4692 Organometallics, Vol. 20, No. 22, 2001
Dunn et al.
Sch em e 1
Fe(CO)5 in refluxing acetone to yield iron-cobalt or di-
iron clusters, 15 or 16, respectively, was originally
reported by Victor in 1977.8 It was described as a
“dehydroxylation process”, and several such iron/cobalt9
or di-iron12 products have now been crystallographically
characterized.
Although the mechanism of this “dehydroxylation”
procedure remains unclear, speculation has focused on
the possible role of metal hydrides.13 For example, Chini
demonstrated that Co2(CO)8 and Fe(CO)5 in acetone
react to form HFeCo3(CO)12.14 Moreover, Geoffroy has
reported that MeCCo3(CO)9 and [Fe(CO)4]2- yield, upon
protonation, the hydride cluster MeCCo2FeH(CO)9.15
Thus the transformation of (HCtC-CH2OH)Co2(CO)6
into (HCdCdCH2)FeCo(CO)6 may involve the formation
of an intermediate metal hydride, 17; as depicted in
Scheme 1, subsequent elimination of water could give
the neutral iron-cobalt cluster 15.
However, the isolation of the trimetallic species 9 and
10 suggests another possibility: nucleophilic attack by
the hydroxyl group on a carbonyl ligand of Fe(CO)5 could
yield the hydrido metal carboxylate 18. It is, of course,
well known that the formation of the [HFe(CO)4]- anion
from Fe(CO)5 and KOH proceeds through an intermedi-
ate carboxylate species.16 It is relevant to note that
Osella has obtained the carboxylate-containing cluster
Fe2(CO)6[(EtCdCEt)C(dO)18O] from hex-3-yne, Fe2-
(CO)9, and 18OH2.17 Moreover, Sappa has convincingly
demonstrated that “deoxygenation” of the propargylic
alcohol HCtCC(Me)(Ph)OH by Fe3(CO)12 is really a
decarboxylation and proceeds with elimination of carbon
dioxide.18 It was tentatively suggested that the hydroxyl
substituent from the alkyne attacks a carbonyl ligand
on iron, and the current observations provide strong
support for such a proposal.
of the resulting loss of aromaticity in the six-membered
rings. Instead, one might invoke expansion of the
tetrahedral cluster to a square-based pyramidal struc-
ture 20. While the apical and basal plane cobalts in 20
would formally be assigned 19 and 17 electrons, respec-
tively, the overall skeletal electron count is appropriate
for a nido octahedral cluster,19 and numerous examples
of such cluster expansion20 or vertex replacement21
reactions are known. Finally, decarboxylation could
yield the observed mixed metal cluster 21, as typified
by 5 and 6.
Future work in this area will look at examining the
role of the solvent in these reactions to gain further
understanding of these fascinating processes.
Exp er im en ta l Section
Gen er a l Meth od s. All reactions were carried out under an
atmosphere of dry nitrogen employing conventional benchtop
and glovebag techniques. Silica gel (particle size: 20-45 µm)
was employed for flash column chromatography. 1H and 13C
NMR spectra were acquired on a Bruker DRX 500 spectrom-
eter and were referenced to the residual proton or 13C solvent
signal. Mass spectra were obtained using a Finnigan 4500
spectrometer by direct electron impact (DEI) or direct chemical
ionization (DCI) with NH3. Elemental analyses were performed
by Guelph Chemical Laboratories, Guelph, Ontario.
As depicted in Schemes 2 and 3, one can envision two
competing pathways. In the case of a cyclopentadiene
ligand, decarbonylation and η4-coordination to the five-
membered ring could precede decarboxylation to yield
19, which possesses an (η5-C5R4)Fe(CO)2H moiety and
corresponds to the observed product 9 or 10.
The alkynylcyclopentadienol)Co2(CO)6 complexes 22 and 23
were prepared as previously described.2
Rea ction of {5-(Tr im eth ylsilyl)eth yn yl)-1,4-d ieth yl-2,3-
diph en ylcyclopen tadien -5-ol}Co2(CO)6, 22, with Ir on P en -
ta ca r bon yl. Over a 25 min period, Fe(CO)5 (9.90 mL, 75.3
mmol) was added to 22 (5.0 g, 7.44 mmol) dissolved in acetone
(125 mL). The solution was heated to reflux and monitored
by TLC approximately every 5 h. Samples of the effluent gases
were shown by IR and mass spectrometry to contain traces of
CO2. After 60 h, the solution was allowed to cool and, after
removal of solvent, yielded a brown oil. The residue was
subjected to flash chromatography on silica gel with CH2Cl2
as eluent to remove inorganic salts arising from decomposed
metal carbonyls. A second purification (flash chromatography)
In contrast, coordination to the five-membered ring
of an indenyl or fluorenyl ligand is disfavored because
(12) (a) Gerlach, J , N.; Wing, R. M.; Ellgen, P. C. Inorg. Chem. 1976,
15, 2959. (b) Eigemann, S.-E.; Fo¨rtsch, W.; Hampel, F.; Schobert, R.
Organometallics 1996, 15, 1511. (c) Bright, D.; Mills, O. S. J . Chem.
Soc., Dalton Trans. 1972, 2465.
(13) McGlinchey, M. J .; Girard, L.; Ruffolo, R. Coord. Chem. Rev.
1995, 143, 331.
(14) Chini, P.; Colli, L.; Peraldo, M. Gazz. Chim. Ital. 1960, 90, 1005.
(15) Epstein, R. A.; Withers, H. W.; Geoffroy, G. L. Inorg. Chem.
1979, 18, 942.
(16) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA, 1987; p 403.
(17) Milone, L.; Osella, D.; Ravera, M.; Stanghellini, P. L.; Stein,
E. Gazz. Chim. Ital. 1992, 122, 451.
(19) Wade, K. Adv. Inorg. Chem. Radiochem. 1976, 18, 1.
(20) Adams, R. D. In The Chemistry of Metal Cluster Complexes;
Shriver, D. F., Kaesz, H. D., Adams, R. D., Eds.; VCH: New York,
1989; Chapter 3, pp 121-170.
(18) Gervasio, G.; Sappa, E. Organometallics 1993, 12, 1458.
(21) Vahrenkamp, H. Adv. Organomet. Chem. 1983, 22, 169.