The mechanism of carbon-carbon bond activation in cationic 6- alkylcyclohexadienyl ruthenium hydride complexes

Article abstract of DOI:10.1021/ja992987e

Carbon-carbon bond activation in cationic 6-endo-methyl-η⁵- cyclohexadienyl and 6-exo-methyl-η⁵-cyclohexadienyl ruthenium hydride complexes has been investigated. Contrary to expectations, it is the 6-exo- methyl complex and not the stereoisomeric 6-endo-methyl complex that undergoes selective carbon-carbon bond activation under exceptionally mild conditions, quantitatively converting the 6-exo-methyl substituent and the hydride ligand to methane. The mechanism of the activation reaction involves dissociation of protic acid from the agostic starting complex by reaction with a weak base (typically water), followed by protolytic activation of the alkyl group, with 'back-side' assistance from the nucleophilic metal center. Under the same conditions, the corresponding 6-endo-methyl isomer undergoes selective dehydrogenation rather than demethylation, despite the proximity of the endo-methyl substituent to the metal center. For both exo and endo isomers, the cationic ruthenium hydride intermediates were determined by spectroscopic analysis to adopt fluxional agostic structures. The agostic complexes are kinetically stable at room temperature under rigorously anhydrous conditions but convert quantitatively to cationic η⁶-arene products in the presence of a Bronsted base. The rates of both carbon-carbon bond activation and dehydrogenation are dependent on the identity and concentration of the base and suppressed in the presence of excess acid. The protolytic mechanism for carbon-carbon bond activation is supported by deuterium-labeling studies and by the reactivity of the neutral complexes toward Lewis acids and one-electron oxidants. This mechanism is shown to be relevant to carbon-carbon bond activation reactions observed in less- substituted 6-exo-methyl-η⁵-cyclohexadienyl complexes and in a steroid- derived 6,6-disubstituted-η⁵-cyclohexadienyl complex, representative of previously reported cases of dealkylative ligand aromatization. The low kinetic barrier for the protolytic dealkylation mechanism is contrasted to the comparatively high activation barriers reported for carbon-carbon bond activation reactions that occur in structurally related systems that cannot access a protolytic pathway. This investigation provides a consistent basis for rationalizing this potentially important but poorly understood class of metal-mediated reactions.

Full text of DOI:10.1021/ja992987e

2784
J. Am. Chem. Soc. 2000, 122, 2784-2797
The Mechanism of Carbon-Carbon Bond Activation in Cationic
6-Alkylcyclohexadienyl Ruthenium Hydride Complexes
Christina M. Older and Jeffrey M. Stryker*
Contribution from the Department of Chemistry, UniVersity of Alberta, Edmonton,
Alberta, T6G 2G2 Canada
ReceiVed August 17, 1999
Abstract: Carbon-carbon bond activation in cationic 6-endo-methyl-η5-cyclohexadienyl and 6-exo-methyl-
η5-cyclohexadienyl ruthenium hydride complexes has been investigated. Contrary to expectations, it is the
6-exo-methyl complex and not the stereoisomeric 6-endo-methyl complex that undergoes selective carbon-
carbon bond activation under exceptionally mild conditions, quantitatively converting the 6-exo-methyl
substituent and the hydride ligand to methane. The mechanism of the activation reaction involves dissociation
of protic acid from the agostic starting complex by reaction with a weak base (typically water), followed by
protolytic activation of the alkyl group, with “back-side” assistance from the nucleophilic metal center. Under
the same conditions, the corresponding 6-endo-methyl isomer undergoes selective dehydrogenation rather than
demethylation, despite the proximity of the endo-methyl substituent to the metal center. For both exo and
endo isomers, the cationic ruthenium hydride intermediates were determined by spectroscopic analysis to adopt
fluxional agostic structures. The agostic complexes are kinetically stable at room temperature under rigorously
anhydrous conditions but convert quantitatively to cationic η6-arene products in the presence of a Brønsted
base. The rates of both carbon-carbon bond activation and dehydrogenation are dependent on the identity and
concentration of the base and suppressed in the presence of excess acid. The protolytic mechanism for carbon-
carbon bond activation is supported by deuterium-labeling studies and by the reactivity of the neutral complexes
toward Lewis acids and one-electron oxidants. This mechanism is shown to be relevant to carbon-carbon
bond activation reactions observed in less-substituted 6-exo-methyl-η5-cyclohexadienyl complexes and in a
steroid-derived 6,6-disubstituted-η5-cyclohexadienyl complex, representative of previously reported cases of
dealkylative ligand aromatization. The low kinetic barrier for the protolytic dealkylation mechanism is contrasted
to the comparatively high activation barriers reported for carbon-carbon bond activation reactions that occur
in structurally related systems that cannot access a protolytic pathway. This investigation provides a consistent
basis for rationalizing this potentially important but poorly understood class of metal-mediated reactions.
Introduction
η4-cyclopentadiene and η5-cyclohexadienyl complexes are
converted into dealkylated η5-cyclopentadienyl and η6-arene
complexes, respectively.8,9 Similar reactions of η5-cyclohexa-
dienyl intermediates are presumably also involved in coordina-
Although the direct activation of alkane carbon-carbon bonds
by soluble transition metal complexes remains elusive, transition
metals mediate the activation of carbon-carbon bonds in a range
of contexts.1 Both oxidative and nonoxidative processes are
relatively common, typically driven by a combination of kinetic
and thermodynamic factors such as release of ring strain2 and
coordination-induced proximity. Proximity effects function to
overcome the generally higher kinetic barrier for carbon-carbon
bond activation over competitive carbon-hydrogen bond activa-
tion and promote selectivity in both oxidative3 and nonoxidative
reactions, the latter of which can generally be classified as
â-alkyl elimination reactions.4-9
(3) Oxidative carbon-carbon activation: (a) Suggs, J. W.; Jun, C.-H. J.
Am. Chem. Soc. 1986, 108, 4679 and references therein. Jun, C.-H.
Organometallics 1996, 15, 895. Jun, C.-H.; Lee, H. J. Am. Chem. Soc. 1999,
121, 880. (b) Rybtchinski, B.; Milstein, D. J. Am. Chem. Soc. 1999, 121,
4528. Rybtchinski, B.; Milstein, D. Angew. Chem., Int. Ed. 1999, 38, 870.
van der Boom, M. E.; Ben- David, Y.; Milstein, D. J. Chem. Soc., Chem.
Commun. 1998, 917. Liou, S.-Y.; van der Boom, M. E.; Milstein, D. J.
Chem. Soc., Chem. Commun. 1998, 687. van der Boom, M. E.; Liou, S.-
Y.; Ben-David, Y.; Gozin, M.; Milstein, D. J. Am. Chem. Soc. 1998, 120,
13415. Gandelman, M.; Vigalok, A.; Ben-David, Y.; Milstein, D. Orga-
nometallics 1997, 16, 3981. Rybtchinski, B.; Vigalok, A.; Ben-David, Y.;
Milstein, D. J. Am. Chem. Soc. 1996, 118, 12406 and references therein.
(4) â-Alkyl elimination by “low-electrophilicity” late transition metals
(selected recent references): (a) McNeill, K.; Andersen, R. A.; Bergman,
R. G. J. Am. Chem. Soc. 1997, 119, 11244. Hartwig, J. A.; Andersen, R.
A.; Bergman, R. G. J. Am. Chem. Soc. 1989, 111, 2717. Kaplan, A. W.;
Bergman, R. G. Organometallics 1997, 16, 1106. (b) Takemori, T.; Suzuki,
H.; Tanaka, M. Organometallics 1996, 15, 4346. Suzuki, H.; Takaya, Y.;
Takemori, T. J. Am. Chem. Soc. 1994, 116, 10779 and references therein.
(c) Flood, T. C.; Statler, J. A. Organometallics 1984, 3, 1795. Flood, T.
C.; Bitler, S. P. J. Am. Chem. Soc. 1984, 106, 6076. Ermer, S. P.; Struck,
G. E.; Bitler, S. P.; Richards, R.; Bau, R.; Flood, T. C. Organometallics
1993, 12, 2634. (d) Thompson, S. K.; Young, G. B. Organometallics 1989,
8, 2068. Ankianiec, B. C.; Christou, V.; Hardy, D. T.; Thomson, S. K.;
Young, G. B. J. Am. Chem. Soc. 1994, 116, 9963. (e) Thomas, B. J.; Noh,
S. K.; Schulte, G. K.; Sendlinger, S. C.; Theopold, K. H. J. Am. Chem.
Soc. 1991, 113, 893.
Ligand aromatization provides the driving force for one
general class of â-alkyl elimination processes, in which alkylated
(1) General reviews: Rybtchinski, B.; Milstein, D. Angew. Chem., Int.
Ed. 1999, 38, 870. Herrmann, W. A.; Cornils, B. Angew. Chem., Int. Ed.
Engl. 1997, 36, 1048. Merola, J. S. Curr. Org. Chem. 1997, 1, 235. Crabtree,
R. H. Chem. ReV. 1985, 85, 245.
(2) Oxidative activation driven by relief of ring strain. Recent reviews:
(a) Jennings, P. W.; Johnson, L. L. Chem. ReV. 1994, 94, 2241. Hirao, T.
Top. Curr. Chem. 1996, 178, 99. Recent lead references: (b) Yeh, W.-Y.;
Hsu, S. C. N.; Peng, S.-M.; Lee, G.-H. Organometallics 1998, 17, 2477.
(c) Perthuisot, C.; Edelbach, B. L.; Zubris, D. L.; Jones, W. D. Organo-
metallics 1997, 16, 2016. Edelbach, B. L.; Lachicotte, R. J.; Jones, W. D.
J. Am. Chem. Soc. 1998, 120, 2843.
10.1021/ja992987e CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/14/2000

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2785
tion-induced hexaalkylbenzene dealkylation, a process mediated
by aluminum trichloride and proceeding only in the presence
of the transition metal.10
6,6-dimethyl-η5-cyclohexadienyl complexes of ruthenium (eq
1). Despite the presence of the highly labile dichloromethane
The mechanisms of 6-alkyl-η5-cyclohexadienyl dealkylation
reactions remain elusive. In the most rigorous previous inves-
tigation, Dimauro and Wolczanski9d clearly demonstrated that
generation of a coordinatively unsaturated metal center is
required11 to induce â-methyl activation in phosphine-supported
ligand, however, the demethylation proceeds slowly even at
elevated temperature. In this investigation, it was not possible
to determine whether it is the endo- or exo-methyl substituent
that is activated in the reaction. Chaudret and co-workers9a
demonstrated a similarly high activation barrier for the demeth-
ylation of 4,4-dimethyl-2-cyclohexenone, using an unsaturated
cationic ruthenium complex to effect dehydrogenation and
aromatization (eq 2). The formation of the 4-methylanisole
(5) â-Alkyl elimination by electrophilic lanthanide and early transition
metals: (a) Watson, P. L.; Roe, D. C. J. Am. Chem. Soc. 1982, 104, 6471.
Watson, P. L. J. Am. Chem. Soc. 1983, 105, 6471. Watson, P. L.; Parshall,
G. W. Acc. Chem. Res. 1985, 18, 51. (b) Bunel, E.; Burger, B. J.; Bercaw,
J. E. J. Am. Chem. Soc. 1988, 110, 976. Bercaw, J. E. Pure Appl. Chem.
1990, 62, 1151. Burger, B. J.; Thompson, M.; Cotter, D.; Bercaw, J. J.
Am. Chem. Soc. 1990, 112, 1566. (c) Horton, A. D. Organometallics 1996,
15, 2675. Etienne, M.; Mathieu, R.; Donnadieu, B. J. Am. Chem. Soc. 1997,
119, 3218. â-Carbon elimination in olefin polymerization catalysis: (d)
Eshuis, J.; Tan, Y.; Teuben, J. J. Mol. Catal. 1990, 62, 277. Eshuis, J. J.
W.; Tan, Y. Y.; Meetsma, A.; Teuben, J. H.; Renkema, J.; Evens, G. G.
Organometallics 1992, 11, 362. (e) Mise, T.; Kageyama, A.; Miya, S.;
Yamazaki, H. Chem. Lett. 1991, 1525. (f) Resconi, L.; Piemontesi, F.;
Franciscono, G.; Abis, L.; Fiorani, T. J. Am. Chem. Soc. 1992, 114, 1025.
(g) Kesti, M. R.; Waymouth, R. M. J. Am. Chem. Soc. 1992, 114, 3565.
(h) Guo, Z.; Swenson, D. C.; Jordan, R. F. Organometallics 1994, 13, 1424.
(i) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10015.
(j) Hajela, S.; Bercaw, J. E. Organometallics 1994, 13, 1147. (k) Yang, X.;
Jia, L.; Marks, T. J. J. Am. Chem. Soc. 1993, 115, 3392.
(6) Carbon-carbon bond activation by electrophilic late metal com-
plexes: (a) Terheijden, J.; van Koten, G.; Vinke, I. C.; Spek, A. L. J. Am.
Chem. Soc. 1985, 107, 2891. (b) Nicholls, J. C.; Spencer, J. L. Organo-
metallics 1994, 13, 1781. Cracknell, R. B.; Nicholls, J. C.; Spencer, J. L.
Organometallics 1996, 15, 446. (c) Dzwiniel, T. D.; Etkin, N.; Stryker, J.
M., J. Am. Chem. Soc. 1999, 121, 10640.
(7) Activation reactions driven by a combination of factors: (a) Mu-
rakami, M.; Amii, H.; Kunji, S.; Ito, Y. J. Am. Chem. Soc. 1996, 118, 8285.
Murakami, M.; Amii, H.; Ito, Y. Nature 1994, 370, 540. (b) Kondo, T.;
Kodoi, K.; Nishinaga, E.; Okada, T.; Morisaki, Y.; Watanabe, Y.; Mitsudo,
T. J. Am. Chem. Soc. 1998, 120, 5587.
(8) Carbon-carbon bond activation driven by ligand aromatization.
Alkylated η4-cyclopentadiene conversions: (a) Kang, J. W.; Maitlis, P. M.
J. Am. Chem. Soc. 1968, 90, 3259. Kang, J. W.; Moseley, K.; Maitlis, P.
M. J. Am. Chem. Soc. 1969, 91, 5970. King, R. B.; Efraty, A. J. Am. Chem.
Soc. 1972, 94, 3773. King, R. B.; Douglas, W. M.; Efraty, A. J. Organomet.
Chem. 1974, 69, 131. (b) Eilbracht, P. Chem. Ber. 1976, 109, 1429, 3136.
Eilbracht, P. J. Organomet. Chem. 1976, 120, C37. Eilbracht, P.; Dahler,
P. J. Organomet. Chem. 1977, 125, C23. Eilbracht, P.; Dahler, P. J.
Organomet. Chem. 1977, 127, C48. Eilbracht, P.; Mayser, U. J. Organomet.
Chem. 1977, 135, C26. Eilbracht, P.; Dahler, P. Chem. Ber. 1980, 113,
542. Eilbracht, P.; Dahler, P.; Mayser, U.; Henkes, E. Chem. Ber. 1980,
113, 1033. Eilbracht, P.; Mayser, U.; Tiedtke, G. Chem. Ber. 1980, 113,
1420. Eilbracht, P.; Mayser, U. Chem. Ber. 1980, 113, 2211. (c) Benfield,
F. W. S.; Green, M. L. H. J. Chem. Soc., Dalton Trans. 1974, 1324. Barretta,
A.; Cloke, F. G., N.; Feigenbaum, A.; Green, M. L. H.; Gourdon, A.; Prout,
K. J. Chem. Soc., Chem. Commun. 1981, 156. Baretta, A.; Chong, K. S.;
Cloke, F. G. N.; Feigenbaum, A.; Green, M. L. H. J. Chem. Soc., Dalton
Trans. 1983, 861. Green, M. L. H.; Wong, L.-L. J. Chem. Soc., Chem.
Commun. 1984, 1442. Gladyz, J. A.; Fulcher, J. G.; Ugolick, R. C.; Hanlan,
A. J. L.; Bocarsly, A. B. J. Am. Chem. Soc. 1979, 101, 3388. Fu, Y.-T.;
Chao, P.-C.; Liu, L.-K. Organometallics 1998, 17, 221. (d) Crabtree, R.
H.; Dion, R. P. J. Chem. Soc., Chem. Commun. 1984, 1260. Crabtree, R.
H.; Dion, R. P.; Gibboni, D. J.; McGrath, D. V.; Holt, E. H. J. Am. Chem.
Soc. 1986, 108, 7222. (e) Hemond, R. C.; Hughes, R. P.; Locker, M. B.
Organometallics 1986, 5, 2392. (f) Jones, W. D.; Maguire, Organometallics
1987, 6, 1301.
complex is presumed to arise from reaction with the methanol
liberated upon protonation of [(C5Me5)Ru(OMe)]2 with trifluo-
rosulfonic acid. The methyl fragment is released principally as
methane, although a minor amount of ethane is also formed.
On the basis of this observation, the dealkylation was proposed
to proceed by extrusion of methyl radical from a cationic η5-
cyclohexadienyl hydride intermediate. The ruthenium fragment
also aromatizes unsaturated steroid substrates upon prolonged
thermolysis at high temperature (e.g., eq 3); a partially
characterized intermediate was observed in these reactions,
spectroscopically consistent with the proposed η5-cyclohexa-
dienyl hydride complex.9a,b The high activation barrier for
demethylation was attributed to the stereochemistry of ruthenium
binding: the reaction clearly proceeds by activation of a methyl
group on the exo face of the coordinated ring.
In contrast, the groups of both Chaudret and Itoh have
reported closely related dealkylation reactions that proceed under
surprisingly mild conditions, although the dealkylations occur
only as minor reaction pathways (Scheme 1). Thus, the
aromatization of 3-methylcyclohexene proceeds under Chau-
dret’s conditions to yield a minor fraction of the η6-benzene
complex along with the major η6-toluene product at room
temperature.9a Itoh observes demethylation as a minor pathway
in ruthenium-mediated diene/alkyne cycloaddition at low tem-
perature, a reaction that leads principally to the formation of
double alkyne adducts.9c Both reactions were rationalized by
proposing the formation of stereoisomeric η5-cyclohexadienyl
intermediates 1-exo and 1-endo, with the dealkylation pathway
arising from the sterically less favorable minor endo-methyl
intermediate. It is attractive to attribute the difference in
activation barriers between this system and Wolczanski’s
(9) Carbon-carbon bond activation driven by ligand aromatization. η5-
Cyclohexadienyl conversions: (a) Rondon, D.; Chaudret, B.; He, X.-D.;
Labroue, D. J. Am. Chem. Soc. 1991, 113, 5671. Urbanos, F.; Halcrow, M.
A.; Fernandez-Baeza, J.; Dahan, F.; Labroue, D.; Chaudret B. J. Am. Chem.
Soc. 1993, 115, 3484. Chaudret, N. Bull. Soc. Chim. Fr. 1995, 132, 268.
(b) Chaudret, B.; Halcrow, M. A.; Urbanos, F. Organometallics 1993, 12,
955. (c) Masuda, K.; Ohkita, H.; Kurumatani, S.; Itoh, K. Organometallics
1993, 12, 2221. (d) Dimauro, P. T.; Wolczanski, P. T. Polyhedron 1995, 1,
149.
(10) (a) Fischer, E. O.; Elschenbroich, C. Chem. Ber. 1970, 103, 162.
(b) Roman, E.; Astruc, D. Inorg. Chim. Acta 1979, 37, L465. Astruc, D.
Tetrahedron 1983, 39, 4027.
(11) The reaction of 1,1-dimethylcyclohexane with [L2Ir(acetone)2]+
[L ) (p-FC6H4)3P], for example, leads to the coordinatively saturated [(6,6-
dimethyl-η5-cyclohexadienyl)Ir(H)L2]+, which does not undergo subsequent
demethylation, even under harsh reaction conditions.8d

2786 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
Older and Stryker
Scheme 1
Figure 1.
of the hydride ligand to the η6-arene is adequately precedented,15
but this leads to a 6-exo-methyl-η5-cyclohexadienyl intermediate
III that appears to be a poor candidate for low-temperature
carbon-carbon bond activation. While the 6-endo-methyl
stereoisomer IV constitutes an attractive precursor to demeth-
ylation, it is not at all obvious how such an intermediate might
arise.16
In this report, the synthesis and aromatization reactions of a
series of stereochemically unambiguous ruthenium 6-methyl-
η5-cyclohexadienyl complexes are described. The low-temper-
ature carbon-carbon bond activation is shown to proceed by
an unexpected mechanistic pathway that considerably clarifies
prior investigations into the dealkylation of η5-cyclohexadienyl
complexes and our own report of hexamethylbenzene demeth-
ylation.
presumed endo-dealkylations to the possibility of a concerted
extrusion of methane from the coordinatively saturated and
sterically crowded intermediate 1-endo. This rationalization is
inconsistent, however, with Chaudret’s high-temperature de-
methylation of 4,4-dimethyl-2-cyclohexenone, which is proposed
to proceed via an analogous intermediate.
Our interest in the aromatization of η5-cyclohexadienyl
complexes was engendered by the serendipitous discovery of a
general new class of arene dealkylation reactions that converts
coordinated hexamethylbenzene into pentamethylbenzene in
high yield under exceptionally mild conditions (eq 4).12 The
Results and Discussion
Synthesis of 6-Methyl-η5-cyclohexadienyl Ruthenium Com-
plexes. To evaluate stereochemical effects on carbon-carbon
bond activation in 6-methyl-η5-cyclohexadienyl aromatization
reactions, a stereochemically unambiguous synthesis of cationic
ruthenium complexes modeling putative intermediates III and
IV was required. This was accomplished by protonation of the
neutral (η5-cyclopentadienyl)ruthenium η5-cyclohexadienyl pre-
cursors, which in turn were prepared by regioselective nucleo-
philic addition to the corresponding cationic η6-arene complexes.
Such reactions generally proceed by stereoselective alkylation
of the exo-face of the coordinated arene.17,18
dealkylation is integrated into an overall [3 + 2] cycloaddition
that converts (η6-hexamethylbenzene)ruthenium η3-allyl com-
plexes and disubstituted alkynes into (η6-pentamethylbenzene)-
ruthenium η5-dialkylcyclopentadienyl complexes and methane.
Allyl/alkyne [3 + 2] cycloaddition reactions have been previ-
ously reported for η5-cyclopentadienyl,13 η5-pentamethylcyclo-
pentadienyl,14 and η6-arene13 complexes of the late transition
metals, but the process normally proceeds via “oxidative”
extrusion of dihydrogen. While no mechanistic investigations
have been reported, the transformation can be rationalized by
proposing the intermediacy of a cationic η4-cyclopentadiene
hydride complex I (Figure 1). This intermediate can, in principle,
undergo dehydrogenation to give the expected η5-cyclopenta-
dienyl product II or demethylation of the ancillary hexameth-
ylbenzene ligand, as observed. A satisfying mechanistic rationale
for the dealkylation pathway, however, is problematic. Migration
Thus, the known19 cation [(η5-C5Me5)Ru(η6-hexamethyl-
benzene)]+BF4- (2) was prepared from [(η5-C5Me5)RuCl]4 and
converted to the η5-1,2,3,4,5,6-endo-hexamethylcyclohexadienyl
complex 3 in moderate yield by treatment with lithium trieth-
ylborohydride in tetrahydrofuran at low temperature (Scheme
2). The addition was confirmed by the presence of a high-
(15) (a) Brookhart, M.; Pinhas, A. R.; Lukacs, A. Organometallics 1982,
1, 1730. (b) Rush, P. K.; Noh, S. K.; Brookhart, M. Organometallics 1986,
5, 1745. (c) Kowalski, A. S.; Ashby, M. T. J. Am. Chem. Soc. 1995, 117,
12639.
(16) It is possible to propose a pathway involving protonation of a neutral
η6-arene intermediate on the exo face of the arene ligand to give a 6-endo-
methyl complex. This requires deprotonation of a cationic η4-cyclopenta-
diene intermediate by a catalytic amount of an adventitious base to generate
the neutral substrate. It is also possible to consider an unprecedented
unimolecular mechanism involving a 1,2-methyl migration on the exo-face
of the 6-methyl-η5-cyclohexadienyl intermediate, leading to the formation
of a 6,6-dimethyl-η5-cyclohexadienyl intermediate that undergoes activation
of the endo substituent. These rationales, along with synthetic and
mechanistic investigations specifically relevant to hexamethylbenzene
dealkylation, will be more completely discussed in a subsequent report.12b
(17) Alkylation of [(η5-cyclopentadienyl)Ru(η6-benzene)]+: Vol’kenau,
N. A.; Bolesova, I. N.; Shul’pina, L. S.; Kitaigorodskii, A. N. J. Organomet.
Chem. 1984, 267, 313.
(12) (a) Preliminary communication: Older, C. M.; Stryker, J. M.
Organometallics 1999, 17, 5596. (b) Synthetic and mechanistic details will
be presented in a complete account: Older, C. M.; Stryker, J. M., manuscript
in preparation.
(13) (a) Lutsenko, Z. L.; Aleksandrov, G. G.; Petrovskii, P. V.; Shubina,
E. S.; Andrianov, V. G.; Struchkov, Yu. T.; Rubezhov, A. Z. J. Organomet.
Chem. 1985, 281, 349. (b) Lutsenko, Z. L.; Petrovskii, P. V.; Bezrukova,
A. A.; Rubezhov, A. Z. Bull. Acad. Sci. USSR, DiV. Chem. Sci. 1988, 37,
735; IzV. Akad. Nauk SSSR, Ser. Khim. 1988, 855.
(18) Alkylation of [(η5-cyclopentadienyl)Fe(η6-RnC6H6-n)]+: (a) Khand,
I. U.; Pauson, P. L.; Watts, W. E. J. Chem. Soc. (C) 1968, 2257. (b) Khand,
I. U.; Pauson, P. L.; Watts, W. E. J. Chem. Soc. (C) 1969, 2024.
(19) Fagan, P. J.; Ward, M. D.; Calabrese, J. C. J. Am. Chem. Soc. 1989,
111, 1698.
(14) (a) Iridium: Schwiebert, K. E.; Stryker, J. M. Organometallics 1993,
12, 600. (b) Cobalt: Nehl, H. Chem. Ber. 1993, 126, 1519.

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2787
Scheme 2a
Scheme 3a
a Conditions: (i) HBF4‚Et2O, acetone-d6, 20 °C, 10 min; (ii) HOTf,
CH2Cl2 (anhydrous), -35 °C f room temperature; (iii) MeOH
(anhydrous), room temperature, 3 h (8) or 5 d (9).
a Conditions: (i) C6Me6, AqBF4, CH2Cl2, 35 °C; (ii) LiEt3BH, THF,
0 °C f room temperature; (iii) C5H6, Na2CO3, EtOH, ∆; (iv) C6Me5H,
AgBF4, CH2Cl2, 35 °C; (v) MeLi, THF, -78 °C f room temperature.
sulfonic acid (triflic acid) in rigorously anhydrous dichlo-
romethane-d2, however, leads to the quantitative formation of
an intermediate, identified spectroscopically as the cationic
agostic hydride complex [(C5H5)Ru(η4-1,2,3,4,5-exo-6-endo-
hexamethylcyclohexadiene)]+OTf- (8). The complex exhibits
an upfield doublet of septets at δ -4.81 (J ) 13.3, 2.6 Hz,
respectively), characteristic of a fluxional agostic hydride ligand,
rapidly equilibrating between the ends of the unsaturated moiety
via a symmetric η5-cyclohexadienyl ruthenium hydride transition
state or intermediate. The large vicinal coupling constant
confirms the trans disposition of the methyl substituents,
supporting a mechanism involving direct protonation at the
ruthenium center. Consistent with this fluxional structure, the
1
field18,20 methyl doublet (δ 0.95) in the H NMR spectrum,
accompanied by a quartet resonance for the methine position
(δ 2.44). While structurally related complexes bearing a similar
exo-methine hydrogen generally show an anomalously low-
frequency C-H stretch in the infrared spectrum (ca. 2750-
2820 cm-1),20 the lowest frequency C-H absorption in the
infrared spectrum of complex 3 appears at 2847 cm-1, only
marginally lower in energy than the 2852 cm-1 absorption
observed for the stereoisomeric complex 7 (vide infra). The
corresponding cyclopentadienyl complex 5 was similarly pre-
pared from [(η5-C5H5)Ru(η6-hexamethylbenzene)]+Cl- (4),21
which was itself synthesized from [(η6-hexamethylbenzene)-
22a
1
RuCl2]2
and cyclopentadiene using a modification of a
complex shows time-averaged symmetry in both the H and
literature procedure (Scheme 2).22 The off-white solid was
characterized spectroscopically and used for subsequent trans-
formations without further purification.
13C NMR spectra at room temperature and a methyl resonance
(δ 1.67, d, 6H) coupled to the agostic hydride by an unusually
small vicinal coupling constant (J ) 2.7 Hz), as established by
homonuclear decoupling experiments. This dynamic behavior
and the low barrier to the 1,5-intraligand hydride transfer are
closely analogous to several extensively investigated agostic
complexes formed by protonation of ruthenium η5-pentadienyl
complexes.23,24 Attempted isolation of this intermediate results
in the development of product mixtures, and no further
characterization was pursued.
Stereoisomeric η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl
complex 7 was prepared from [(η5-C5Me5)Ru(η6-pentamethyl-
benzene)]+BF4 (6)19 by regioselective alkylation with meth-
-
yllithium at low temperature (Scheme 2). The selective addition
to the single unsubstituted position on the pentamethylbenzene
ligand is remarkable but consistent with the results reported by
Pauson et al. for alkylation of the analogous pentamethylbenzene
complexes of iron.18 Complex 7 was purified by chromatography
on neutral alumina and isolated as a crystalline solid in moderate
yield. The 6-exo-methyl substituent in complex 7 is strongly
shielded by the ruthenium center, resonating at δ 0.11 in the
1H NMR spectrum.
Protonation of 6-endo-Hexamethylcyclohexadienyl Com-
plexes. The reactivity of the 6-endo-methyl complexes 3 and 5
was first investigated. Contrary to expectations, protonation of
cyclopentadienyl complex 5 with tetrafluoroboric acid in
acetone-d6 at room temperature leads to the rapid regeneration
of the cationic η6-hexamethylbenzene complex 4, presumably
via loss of dihydrogen (Scheme 3). The reaction is clean and
quantitative, as determined both by in situ spectroscopic analysis
and by product isolation. Protonation using trifluoromethane-
Protonation of the corresponding pentamethylcyclopentadi-
enyl complex 3 with triflic (or tetrafluoroboric) acid in dichlo-
romethane-d2 under strictly anhydrous conditions provides the
isostructural fluxional agostic hydride complex 9 (Scheme 3),
as deduced by analysis of the closely analogous spectroscopic
1
data. The agostic hydride resonance in the H NMR spectrum
is shifted to higher field (δ -5.82), consistent with the greater
shielding induced by the pentamethylcyclopentadienyl ring. This
signal shows the expected large vicinal coupling constant to
the 6-exo proton, but the coupling to the flanking methyl
substituents is unresolved at room temperature, despite the
(23) (a) Cox, D. N.; Roulet R. J. Chem. Soc., Chem. Commun. 1988,
951. (b) Cox, D. N.; Roulet R. J. Chem. Soc., Chem. Commun. 1989, 175.
(c) Lumini, T.; Cox, D. N.; Roulet R.; Chapuis, G.; Nicolo, F. HelV. Chim.
Acta 1990, 73, 1931. (d) Lumini, T.; Cox, D. N.; Roulet R.; Schenk, K. J.
Organomet. Chem. 1992, 434, 363. (e) Cox, D. N.; Lumini, T.; Roulet R.
J. Organomet. Chem. 1992, 438, 195.
(24) (a) Gleiter, R.; Hyla-Kryspin, I.; Ziegler, M. L.; Sergeson, G.; Green,
J. C.; Stahl, L.; Ernst, R. D. Organometallics 1989, 8, 298. (b) Newbound,
T. D.; Stahl, L.; Ziegler, M. L.; Ernst, R. D. Organometallics 1990, 9, 2962.
(c) Trakarnpruk, W.; Hyla-Kryspin, I.; Arif, A. M.; Gleiter, R.; Ernst, R.
D. Inorg. Chim. Acta 1997, 259, 197. (d) Bosch, H. W.; Hund, H.-U.;
Nietlispach, D.; Salzer, A. Organometallics 1992, 11, 2087.
(20) See: Faller, J. W. Inorg. Chem. 1980, 19, 2857 and references
therein.
(21) Complex 2 was previously prepared by the treatment of [(η6-
hexamethylbenzene)RuCl2]2 with (C5H5)Tl: Baird, M. C. Zelonka, R. A.
J. Organomet. Chem. 1972, 44, 383.
(22) (a) Bennett, M. A.; Huang, T.-N.; Matheson, T. W.; Smith, A. K.
Inorg. Synth. 1982, 21, 74. (b) Bennett, M. A.; Matheson, T. W. J.
Organomet. Chem. 1978, 153, C25.

2788 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
Older and Stryker
Scheme 4a
Scheme 5
a Conditions: (i) HOTf, CH2Cl2 (anhydrous), room temperature; (ii)
H2O, room temperature, 1 h.
signal. Similarly, irradiation of the upfield 6-exo-methyl reso-
nance leads to saturation transfer to the remaining (broad) methyl
signals at δ 2.25 and 1.71. Taken together, the spectroscopic
data are consistent with the occurrence of two distinct fluxional
processes operating at very different rates: facile 1,5-intraligand
transfer of the agostic hydrogen, similar to that observed in the
endo-methyl series, and a higher activation energy process that
shifts the agostic interaction to the adjacent endo-methine
hydrogen (Scheme 5). The latter process is slow below room
temperature but at higher temperature provides a mechanism
for complete equilibration of the cyclohexadiene methyl groups
and methine hydrogen atoms. Structurally similar but nonflux-
ional exo-methyl intermediates have been previously observed
spectroscopically by Chaudret et al., although the complexes
were described as classical cationic η5-cyclohexadienyl ruthe-
nium hydrides.9a,b
Agostic exo-methyl cyclohexadiene complex 10 persists
indefinitely in solution under strictly anhydrous conditions. Upon
addition of distilled water (6 equiv) to a solution of the agostic
complex prepared from the protonation using 1 equiv of acid,
quantitative conversion to the cationic η6-pentamethybenzene
complex 6 and methane (δ 0.2 in CD2Cl2, 22 °C) is observed
within 1 h at room temperature (Scheme 4). As observed in the
6-endo-methyl complex, this extrusion is also suppressed in the
presence of excess acid, provided that the amount of acid is
sufficient to scavenge all of the base present in solution.25 No
other organic or organometallic intermediates or byproducts are
detected by spectroscopic analysis.
Isotopic Labeling Studies. Two deuterium labeling experi-
ments confirm the exchange equilibria suggested by variable-
temperature NMR spectroscopy and provide additional insight
into the mechanism of carbon-carbon bond activation in this
system. Treatment of 6-exo-hexamethylcyclohexadienyl complex
7 with deuterium-labeled26 triflic acid (CF3SO3D, 1.2 equiv) in
nominally anhydrous acetone-d6 leads to the instantaneous
formation of the deuterium-labeled agostic cyclohexadiene
complex 10, as determined by analysis of the 1H NMR spectrum.
This reaction is followed by slow formation of deuterium-labeled
η6-pentamethylbenzene complex 6-d1 and evolution of methane
over a period of 24 h (Scheme 6). To the limits of spectroscopic
detection, the deuterium label is located exclusively in the η6-
pentamethylbenzene ligand methine position, although the
methane produced also contains an unquantified but substantial
amount of methane-d1 (CH3D, δ 0.13, 1:1:1 triplet). These
results are consistent with a mechanism in which the kinetically
formed cationic intermediate 10-d1, with a C-D agostic
appearance of resolved coupling (J ) 2.4 Hz) in the corre-
sponding methyl resonance.
Complexes 8 and 9 persist indefinitely in solution, so long
as anhydrous conditions are maintained. Upon addition of water
or methanol (5-7 equiv), however, both complexes extrude
hydrogen and convert to the corresponding η6-hexamethylben-
zene complexes 2 and 4 (Scheme 3). The reaction of cyclo-
pentadienyl complex 8 with water is quantitative and complete
within minutes at room temperature, but the alcohol is a much
less effective catalyst: upon treatment with methanol, the
dehydrogenation of complex 8 and pentamethyl complex 9
requires hours and days, respectively, to go to completion. As
determined in a series of qualitative experiments, both dehy-
drogenation reactions are inhibited by excess acid and sup-
pressed when the molar amount of protic acid exceeds the
amount of base present in or added to the system.
Protonation of the 6-exo-Hexamethylcyclohexadienyl Com-
plex. Carbon-Carbon Bond Activation. Consistent with the
reactivity observed in the 6-endo series, but contrary to our
initial assumptions, protonation of 6-exo-hexamethylcyclohexa-
dienyl complex 6 with either triflic acid or tetrafluoroboric acid
leads to exclusive activation of the exo-methyl carbon-carbon
bond. Similar to the endo-methyl derivatives, an intermediate
complex can be prepared and characterized spectroscopically
under rigorously anhydrous conditions. Thus, treatment of
complex 7 with a slight excess of triflic acid in anhydrous
dichloromethane-d2 at room temperature results in the formation
of the cationic agostic hydride complex [(C5Me5)Ru(η4-1,2,3,4,5-
exo-6-exo-hexamethylcyclohexadiene)]+OTf- (10) (Scheme 4).
Spectroscopic analysis of this material at room temperature is
relatively uninformative: the complex exhibits resolved but very
1
broad resonances in the H NMR spectrum. The broad signal
at δ -5.83 nonetheless suggests the presence of an agostic
hydride ligand. The 13C NMR spectrum at room temperate is
even less informative, with only the pentamethylcyclopentadi-
enyl resonances evident above the baseline. Reacquisition of
the data at low temperature (-80 °C) resolves the (nonagostic)
6-endo-hydrogen signal at δ 2.53 (br q, J ) 6.6 Hz) coupled to
the 6-exo-methyl resonance at δ 0.22 (d, J ) 6.5 Hz, 3H), which
is shifted considerably further upfield from the 6-endo-methyl
resonances20 in either complex 8 or 9 (δ 1.12 and 1.21,
respectively). The agostic hydride signal, however, remains an
unresolved multiplet even at low temperature, reflecting both
the small vicinal coupling constant to the 6-endo-hydrogen and
the persistent fluxionality arising from the fast 1,5-intraligand
hydride transfer. Both the 1H and 13C NMR spectra at -80 °C
reflect the higher symmetry structure that results from this
fluxional process.
At higher temperatures in 1,1,2,2-tetrachloroethane-d2, the
agostic and free methine signals further broaden, as do the arene
methyl resonances, coalescing at approximately 80 °C. The
higher temperature fluxional process was further probed by
irradiation of the agostic hydride resonance at room temperature,
which results in spin saturation transfer to the broad endo-
hydrogen resonance at δ 2.59 and complete suppression of this
(25) The reaction of 6-exo complex 7 in dichloromethane with 2.7 equiv
of triflic acid in the presence of 6.7 equiv of water, for example, proceeds
50% to completion after 48 h at room temperature, as monitored by NMR
spectroscopy. Under otherwise identical conditions but in the presence of
6 equiv each of triflic acid and water, no detectable conversion is observed
over the same time period. Upon addition of a further 5 equiv of water to
this reaction mixture, however, the conversion to pentamethylbenzene
complex 6 proceeds approximately 60% to completion after an additional
48 h.
(26) Deuterium-labeled triflic acid was prepared in situ by exchange of
triflic acid and acetone-d6. This exchange appears to be rapid at room
temperature, in contrast to the isotopic exchange of tetrafluoroboric acid in
the same medium, which proceeds only slowly.

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2789
Scheme 6
Scheme 7
interaction, equilibrates rapidly to form the thermodynamically
more favored27 intermediate 10′-d1 prior to activation of the
6-exo carbon-carbon bond. This proposal is supported by the
absence of the signal assigned to the 6-endo-hydrogen (δ 2.7
assisted protolytic dealkylation by Brønsted acid with complete
selectivity for activation of the exo substituent (Scheme 7). The
acid is liberated from the intermediate agostic cyclohexadiene
cation in a preequilibrium induced by a base, the source of which
can be the solvent, adventitious or added water, or, perhaps,
the weakly basic counterion. In addition to accommodating the
stereoselectivity of the activation process, this mechanistic
rationale is consistent with and strongly supported by all of the
experimental evidence. The thermal stability of the 6-exo and
6-endo agostic η4-methylcyclohexadiene cations 8-10 under
strictly anhydrous conditions is inconsistent with a mechanism
involving thermal radical scission of an exo-methyl substituent
or any pathway invoking direct metal-mediated activation of
an endo-methyl substituent or concerted extrusion of methane
from a classical hydrido 6-endo-methyl-η5-cyclohexadienyl
intermediate. Standard steady-state or preequilibrium kinetic
analysis of the mechanism proposed in Scheme 7 predicts that
the rate of dealkylation will be first-order dependent on the
concentration of base and independent of the concentration of
acid above that required to produce the agostic complex 10, as
observed experimentally. The inhibition of the reaction observed
in the presence of excess strong acid thus arises from the
efficient scavenging of the available base. Inhibition by excess
acid equally eliminates from consideration pathways involving
protonolysis of an already cationic precursor. Also consistent
with an acid/base mechanism is the qualitative observation of
rate differences for both dealkylation and dehydrogenation as
a function of the Brønsted base (water vs methanol) and the
significantly slower dehydrogenation rate observed for the more
strongly basic pentamethylcyclopentadienyl ligand system in
complex 8 over the parent cyclopentadienyl complex 9.
1
in acetone-d6) in the H NMR spectrum acquired immediately
after addition of the acid. The small amount of methane-d1 can
arise from methane loss from undetected 10-d1 or, more likely,
from the formation of doubly labeled 10-d2 by incorporation of
a second deuterium atom via exchange of the agostic protium
in 10′-d1 with the slight excess of labeled triflic acid used in
the experiment. This exchange requires that the agostic inter-
mediate 10 be in equilibrium with the neutral hexamethylcy-
clohexadienyl complex 7 and free acid during the course of the
demethylation reaction. Importantly, it also suggests an alterna-
tive mechanism for carbon-carbon bond activation of the exo-
methyl complex 10 (vide infra).
The high-temperature exchange equilibria revealed by NMR
spectroscopy were probed chemically by treatment of 6-exo-
28
CD3-pentamethylcyclohexadienyl complex 7-d3 with triflic
acid in wet dichloromethane-d2 (eq 5). This reaction provides
pentamethylbenzene complex 6 with equivalent deuterium
incorporation into all arene methyl positions, confirming
complete endo-hydrogen scrambling in this system. Consistent
with this observation, the methane released into solution is
largely unlabeled, accompanied by a minor fraction of the
tentatively identified CD3H (δ 0.12, br s). Repetition of this
experiment using tetrafluoroboric acid in acetone-d6 also
provides equivalent incorporation of the labeled methyl group
into all arene positions. In addition, however, a small amount
of deuterium appears in the unsubstituted arene position,
presumably a result of the slow H/D exchange between the acid
and acetone-d6.26
The deuterium labeling experiment (Scheme 6) suggests that
protonation of the neutral hexamethylcyclohexadienyl complex
is reversible under the reaction conditions. To confirm that protic
acid is liberated from the cationic intermediate and to probe
for the intervention of alternative dealkylation mechanisms, a
crossover experiment was designed in which no external source
of acid is provided. Thus, the [3 + 2] cycloaddition of (η6-
hexamethylbenzene)Ru(η3-allyl)OTf (11) and diphenylacetylene
was conducted in the presence of the neutral 6-endo-hexameth-
ylcyclohexadienyl complex 3 (eq 6). Without intervention, the
diphenylacetylene cycloaddition proceeds slowly and quantita-
tively to yield methane and (η5-1,2-diphenylcyclopentadienyl)-
ruthenium(η6-pentamethylbenzene)+OTf- under the reaction
conditions, presumably via an intermediate analogous to 6-exo
complex 10.12 The use of the 6-endo-hexamethylcyclohexadienyl
complex 3 in this experiment was based on its established
efficiency as an acid scavenger. In the event, the reaction
proceeds quantitatively to give the crossover products, [(η5-C5-
Me5)Ru(η6-C6Me6)]+OTf- (2-OTf) and the neutral (η5-1,2,3,4,5,6-
Mechanism of the Carbon-Carbon Bond Activation.
Taken together, our experimental observations strongly suggest
that the mechanism for low-temperature carbon-carbon bond
activation of 6-alkylcyclohexadienyl ligands involves a ruthenium-
(27) The thermodynamic preference for CH over CD agostic interactions
has been documented: (a) Brookhart, M.; Green, M. L. H. J. Organomet.
Chem. 1983, 250, 395. (b) Calvert, R. B.; Shapley, J. R. J. Am. Chem. Soc.
1978, 100, 7726.
(28) The 6-exo-CD3-pentamethylcyclohexadienyl complex 7-CD3 was
prepared by alkylation of cationic η6-pentamethybenzene complex 6 with
CD3Li, under conditions otherwise identical to those described for the
synthesis of the unlabeled material.

2790 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
Older and Stryker
Figure 2.
Scheme 8
exo-hexamethylcyclohexadienyl)ruthenium η5-1,2-diphenylcy-
clopentadienyl (12),29 unambiguously establishing the presence
of a Brønsted acid in the medium.30 This reaction also confirms
the competency of exo-methyl complex 12 as an intermediate
in the dealkylative [3 + 2] cycloaddition reaction.12b A related
electrophilic abstraction of an exo carbon-carbon bond was
recently postulated for acid-mediated cyclopropane ring opening
of (η4-spirobicyclo[2.4]hepta-4,6-diene)Fe(CO)3.31
The specificity and low activation barrier for cleavage of the
exo carbon-carbon bond can be rationalized by considering that
significant stabilization of the transition state may be provided
from interaction of a filled metal-centered (or mixed metal/
cyclohexadienyl)32 molecular orbital with the σ*-orbital of the
activating carbon-carbon bond (Figure 2a). This bonding
interaction transforms without discontinuity into a metal-arene
bonding orbital as the bent cyclohexadienyl ring32 converts into
the planar arene ligand. A similar molecular orbital interaction
is considered to be responsible for the low-frequency C-H
vibrational mode observed for the exo carbon-hydrogen bond
in the infrared spectra of other η5-cyclohexadienyl complexes20
and is consistent with the shielding experienced by the exo-
methyl substituent in the 1H NMR spectrum of complex 7. The
metal is, in effect, displacing the methyl substituent by a
concerted nucleophilic substitution with electrophilic assistance
from the proton, which functions to stabilize the anionic alkyl
leaving group. The high activation barrier for direct metal-
mediated endo activation can be attributed to the strong metal
T alkyl antibonding interaction experienced by the endo
substituent as the η5-cyclohexadienyl ligand distorts to position
the substituent in proximity to the metal (Figure 2b).32 This four-
electron repulsion is mitigated by unsaturation at the metal, but
as demonstrated by Dimauro and Wolczanski,9d the barrier to
carbon-carbon bond activation remains high.
Dealkylation Using Lewis Acids. Additional support for the
protolytic mechanism for carbon-carbon bond activation in
6-exo-methylcyclohexadienyl ruthenium complexes is provided
by the reactions of the neutral precursor with Lewis acids.
Although carbon-carbon bond activation is, indeed, observed
under mild conditions, surprising differences are found in the
mechanistic detail. Thus, the reaction of η5-1,2,3,4,5,6-exo-
hexamethylcyclohexadienyl complex 7 with tris(perfluorophen-
yl)borane33 at low temperature under strictly anhydrous condi-
tions leads to instantaneous and quantitative hydride abstraction
from a methyl substituent, providing the 5-methylene-1,3-
cyclohexadiene complex 13, as determined by in situ NMR
spectroscopic analysis (Scheme 8). This complex could not be
isolated without suffering partial further conversion, but spec-
troscopic analysis strongly supports the structural assignment.
The exocyclic methylene moiety is defined by the appearance
of two downfield resonances for the inequivalent methylene
hydrogen atoms (δ 4.33 and 3.54), bonded directly to an sp2-
hybridized carbon (δ 76.2, 1JC-H ) 161 Hz), as determined by
13C-1H heteronuclear correlated NMR spectroscopy. The
chemical shifts of the remaining quaternary carbon centers
suggest strongly that the s-cis,s-trans-triene fragment is η6-
coordinated to the metal.34 The 6-exo-methyl and 6-endo-
hydrogen resonances remain relatively unperturbed at δ 0.66
(d, J ) 6.7 Hz) and δ 1.84 (q, J ) 6.7 Hz), respectively.
Although the hydride abstraction may be reversible under the
reaction conditions, tris(perfluorophenyl)borane does not activate
the carbon-carbon bond until protic acid is present, typically
as a result of the introduction of water. Complex 13 thus persists
indefinitely at room temperature under anhydrous conditions
but converts slowly to η6-pentamethylbenzene cation 6 and
methane in the presence of even a trace of moisture; the reaction
with added water (7.5 equiv) is complete within a few hours at
room temperature. The counterion produced in this reaction is
tentatively identified as the known borate anion (C6F5)3BOH,35
on the basis of the broad hydroxyl absorption at 3670 cm-1 in
the infrared spectrum.
In contrast, the reaction of hexamethylcyclohexadienyl com-
plex 7 with aluminum tribromide under otherwise identical
conditions proceeds directly to η6-pentamethylbenzene complex
6 without evidence of intermediate hydride abstraction (Scheme
8). The counterion in this complex is tentatively identified as
-
MeAlBr3 on the basis of a characteristic broad singlet at δ
0.00 (3H) in the 1H NMR spectrum.36 It is interesting to consider
these results in the context of the partial dealkylation of
(29) The identity of this complex was confirmed by independent synthesis
starting from [(η5-1,2-diphenylcyclopentadienyl)Ru(η6-pentamethylbenzene)]-
OTf12a and MeLi in tetrahydrofuran at low temperature, similar to the
synthesis of complex 7. See Experimental Section.
(30) It is, of course, reasonable to propose that the cationic agostic
complex itself acts as the Brønsted acid to mediate dealkylation or
dehydrogenation of the neutral precursor. In the dealkylation reaction,
however, at least some neutral intermediate must be formed in solution by
loss of the proton to the medium.
(33) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1991, 113,
3623. Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116,
10015.
(34) A similar cyclic s-cis,s-trans triene complex, but embedded in a
steroid framework, has been reported by Chaudret et al.9a
(35) Siedle, A. R.; Newmark, R. A.; Lamanna, W. M. Organometallics
1993, 12, 1491. Schaefer, W. P.; Quan, R. W.; Bercaw, J. E. Acta
Crystallogr. C 1993, 49, 878.
-
-
(36) MeAlBr3 appears to be unknown; the aluminate MeAlCl3 has
been described: Sangokoya, S. A.; Pennington, W. T.; Robinson, G. H. J.
Crystallogr. Spectrosc. Res. 1990, 20, 53.
(31) Fu, Y.-T.; Chao, P.-C.; Liu, L.-K. Organometallics 1998, 17, 221.
(32) Hoffmann, R.; Hofmann, P. J. Am. Chem. Soc. 1976, 82, 598.

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2791
Scheme 9
70 ratio. While the former byproduct clearly results from a one-
electron oxidation of the ruthenium complex, the latter could
arise either from direct abstraction of methyl anion by the trityl
cation or from methyl radical abstraction following a one-
electron oxidation. To constrain the mechanistic pathway to
oxidation only, parallel experiments were performed using
ferricinium (Scheme 9). The oxidation of 6-exo-methyl complex
-
7 with Cp2Fe+PF6 at room temperature leads to the isolation
of a mixture of dehydrogenation and dealkylation products 2
and 6 in a ratio of 20:80, similar to the product mixture obtained
from trityl cation. As previously noted,37 however, the product
distribution is profoundly influenced by the reaction tempera-
ture: when the oxidation is conducted at -78 °C, the ratio of
complexes 2 and 6 changes to 67:33, now favoring endo-
hydrogen rather than exo-methyl abstraction.40 For comparison,
the oxidation of the corresponding endo-methyl complex 3 with
ferricinium at low temperature leads to exclusive loss of the
exo-hydrogen, providing the η6-hexamethylbenzene complex 2
in quantitative yield (Scheme 9).
hexaalkylbenzenes observed during aluminum trichloride-
induced exchange reactions with ferrocene and ruthenocene.10
The highest fraction of dealkylated arene complex is reported
from reactions mediated by wet AlCl3; little or no dealkylation
is obtained with highly purified reagent.10b Although no
mechanistic details were determined, the dealkylations were
attributed to the presence of residual protic acid in the Lewis
acid, a result of contamination by water.
The quantitative demethylation observed in the thermal
reaction of cationic agostic complex 10 at or below room
temperature is thus inconsistent with a mechanistic pathway
involving odd-electron intermediates. The experimental data
converge on the conclusion that the mechanism of methane
extrusion from such complexes proceeds via reversible depro-
tonation and direct methyl abstraction by the Brønsted acid.
6-Alkylcyclohexadienyl Aromatization by Carbon-Car-
bon Bond Activation. Literature Dealkylations Revisited. The
relevance of this protolytic mechanism to the demethylation
reactions reported by Chaudret et al.9a,b and Itoh et al.9c was
addressed by extending our investigation to intermediates
proposed for or observed in the course of these transformations.
Thus, the otherwise unsubstituted 6-exo-methylcyclohexadienyl
complex 14 was prepared from the known complex [(η5-C5-
Activation by One-Electron Oxidants. Dimauro and Wolc-
zanski have established that the kinetic barrier to carbon-carbon
bond activation in 6,6-dimethylcyclohexadienyl ruthenium(II)
complexes is significantly reduced upon one-electron oxidation.9d
It has also been shown that one-electron oxidation of 5-exo-
alkyl-η4-cyclohexadiene complexes of iron leads to activation
of the alkyl substituent at room temperature, although in this
case the substituents (benzyl, dithiane) are activated toward
radical extrusion.37 Oxidation of the same iron complexes,
however, at low temperature (-50 °C) instead affords exclusive
abstraction of an endo-hydrogen rather than carbon-carbon
bond scission. To evaluate the potential relevance of one-
electron oxidation to the 6-exo-methylcyclohexadienyl dealky-
lation mechanism, the reactions of exo- and endo-ruthenium
complexes 3 and 7 with one-electron oxidants were investigated.
The reactivities of both triphenylmethyl (trityl) and ferricinium
cations were evaluated, the former because it can react as either
an outer-sphere oxidant or an alkyl abstraction agent, and the
latter because it functions exclusively as an outer-sphere
oxidant.38
The addition of stoichiometric trityl cation to a solution of
(η5-C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl)
(7) in dichloromethane at -35 °C, followed by warming to room
temperature, results in rapid conversion to a mixture of
η6-hexamethylbenzene and η6-pentamethylbenzene ruthenium
complexes 2 and 6, derived from carbon-hydrogen and
carbon-carbon bond activations, respectively (Scheme 9). The
products are formed in a ratio of 30:70 favoring carbon-carbon
bond cleavage and are accompanied by the formation of a
mixture of 3-benzhydrylidene-6-triphenylmethyl-1,4-cyclohexa-
diene39 (trityl dimer) and 1,1,1-triphenylethane in the same 30:
Me5)Ru(η6-C6H6)]+PF6 (16)41 by alkylation with methyl-
-
lithium. Protonation of the ruthenium complex 14 with strong
acid was then expected to provide the cationic (and presumably
agostic) complex 1-exo (Scheme 10), the proposed intermediate
in both the aromatization of 3-methylcyclohexene induced by
the (C5Me5)Ru+ fragment9a and the ruthenium-mediated [4 +
2] cycloaddition reaction of 1,3-pentadiene and acetylene.9c
The addition of tetrafluoroboric acid or triflic acid to a
solution of 6-exo-methyl complex 14 in acetone results in rapid
formation of aromatized products (Scheme 10). The product
mixture consists of a 91:9 ratio of the known9 η6-toluene and
η6-benzene complexes 15 and 16, arising from competitive
dehydrogenation and demethylation pathways, respectively. No
evidence for the formation of a cationic agostic intermediate is
observed in reactions run in dichloromethane under strictly
anhydrous conditions, and neither the solvent nor the water
content significantly impacts the product ratio. The proportion
of dealkylation to dehydrogenation obtained is indistinguishable
from that reported for the diene cycloaddition and only slightly
higher than that observed in the methylcyclohexene aromati-
zation.9 It is thus unnecessary to propose the formation of a
sterically unfavorable 6-endo-methylcyclohexadienyl intermedi-
ate to account for the minor fraction of dealkylated product:
complex 1-exo is by itself competent to produce both of the
η6-arene products obtained in these prior inVestigations. The
product mixture is readily rationalized by assuming that the
(37) (a) Mandon, D.; Astruc, D. Organometallics 1989, 8, 2372. (b) For
other oxidative carbon-carbon bond activations, see ref 17 and the
following: Nesmeyanov, A. N.; Vol’kenau, N. A.; Shilovtseva, L. S.;
Petrakova, V. A. J. Organomet. Chem. 1975, 85, 365.
(40) The oxidation reactions were not performed in sealed vessels;
consequently, the composition of the volatile fractions remains undeter-
mined.
(41) Bosch, H. W.; Hund, H.-U.; Nietlispach, D.; Salzer, A. Organo-
metallics 1992, 11, 2087.
(38) (a) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877. (b)
Beck, W.; Sunkel, K. Chem. ReV. 1988, 88, 1405.
(39) Spectroscopic data: Volz, H.; Lotsch, W.; Schnell, H.-W. Tetra-
hedron 1970, 26, 5343.

2792 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
Older and Stryker
Scheme 10
Scheme 11a
a Conditions: (i) [(C5Me5)RuCl]4, THF; then Rieke Zn0, room
temperature, 12 h; (ii) HBF4‚Et2O, CD2Cl2, room temperature; (iii)
HBF4‚Et2O (2.2 equiv), H2O (10 equiv), CH2Cl2, room temperature,
4d; (iv) Ph3C+BF4-, CH2Cl2, room temperature, 1 h; (v) H2O (10 equiv),
room temperature.
6-exo-methyl intermediate equilibrates rapidly by migration of
the agostic hydride, analogous to the degenerate fluxionality
determined for the agostic η4-hexamethylcyclohexadiene com-
plex 10. The equilibration provides a mixture of agostic alkyl-
η4-cyclohexadiene complexes statistically and, perhaps, elec-
tronically favoring isomers in which the methyl substituent is
positioned such that it cannot be activated (Scheme 10). Such
equilibration is precedented by previously reported isomeriza-
tions of 6-exo-alkylcyclohexadienyl osmium hydride com-
plexes.42 In contrast to ruthenium, however, the η5-cyclohexa-
dienyl osmium cations assume a classical hydride structure and
do not undergo ligand aromatization, both presumably a
consequence of the greater basicity of the third row metal center.
The 6-exo-methylcyclohexadienyl complex 14 also aromatizes
upon reaction with Lewis acids and one-electron oxidants, but
the proportion of dealkylation observed strongly depends on
the choice of reagent (eq 7). The reaction with tris(perfluo-
Finally, the question of steroid aromatization was addressed,
to assess the relevance of protolytic activation mechanisms in
this ruthenium-mediated demethylation reaction (eq 3).9a,b
Steroid dealkylation proceeds unambiguously by coordination
of the unsaturated metal fragment to the steroid R-face and
activation of the exo-methyl substituent. Based on the high
activation barrier for this reaction and the observation of trace
amounts of ethane in certain A-ring aromatization reactions, a
free radical mechanism was postulated involving homolytic
scission of the carbon-carbon bond.
To address the mechanism of this reaction, the aromatization
of ergosterol was reinvestigated, beginning with an independent
synthesis of Chaudret’s partially characterized hydridoruthenium
intermediate.9b Thus, the complex formed in situ from the
reaction of [(C5Me5)RuCl]4 and ergosterol in tetrahydrofuran
was treated with an excess of zinc dust to yield the neutral η5-
cyclohexadienyl complex, (C5Me5)Ru(η5-9-dehydroergosterol)
(17) in modest, unoptimized, yield (Scheme 11). Protonation
of this complex with tetrafluoroboric acid in anhydrous dichlo-
romethane-d2 at room temperature leads to the quantitative
formation of cationic agostic η4-ergosterol complex 18, identi-
fied and characterized by spectroscopic analysis but not isolated.
The spectroscopic data are fully consistent with the intermediate
proposed by Chaudret et al., but the hydride resonance at δ
-5.00 in CD2Cl2 is closely analogous to the corresponding
signals in agostic cyclohexadiene complexes 8-10, strongly
suggesting that this complex adopts an agostic structure rather
than the previously assigned classical η5-cyclohexadienyl hy-
dride structure. In contrast to the fluxional character of
complexes 8-10, agostic ergosterol complex 18 is structurally
rigid at room temperature on the NMR time scale, presumably
a consequence of the structural dissymmetry of the steroid
framework. The complex is stable indefinitely in solution at
room temperature under anhydrous conditions.
rophenyl)boron at room temperature, for example, provides η6-
toluene complex 15 with high (g98:2) selectivity; only a trace
of η6-benzene complex 16 is detected in the crude product
mixture. This result is consistent with the odd reluctance of tris-
(perfluorophenyl)boron to promote the demethylation of com-
plex 7 under anhydrous conditions. In contrast, the reaction with
aluminum tribromide proceeds to give a 90:10 ratio of hydrogen
to methyl abstraction, closely comparable to the reaction
mediated by protic acid. The highest selectivity for demeth-
ylation, however, is obtained upon one-electron oxidation with
the ferricinium salt, providing the aromatized complexes in a
42:58 ratio favoring the η6-benzene complex 16.43 Given these
results, it is curious that oxidizing agents are generally selected
for reactions requiring the abstraction of an endo-hydrogen.17,18,37b
Upon addition of water, however, the cationic ergosterol
complex undergoes slow conversion to the η6-arene complex
[(C5Me5)Ru(η6-neoergosterol)]+BF4- (19) and methane at room
temperature. The reaction of η5-dehydroergosterol complex 17
with tetrafluoroboric acid (2.2 equiv) in anhydrous dichlo-
romethane containing 10 equiv of water, for example, is
complete in 4 days at room temperature and provides the
aromatized product in quantitative yield. The substantial increase
in the rate of carbon-carbon bond activation obtained upon
addition of water to the reaction medium is inconsistent with a
free radical mechanism for activation of the carbon-carbon
bond. The slow rate of carbon-carbon bond activation under
(42) Werner, R.; Werner, H. Chem. Ber. 1984, 117, 161.
(43) The reaction of 6-exo-methylcyclohexadienyl complex 14 with
Ph3C+BF4- also favors methyl over hydrogen abstraction, but spectroscopic
analysis of the crude mixture suggests that the major product formed in
this reaction is an adduct resulting from trityl alkylation of the cyclohexa-
dienyl ligand. This material has not been isolated in pure form and remains
to be conclusively identified.

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2793
or in a drybox. Toluene, benzene, tetrahydrofuran, diethyl ether, hexane,
and pentane were distilled from sodium/benzophenone ketyl or from
sodium/potassium alloy/benzophenone ketyl. Reagent-grade acetone
was degassed and stored under nitrogen with no further purification.
Dichloromethane, dichloromethane-d2, and acetonitrile were distilled
from calcium hydride and degassed. Unless stated otherwise, all
reactions were carried out under a nitrogen atmosphere. IR spectra were
recorded on a Nicolet Magna IR 750 or a Nicolet 20SX spectropho-
Chaudret’s conditions (TfOH, 0.5 equiv of [(C5Me5)Ru(OMe)]2),
can be attributed to an anhydrous reaction medium containing
only 1 equiv of methanol, a less effective Brønsted base for
mediating protolytic demethylation (vide supra). The detection
of a trace of ethane from some steroid demethylation reactions
(but not in the case of ergosterol9b) can be rationalized, we
believe, by considering that the active demethylation agent under
Chaudret’s reaction conditions is likely to be CH3OH2+, formed
from deprotonation of agostic complex 18 by methanol. In
addition to the acidic protons, this reagent bears an electrophilic
methyl group, which may participate in the abstraction of the
angular methyl group as a minor pathway. Finally, we note that
the optimal method for the demethylation of ergosterol involves
neither Chaudret’s conditions nor our own: the reaction of η5-
9-dehydroergosterol complex 17 with trityl cation proceeds
quantitatively to demethylated complex 19 at room temperature
in less than 1 h!
1
tometer. H NMR and 13C NMR spectra were recorded on a Bruker
AM-400 (1H, 400 MHz; 13C, 100 MHz), a Bruker AM-360 (1H, 360
MHz), a Bruker AM-300 (1H, 300 MHz; 13C, 75 MHz), or a Varian
Unity-Inova 300 (1H, 300 MHz) spectrometer. High-resolution mass
spectra were obtained on a Kratos MS-50 spectrometer, and elemental
analyses were performed by the University of Alberta Microanalysis
Laboratories. The following compounds were prepared according to
published procedures: [(C5Me5)RuCl]4,19 [(C5Me5)Ru(η6-C6H6)]+PF6- 42
[(η6-C6Me6)RuCl2]2,22 [Cp2Fe]+PF6- 37a [(η5-C5H3Ph2)Ru(η6-C6Me5H)]+-
OTf-,12a and (C6Me6)Ru(C3H5)OTf.12a
,
,
[(C5Me5)Ru(η6-C6Me6)]+BF4- (2). A Schlenk flask equipped with
a septum was charged with [(C5Me5)RuCl]4 (0.100 g, 0.368 mmol/
Ru), hexamethylbenzene (0.310 g, 1.90 mmol), and silver tetrafluo-
roborate (0.076 g, 0.390 mmol). Dichloromethane (6 mL) was added
by syringe, and the reaction mixture immediately turned dark green.
The reaction was then heated to a gentle reflux for 5 h, gradually turning
to a light brown with a gray precipitate. The solution was then filtered
over Celite and evaporated to dryness. The tan residue was rinsed with
hexane to remove excess hexamethylbenzene and then dissolved in a
minimum of dichloromethane, and the product was precipitated with
diethyl ether. A white powder was collected (0.131 g, 73%), spectro-
scopically identical to the known [(C5Me5)Ru(η6-C6Me6)]+OTf-.19
(C5Me5)Ru(η5-1,2,3,4,5,6-endo-hexamethylcyclohexadienyl) (3). A
suspension of [(C5Me5)Ru(η6-C6Me6)]+BF4- (2, 0.050 g, 0.103 mmol)
in tetrahydrofuran (3 mL) was cooled to 0 °C, and LiEt3BH (1.0 M in
THF, 0.11 mL, 0.110 mmol, 1.07 equiv) was added via syringe. After
being stirred at 0 °C for 30 min, the reaction was warmed to room
temperature for 1 h before the solvent was removed in vacuo. The
residue was triturated with 2 × 3 mL of pentane, and the pentane
extracts were filtered over Celite and dried to give a white solid. The
crude product was then redissolved in pentane and filtered through a
plug of alumina (5% H2O). After evaporation of pentane, the product
(0.023 g, 56%) was obtained as a white crystalline solid and used
without further purification. IR (CH2Cl2 cast, cm-1): 2967 (s), 2898
(s), 2870 (s), 2847 (s), 1375 (s), 1026 (m). 1H NMR (400 MHz, CD2-
Cl2): δ 2.44 (q, J ) 7.1 Hz, 1H, exo-H), 2.07 (s, 3H, CH3), 1.64 (s,
6H, CH3), 1.61 (s, 15H, C5Me5), 1.05 (s, 6H, CH3), 0.95 (d, J ) 7.1
Hz, 3H, endo-CH3). 13C{1H} NMR (100 MHz, CD2Cl2): δ 96.8, 89.9,
87.0, 41.1, 35.3, 17.4, 15.9, 15.3, 15.0, 10.0. MS m/z (relative
intensity): calculated for C22H34102Ru (M+), 400.1704; found, 400.1655
(28.08); calculated for C22H33102Ru (M - H), 399.1626; found, 399.1625
(100.00); calculated for C21H31102Ru (M - CH3), 385.1469; found,
385.1464 (4.41).
[(C5H5)Ru(η6-C6Me6)]+Cl- (4). A small round-bottom flask equipped
with nitrogen inlet, reflux condenser, and stir bar was charged with
[(η6-C6Me6)RuCl2]2 (0.080 g, 0.239 mmol/Ru), anhydrous sodium
carbonate (0.100 g, 0.943 mmol), and ethanol (8 mL). Freshly cracked
cyclopentadiene (1.0 mL, 14.9 mmol) was then added via syringe, and
the reaction mixture was heated to reflux for 2 h, during which time
the suspended dimer dissolved and the solution turned pale yellow.
Afterward, the volatile liquids were removed in vacuo, and the crude
product was extracted into dichloromethane (2 × 4 mL), which was
then filtered over Celite. Addition of diethyl ether to the filtrate
precipitated the product as a white powder (0.072 g, 83%) which was
spectroscopically identical to the known [(C5H5)Ru(η6-C6Me6)]+Cl-.21
(C5H5)Ru(η5-1,2,3,4,5,6-endo-hexamethylcyclohexadienyl) (5). A
suspension of [(C5H5)Ru(η6-C6Me6)]+Cl- (4, 0.055 g, 0.151 mmol) in
tetrahydrofuran (3 mL) was cooled to 0 °C, and LiEt3BH (1.0 M in
THF, 0.18 mL, 0.180 mmol, 1.2 equiv) was added via syringe. After
being stirred at 0 °C for 30 min, the reaction was warmed to room
temperature for another hour before the solvent was removed in vacuo.
The residue was triturated with 2 × 3 mL of pentane, and the pentane
extracts were filtered over Celite and dried to a white solid. The crude
Conclusion. A novel mechanism for metal-mediated carbon-
carbon bond activation has been defined for the dealkylation
of cationic 6-exo-methylcyclohexadiene ruthenium hydride
complexes at low temperature. The reaction has been determined
to proceed by an acid/base mechanism, in which an exogenous
Brønsted base deprotonates the agostic hydride complex and
the resulting acid activates the exo carbon-carbon bond of the
6-exo-methylcyclohexadienyl complex by a protolytic mecha-
nism. Protolytic demethylation is the exclusive pathway ob-
served in the reactions of 1,2,3,4,5,6-exo-hexamethylcyclohexa-
dienyl complexes and in a steroid-derived 6,6-disubstituted
cyclohexadienyl complex. Under the same conditions, the
stereoisomeric 1,2,3,4,5,6-endo-hexamethylcyclohexadienyl com-
plexes undergo exclusive carbon-hydrogen bond activation,
releasing hydrogen by a similar protolytic mechanism. The facile
nature of the protolytic carbon-carbon bond activation contrasts
dramatically with the previously reported dealkylations of 6,6-
dimethyl-substituted cyclohexadienyl complexes,9d which ac-
tivate what is presumed to be the 6-endo-methyl carbon-carbon
bond by a concerted metal-mediated abstraction, but only by
traversing a substantially higher activation barrier. In the
protolytic mechanism, the transition metal still functions to
stabilize the transition state for carbon-carbon bond scission,
but it does so by a mechanism clearly distinct from typically
invoked â-alkyl elimination pathways. On the basis of this
investigation, the relevance of protolytic carbon-carbon bond
activation to the dealkylation of coordinated hexaalkylbenzenes
appears highly probable, in the context of both Lewis acid-
mediated cyclopentadienyl/arene exchange reactions and the
hexamethylbenzene demethylation we observed in the course
of [3 + 2] allyl/alkyne cycloaddition.12b
For monosubstituted 6-exo-methylcyclohexadienyl complexes
and their cationic conjugate acids, protolytic dealkylation is
observed as a minor reaction pathway, consistent with the minor
fraction of dealkylation product obtained from various previ-
ously reported metal-mediated cycloaddition and cyclohexene
aromatization reactions. As such, cationic exo-substituted agostic
η4-cyclohexadiene complexes thus constitute competent inter-
mediates in these poorly understood processes. While there
remains a number of obvious opportunities for quantification
of the kinetic relationships defined by this investigation, the
metal-assisted protolytic pathway constitutes a simple and
potentially exploitable alternative strategy for selective carbon-
carbon bond activation.
Experimental Section
General. All manipulations of air-sensitive compounds were per-
formed under prepurified nitrogen using standard Schlenk techniques

2794 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
Older and Stryker
product was redissolved in pentane and filtered through a plug of
alumina (5% H2O). After evaporation of pentane the product (0.031 g,
62%) was obtained as an off-white solid and used without further
purification. 1H NMR (400 MHz, CD2Cl2): δ 4.36 (s, 5H, C5H5), 2.41
(s, 3H, CH3), 2.21 (q, J ) 7.1 Hz, 1H, exo-H), 1.94 (s, 6H, CH3), 1.35
(s, 6H, CH3), 1.12 (d, J ) 7.1 Hz, 3H, endo-CH3). 13C{1H} NMR (100
MHz, CD2Cl2): δ 91.9, 91.2, 78.3, 39.9, 37.0, 18.9, 17.6, 17.4, 17.1.
6H, CH3), 1.31 (d, J ) 6.7 Hz, 3H, endo-CH3), -4.81 (d of sept, J )
13.3, 2.6 Hz, 1H, Hag).
Addition of MeOH to 8. Methanol (0.008 mL, 6.5 equiv) was added
via microsyringe to the solution of [(C5H5)Ru(η4-1,2,3,4,5-exo-6-endo-
hexamethylcyclohexa-1,3-diene)]+BF4 (8) prepared above and the
-
reaction monitored spectroscopically. Complete conversion to [(C5H5)-
Ru(η6-C6Me6)]+BF4- (4) was observed within 4 h at room temperature,
accompanied by a trace of decomposition products.
[(C5Me5)Ru(η6-C6Me5H)]+BF4 (6). A procedure similar to that
-
4
5
[(C Me5)Ru(η -1,2,3,4,5-exo-6-endo-hexamethylcyclohexa-1,3-
diene)]+OTf- (9). A solution of (C5Me5)Ru(η5-1,2,3,4,5,6-endo-hexa-
methylcyclohexadienyl) (3, 0.030 g, 0.075 mmol) in anhydrous CD2Cl2
was placed into an NMR tube equipped with a rubber septum. Triflic
acid (0.007 mL, 0.079 mmol) was then added via microsyringe and
the resulting solution analyzed by 1H and 13C{1H} NMR spectroscopy.
1H NMR (400 MHz, CD2Cl2): δ 2.44 (dq, J ) 13.0, 6.6 Hz, 1H, exo-
H), 2.40 (s, 3H, CH3), 1.85 (s, 6H, CH3), 1.77 (s, 15H, C5Me5), 1.44
(d, J ) 2.4 Hz, 6H, CH3), 1.21 (d, J ) 6.6 Hz, 3H, endo-CH3), -5.82
(br d, J ) 13.0 Hz, Hag). 13C{1H} NMR (100 MHz, CD2Cl2): δ 101.9,
99.9, 96.8, 67.1, 42.1, 16.9, 16.0, 14.9, 14.2, 9.7 (triflate carbon not
observed).
described for complex 2 above was followed, except that pentameth-
ylbenzene was used in place of hexamethylbenzene. Thus, a mixture
of [(C5Me5)RuCl]4 (0.300 g, 1.104 mmol/Ru), pentamethylbenzene
(0.840 g, 5.66 mmol), and silver tetrafluoroborate (0.218 g, 1.12 mmol)
in 12 mL of dichloromethane was maintained at a gentle reflux for 5
h. A bright white powder (0.483 g, 93%) was isolated after recrystal-
lization from dichloromethane/diethyl ether. 1H NMR (400 MHz, CD2-
Cl2): δ 5.46 (s, 1H, C6Me5H), 2.10 (s, 6H, C6Me5H), 2.06 (s, 3H,
C6Me5H), 2.03 (s, 6H, C6Me5H), 1.71 (s, 15H, C5Me5). 13C{1H} NMR
(100 MHz, CD2Cl2): δ 99.2, 99.1, 98.7, 93.1, 91.5, 17.9, 14.7, 14.2,
9.4. Anal. Calcd for C21H31RuBF4: C, 53.51%; H, 6.63. Found: C,
53.38; H, 6.52.
Addition of MeOH to 9. Methanol (0.011 mL, 6.0 equiv) was added
via microsyringe to the CD2Cl2 solution of [(C5Me5)Ru(η4-1,2,3,4,5-
exo-6-endo-hexamethylcyclohexa-1,3-diene)]+OTf- (9) prepared above
and the reaction monitored spectroscopically. After 24 h, only 21%
conversion to [(C5Me5)Ru(η6-C6Me6)]+OTf- (2) was observed. The
reaction was complete after about 5 days at room temperature.
[(C5Me5)Ru(η4-1,2,3,4,5-exo-6-exo-hexamethylcyclohexa-1,3-di-
ene)]+OTf- (10). A solution of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexa-
methylcyclohexadienyl) (7, 0.015 g, 0.038 mmol) in CD2Cl2 was placed
into an NMR tube equipped with a septum. Triflic acid (0.004 mL,
0.045 mmol) was added via syringe and the resulting solution analyzed
(C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl) (7). A
suspension of [(C5Me5)Ru(η6-C6Me5H)]+BF4- (6, 0.165 g, 0.350 mmol)
in 6 mL of tetrahydrofuran was cooled to -78 °C using a dry ice/
acetone bath, and MeLi (1.4 M in diethyl ether, 0.75 mL, 1.050 mmol,
3.0 equiv) was added via syringe. The reaction was stirred at -78 °C
for several hours before gradually being warmed to room temperature.
The resulting yellow solution was then evaporated under low pressure,
and the residue was triturated with several portions of pentane. The
combined pentane extracts were filtered over Celite, and the yellow
filtrate was concentrated to a solid. After thorough drying in vacuo,
the yellow solid was dissolved in pentane and filtered through a plug
of alumina (5% H2O), giving a colorless filtrate. This filtrate was
concentrated under low pressure to give 0.058 g (42%) of white
crystalline material, which was used without further purification. IR
(CH2Cl2 cast, cm-1): 2959 (s), 2937 (s), 2902 (s), 2881 (s), 2852 (s),
1376 (m), 1265 (m), 1017(m), 740 (s). 1H NMR (400 MHz, CD2Cl2):
δ 2.02 (s, 3H, CH3), 1.94 (q, J ) 6.3 Hz, 1H, endo-H), 1.60 (s, 21H,
CH3 and C5Me5), 1.31 (s, 6H, CH3), 0.11 (d, J ) 6.3 Hz, 3H, exo-
CH3). 13C{1H} NMR (100 MHz, CD2Cl2): δ 88.1, 86.7, 86.5, 50.2,
39.9, 20.3, 20.2, 15.3, 14.4, 10.0. MS m/z (relative intensity): calculated
for C22H34102Ru (M+), 400.1704; found, 400.1687 (3.81); calculated
for C22H33102Ru (M - H), 399.1626; found, 399.1644 (6.55); calculated
for C21H31102Ru (M - CH3), 385.1469; found, 385.1464 (100.00).
1
by H and 13C{1H} NMR spectroscopy at room temperature and at
-80 °C. 1H NMR (400 MHz, CD2Cl2, 20 °C): δ 2.59 (br s, 1H, endo-
H), 2.37 (br s, 3H, CH3), 1.79 (br s, 12H, CH3), 1.77 (s, 15H, C5Me5),
1
0.33 (br s, 3H, exo-CH3), -5.73 (br s, 1H, Hag). H NMR (400 MHz,
CD2Cl2, -80 °C): δ 2.53 (br q, J ) 6.6 Hz, 1H, endo-H), 2.25 (s, 3H,
CH3), 1.71 (s, 12H, CH3), 1.67 (s, 15H, C5Me5), 0.22 (d, J ) 6.5 Hz,
3H, exo-CH3), -5.40 (br s, 1H, Hag). 13C{1H} NMR (100 MHz, CD2-
Cl2, 20 °C): δ 96.8, 9.8 (all other signals broadened into the baseline).
13C{1H} NMR (100 MHz, CD2Cl2, -80 °C): δ 98.6, 96.1, 95.6, 74.0,
51.2, 22.4, 19.8, 14.0, 12.8, 9.2 (triflate carbon not observed).
Qualitative spin saturation transfer experiments (400 MHz, CD2Cl2,
20 °C): irradiation of δ -5.73 (Hag) T suppression of the signal at
2.59 ppm (endo-H); irradiation of δ 0.33 (exo-CH3) T suppression of
the signal at 1.79 ppm (12H, CH3) and diminution of the signal at 2.37
ppm (3H, CH3).
(C5Me5)Ru(η5-1,2,3,4,5-pentamethyl-6-exo-CD3-cyclohexadien-
yl) (7-d3). A procedure similar to that described for complex 7 was
followed. A cold (-78 °C) suspension of (C5Me5)Ru(η6-C6Me5H)]+-
Addition of Water to 10. Excess distilled water (0.004 mL, 5.9
equiv) was added via microsyringe to the solution of [(C5Me5)Ru(η4-
1,2,3,4,5-exo-6-exo-hexamethylcyclohexa-1,3-diene)]+OTf- (10) pre-
pared above. The NMR tube was shaken for 1 min, after which time
the reaction was monitored by 1H NMR spectroscopy at 10-min
intervals. After 50 min, the conversion to [(C5Me5)Ru(η6-C6Me5H)]+-
OTf- (6) and methane (δ 0.20 ppm in CD2Cl2) was complete.
Protonation of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclo-
hexadienyl) (7) with CF3SO3D (Prepared in Situ by Exchange with
Acetone-d6). In the drybox, a solution of (C5Me5)Ru(η5-1,2,3,4,5,6-
exo-hexamethylcyclohexadienyl) (7, 0.015 g, 0.038 mmol) in acetone-
d6 was placed into an NMR tube equipped with a rubber septum and
sealed. After removal of the solution from the drybox, triflic acid (0.004
mL, 0.045 mmol) was then added via syringe and the reaction monitored
spectroscopically. After 20 min at room temperature, 1H NMR showed
complete conversion of starting material to [(C5Me5)Ru(η4-1,2,3,4,5-
exo-6-exo-hexamethylcyclohexa-1,3-diene)]+OTf- (10) along with
traces of [(C5Me5)Ru(η6-C6Me5D)]+OTf- (6-d1). After 24 h at room
temperature, conversion to complex 6-d1 was complete, accompanied
by the formation of methane (δ 0.15 in acetone-d6) and methane-d1
(1:1:1 triplet at δ 0.13 in acetone-d6).
-
BF4 (6, 0.050 g, 0.106 mmol) in tetrahydrofuran was prepared, and
CD3Li (approximately 0.3 M in diethyl ether, 1.06 mL, 3.0 equiv) was
added by syringe. The flask was then warmed to room temperature
and the product isolated as described. Pale yellow crystals were
1
collected (0.015 g, 37%). H NMR (400 MHz, CD2Cl2): δ 2.02 (s,
3H, CH3), 1.93 (br s, 1H, endo-H), 1.60 (s, 21H, CH3 and C5Me5),
1.31 (s, 6H, CH3).
Protonation of Complex 5 with HBF4‚Et2O in Acetone-d6. In the
drybox, a solution of (C5H5)Ru(η5-1,2,3,4,5,6-endo-hexamethylcyclo-
hexadienyl) (5, 0.010 g, 0.030 mmol) in acetone-d6 was placed into an
NMR tube equipped with a rubber septum. Tetrafluoroboric acid
(diethyl ether complex, 85% in diethyl ether) (0.004 mL, 0.045 mmol)
was then added via microsyringe and the resulting solution analyzed
by 1H NMR spectroscopy. Clean conversion to [(C5H5)Ru(η6-C6Me6)]+-
-
BF4 (4) was observed within 10 min at room temperature.
[(C5H5)Ru(η4-1,2,3,4,5-exo-6-endo-hexamethylcyclohexa-1,3-di-
ene)]+BF4 (8). In the drybox, a solution of (C5H5)Ru(η5-1,2,3,4,5,6-
-
endo-hexamethylcyclohexadienyl) (5, 0.016 g, 0.049 mmol) in anhy-
drous CD2Cl2 was placed into an NMR tube equipped with a rubber
septum. Tetrafluoroboric acid (diethyl ether complex, 85%) was then
added (0.008 mL, 0.064 mmol) via microsyringe and the resulting
Protonation of (C5Me5)Ru(η5-1,2,3,4,5-pentamethyl-6-exo-CD3-
cyclohexadienyl) (7-d3) with CF3SO3H. A dichloromethane solution
of (C5Me5)Ru(η5-1,2,3,4,5-pentamethyl-6-exo-CD3-cyclohexadienyl) (7-
d3, 0.015 g, 0.037 mmol) was placed into an NMR tube and capped
1
1
yellow solution analyzed by H NMR spectroscopy. H NMR (400
MHz, CD2Cl2): δ 5.14 (s, 5H, C5H5), 2.82, (s, 3H, CH3), 2.50 (dq,
J ) 13.2, 6.6 Hz, 1H, exo-H), 2.17 (s, 6H, CH3), 1.67 (d, J ) 2.7 Hz,

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2795
with a rubber septum. The solution was frozen using liquid N2 (-196
°C), and triflic acid (0.004 mL, 0.045 mmol) was added via microsy-
ringe, followed quickly by the addition of excess distilled water (0.004
mL, 0.222 mmol, 6 equiv). The NMR tube was then warmed rapidly
to room temperature and allowed to stand for a further 2 h. Subsequent
spectroscopic analysis of the solution by 2H NMR spectroscopy revealed
the incorporation at deuterium at δ 2.07, 2.03, and 2.00 ppm in a ratio
of 2:1:2, respectively, consistent with equal incorporation of CD3 at
each of the arene methyl positions. Also present is a broad upfield
signal at approximately 0.12 ppm, consistent with the formation of a
small amount of CD3H. Similar results were obtained by protonation
of 7-d3 using 1 equiv of tetrafluoroboric acid in acetone-d6.
Hz, CH3), 13.4 (q, 1JCH ) 130 Hz, CH3), 13.0 (q, 1JCH ) 129 Hz, CH3),
1
9.7 (q, JCH ) 127 Hz, C5Me5). HMQC (300 MHz, CD2Cl2): δ 76.2
(CH2) T δ 4.33, 3.54 (CH2); δ 40.2 (CH) T δ 1.84 (CH); δ 21.0
(CH3) T δ 0.66 (CH3); δ 18.9 (CH3) T δ 1.39 (CH3); δ 14.9 (CH3) T
δ 1.48 (CH3); δ 13.4 (CH3) T δ 1.88 (CH3); δ 13.0 (CH3) T δ 2.02
(CH3); δ 9.7 (C5Me5) T δ 1.67 (C5Me5).
Addition of Water to 13. Excess distilled water (0.004 mL, 7.5
equiv) was added via microsyringe to the anhydrous CD2Cl2 solution
of [(C5Me5)Ru(η6-1,2,3,4,6-exo-pentamethyl-5-methylenecyclohexa-1,3-
diene)]+HB(C6F5)3 (13) prepared above. The NMR tube was shaken
-
for 1 min, and then the reaction was monitored by 1H NMR
spectroscopy. After 4 h the conversion of starting material to [(C5-
Me5)Ru(η6-C6Me5H)]+HOB(C6F5)3- (6-B(Arf)3OH) and methane (0.20
ppm in CD2Cl2) was complete. IR (CH2Cl2 cast, cm-1): 3670 (br),
1644 (m), 1515 (s), 1465 (s), 1089 (s), 976 (s).
Reaction of (η6-C6Me6)Ru(η3-C3H5)OTf (11),12a (C5Me5)Ru(η5-
1,2,3,4,5,6-endo-hexamethylcyclohexadienyl) (3), and Diphenylacet-
ylene. Crossover Experiment. A solution of (C5Me5)Ru(η5-1,2,3,4,5,6-
endo-hexamethylcyclohexadienyl) (3, 0.014 g, 0.035 mmol) and di-
phenylacetylene (0.008 g, 0.045 mmol) in anhydrous CD2Cl2 was
prepared and placed in an NMR tube capped with a rubber septum.
The solution was cooled in the drybox freezer (-35 °C) for 20 min,
after which time an orange solution of (C6Me6)Ru(η3-C3H5)OTf (11,
0.017 g, 0.037 mmol) in ∼0.2 mL of CD2Cl2 was added. The reaction
mixture was warmed to room temperature and the resulting yellow
Reaction of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexa-
dienyl) (7) with AlBr3. A solution of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-
hexamethylcyclohexadienyl) (7, 0.010 g, 0.025 mmol) in CD2Cl2 was
loaded into an NMR tube, and pure AlBr3 (0.008 g, 0.030 mmol) was
added. The NMR tube was capped and shaken for 1 min. After the
1
solution was left to stand for 10 min at room temperature, H NMR
spectroscopic analysis indicated only a single product in quantitative
1
-
solution analyzed by H NMR spectroscopy. The spectrum revealed
yield, [(C5Me5)Ru(η6-C6Me5H)]+MeAlBr3 (6-MeAlBr3), based on
quantitative formation of [(C5Me5)Ru(η6-C6Me6)]+OTf- (2) and (η5-
C5H3Ph2)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl) (12), con-
firmed by comparison to authentic material prepared by independent
synthesis (see below). The solvent was then removed under reduced
pressure, the residue triturated with 3 mL of pentane, and the pentane
extract filtered over Celite and dried to afford a pale yellow solid (12,
0.012 g, approximately 71%; contaminated with a small amount of
diphenylacetylene).
spectroscopic comparison with the well-characterized triflate analogue.
1H NMR (360 MHz, CD2Cl2): δ 5.37 (s, 1H, C6Me5H), 2.10 (s, 6H,
C6Me5H), 2.06 (s, 3H, C6Me5H), 2.03 (s, 6H, C6Me5H), 1.71 (s, 15H,
C5Me5), 0.00 (br s, 3H, MeAlBr3).
Reaction of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexa-
dienyl) (7) with Ph3C+BF4-. A solution of (C5Me5)Ru(η5-1,2,3,4,5,6-
exo-hexamethylcyclohexadienyl) (7, 0.010 g, 0.025 mmol) in anhydrous
CD2Cl2 was cooled to -35 °C and triphenylcarbenium tetrafluoroborate
(0.009 g, 0.027 mmol) added. The solution immediately turned bright
yellow and was placed into an NMR tube. Subsequent 1H NMR analysis
(η5-C5H3Ph2)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl) (12,
Independent Synthesis). A suspension of [(η5-C5H3Ph2)Ru(η6-C6-
Me5H)]+OTf- 12a (0.055 g, 0.089 mmol) in 4 mL of tetrahydrofuran
was cooled to -78 °C using a dry ice/acetone bath, and MeLi (1.4 M
in diethyl ether, 0.15 mL, 0.210 mmol, 2.36 equiv) was added via
syringe. The reaction was stirred at -78 °C for several hours before
being gradually warmed to room temperature. The resulting yellow
solution was then evaporated under low pressure and the residue
triturated with several portions of hexane (3 × 2 mL). The combined
hexane extracts were filtered over Celite, and the yellow filtrate was
concentrated to a solid. After thorough drying in vacuo the solid was
then dissolved in pentane and filtered through a plug of alumina (5%
H2O). The filtrate was concentrated under low pressure to yield 0.016
indicated the formation of a mixture of [(C5Me5)Ru(η6-C6Me6)]+BF4
-
(2) and [(C5Me5)Ru(η6-C6Me5H)]+BF4- (6); no intermediate formation
is observed. The integrated ratio of 2:6 was 30:70 at long pulse delay.
Also present are trityl dimer [1H NMR, (300 MHz, CD2Cl2): δ 7.33-
7.15 (m, 21H), 6.95 (dd, J ) 8.1, 1.8 Hz, 4H), 6.23 (dd, J ) 10.5, 2.0
Hz, 2H), 5.97 (dd, J ) 10.5, 3.8 Hz, 2H), 5.13 (br s, 1H)] and Ph3-
CMe [1H NMR, (300 MHz, CD2Cl2): δ 7.37-7.06 (m, 15H, HPh), 2.17
(s, 3H, Me)] in a ratio of 30:70.
Reaction of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexa-
dienyl) (7) with [Cp2Fe]+PF6-. (i) At -78 °C. A blue suspension of
[Cp2Fe]+PF6- (0.009 g, 0.027 mmol) in 2 mL of dichloromethane was
cooled to -78 °C in a dry ice/acetone bath, and a solution of (C5Me5)-
Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl) (7, 0.011 g, 0.027
mmol) in 1 mL of dichloromethane was added slowly via syringe. The
reaction mixture turned from blue to orange within 5 min, and the cold
bath was removed. Upon warming to room temperature the reaction
lightened to pale yellow; the solvent was then removed under reduced
pressure. 1H NMR analysis of the yellow residue revealed the formation
of [(C5Me5)Ru(η6-C6Me6)]+PF6- (2) and [(C5Me5)Ru(η6-C6Me5H)]PF6
(6) in a ratio of 67:33, along with ferrocene.
1
g (37%) of pale yellow crystalline complex 12. H NMR (400 MHz,
CD2Cl2): δ 7.19-7.10 (m, 10H, HPh, partially obscured), 4.70 (t, J )
2.5 Hz, 1H), 4.65 (d, J ) 2.5 Hz, 2H), 2.23 (q, J ) 6.4 Hz, 1H), 2.12
(s, 3H, CH3), 1.63 (s, 6H, CH3), 1.44 (s, 6H, CH3), 0.15 (d, J ) 6.4
Hz, 3H, CH3). 13C{1H} NMR (100 MHz, CD2Cl2): δ 137.2, 129.4,
128.0, 125.9, 94.4, 90.9, 88.3, 81.3, 78.4, 50.3, 43.2, 22.1, 20.8, 16.2,
15.4. MS m/z (relative intensity): calculated for C29H32102Ru (M+),
482.1548; found, 482.1543 (1.54); calculated for C29H31102Ru (M -
H), 481.1469; found, 481.1458 (2.80); calculated for C28H29102Ru
(M - CH3), 467.1313; found, 467.1325 (100.00).
(ii) At 22 °C. The reaction was performed as described above, except
that no cold bath was used. Upon addition of complex 7, the reaction
mixture immediately turned pale yellow as stirring was continued for
a further 10 min. The solvent was then removed in vacuo and the residue
analyzed by 1H NMR spectroscopy. [(C5Me5)Ru(η6-C6Me6)]+PF6- (2)
[(C5Me5)Ru(η6-1,2,3,4,6-exo-pentamethyl-5-methylenecyclohexa-
-
1,3-diene)]+HB(C6F5)3 (13). An NMR tube was charged with a
solution of (C5Me5)Ru(η5-1,2,3,4,5,6-exo-hexamethylcyclohexadienyl)
(7, 0.012 g, 0.030 mmol) in anhydrous CD2Cl2 and the tube cooled to
-35 °C. Immediately upon addition of tris(perfluorophenyl)borane
(0.016 g, 0.031 mmol), the colorless solution turned bright yellow.
Spectroscopic analysis (1H, 13C NMR, HMQC) of the reaction mixture
showed quantitative formation of product (complex 13) within 10 min.
1H NMR (400 MHz, CD2Cl2): δ 4.33 (s, 1H, HC6Me5CH2), 3.54 (s,
1H, HC6Me5CH2), 2.02 (s, 3H, CH3), 1.88 (s, 3H, CH3), 1.84 (q, J )
6.7 Hz, 1H, endo-H), 1.67 (s, 15H, C5Me5), 1.48 (s, 3H, CH3), 1.39 (s,
3H, CH3), 0.66 (d, J ) 6.7 Hz, 3H, exo-CH3). 13C NMR (100 MHz,
CD2Cl2, 1JCH values obtained from the HMQC spectrum): δ 106.4 (s,
HC6Me5CH2), 98.7 (s, HC6Me5CH2), 96.3 (s, C5Me5), 94.5 (s, HC6-
Me5CH2), 89.6 (s, HC6Me5CH2), 76.2 (d, 1JCH ) 161 Hz, HC6Me5CH2),
and [(C5Me5)Ru(η6-C6Me5H)]+PF6 (6) were formed in a ratio of 30:
-
70, accompanied by ferrocene.
Reaction of (C5Me5)Ru(η5-1,2,3,4,5,6-endo-hexamethylcyclohexa-
dienyl) (3) with [Cp2Fe]+PF6-. To a blue suspension of Cp2Fe+PF6
-
(0.009 g, 0.027 mmol) in 2 mL of dichloromethane cooled to -78 °C
in a dry ice/acetone bath was added a solution of (C5Me5)Ru(η5-
1,2,3,4,5,6-endo-hexamethylcyclohexadienyl) (3, 0.011 g, 0.027 mmol)
in 1 mL of dichloromethane slowly via syringe. The reaction mixture
immediately faded from blue to pale yellow, and the cold bath was
removed. After the mixture warmed to room temperature, the solvent
was removed under reduced pressure. 1H NMR spectroscopic analysis
of the yellow residue revealed complete and quantitative conversion
1
1
55.7 (s, HC6Me5CH2), 40.2 (d, JCH ) 133 Hz, CH), 21.0 (q, JCH
)
133 Hz, exo-CH3), 18.9 (q, 1JCH ) 129 Hz, CH3), 14.9 (q, 1JCH ) 130
to [(C5Me5)Ru(η6-C6Me6)]+PF6 (2) and ferrocene.
-

2796 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
Older and Stryker
(C5Me5)Ru(η5-6-exo-methylcyclohexadienyl) (14). To a suspension
of [(C5Me5)Ru(η6-C6H6)]+PF6- (16,39 0.210 g, 0.457 mmol) in 20 mL
of diethyl ether cooled to 0 °C using an ice bath was added MeLi (1.4
M in diethyl ether, 0.36 mL, 0.504 mmol, 1.1 equiv) via syringe. No
reaction was observed until 3 mL of tetrahydrofuran was added. The
reaction mixture was stirred at 0 °C for 1 h and gradually warmed to
room temperature over another 4 h. The bright orange solution was
evaporated under low pressure and the residue triturated with several
portions of hexane (3 × 2 mL). The combined hexane extracts were
filtered over Celite, and the orange filtrate was concentrated to an oily
residue. After thorough drying in vacuo, the red residue was dissolved
in pentane and filtered through a plug of alumina (5% H2O), giving a
nearly colorless filtrate. This filtrate was concentrated under reduced
pressure to yield 0.086 g (57%) of a pale yellow crystalline complex
room temperature for 6 h, gradually turning deep green. Afterward,
excess activated zinc (Rieke zinc,
44 0.100 g) was added, and the reaction
mixture was stirred for a further 10 h, during which time the green
solution became nearly colorless. The solvent was then evaporated under
reduced pressure, the residue triturated with several portions of hexanes
(3 × 2 mL), and the combined extracts filtered over Celite. The filtrate
was evaporated to give a brown foam and then redissolved in
tetrahydrofuran and filtered through a plug of alumina (5% H2O). The
solvent was removed under reduced pressure to yield 0.157 g of
ergosterol complex 17 (52%) as a tan solid, which was used without
1
further purification. H NMR (400 MHz, C6D6, selected assignments
only): δ 5.28, 5.21 (AB quartet, J ) 7.5 Hz, 2H, H22/H23), 5.11 (d,
J ) 4.8 Hz, 1H, H6/H7), 3.59 (d, J ) 4.8 Hz, 1H, H6/H7), 3.97-3.85
(m, 1H, H3) 1.70 (s, 15H, C5Me5), 1.10 (d, J ) 6.6 Hz, 3H), 1.01 (d,
J ) 6.9 Hz, 3H), 0.91 (d, J ) 6.7 Hz, 3H), 0.90 (d, J ) 6.7 Hz, 3H),
0.75 (s, 3H), 0.54 (s, 3H). 13C NMR (100 MHz, C6D6): δ 136.2 (C22/
C23), 132.3 (C22/C23), 95.8 (C5/C8), 92.1 (C5/C8), 87.2 (C5Me5), 79.3
(C6/C7), 78.5, (C6/C7), 75.5 (C9), 69.9 (C3), 55.2, 51.4, 44.2, 42.4, 43.3,
41.0, 38.6, 37.7, 36.4, 33.4, 32.9, 29.8, 25.2, 23.7, 22.6, 21.3, 20.2,
19.9, 17.9, 11.3, 10.9 (C5Me5). MS m/z (relative intensity): calculated
for C38H58102RuO (M+), 632.3534; found, 632.3405 (2.83); calculated
for C37H55102RuO (M - CH3), 617.3297; found, 617.3288 (100.00).
1
14, which was used without further purification. H NMR (400 MHz,
CD2Cl2): δ 5.13 (td, J ) 4.7, 0.9 Hz, 1H), 3.89 (tt, J ) 4.8, 1.3 Hz,
2H), 2.22-2.20 (m, 3H), 1.88 (s, 15H), 0.12 (d, J ) 5.9 Hz). 13C{1H}
NMR (100 MHz, CD2Cl2): δ 89.2, 80.5, 79.3, 36.7, 34.6, 28.0, 11.5.
MS m/z (relative intensity): calculated for C17H24102Ru (M+), 330.0922;
found, 330.0906 (5.97); calculated for C17H23102Ru (M - H), 329.0843;
found, 329.0869 (8.77); calculated for C16H21102Ru (M - CH3),
315.0687; found, 385.0686 (100).
Protonation of (C5Me5)Ru(η5-6-exo-methylcyclohexadienyl) (14).
A solution of (C5Me5)Ru(η5-6-exo-CH3C6H6) (14, 0.015 g, 0.046 mmol)
in anhydrous CD2Cl2 was placed into an NMR tube and the tube capped
with a rubber septum. Triflic acid (0.004 mL, 0.045 mmol) was added
via microsyringe and the resulting solution analyzed by 1H NMR
spectroscopy. The reaction was complete within 10 min at room
temperature, giving [(C5Me5)Ru(η6-CH3C6H5)]+OTf- (15), spectro-
scopically identical to known compound9 and the demethylated product,
[(C5Me5)Ru(η6-C6H6)]+OTf- (16), also based on spectroscopic com-
parison to authentic material,9 in a ratio of approximately 91:9.
Comparable results were obtained from the analogous reaction of (C5-
Me5)Ru(η5-6-exo-methylcyclohexadienyl) (14) tetrafluoroboric acid in
acetone-d6.
Reaction of (C5Me5)Ru(η5-6-exo-methylcyclohexadienyl) (14)
with B(C6F5)3. A solution of (C5Me5)Ru(η5-6-exo-methylcyclohexa-
dienyl) (0.020 g, 0.060 mmol) in anhydrous CD2Cl2 was loaded into
an NMR tube and tris(perfluorophenyl)borane (0.035 g, 0.068 mmol)
added in one portion. The NMR tube was then capped and shaken for
1 min. After 10 min at room temperature, 1H NMR spectroscopic
analysis of the colorless solution indicated the formation of the known
cation [(C5Me5)Ru(η6-CH3C6H5)]+ (15)9 as the major product ac-
companied by only a trace of [(C5Me5)Ru(η6-C6H6)]+ (16)9 in a ratio
of >98:2. The counterions were not identified.
[(C5Me5)Ru(η4-ergosterol)]+BF4 (18). (i) From HBF4‚Et2O. A
-
solution of (C5Me5)Ru(η5-9-dehydoergosterol) (17, 0.027 g, 0.043
mmol) in anhydrous dichloromethane-d2 was placed into an NMR tube
and capped with a rubber septum. Tetrafluoroboric acid (diethyl ether
complex, 85% in diethyl ether, 0.008 mL, 0.064 mmol) was then added
via microsyringe and the resulting yellow solution analyzed by 1H and
13C{1H} NMR spectroscopy. (Chemical shifts vary slightly depending
on the concentration of excess acid.)
(ii) From HOTf. A solution of (C5Me5)Ru(η5-dehydroergosterol)
(17, 0.022 g, 0.035 mmol) in distilled diethyl ether (3 mL) was prepared
in a small Schlenk flask. Triflic acid (0.004 mL, 0.045 mmol) was
added via microsyringe and the resulting solution stirred for 1 h. Within
20 min an oily yellow residue deposited from the solution. The solution
was decanted, and the residue was rinsed with several portions of diethyl
ether and dried under vacuum. The resulting yellow foam was analyzed
spectroscopically without further purification. 1H NMR (400 MHz, CD2-
Cl2, selected assignments only): δ 6.26 (d, J ) 5.7 Hz, 1H, H6/H7),
4.65 (d, J ) 5.7 Hz, 1H, H6/H7), 5.30, 5.14 (AB quartet, J ) 7.2 Hz,
2H, H22/H23), 4.08-3.90 (m, 1H, H3), 1.91 (s, 15H, C5Me5), 1.00
(d, J ) 6.6 Hz, 3H), 0.93 (d, J ) 6.6 Hz, 3H), 0.85 (d, J ) 6.6 Hz,
3H), 0.83 (d, J ) 6.8 Hz, 3H), 0.65 (s, 3H), 0.55 (s, 3H), -5.00 (br s,
1H, Hag). 13C{1H} NMR (100 MHz, CD2Cl2): δ 134.9 (C22/C23), 133.4
(C22/C23), 103.8 (C8), 97.6 (C5Me5), 88.7 (C6/C7), 88.6 (C6/C7), 81.6
(C5), 79.7 (C3), 68.1 (br C9), 54.7, 51.0, 45.8, 42.2, 43.2, 40.5, 33.4,
39.2, 36.3, 31.0, 29.0, 24.9, 24.0, 26.1, 21.0, 20.0, 19.8, 17.6, 10.8
(C18), 10.6 (C5Me5). The triflate carbon was not observed.
Reaction of (C5Me5)Ru(η5-6-exo-methylcyclohexadienyl) (14)
with AlBr3. A solution of (C5Me5)Ru(η5-6-exo-methylcyclohexadienyl)
(14, 0.025 g, 0.076 mmol) in anhydrous CD2Cl2 was loaded into an
NMR tube and AlBr3 (0.021 g, 0.079 mmol) added. The NMR tube
was then capped and shaken for 1 min. After 10 min at room
temperature, 1H NMR spectroscopic analysis of the colorless solution
indicated the formation of a mixture of the known9 cations [(C5Me5)-
Ru(η6-CH3C6H5)]+ (15) and [(C5Me5)Ru(η6-C6H6)]+ (16) in a ratio of
ca. 90:10, with no trace of the starting material. The counterions,
Dealkylation of (C5Me5)Ru(η5-9-dehydroergosterol) (17). (i) By
Protic Acid. A solution of (C5Me5)Ru(η5-9-dehydroergosterol) (0.044
g, 0.032 mmol) in dry dichloromethane (2 mL) was prepared in a small
Schlenk flask. Tetrafluoroboric acid (diethyl ether complex, 85% in
diethyl ether, 0.013 mL, 0.075 mmol) was added via syringe, and the
reaction was stirred for 5 min before addition of distilled water (0.006
mL, 0.333 mmol, 10.4 equiv). After 4 days at room temperature, the
solvent was removed under reduced pressure, leaving a solid tan residue.
Analysis of the residue by 1H NMR spectroscopy showed clean
conversion to the known complex, [(C5Me5)Ru(η6-neoergosterol)]+
-
presumably HAlBr3 and MeAlBr3-, respectively, were not further
characterized.
Reaction of (C5Me5)Ru(η5-6-exo-methylcyclohexadienyl) (14)
-
with Cp2Fe+PF6-. A blue suspension of Cp2Fe+PF6 (0.025 g, 0.076
mmol) in 2 mL of dichloromethane was prepared, and a solution of
(C5Me5)Ru(η5-6-exo-methylcyclohexadienyl) (0.025 g, 0.076 mmol) in
2 mL of dichloromethane was added slowly via cannula. The blue
suspension immediately turned pale yellow and became homogeneous.
After the suspension was stirred at room temperature for 30 min, the
-
BF4 (19), which was spectroscopically consistent with the complex
reported by Chaudret et al.9b
(ii) By Ph3CBF4. In the drybox, a solution of (C5Me5)Ru(η5-
dehydroergosterol) (0.015 g, 0.024 mmol) in 3 mL of tetrahydrofuran
was placed in a Schlenk flask. A yellow slurry of triphenylcarbenium
tetrafluoroborate (0.009 g, 0.027 mmol) was added dropwise over 5
min; the reagent rapidly dissolved and the solution turned light brown.
The reaction was stirred for 1 h, and the solvent was then removed in
1
solvent was removed under reduced pressure. H NMR spectroscopic
analysis of the yellow residue indicated conversion to [(C5Me5)Ru(η6-
CH3C6H6)]+PF6- (15)9 and [(C5Me5)Ru(η6-C6H6)]+PF6- (16)9 in a ratio
of 42:58, accompanied by the formation of ferrocene.
(C5Me5)Ru(η5-9-dehydroergosterol) (17). In the drybox, a Schlenk
flask was charged with [(C5Me5)RuCl]4 (0.130 g, 0.478 mmol Ru
content) and ergosterol (0.200 g, 0.504 mmol, 1.05 equiv). Tetrahy-
drofuran (6 mL) was added, and the reaction mixture was stirred at
(44) Rieke, R. D.; Sell, M. S.; Klein, W. R.; Chen, T.; Brown, J. D.;
Hanson, M. V. In ActiVe Metals; Fu¨rstner, A., Ed.; VCH: New York, 1996;
Chapter 1.

Mechanism of Carbon-Carbon Bond ActiVation
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2797
vacuo. Analysis of the residue by 1H NMR spectroscopy in acetone-d6
revealed clean conversion to the known [(C5Me5)Ru(η6-neoergosterol)]+-
BF4 (19) and Ph3CCH3.
University of Alberta is gratefully acknowledged. C.M.O.
acknowledges the support of an NSERC Post-graduate Scholar-
ship and a Dissertation Year Fellowship from the University of
Alberta.
-
Acknowledgment. Financial support from the Natural Sci-
ences and Engineering Research Council of Canada and the
JA992987E