. G.; Brammer, L.; Allen, F. H.; Kennard, O.; Watson,
D. G.; Taylor, R. J. Chem. Soc., Dalton Trans. 1989, 51.

3718 Organometallics, Vol. 24, No. 15, 2005
Ristic-Petrovic et al.
Table 3. Selected Distances and Angles for
Compound 2e
for 2e would be expected to give rise to two resonances
in the 31P{1H} NMR spectrum, corresponding to the two
chemically inequivalent ends of the dppm ligands, only
one complex multiplet is observed, resulting from ac-
cidental overlap of the two expected signals.
(a) Distances (Å)
molecule A
distance
molecule B
distance
atom 1
atom 2
(ii) Methylallene. The reaction of 1 with methylal-
lene very much parallels that of allene, as described
above. At -78 °C the reaction mixture contains two
major products in approximately equal proportions, as
shown by the 31P{1H} NMR spectrum, together with a
third minor species. The two major species, 3a and 3b,
having 31P{1H} resonances at δ -0.4 and -6.9 and at δ
-2.6 and -8.0 are also hydrides, with resonances in the
1H NMR spectrum at δ -12.0 and -12.6, respectively.
Unfortunately, additional features could not be estab-
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(2)
Ir(2)
Ir(2)
Ir(2)
Ir(2)
Ir(2)
O(1)
O(2)
C(3)
C(4)
Ir(2)
P(1)
P(3)
C(1)
C(4)
C(6)
P(2)
P(4)
C(2)
C(3)
C(4)
C(5)
C(1)
C(2)
C(4)
C(5)
2.8438(10)
2.365(5)
2.371(5)
1.90(2)
2.09(2)
2.19(2)
2.317(4)
2.329(5)
1.88(2)
2.20(2)
2.12(2)
2.21(2)
1.13(2)
1.17(2)
1.41(2)
1.44(2)
2.8353(9)
2.361(5)
2.335(5)
1.91(2)
2.10(2)
2.20(2)
2.341(5)
2.336(5)
1.90(2)
2.21(2)
2.13(2)
2.21(2)
1.14(2)
1.12(2)
1.45(2)
1.44(2)
1
lished from the H NMR spectrum owing to overlap,
particularly in the methylene region. However, each
species 3a and 3b also displays a pair of resonances in
the 13C NMR spectrum at approximately δ 190 and 176,
corresponding to terminal carbonyls. The close spectral
similarities of these products and their resemblance to
2a and 2b suggest the isomers shown in Scheme 2,
having the formulation [Ir2H(η2-H2CdCdC(H)Me)(CO)2-
(µ-CH2)(dppm)2][CF3SO3], in which η2 coordination of
methylallene at one metal has resulted in C-H activa-
tion of the methyl ligand at the adjacent metal. Again,
we were unable to determine which isomer corresponds
to which set of NMR resonances.
(b) Angles (deg)
molecule A
molecule B
angle
atom 1
atom 2
atom 3
angle
Ir(2)
Ir(2)
Ir(2)
P(1)
C(1)
C(1)
C(4)
P(2)
Ir(1)
Ir(2)
Ir(2)
Ir(1)
Ir(1)
Ir(1)
Ir(2)
Ir(2)
C(3)
Ir(2)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(2)
C(1)
C(2)
C(3)
C(4)
C(4)
C(4)
C(4)
C(4)
C(4)
C(5)
C(1)
C(4)
C(6)
P(3)
C(4)
C(6)
C(6)
P(4)
O(1)
O(2)
C(4)
Ir(2)
C(3)
C(5)
C(3)
C(5)
C(5)
C(4)
121.1(5)
48.0(4)
148.6(4)
167.8(2)
169.1(7)
90.2(7)
100.6(6)
108.5(2)
180(1)
179(2)
68(1)
84.9(5)
119(1)
120(1)
117.9(5)
48.5(5)
151.6(5)
173.1(2)
166.4(7)
90.5(7)
103.1(7)
114.5(2)
178(2)
176(2)
67.6(9)
84.1(5)
121(1)
119(1)
74(1)
As the temperature is raised, the proportion of the
minor component (3c) increases at the expense of both
3a and 3b, and between -40 and -20 °C compound 3c
1
is the predominant species. In the H NMR spectrum
at -40 °C, compound 3c gives rise to a triplet corre-
sponding to an Ir-bound methyl ligand at δ 0.23
(3JPH ) 6.6 Hz) and the dppm methylenes at δ 3.45. In
this spectrum resonances for the methylallene ligand
appear at δ 1.76, 3.80, and 2.07 in a 3:1:2 ratio,
corresponding to the methyl substituent, the adjacent
unique hydrogen, and the geminal CH2 protons, respec-
tively. The resonance corresponding to the CH2 moiety
appears as a triplet which collapses to a singlet when
the low-field 31P resonance is irradiated. Decoupling the
high-field 31P resonance results in collapse of the
resonance for the methyl ligand to a singlet and at the
74.1(9)
73.8(9)
108(2)
73(1)
106(2)
68(1)
67.5(8)
{1H} NMR spectrum displays resonances at δ 179.2 and
171.2 for the terminal carbonyls, a multiplet at δ 61.3
for the terminal carbons of the allene group, and a
triplet at δ -31.1 for the methyl ligand. An INEPT
experiment shows the central allene carbon as a mul-
tiplet at δ 127.3, which is typical for the central carbon
of η3-allyl groups.25 The 1H NMR spectrum shows
signals for the allene protons at δ 4.23 and 3.52 arising
from the anti and syn protons, respectively, as estab-
lished by NOE experiments. An HMQC experiment
shows that both of these signals arise from protons on
two carbons that are chemically equivalent. The reso-
nance for the Ir-bound methyl group appears as a triplet
at δ 1.18 showing 5 Hz coupling to two adjacent
phosphorus nuclei. Although the structure established
1
same time results in sharpening of the H signals at δ
1.76 and 3.80. These selective 31P-decoupling experi-
ments establish that the methylallene ligand in 3c
bridges the metals, much as proposed for the allene-
bridged 2c. Support for this formulation was obtained
by a ROESY experiment in which NOE was observed
between the unique hydrogen of methylallene and the
Ir-bound methyl group but not between the two methyl
groups. NOE was also observed between the geminal
CH2 protons of the methylallene ligand and one pair of
dppm methylene protons.
(24) (a) Davis, R. E. Chem. Commun. 1968, 248. (b) Ku¨hn, A.;
Burschka, Ch.; Werner, H. Organometallics 1982, 1, 496. (c) Arce, A.
J.; DeSanctis, Y.; Deeming, A. J.; Hardcastle, K. I.; Lee, R. J.
Organomet. Chem. 1991, 406, 209. (d) Chetcuti, M. J.; Fanwick, P. E.;
McDonald, S. R.; Roth, N. N. Organometallics 1991, 10, 1551. (e)
Chacon, S. T.; Chisholm, M. H.; Folting, K.; Huffman, J. C.; Hampden-
Smith, M. J. Organometallics 1991, 10, 3722. (f) Seyferth, D.; Anderson,
L. L.; Davis, W. B.; Cowie, M. Organometallics 1992, 11, 3736. (g)
Chisholm, M. H.; Folting, K.; Lynn, M. A.; Streib, W. E.; Tiedtke, D.
B. Angew. Chem., Int. Ed. Engl. 1997, 36, 52. (h) Chisholm, M. H.;
Streib, W. E.; Tiedtke, D. B.; Wu, D.-D. Chem. Eur. J. 1998, 4, 1470.
(25) Chu, C. S.; Shin, S. Y.; Lee, C. J. Chem. Soc., Dalton Trans.
1992, 1323.
At temperatures above -20 °C two new species (3d
and 3e) begin to appear at approximately the same rate
until 3c disappears, after which 3d slowly disappears,
leaving only 3e, which increases by a corresponding
amount. Although many of the 1H resonances of 3d
overlap with those of 3e, the other spectral parameters
allow us to identify this intermediate with some confi-
dence. An iridium-bound methyl resonance appears as
1
a triplet at δ 1.25 in the H NMR spectrum, and no

Binuclear Activation of Cumulenes
Organometallics, Vol. 24, No. 15, 2005 3719
Scheme 3
hydride resonance is observed. In addition, the methyl
group of the methylallene ligand appears as a doublet
at δ 0.90 (presumably coupled to the hydrogen geminal
to it, although this resonance could not be observed to
confirm this assignment), shifted upfield from that
observed in 3c (δ 1.76), and the 13C{1H} NMR spectrum
shows a pair of carbonyls at δ 208.2 and 207.3. As noted
earlier for 2d, we propose a fluxional process that
rapidly interconverts the bridging and terminal carbon-
yls of 3d. We did not acquire IR data for this intermedi-
ate to confirm this assignment; however this has been
done for the closely comparable ethylene4d and 1,1-
dimethylallene (4d) adducts. Taken together, these data
suggest a structure analogous to these other olefin
adducts, in which the methylallene ligand is bound to
one metal opposite the Ir-Ir bond from the methyl
ligand.
The spectral parameters for 3e are consistent with
the structure shown for [Ir2(CH3)(CO)2(µ-η1:η3-CH2CCH-
(Me))(dppm)2][CF3SO3] in Scheme 2. Although the aver-
age 31P chemical shift for the diphosphine ligands is
closely comparable to that of 2e, the spectrum for 3e is
more complex in keeping with the ABCD spin system
that results from the lower symmetry of the complex
by virtue of the methyl substituent on one end of the
bridging cumulene ligand. Three complex 31P signals
appear in a 1:2:1 intensity ratio centered at δ -14.2,
-16.2, and -17.5; the two accidentally equivalent 31P
nuclei are presumably those adjacent to the methyl
ligand, remote from the inequivalent ends of the meth-
ylallene ligand. The loss of mirror symmetry bisecting
the methylallene ligand upon substitution by the methyl
group gives rise to four dppm methylene signals in the
1H NMR spectrum (δ 5.79, 5.38, 5.19, and 4.87). In this
spectrum the pair of geminal protons of the methylal-
lene ligand appear as multiplets at δ 5.17 and 3.78,
while the unique ligand proton appears as a quartet of
multiplets at δ 4.70. The allene methyl group appears
as a pseudo triplet (3JHH ) 6 Hz, 4JPH ) 6 Hz) at δ 1.07,
showing coupling to its adjacent hydrogen and to one
phosphorus nucleus (presumably the one that is in a
pseudo-trans position), and the Ir-bound methyl appears
as the usual triplet at δ 1.16. In the 13C{1H} NMR
spectrum (natural abundance) the pair of terminal
allene carbons appear as multiplets at δ 78.7 and 55.6,
the allene methyl group appears as a multiplet at δ 18.6,
and the Ir-bound methyl appears as a triplet at δ -32.2.
The location of the allylic methyl group in the endo site,
as diagrammed in Scheme 2, was confirmed by NOE
experiments in which NOE contacts were detected
between this methyl group and both the hydrogen
geminal to it and one of the hydrogens on the other end
of the allyl unit. NOE contacts between the pair of
geminal hydrogens were also observed.
resonances at δ 191.1 and 178.2. Compound 4a is
unambiguously identified as a methylene hydride spe-
cies, as shown in Scheme 3, by the hydride resonance
at δ -12.1 and the methylene resonance at δ 5.96 in
the 1H NMR spectrum. In this spectrum the dppm
methylene protons appear as their usual pattern at δ
5.16 and 3.48, while the methylene protons of the
cumulene ligand appear as a signal at δ 1.90. Two
distinct signals appear for the two methyl substituents
of the cumulene at δ 1.13 and 0.19. The appearance of
only one CH2 resonance and two methyl signals indi-
cates that the dimethylallene ligand is bound to the
metal by the less sterically encumbered end, since in
such a binding mode the dCH2 protons are chemically
equivalent, lying above and below the cumulene plane,
while the syn- and anti-methyl groups are inequivalent.
Warming a CH2Cl2 solution of 4a results in a trans-
formation to [Ir2(CH3)(η2-CH2dCdCMe2)(CO)2(dppm)2]-
[CF3SO3] (4d) with no other species being observed. This
product is analogous to the ethylene adduct previously
described4d and to compounds 2d and 3d. The spectral
parameters of these species are closely comparable. In
the 1H NMR spectrum the dppm methylene protons
appear at δ 3.48 and 3.19, the geminal protons and the
two methyl groups of the cumulene appear as a singlet
at δ 1.58 and as broad singlets at δ 1.22 and 0.98,
respectively, and the Ir-bound methyl group appears as
a triplet at δ 1.17. The appearance of one signal for the
pair of geminal protons and two signals for the methyl
groups again indicates that the dimethylallene group
is bound through the less sterically encumbered “H2Cd
C” end of the molecule. Two signals (broad triplets) are
seen for the carbonyls at δ 206.4 and 204.8, and selective
decoupling suggests that they appear to be primarily
bound to the same metal as the methyl ligand, while
also interacting weakly with the adjacent metal. How-
ever, the IR spectrum, which displays carbonyl stretches
at 1793 and 1963 cm-1, suggests a slightly different
interpretation in which these carbonyls exchange rap-
idly between terminal and bridging sites as diagrammed
in Scheme 3. Facile exchange of the carbonyls from
terminal to bridging gives average resonances in the
13C{1H} NMR spectra that appear as a pair of semibridg-
ing groups, the environments of which differ slightly
owing to the orientation of the cumulene ligand. This
fluxionality is analogous to that proposed for the
analogous ethylene,4d allene (2d), and methylallene (3d)
adducts and is rapid on the NMR time scale, even at
-80 °C, but resolvable on the much faster IR time scale.
(iii) 1,1-Dimethylallene. The reaction of 1 with 1,1-
dimethylallene at -78 °C yields only one product, [Ir2H-
(η2-CH2dCdCMe2)(CO)2(µ-CH2)(dppm)2][CF3SO3] (4a),
exactly analogous to the initial products in the low-
temperature reactions of 1 with ethylene, allene, and
methylallene. The 31P{1H} NMR spectrum displays two
resonances at δ 1.3 and -13.7, having the classic
pattern for an AA′BB′ spin system in these dppm-
bridged complexes, while the 13C{1H} NMR spectrum
of a 13CO-enriched sample shows two terminal-carbonyl

3720 Organometallics, Vol. 24, No. 15, 2005
Ristic-Petrovic et al.
Scheme 4
Figure 2. Perspective view of the [Ir2(CH3)(CO)2(µ-η1:η1-
CH2dCdCF2)(dppm)2]+ cation of compound 5c. Thermal
ellipsoids as for Figure 1. Only the ipso phenyl carbons are
shown.
(iv) 1,1-Difluoroallene. As with all other allenes in
this study, 1,1-difluoroallene reacts with 1 at -78 °C
to yield a methylene-bridged hydride product, [Ir2H(η2-
H2CdCdCF2)(CO)2(µ-CH2)(dppm)2][CF3SO3] (5a), as dia-
grammed in Scheme 4. The hydride resonance in the
1H NMR spectrum appears as a broad signal at δ -12.5,
while the bridging methylene and the allene proton
resonances (also broad) appear at δ 5.22 and 1.62,
respectively. In the 13C{1H} NMR spectrum the carbonyl
resonances at δ 187.4 and 174.4 are typical for termi-
nally bound groups. The 19F NMR spectrum displays
the inequivalent fluoroallene resonances as doublets at
δ -80.5 and -97.5, with a mutual geminal coupling of
84 Hz. The triflate resonance appears as a singlet at δ
-79.2. As was the case for 1,1-dimethylallene, but
unlike those for allene and methylallene, no other
isomer for the η2-bound cumulene adduct is observed
at this temperature.
Warming 5a above -20 °C results in conversion to
[Ir2(CH3)(CO)2(µ-H2CdCdCF2)(dppm)2][CF3SO3] (5c),
and by -10 °C this is the only product remaining. If 5a
is prepared in the presence of excess difluoroallene,
warming results in the appearance of other unidentified
products. To prepare 5c without these additional prod-
ucts, less than 1 equiv of the allene was used, leaving
some unreacted 1 obvious in the NMR spectra.
bond lengths and angles given in Table 4. Compound
5c has the normal dppm-bridged arrangement in which
the bridging dppm groups are essentially trans on both
metals. The difluoroallene moiety bridges the metals,
as proposed above, through the “H2CdC” moiety, having
the difluoromethylene moiety as a pendent group. It is
noteworthy that the orientation of the bridging cumu-
lene group in 5c is reversed from that proposed in 2c
and 3c, having the pendent vinyl group oriented away
from the methyl ligand at the adjacent metal and
instead pointed toward the less crowded end of the
complex, adjacent to the vacant site on Ir(2). We earlier
noted that for compounds 2c and 3c NOE contacts were
1
observed in the appropriate H NMR spectra between
the “H2CdC” moiety of the bridging cumulene and one
pair of dppm methylene protons. Figure 2 clearly shows
close contacts of this type involving the protons on C(3)
and the equatorial protons of the dppm methylene
groups. These contacts (ca. 2.24 Å) are normal van der
Waals separations and clearly short enough to give rise
to the observed NOE effects.
At Ir(1), which has roughly an octahedral environ-
ment, one carbonyl lies almost opposite the methylene
group of the bridging allene group, while the methyl
ligand is bound almost opposite the Ir-Ir bond. At the
other metal the carbonyl lies opposite the other end of
the allene bridge. The Ir-Ir separation (2.8844(5) Å)
corresponds to a normal single bond and can be viewed
The 1H NMR spectrum of 5c at 27 °C shows, in
addition to the dppm resonances, the methyl signal at
δ 0.10 as a triplet, having 8.6 Hz coupling to a pair of
31P nuclei, and the allene methylene signal at δ 2.51,
also as a triplet (3JPH ) 12.0 Hz). Selective 31P decou-
pling experiments establish that both CH3 and CH2
resonances couple to the same pair of 31P nuclei, so these
groups are presumably bound to the same metal. The
carbonyl ligands appear at δ 192.4 and 183.0 in the 13C-
{1H} NMR spectrum, while the pair of fluorine reso-
nances in the 19F NMR spectrum appear as doublets at
δ -66.9 and -81.6 with mutual coupling of 56 Hz, again
consistent with the difluoroallene being bound through
the “H2CdC” moiety. Overall, the 31P, 13C, and 1H spec-
troscopic parameters are similar to those of 2c and 3c,
suggesting a bridging arrangement for the difluoroal-
lene group as diagrammed in Scheme 4. However,
unlike the allene- and methylallene-bridged intermedi-
ates, 5c does not rearrange further upon warming.
The difluoroallene-bridged formulation is confirmed
by an X-ray structure determination, and a representa-
tion of the compound is shown in Figure 2, with selected
2
as resulting from donation of the dz pair of electrons
on the square-planar center of Ir(2) to Ir(1), giving the
latter an octahedral geometry. In this bonding descrip-
tion the metals are in the +1 and +3 oxidation states,
respectively.
The bridging difluoroallene group is essentially sym-
metrically bound to both metals, having comparable
Ir-C distances (2.120(8), 2.110(8) Å), both of which are
somewhat shorter than the Ir-CH3 distance (2.14(1) Å).
Binding of the “H2CdC” moiety between the metals has
resulted in substantial lengthening of the C(3)-C(4)
bond to 1.50(1) Å, consistent with a single bond,25 while
at the unbound end the C(4)-C(5) distance (1.31(1) Å)
remains as a double bond.
(b) Vinyl Carbene Products. Over a 7-day period
at ambient temperature compound 2e rearranges to the
vinyl carbene-bridged product [Ir2(H)(CO)2(µ-η1:η3-HCC-

Binuclear Activation of Cumulenes
Table 4. Selected Interatomic Distances and Angles for Compound 5c
(a) Distances (Å)
Organometallics, Vol. 24, No. 15, 2005 3721
atom 1
atom 2
distance
atom 1
atom 2
distance
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(2)
Ir(2)
Ir(2)
P(1)
P(3)
C(1)
C(3)
C(6)
P(2)
P(4)
C(2)
2.8844(5)
2.342(3)
2.343(2)
1.894(9)
2.120(8)
2.14(1)
2.314(3)
2.335(2)
1.867(8)
Ir(2)
P(1)
P(3)
O(1)
O(2)
F(1)
F(2)
C(3)
C(4)
C(4)
P(2)
P(4)
C(1)
C(2)
C(5)
C(5)
C(4)
C(5)
2.110(8)
3.002(4)a
2.986(3)a
1.14(1)
1.16(1)
1.35(1)
1.34(1)
1.50(1)
1.31(1)
(b) Angles (deg)
atom 1
atom 2
atom 3
angle
atom 1
atom 2
atom 3
angle
Ir(2)
Ir(2)
Ir(2)
P(1)
C(1)
C(1)
C(3)
Ir(1)
Ir(1)
P(2)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(1)
Ir(2)
Ir(2)
Ir(2)
C(1)
C(3)
C(6)
P(3)
C(3)
C(6)
C(6)
C(2)
C(4)
P(4)
114.3(3)
71.8(2)
C(2)
Ir(1)
Ir(2)
Ir(1)
Ir(2)
Ir(2)
C(3)
F(1)
F(1)
F(2)
Ir(2)
C(1)
C(2)
C(3)
C(4)
C(4)
C(4)
C(5)
C(5)
C(5)
C(4)
O(1)
O(2)
C(4)
C(3)
C(5)
C(5)
F(2)
C(4)
C(4)
164.5(4)
171(1)
159.5(3)
173.77(8)
173.8(4)
86.2(5)
173.5(8)
107.0(6)
111.1(6)
129.0(7)
119.9(8)
106.8(7)
124.5(8)
128.6(8)
87.7(4)
94.5(3)
70.1(3)
160.86(8)
a Nonbonded distance.
Scheme 5
With time, compound 6 decomposes into several uni-
dentified species, and it appears that at the same time
13C label becomes scrambled through different sites in
what remains of 6. We have no explanation of this
apparent subsequent scrambling.
Compound 3e also transforms over a several-day
period to two isomers of [Ir2H(CO)2(µ-η1:η3-CHC(CH3)C-
(H)CH3)(dppm)2][CF3SO3] (7a, 7b) as the major products
together with starting material and minor amounts of
several unidentified species. Isomer 7a was unambigu-
ously identified on the basis of a comparison of its 31P-
{1H} and 1H NMR spectra with that of a sample
previously prepared from the reaction of 1 with 2-bu-
tyne.9 In this isomer the two methyl substituents are
mutually cis, as confirmed by 2D 1H NMR experiments
in which NOE contacts are established between the two
methyl signals at δ 2.72 and 1.42. Isomer 7b is assumed
to have a trans arrangement of methyl groups. The
isomeric relationship between 7a and 7b seems clear
on the basis that most 1H NMR signals for the two
species overlap; only the proton on the bridging alky-
lidene carbon, at δ 8.96, and the hydride signal at δ
-11.18 are resolvable from the signals for 7a (δ 8.79,
-11.25). In addition the 31P{1H} NMR spectrum for 7b
is closely comparable with that of 7a.
(CH3)CH2)(dppm)2][CF3SO3] (6), as diagrammed in
Scheme 5. Although the yield of 6 is highly variable, it
has been obtained in up to approximately 80% yield as
determined from 31P{1H} NMR spectroscopy. Only
minor amounts of other related unidentified species,
along with decomposition products which remain at the
end, were observed by 31P{1H} NMR at intermediate
times. The spectroscopy of this product closely resembles
that of a series of complexes previously obtained in the
reaction of 1 with a number of internal alkynes,8,9 so
we propose an analogous structure for 6. The structure
of one of these products, from the reaction of 1 with
3-hexyne, was previously established by X-ray crystal-
lography.8,9 The 31P{1H} NMR spectrum of 6 is consis-
tent with the structure proposed, appearing as an ABCD
spin system in which three resonances in a 2:1:1
intensity ratio appear at δ -5.1, -30.2, and -32.4; the
low-field signal of double intensity results from the
Discussion
1
accidental overlap of two resonances. In the H NMR
spectrum the single proton on the bridging alkylidene
carbon appears as a broad, unresolved resonance at
typically low field (δ 8.72),9 while the olefinic protons
are also broad and appear at δ 3.43 and 3.11. The
methyl protons of the vinylcarbene ligand appear as a
doublet at δ 2.89, and a hydride signal appears as a
multiplet at δ -11.38.
To establish the fate of the Ir-bound methyl group of
2e, the reaction of allene with 13CH3-enriched compound
1 was repeated at ambient temperature. The 13C{1H}
NMR spectrum of the product showed 13C incorporation
into only the 2-methyl site of the vinyl carbene group.
In an earlier study we reported that the methyl
complex [Ir2(CH3)(CO)(µ-CO)(dppm)2][CF3SO3] (1) re-
acted with a variety of ligands, resulting in C-H
activation of the methyl ligand to yield methylene-
bridged hydrido products, [Ir2H(L)(CO)2(µ-CH2)(dppm)2]-
[CF3SO3], in which the added ligand is adjacent to the
bridging methylene group.6 This observation prompted
us to attempt analogous reactions using unsaturated
organic substrates with the goal of effecting C-C bond
formation, by insertion of these substrates into an Ir-
CH2 bond. In the current study we find that cumulenes
react with 1, even at -78 °C, to give the targeted

3722 Organometallics, Vol. 24, No. 15, 2005
Ristic-Petrovic et al.
Scheme 6
favored instead. With allene and methylallene, the first
rearrangement products observed, at between -60 and
-30 °C, are the cumulene-bridged products 2c and 3c,
respectively. These rearrangements can occur in a
couple of ways. In the first case, a partial “merry-go-
round” motion of the ligands, in which movement of the
cumulene in isomers 2b or 3b into the bridging site is
accompanied by migration of the hydride ligand back
to the methylene group regenerating a methyl ligand,
can occur. The second possible pathway involves cumu-
lene dissociation from both isomers 2a, 2b or 3a, 3b,
regenerating 1 followed by cumulene attack at either
metal in site 2, with subsequent movement to the
bridging position to give 2c and 3c, respectively. We
favor the latter mechanism, since otherwise it is difficult
to rationalize why isomers 2a and 3a do not also rotate
into a bridging geometry yielding an isomer analogous
to the 1,1-difluoroallene adduct, 5c. Furthermore, this
proposal of cumulene dissociation and subsequent re-
coordination is supported by the formation of mixtures
of 2a/2b together with 3a/3b when solutions of either
2a/2b or 3a/3b are reacted with the other cumulene
(methylallene or allene, respectively). As described in
the Experimental Section, warming such solutions gives
rise to all isomers of compounds 2 and 3 at the
appropriate temperatures.
In the case of 1,1-difluoroallene, a similar rearrange-
ment occurs, except that the cumulene in this case is
coordinated such that the pendent difluorovinyl moiety
is adjacent to the metal remote from the methyl ligand.
Presumably, the alternate geometry, corresponding to
isomers 2c and 3c, for allene and methylallene, respec-
tively, is unfavorable due to nonbonded contacts be-
tween a fluorine substituent and the methyl ligand or
due to contacts of these groups with the intervening
phenyl groups. In this context, it is worth noting that
in the methylallene adduct, 3c, only the isomer having
the allene methyl substituent remote from the methyl
ligand, and with the geminal hydrogen aimed toward
it instead, is observed. Although the bridged η1:η1
binding mode for a cumulene is analogous to that
observed for bridged mono-olefin20 and alkyne-bridged9,19
complexes, it has apparently never been observed for
cumulenes; it has however been proposed as the bonding
mode in Os2(CO)8(µ-C3H4).18 In all other previous re-
ports of cumulene-bridged structures, both cumulene π
bonds are involved in binding.24,27 This is believed to
be the first unambiguous confirmation of cumulene
binding in which a pair of metals bind through only one
end of the cumulene, leaving the remaining CdCR2
moiety as a pendent group.
methylene hydride products, analogous to those referred
to above (with L ) η2-cumulene). However, for the
cumulenes investigated (allene, methylallene, 1,1-di-
methylallene, and 1,1-difluoroallene), no subsequent
insertion into an Ir-CH2 bond is observed. Instead, a
series of rearrangements, in which the cumulene ligand
appears to migrate over the bimetallic framework, is
observed. Differences in reactivity of the different cu-
mulenes, as the temperature is increased, can be
attributed to either steric or electronic factors associated
with the cumulenes, as will be discussed in what follows.
As has been previously discussed,4d ligand attack in
1 can occur at one of the three sites shown in Scheme 6
(dppm groups above and below the plane of the drawing
are omitted), and several of the isomers observed in this
study are consistent with ligand coordination at the
different sites being favored at different temperatures.
With the four cumulenes studied the kinetic product,
observed at -78 °C, results from cumulene attack at
site 1, adjacent to the methyl ligand. This is the same
site of attack deduced in previous studies for a variety
of ligands and results in movement of the methyl ligand
toward the adjacent metal at which C-H activation
occurs. For allene and methylallene, two isomers are
observed for this kinetic product, which differ merely
by rotation of the cumulene around the Ir-cumulene
bond (only isomer a is shown in Scheme 6). There is
little preference of one isomer over the other since both
isomers for compounds 2 and 3 are present in almost
equal proportions. In the case of both disubstituted
allenes, only one isomer is observed, presumably owing
to unfavorable interactions in the unobserved isomer
involving the pair of substituents and the dppm phenyl
rings. Although we have no evidence to indicate which
isomer is observed, we suggest that both disubstituted
allene adducts have geometries like those of 2a and 3a,
which allow the larger substituents to avoid the bridging
methylene group while also presumably minimizing
contacts with the phenyl groups. The anticipated inser-
tion of the cumulenes into the adjacent Ir-CH2 bond
does not occur in the low-temperature regime of these
products owing presumably to the strong bonds involv-
ing the third-row metal.
(26) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orgpen,
A. G.; Taylor, R. J. Chem. Soc., Perkin Trans. 2 1987, 51.
(27) (a) Chisholm, M. H.; Rankel, L. A.; Bailey, W. E., Jr.; Cotton,
F. A.; Murillo, C. A. J. Am. Chem. Soc. 1977, 99, 1261. (b) Bailey, W.
I., Jr.; Chisholm, M. H.; Cotton, F. A.; Murillo, C. A.; Rankel, L. A. J.
Am. Chem. Soc. 1978, 100, 802. (c) Lewis, L. N.; Huffman, J. C.;
Caulton, K. G. J. Am. Chem. Soc. 1980, 102, 403. (d) Lewis, L. N.;
Huffman, J. C.; Caulton, K. G. Inorg. Chem. 1980, 19, 1246. (e) Al-
Obaidi, Y. N.; Baker, P. K.; Green, M.; White, N. D.; Taylor, G. E. J.
Chem. Soc, Dalton Trans. 1981, 2321. (f) Hoel, E. L.; Ansell, G. B.;
Leta, S. Organometallics 1986, 5, 585. (g) Cayton, R. H.; Chisholm,
M. H.; Hapden-Smith, M. J. J. Am. Chem. Soc. 1988, 110, 4438. (h)
Cayton, R. H.; Chacon, S. T.; Chisholm, M. H.; Hapden-Smith, M. J.;
Huffman, J. C.; Folting, K.; Ellis, P. D.; Huggins, B. A. Angew Chem.,
Int. Ed. Engl. 1989, 28, 1523. (i) Wu, I.-Y.; Tseng, T.-W.; Chen. C.-T.;
Cheng, M.-C.; Lin, Y.-C.; Wang, Y. Inorg. Chem. 1993, 32, 1539.
Warming these products also does not lead to inser-
tion since the aforementioned rearrangements are

Binuclear Activation of Cumulenes
Organometallics, Vol. 24, No. 15, 2005 3723
Scheme 7
Scheme 8
In the case of difluoroallene migration, the transfor-
mation of 5a to 5c apparently occurs by a merry-go-
round movement of the ligands, as outlined in Scheme
7 (dppm groups omitted) instead of by cumulene dis-
sociation and recoordination as described for compounds
2a/2b and 3a/3b. Addition of an excess of allene to a
solution of 5a at -78 °C and warming to ambient
temperature yields none of the isomers of compound 2;
only 5a transforming to 5c is observed. We assume that
the electronegative fluorine substituents lead to more
effective binding of this cumulene, favoring migration
instead of dissociation and recoordination.
It is interesting that in all 1,1-difluoroallene com-
pounds structurally characterized, this group binds
preferentially to metals via the hydrogen-substituted
double bond.28 In mononuclear species this has been
rationalized in frontier orbital terms as resulting from
the LUMO being localized at this end of the molecule,
which owing to its strong π-acid nature favors binding
to electron-rich metals at this end.29 In the initial adduct
(5a) the cumulene binding at one metal through the
“H2CdC” functionality is presumably favored for the
reasons proposed above in mononuclear complexes. The
merry-go-round movement of the ligands then generates
5c which retains binding through the unsubstituted
double bond. In any case, the electronegative fluorine
substituents clearly favor coordination of the “H2CdC”
moiety in the bridging position, since subsequent warm-
ing does not result in further rearrangement.
at the expense of 4a, and at -10 °C 2d and 2e also begin
to appear. At ambient temperature the major species
is 2e with only minor amounts of 4d present. Upon
dimethylallene dissociation, the smaller allene molecule
competes more effectively for the iridium coordination
sites.
Warming compound 2c or 3c results in two parallel
transformations yielding simultaneously the η2-cumu-
lene adducts 2d or 3d together with the η1:η3-bridged
product 2e or 3e. The transformation of compound 2c
or 3c into 2e or 3e, respectively, can occur by a twist
about the bond between one metal and the central
cumulene carbon, accompanied by bending back of the
phosphines at the adjacent metal. However, the trans-
formation of compound 2c or 3c into either 2d or 3d
must occur by cumulene dissociation and subsequent
attack at site 3 between the two carbonyls. Presumably,
compounds 2d and 3d do not transform directly into the
respective species 2e and 3e, since this would require
substantial ligand reorganization. Instead, we propose
a slow isomerization back to 2c and 3c by cumulene
dissociation and recoordination, followed by isomeriza-
tion to 2e and 3e as described above. Again, we have
confirmed that cumulene dissociation is involved at
some stage in the transformations involving 2c and 3c
by the addition of methylallene and allene, respectively,
generating mixtures of isomers of both compounds 2 and
3 in both cases.
Increasing the steric bulk of the substituents even
further to the dimethylallene adduct results in no
cumulene-bridged isomer being observed; presumably,
the steric contacts in such a bridged species are too
highly unfavorable. Instead, warming a solution of 4a
results in rearrangement to isomer 4d (Schemes 3 and
6), which also contains an η2-bound cumulene bound
terminally to one metal. Clearly in this case rearrange-
ment must occur by ligand dissociation since in 4a the
cumulene is flanked by the methylene group and a
carbonyl, whereas in 4d it is flanked by both carbonyls.
This has been verified by warming a solution of 4a in
the presence of allene. Although in this case no reaction
takes place at -78 °C, warming to -60 °C gives rise to
small amounts of 2c together with 4a. Further temper-
ature increases result in the increased formation of 2c
Subsequent rearrangement of 2e and 3e to their
respective vinylcarbene (allylidene) products 6 and 7a/b
occurs slowly at ambient temperature. We propose that
this transformation occurs by migration of the adjacent
methyl ligand to the central carbon of the bridging
cumulene group in 2e and 3e to give an η3-allyl
intermediate (A) as shown in Scheme 8 (all ancillary
ligands are omitted for clarity). This proposed methyl
migration is supported by labeling studies in which it
was established that the labeled methyl ligand in 2e
ended up exclusively as the methyl group in 6. Rotation
of the allyl group at one metal would bring the two ends
of this group into the vicinity of the adjacent metal and
C-H activation at either end of the monosubstitued
allyl group (RdH) would yield 6, while C-H activation
at the less substituted end of the disubstituted allyl
group (R ) Me) would yield 7b. However, in the
transformation of 3e, two isomers (7a and 7b) are
observed. We propose that the second isomer (7a) arises
(28) (a) Lentz, D.; Nichelt, N.; Willemsen, S. Chem. Eur. J. 2002, 8,
1205. (b) Lentz, D. J. Fluorine Chem. 2004, 125, 853.
(29) Dixon, D. A.; Smart, B. E. J. Phys. Chem. 1989, 93, 7772.

3724 Organometallics, Vol. 24, No. 15, 2005
Ristic-Petrovic et al.
via an η1-allyl intermediate (C), as also diagrammed in
Scheme 8. Rotation about the vinylic C-C bond followed
by generation of the η3-allyl species (D) results in an
isomerization from a geometry in which both methyl
substituents have a mutually anti arrangement to one
in which they are syn. Carbon-hydrogen activation at
the less substituted end of this allyl group would
generate 7a. The η3-η1 conversion is well documented
in allyl complexes and is the process proposed for the
facile equilibration often seen between syn and anti
protons in these systems.30 Although intermediate C has
an η1-allyl group that is bound to the metal via the more
substituted carbon, this process cannot be unfavorable
since 7a is the more abundant isomer (in a 2:1 ratio).
We suggest that this is due to more unfavorable
interactions of the terminal methyl group in the product
7b with the rest of the complex, since in this arrange-
ment this methyl group is thrust into the vicinity of the
metal-metal bond. In 7a this methyl group avoids these
unfavorable contacts. This was clearly demonstrated in
the structure determination of the analogous [Ir2H(CO)2-
(µ-η1:η3-HCC(Et)C(H)Et)(dppm)2][CF3SO3], in which the
methyl groups of 7a are replaced by ethyl substituents.9
In this structure it is clear that the mutually cis
arrangement of ethyl groups is favorable, while a trans
arrangement (as in 7b) would not be.
CdCRR′)(CO)2(µ-CH2)(dppm)2][CF3SO3] (R, R′ ) H,
CH3; R ) R′ ) H, CH3, F), in no case does cumulene
insertion into this Ir-CH2 bond occur. Instead, as the
temperature is increased, facile cumulene dissociation
and recoordination at an alternate site is observed
accompanied by regeneration of the methyl ligand from
the methylene and hydride groups. Subsequent in-
creases in temperature give rise to accompanying rear-
rangements which are more extensive for the allene and
methylallene adducts than for the disubstituted cumu-
lene adducts, which display similar although more
limited reactivity owing either to steric or electronic
effects. The detailed variable-temperature NMR study
on the four cumulenes allows the identification of the
cumulene coordination modes at different stages of the
reactions and establishes that only the µ-η1:η3-binding
mode involving allene and methylallene results in
coupling involving the methyl ligand accompanied by
C-H activation of the resulting fragment. No other
cumulene coordination mode resulted in ligand coupling,
even when groups were adjacent. Neither the 1,1-
dimethylallene nor the 1,1-difluoroallene ligands result
in ligand coupling reactions since neither achieves the
necessary µ-η1:η3-coordination mode. The dimethyl-
substituted molecule appears unable to access the
bridging site owing to steric repulsion involving the
methyl substituents, and the electronegative substitu-
ents on the difluoro analogue appear to favor bridging
through the “H2CdC” moiety, which is subsequently
unreactive.
Conclusions
This study has established another example involving
facile C-H activation of the methyl ligand in [Ir2(CH3)-
(CO)(µ-CO)(dppm)2][CF3SO3] upon coordination of a
ligand (in this case a series of cumulene molecules) at
the adjacent metal. It also clearly demonstrates how
ligand reactivity can change dramatically depending
upon the ligand binding mode and the extent of involve-
ment of the adjacent metal centers. Although with each
cumulene investigated the cumulene is bound adjacent
to an Ir-CH2 bond in the kinetic product [Ir2H(η2-H2Cd
Acknowledgment. We thank the Natural Sciences
and Engineering Research Council of Canada (NSERC)
and the University of Alberta for financial support of
the research and for scholarships (to D.J.A. and J.R.T.),
and NSERC for funding the X-ray diffractometer. X-ray
data for 2e were collected by Dr. James F. Britten at
McMaster University, Hamilton, ON, Canada.
Supporting Information Available: Tables of X-ray
experimental details, atomic coordinates, interatomic distances
and angles, anisotropic thermal parameters, and hydrogen
parameters for compounds 2e and 5c. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(30) (a) Corradini, P.; Maglio, G.; Musco, A.; Paiaro, G. Chem.
Commun. 1966, 618. (b) Vrieze, K.; Volger, H. C. J. Organomet. Chem.
1967, 9, 537. (c) Faller, J. W.; Thomsen, M. E.; Mattina, M. J. J. Am.
Chem. Soc. 1071, 93, 2642. (d) Nixon, J. F.; Wilkins, B.; Clement, D.
A. J. Chem. Soc., Dalton Trans. 1974, 1993. (e) Osakada, K.; Chiba,
T.; Nakamura, Y.; Yamamoto, T.; Yamamoto, A. Organometallics 1989,
8, 2602. (f) Consiglio, G.; Waymouth, R. M. Chem. Rev. 1989, 89, 257.
(g) Solin, N.; Szabo, K. J. Organometallics 2001, 20, 5464.
OM0580102