Photochemical reactivity of group 9 metal pinacolate complexes

Article abstract of DOI:10.1021/om0010870

Coordinatively unsaturated pentamethylcyclopentadienyl pinacolate complexes of the group 9 transition metals (4-6) have been prepared and characterized. Photolysis of either the cobalt complex 4 or the rhodium complex 5 results in cleavage of the central

Full text of DOI:10.1021/om0010870

2250
Organometallics 2001, 20, 2250-2261
P h otoch em ica l Rea ctivity of Gr ou p 9 Meta l P in a cola te
Com p lexes
Andrew W. Holland, David S. Glueck, and Robert G. Bergman*
Department of Chemistry, University of California, Berkeley, and Center for New Directions in
Organic Synthesis, Berkeley, California 94720-1460
Received December 26, 2000
Coordinatively unsaturated pentamethylcyclopentadienyl pinacolate complexes of the group
9 transition metals (4-6) have been prepared and characterized. Photolysis of either the
cobalt complex 4 or the rhodium complex 5 results in cleavage of the central carbon-carbon
bond in the diolate, generating acetone. Various trapping studies demonstrate that an intact
[Cp*M] fragment is produced in these reactions, and in the absence of added traps this
fragment reacts either with aromatic solvents or with an intact molecule of the starting
pinacolate complex. The oxidation of the resulting rhodium(II) product 11 by air (or O2) in
the presence of pinacol regenerates the rhodium(III) pinacolate complex 5. Photolysis of
rhodium complex 5 in the presence of pinacol and oxidant (either O2 or N2O) results in the
catalytic conversion of pinacol to acetone.
Sch em e 1
In tr od u ction
Because of their relevance to both biological and
synthetic catalytic systems,1-5 organometallic com-
pounds containing late-metal-oxygen bonds have been
studied increasingly in recent years. Advances in their
syntheses have allowed for thorough investigations of
many metal-alkoxo and -aryloxo species,1,6 but this
realm of study has only very rarely been extended to
include chelating bis(alkoxo) complexes (1). This omis-
sion is surprising, given the diverse reactivity and
synthetic relevance of those metal glycolate complexes
that have been prepared. Early-metal glycolates can be
formed by the coupling of organic carbonyl compounds
at low-valent metal centers7 (Scheme 1, eq 1 reverse
reaction), and this reaction constitutes the key step in
the metal-mediated preparation of unsymmetrical diols7
(2) and alkenes8 (3) by reductive coupling. Diolate
complexes with sufficiently oxophilic metal centers have
been observed to undergo a cycloreversion reaction in
which the carbon chain of the diolate is extruded as an
olefin, leaving behind a metal dioxo species (Scheme 1,
eq 2).9-12 This transformation plays a central role in
both reductive couplings by early metals and the
catalytic reduction of glycols and epoxides to olefins by
metals as late as group 7.13,14 The reverse reaction, the
dihydroxylation of alkenes, is well established to involve
osmium glycolate intermediates formed by a cycloaddi-
tion reaction.15
In the only report to date of the chemistry of glyco-
lates that bear metals beyond group 8 of the transition
series, Andrews and co-workers reported that dppePt-
(glycolate) complexes undergo a photochemical trans-
formation of the type shown in eq 1 of Scheme 1.16 The
diolate is cleaved to produce the corresponding organic
carbonyl compounds, generating an unsaturated [dppePt]
fragment that can be trapped convincingly by several
added reagents. This reaction (postulated to take place
in a [2 + 2 + 2] cycloreversion) also echoes the
mechanism of oxidative diol cleavage by HIO4.17
(1) Bryndza, H. E.; Tam, W. Chem. Rev. 1988, 88, 1163.
(2) Masters, C. Homogeneous Transition Metal Catalysis; Chapman
and Hall: New York, 1981.
(3) Sawyer, D. T. Oxygen Chemistry; Oxford University Press: New
York, 1991.
(4) Sheldon, R. A.; Kochi, J . K. Metal Catalyzed Oxidation of Organic
Compounds; Academic Press: New York, 1981.
(5) Simandi, L. I. Catalytic Activation of Dioxygen by Metal Com-
plexes; Kluwer: Dordrecht, The Netherlands, 1992.
(6) Bergman, R. G. Polyhedron 1995, 14, 3227.
(7) Steinhuebel, D. P.; Lippard, S. J . J . Am. Chem. Soc. 1999, 121,
11762.
(8) McMurry, J . E. Chem. Rev. 1989, 89, 1513.
(9) Schrock, R. R.; Pedersen, S. F.; Churchill, M. R.; Ziller, J . W.
Organometallics 1984, 3, 1574.
(10) Herrmann, W. A.; Marz, D.; Herdtweck, E.; Schafer, A.; Wagner,
W.; Kneuper, H.-J . Angew. Chem., Int. Ed. Engl. 1987, 26, 462.
(11) Zhu, Z.; Al-Ajlouni, A. M.; Espenson, J . H. Inorg. Chem. 1996,
35, 1408.
(12) Gable, K. P.; J uliette, J . J . J . J . Am. Chem. Soc. 1996, 118,
2625.
(13) Gable, K. P.; Brown, E. C. Organometallics 2000, 19, 944.
(14) Cook, G. K.; Andrews, M. A. J . Am. Chem. Soc. 1996, 118, 9448.
(15) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 2483-2547.
(16) Andrews, M. A.; Gould, G. L. Organometallics 1991, 10, 387.
10.1021/om0010870 CCC: $20.00 © 2001 American Chemical Society
Publication on Web 04/27/2001

Group 9 Metal Pinacolate Complexes
Organometallics, Vol. 20, No. 11, 2001 2251
F igu r e 1. ORTEP diagram of 4 (displacement ellipsoids
at 50% probability). Hydrogen atoms have been omitted
for clarity. Selected bond distances (Å) and angles (deg):
Co-O(1), 1.806(3); Co-O(2), 1.823(3); C(2)-O(2), 1.438(6);
C(1)-O(1), 1.452(5); C(1)-C(2), 1.533(6); O(1)-Co-O(2),
88.64(11); Co-O(1)-C(1), 112.5(3); Co-O(2)-C(2), 110.6-
(3); O(1)-C(1)-C(2)-O(2), 41.9(4).
F igu r e 2. ORTEP diagram of 5 (displacement ellipsoids
at 50% probability). Hydrogen atoms have been omitted
for clarity. Selected bond distances (Å) and angles (deg):
Rh-O(1), 1.9752(28); Rh-O(2), 1.9612(24); C(1)-O(1),
1.4056(46); C(2)-O(2), 1.4324(41); C(1)-C(2), 1.5278(47);
O(1)-Rh-O(2), 82.90(8); Rh-O(1)-C(1), 111.33(24); Rh-
O(2)-C(2), 113.08(20).
This chemistry is instructive, given its place in the
pattern of reactivity known for other metal glycolates.
Additionally, though, it offers the real prospects of both
effecting a useful synthetic transformation via late
metal glycolates and generating coordinatively unsatur-
ated late-metal centers via this cycloreversion. An
alternative to the relatively harsh periodate and lead
tetraacetate methodologies for oxidative cleavage of
diols would be welcome, and coordinatively unsaturated
late-metal centers have consistently proven to exhibit
remarkable reactivity. For these reasons, we have
expanded the family of late-metal glycolates to group 9
of the transition series and explored the reactivity of
these complexes on both of the fronts mentioned above.
Herein, we report the syntheses of pinacolate complexes
of cobalt, rhodium, and iridium, their photochemistry,
and their role in the catalysis of oxidative glycol cleav-
age.
ment of the corresponding dichloro dimers with pinacol
and base in THF (eq 4). They were isolated in good yield
(ca. 75%) as green, purple, and maroon crystals, respec-
tively, following benzene extraction and recrystallization
from diethyl ether. The yields and purities of 4 and 5
were substantially diminished when solutions were
exposed to light for significant periods of time, and
rigorous protection of the compound from light was
required for the isolation of the rhodium complex 5 in
high purity. The structures of all three complexes were
confirmed by single-crystal X-ray diffractometry, and
the structures and selected bond distances and angles
are given in Figures 1-3. The complexes were found to
be monomers and to exhibit twisted dioxametallocyclo-
pentane rings similar to those reported for other metal
glycolate complexes.18,19 This twisting resulted in dis-
order in the crystal structures, as both “left-handed” and
Resu lts
Syn th esis a n d Th er m a l Sta bility of P in a cola te
Com p lexes. The group 9 pentamethylcyclopentadienyl
pinacolate complexes 4-6 were synthesized by treat-
(17) Buist, G. J .; Bunton, C. A.; Hipperson, W. C. J . Chem. Soc. B
1971, 2128.
(18) Andrews, M. A.; Voss, E. J .; Gould, G. L.; Klooster, W. T.;
Koetzle, T. F. J . Am. Chem. Soc. 1994, 116, 5730.
(19) Klooster, W. T.; Voss, E. J . Inorg. Chim. Acta 1999, 285, 10.
(20) Buchman, B.; Piantini, U.; Philipsborn, W.; Salzer, A. Helv.
Chim. Acta 1987, 70, 1487.

2252 Organometallics, Vol. 20, No. 11, 2001
Holland et al.
adducts but do undergo clean thermal reactions with
CO and various heterocumulenes, and this chemistry
will be reported separately.
P h otoch em istr y of Cp *Co(p in ) (4). Irradiation of
the cobalt complex 4 in C6D6 (500 W mercury lamp,
through Pyrex) yielded 2 equiv of acetone (Scheme 2).
Immediately upon irradiation, this was the only reso-
1
nance evident in the H NMR spectrum. Over time in
the dark at room temperature, the initially purple
product solution turned brown, and all the resonances
in the 1H NMR spectrum of the sample became ex-
tremely broad. The secondary products of this reaction
have not been explored at length. Efforts to isolate the
purple intermediate were also unsuccessful, as the
complex does not precipitate from solution as clean
material and is unstable at higher concentrations. The
exposure of C6D6 solutions of 4 to ambient light, as well
as mercury-lamp irradiation of 4 in C6D12, yielded
similarly complex results, although in these cases a
dinuclear Co(II) pinacolate complex, 7, was also gener-
ated in modest and variable yield (<50%). This complex
could be produced in 80% spectroscopic yield by adding
a second equivalent of 4 to a freshly irradiated solution
of 4 in C6D6 and allowing the mixture to stand in the
dark for 1 day. The identity of the dinuclear cobalt
pinacolate 7 was confirmed by independent synthesis
from pinacol, base, and [Cp*CoCl]2 (Scheme 2).
We sought to simplify this reaction by trapping the
[Cp*Co] fragment to form a single product. Various
phosphines, isonitriles, silanes, and hydrogen reacted
thermally with the starting material, and bipyridine and
several dienes reacted upon irradiation with cobalt
pinacolate 4 to give multiple products. In the presence
of excess 2,3-dimethylbutadiene, irradiation of 4 in C6D6
still initially yielded 2 equiv of acetone and purple,
NMR-silent material. Over a matter of hours, however,
the known diene adduct 820 was formed in >90% yield
as determined by 1H NMR spectroscopy. In C6D12, 8 was
produced in >95% yield immediately upon irradiation
F igu r e 3. ORTEP diagram of 6 (displacement ellipsoids
at 50% probability). Hydrogen atoms have been omitted
for clarity. Selected bond distances (Å) and angles (deg):
Ir-O(1), 1.936(3); Ir-O(2), 1.936(3); C(1)-O(1), 1.51(1);
C(2)-O(2), 1.50(1); C(1)-C(2), 1.51(1); O(1)-Ir-O(2), 80.2-
(2); Ir-O(1)-C(1), 112.6(4); Ir-O(2)-C(2), 111.6(5).
“right-handed” conformations cocrystallized owing to the
essentially identical positions of the four methyl groups
in the two twist isomers. Figures 1-3 show the results
of modeling of the central carbon and oxygen atoms of
the pinacolate as half-occupancy pairs of atoms with
1
1
isotropic thermal parameters. The H NMR spectra of
(as determined by H NMR spectroscopy) and isolated
all three complexes at room temperature each showed
only one signal for the four pinacolate methyl groups,
indicating a low barrier to interconversion of the two
possible ring-twist isomers.
in 70% yield by recrystallization from ether (eq 5).
All the complexes were found to be moderately air-
and water-sensitive but were thermally stable in solu-
tion at room temperature and to varying extents at
elevated temperatures. Upon extended heating, all three
compounds ultimately decomposed to yield numerous
unidentified products. While the complexes are formally
16 e- species, they show no propensity to form stable
18 e- adducts. Addition of 1 equiv of trimethylphosphine
to iridium pinacolate 6 did yield a new, yellow compound
1
with features in its H and 31P NMR spectra consistent
with the formulation Cp*Ir(PMe3)(pin) (δ 1.49 (d, 2.0
Hz, 15H) 1.43 (s, 6H) 1.42 (s, 6H) 1.35 (d, 10.4 Hz, 9 H)
ppm). However, this complex decomposed at room
temperature within several minutes to give multiple
products, including the starting pinacolate 6. Similarly,
addition of various trialkylphosphines, triarylphos-
phines, or isonitriles to these complexes led to decom-
position to multiple products, and no adducts were
observed at room temperature (as monitored by 1H
NMR spectroscopy) with imines, ethers, dienes, amines,
or thiophenes. These complexes do not form stable
P h otoch em istr y of Cp *Rh (p in ) (5). Irradiation of
5 in C6D6 for 30-40 min yielded 2 equiv of acetone along
with the di- and trinuclear, solvent-incorporated prod-
ucts 9 and 10 (Scheme 3; Table 1, entry 2). These
products typically accounted for >90% of the total
1
cyclopentadienyl resonances in the H NMR spectra of
the products, and their characterization is described
below. The ratio of species 9 and 10 remained un-
changed when the product mixture was subjected to the
photolysis conditions for an additional 24 h or more.

Group 9 Metal Pinacolate Complexes
Organometallics, Vol. 20, No. 11, 2001 2253
Sch em e 2
Sch em e 3
Ta ble 1. P h otolysis of 5
[5] (C6D6)
light
filter cutoff
time
amt of acetone,
entry
((0.5 mM)
source
filter
(nm)
required
equiv ((0.1)
% 11 ((5)
% 10 ((5)
% 9 ((5)
1
2
3
4
5
6
7
20
20
20
20
20
0.1
Hg lamp
Hg lamp
Hg lamp
Hg lamp
ambient
ambient
Hg lamp
quartz
Pyrex
U-glass
amber
Pyrex
Pyrex
Pyrex
<150
300
334
402
300
300
300
<20 min
2
2
0
<5
10
15
50
<5
>95
30
35
40
40
45
30
0
65
55
45
40
5
30-40 min
40-50 min
70-80 min
36-48 h
1.9
1.8
1.5
na
1.0
1-2 h
30-60 min
50
0
20 (C6D12
)
Identically prepared samples of 5 exposed to ambient
fluorescent light for 1-2 days also reacted to completion
but produced somewhat less than 2 equiv of acetone and
yielded a dramatically different product ratio (Table 1,
entry 5). Dinuclear benzene adduct 9 was almost
entirely absent, and the new dinuclear pinacolate 11
was observed in addition to trinuclear complex 10. This
product mixture was stable to irradiation for at least
24 h. Early in the course of both ambient and mercury-
lamp photolyses (<5% conversion), only products 9 and
11 were observed. The concentration of dinuclear ben-
zene adduct 9 was actually observed by 1H NMR to
decrease when a briefly irradiated sample containing
∼50% starting material was exposed to ambient light
for 1 day.
The use of filters with the mercury lamp (Table 1,
entries 3 and 4) could not reproduce the ambient-light
product ratio. When filters were used, dinuclear pina-
colate 11 was produced in significant quantities, the
production of benzene adduct 9 was noticeably dimin-
ished, and the time required for complete disappearance
of starting material increased. Irradiation through
quartz (entry 1) produced the opposite result; a product
mixture unusually rich in 9 and poor in 11 was
observed.
When a very dilute (0.1 mM) benzene solution of 5
was exposed to ambient light (Table 1, entry 6), com-
plete conversion to a product mixture similar to that
obtained through mercury-lamp irradiation was ob-
served within 2 h.

2254 Organometallics, Vol. 20, No. 11, 2001
Holland et al.
Sch em e 4
The irradiation of 5 in cyclohexane-d12 generated only
1 equiv of acetone and a nearly quantitative yield of the
dinuclear Rh(II) pinacolate 11 (Scheme 4; Table 1, entry
7). This species was thus obtained on a larger scale in
54% isolated yield by irradiation of 5 in pentane and
subsequent recrystallization from ether. The 1H and 13C-
{1H} NMR spectra of the compound suggest a single
pinacolate ligand and two equivalent Cp* groups, and
both mass spectrometry and elemental analysis con-
firmed the formulation (Cp*Rh)2(pin). This species was
found to undergo no exchange with acetone-d6 upon
either irradiation or mild heating. Stirring 11 in metha-
nol for 3 days yielded the known compound (Cp*Rh)2-
(OMe)2 (12)21 and pinacol (both in 60% yield as mea-
sured by 1H NMR spectroscopy). Treatment of the
dimethoxo species 12 with pinacol produced 11 in
quantitative NMR yield after 1 h at 45 °C, and the
product was isolated in 39% yield. These results suggest
a structure in which a single, intact pinacolate ligand
bridges two [Cp*Rh] fragments, as illustrated in Scheme
4.
photolyses in C6D6 were also recrystallized from ether
(<10% yield) to give similar black material that lacked
1
any downfield H NMR resonances. A mass spectrum
of this partially deuterated 10 included an M+ peak at
m/z 798, which corresponds to (Cp*Rh)3(C6D6). Although
efforts to cleanly isolate the dinuclear species 9 have
been unsuccessful, spectra of 9 and 10 together include
a singlet at δ 1.75 ppm (30H) and 3 multiplets downfield
(6H) that are not associated with trinuclear adduct 10.
While we lack sufficient evidence to unambiguously
assign structures to 9 or 10, the NMR data suggest that
the benzene unit is intact (no typical rhodium-hydride
signals are observed) and that in both cases the benzene
most likely is bound to each rhodium center in an η2 or
η4 manner. The NMR spectra of such bridging arene
complexes22-26 sometimes exhibit large shifts in the
bound arene resonances, as do those reported here.
Unfortunately, it appears that no known complexes
exhibit spectroscopic features sufficiently similar to
those observed here to offer particularly useful com-
parisons. Possible structures for complexes 9 and 10 are
illustrated in Figure 4.
Products 9 and 10 were prepared on a larger scale
by mercury-lamp irradiation of 5 in C6H6 through Pyrex.
Recrystallization of the photoproducts from diethyl ether
provided compound 10 as a black solid (26% yield) which
Again we sought to simplify the system by the use of
an appropriate trap and found that irradiation of 5 in
the presence of excess 2,3-dimethylbutadiene produced
2 equiv of acetone and the diene adduct Cp*Rh(diene)
(13) in >95% yield by NMR spectroscopy (eq 6) and 66%
isolated yield. The products of reaction showed no
dependence on solvent or light source.
1
exhibited H NMR singlets at δ 2.06 and 1.78 ppm in a
15:30 ratio, and three additional multiplets downfield
integrated to two protons each. Products from various
The iridium analogue 6 proved stable to 8 h mercury-
lamp irradiation through quartz both with and without
added 2,3-dimethylbutadiene. Extended irradiation (>24
(21) Koelle, U.; Hoernig, A.; Englert, U. J . Organomet. Chem. 1992,
438, 309.
(22) Budzelaar, P. H. M.; Moonen, N. N. P.; Gleder, R.; Smits, J . M.
M.; Gal, A. W. Chem. Eur. J . 2000, 6, 2740.
(23) Muller, J .; Gaede, P. E.; Qiao, K. Angew. Chem., Int. Ed. Engl.
1993, 32, 1697.
(24) Schneider, J . J .; Wolf, D.; J aniak, C.; Heinemann, O.; Rust, J .;
Kruger, C. Chem. Eur. J . 1998, 4, 1982.
(25) Schneider, J . J .; Denninger, U.; Heinemann, O.; Kruger, C.
Angew. Chem., Int. Ed. Engl. 1995, 34, 592.
(26) Wadepohl, H. Angew. Chem., Int. Ed. Engl. 1992, 31, 247.
F igu r e 4. Possible structures for complexes 9 and 10.

Group 9 Metal Pinacolate Complexes
Organometallics, Vol. 20, No. 11, 2001 2255
Ta ble 2. Ca ta lysis Ru n s, u sin g O2
h) under the same conditions produced very slight (<5%)
decomposition to numerous products but did not produce
acetone.
[Rh] pinacol
(mM (equiv O2 press
turnovers
per Rh
((0.2)
entry
catalyst
solvent (1) (10%) ((1 Torr)
Oxid a t ion of (Cp *R h )2(p in ) (7). In the course of
exploring the stability of dinuclear rhodium species 11,
it was observed that traces of the mononuclear pinaco-
late 5 were sometimes generated in impure solutions
containing pinacol. Neither the heating of 11 in neat
pinacol at 75 °C nor the irradiation of 11 in the presence
of >200 equiv of pinacol resulted in any reaction. The
exposure of 11 to air yielded only slow decomposition.
The reaction of 11 with both the atmosphere and an
excess of pinacol, however, yielded 5 in 75% spectro-
scopic yield (based on rhodium) after 48 h in the dark
at room temperature (eq 7). The use of pure O2 instead
1
2
3
4
5
6
5
11
5
5
11
C6D6
C6D6
C6D12
C6D6
C6D6
17
13
17
14
14
14
200
20
20
20
20
20
750
746
375
389
380
392
9.8
2.8
3.2
3.4
2.4
2.7
photolyzed 5 C6D6
approximately 40 turnovers (20 TO/Rh) were observed
over the course of 18 h (Table 2). The pale yellow
solution and red precipitate remaining at the end of this
time showed no catalytic activity. When the reaction
was repeated under otherwise identical conditions using
0.02 M 5, 28 turnovers (28 TO/Rh) were observed in the
same time period, and the organometallic products were
similarly unreactive. The 1H NMR spectrum of the
catalysis reaction mixture after 40 min of irradiation
revealed that 5 and a small amount of 9 were the
dominant species present. Complex 5 was found to be a
better catalyst than a mixture of its photoproducts.
While catalysis using oxygen as the oxidant required
high concentrations of pinacol and did not achieve better
than 25% conversion, catalysis using N2O proved less
demanding and more efficient. When an NMR tube was
loaded with 0.04 M 5, 0.80 M pinacol, and 1 atm N2O,
several hours of irradiation with periodic shaking
produced acetone in 85% yield and consumed all the
starting pinacol.
Rela ted Diola te Com p lexes. As is the case with
other metal glycolates, the pinacolate ligand in this
system was observed by 1H NMR spectroscopy to
undergo facile exchange with free diol (the chemical
shifts of the pinacolate ligand and free pinacol converge
as their concentrations increase).18 In addition, treat-
ment of rhodium pinacolate 5 with catechol in THF
generated the known catecholate species 1427 and
pinacol quantitatively upon mixing (Scheme 5). Neither
14 nor its phosphine adduct 15 undergo any observable
reaction upon irradiation in the presence of excess
dimethylbutadiene. Treatment of 4, 5, or 6 with second-
ary and primary diols (butanediol, 1-phenyl-1,2-
ethanediol, or ethylene glycol) invariably led to rapid
decomposition, and complexes with such diolates also
proved inaccessible by direct synthesis from [Cp*RhCl2]2
or Cp*Rh(PMe3)Cl2.
of air gave similar results, although the yield could not
be improved from 75% by either increasing or decreasing
the O2 pressure. Complex 5 decomposed extremely
slowly when exposed to O2 and H2O in the presence of
pinacol; therefore, decomposition of the products under
the reaction conditions is likely not the cause of the
limited yield.
Other oxidizing agents have proven ineffective in
achieving this transformation. Nitrous oxide, pyridine
N-oxide, trimethylamine N-oxide, diphenyl hydrazine,
and 2-methyl-2-nitrosopropane all either failed to react
or decomposed upon extended heating or irradiation
with 11 and pinacol. The simultaneous use of a ferro-
cenium salt and base yielded only decomposition of 11,
as did an excess of nitrosobenzene. While oxidation by
tert-butyl peroxide eventually produced traces of 5, it
did so only after extended heating. The addition of tert-
butyl hydroperoxide to a mixture of 11 and pinacol did
afford a reasonable amount of 5 and tert-butyl alcohol
within 5 min of mixing. The yield of 5 in this reaction
was typically about 70%, as determined by 1H NMR
spectroscopy. With or without pinacol present, the cobalt
analogue 7 reacted with oxidizing agents to give 4 in
no greater than 50% yield and also produced dark,
highly insoluble material.
Discu ssion
The thermal stability of the pinacolate complexes
4-6, even though they are formally 16 e- monomeric
species, suggests that they are stabilized by significant
π donation from oxygen to the metal center. Although
there is little useful basis for comparison of the metal-
oxygen bond lengths, this notion is further supported
by the instability of these complexes toward good Lewis
bases. The pinacolate could not be prepared in good yield
from the corresponding Cp*M(PMe3)Cl2 complexes, and
attempts to prepare 18 e- adducts by treatment of the
isolated pinacolates with phosphines were similarly
unsuccessful. All such efforts led to decomposition, but
Ca ta lysis. Irradiation of rhodium pinacolate 5 in the
presence of excess pinacol and O2 resulted in the weakly
catalytic conversion of pinacol to acetone (eq 8). When
the reaction was carried out in an NMR tube with 0.01
M 11, 2.0 M pinacol, and bubbling O2 in C6D6 or C6D12,
(27) Espinet, P.; Bailey, P. M.; Maitlis, P. M. J . Chem. Soc., Dalton
Trans. 1979, 1542.

2256 Organometallics, Vol. 20, No. 11, 2001
Holland et al.
Sch em e 5
two Cp*Rh centers. This species is presumably formed
by reaction of the [Cp*Rh] fragment with 5, and its
structure is well precedented by numerous [Cp*RhX]2
dimers.32-35 In addition to this dinuclear pinacolate, two
new products that incorporate solvent molecules are
produced when the photolysis is conducted in benzene.
On the basis of the facts that 10 is not produced early
in the reaction under any circumstances, 9 is consumed
under appropriate circumstances, and 9 is not directly
converted to 10 upon irradiation, it seems likely that 9
reacts with the [Cp*Rh] fragment to produce the tri-
nuclear species 10. Product 9, then, is likely a dinuclear
species that results from the reaction of two [Cp*Rh]
fragments with solvent (or, perhaps more likely, two
short-lived Cp*Rh-solvent adducts).
The variable ratio of these products can be rational-
ized based on Scheme 6. The product ratio correlates
most directly with the duration of the photolysis (Table
1). During photolysis reactions which are complete after
short times, the short-lived [Cp*Rh] fragment should
be more abundant relative to other species; therefore,
the formation of 9 should be favored. Those reactions
in which the starting material persists longer, contain-
ing less [Cp*Rh] relative to other species, should favor
11 and 10 relative to 9. This indeed was observed. This
explanation is precedented by the work of Mu¨ller and
co-workers, who have observed similar reactivity on the
part of the unsubstituted [CpRh] fragment, which goes
on to form di- and trinuclear arene adducts when
generated by photolysis of CpRh(C2H4)2 in benzene.23
It is worth noting, however, that the NMR spectra of
complexes 9 and 10 are inconsistent with the basic
structural features reported for the analogues bearing
unsubstituted Cp ligands.
The preparation of the catecholate analogue 14 dem-
onstrates the viability of synthesizing various glycolate
complexes by simple exchange, but unfortunately this
approach has yet to yield any species in which C-C
cleavage can be observed. A [2 + 2 + 2] cycloreversion
in 14 would disrupt aromaticity; therefore, it is unsur-
prising that this species is stable to irradiation. Equally
unsurprising is the rapid decomposition of the less
substituted glycolate species presumably formed upon
exchange with the pinacolate. They, like the parent
compound, would be formally 16 e- species and have
an open site available for decomposition by â-hydride
elimination.36
in the case of iridium complex 6, NMR evidence indi-
cated that this decay took place via an intermediate
Cp*Ir(PMe3)(pin) species. The instability of such a
species is consistent with favorable π donation in the
parent pinacolate, which would be replaced by desta-
bilizing dπ-pπ repulsion in the adduct.28
Trapping experiments with dimethylbutadiene in
cyclohexane-d12 clearly illustrate that the cobalt and
rhodium species 4 and 5 undergo a photochemical
transformation analogous to that observed by Andrews
for similar types of platinum complexes. In C6D6, it
appears that the chemistry of the [Cp*Co] fragment
generated by photolysis of 4 is complicated by the
prompt formation of a long-lived paramagnetic species.
Given that this species appears to be formed only in
benzene, it seems most likely that it is some sort of
(Cp*Co)n(C6D6)m arene adduct. However, while such
complexes are well-known, they are either diamag-
netic24,25,29 or charged30,31 species. The exact nature of
the kinetic product thus remains unclear. In the pres-
ence of a better trap, this species went on to eventually
produce the expected thermodynamic products (Scheme
4) and generated either the diene adduct 8 or the
dinuclear pinacolate 7 (when mononuclear pinacolate
4 was added as the trap). The reactivity of the (Cp*Co)n-
(C6D6)m species was not further explored because it
could not be isolated and tended to produce multiple
products, including paramagnetic material.
Given the results for the cobalt pinacolate 4 and the
rhodium analogue 5, the iridium pinacolate 6 is remark-
ably stable even to high-energy irradiation. Like the
other pinacolates, it has strong absorbances in both the
ultraviolet and visible regions; thus, it is unlikely that
it simply lacks an accessible chromophore. While third-
row-metal complexes are generally less reactive than
their first- and second-row counterparts, it is nonethe-
less surprising that even the high energy available from
UV radiation does not effect cycloreversion or other
decomposition.
The fate of the [Cp*Rh] fragment without an added
diene was dependent on the reaction conditions (Scheme
6). Irradiation in the absence of aromatic solvents
yielded a single product identified as dinuclear Rh(II)
complex 11, in which a single pinacolate ligand bridges
(32) Caulton, K. G.; Cotton, F. A. J . Am. Chem. Soc. 1969, 91, 6517.
(33) Caulton, K. G.; Cotton, F. A. J . Am. Chem. Soc. 1971, 93, 1914.
(34) Felthouse, T. R. Prog. Inorg. Chem. 1982, 29, 73.
(35) Werner, H.; Klingert, B.; Rheingold, A. L. Organometallics 1988,
7, 911.
(36) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, 5th
ed.; Wiley: New York, 1988.
(28) Caulton, K. G. New J . Chem. 1994, 18, 25.
(29) Schneider, J . J . Z. Naturforsch., B 1995, 50, 1055.
(30) Lai, Y.-H.; Tam, W.; Vollhardt, K. P. C. J . Organomet. Chem.
1981, 216, 97.
(31) Koelle, U.; Fuss, B.; Rajasekharan, M. V.; Ramakrishna, B. L.;
Ammeter, J . H.; Bohm, M. C. J . Am. Chem. Soc. 1984, 106, 4152.

Group 9 Metal Pinacolate Complexes
Organometallics, Vol. 20, No. 11, 2001 2257
Sch em e 6
The regeneration of 5 from the dinuclear rhodium
pinacolate 11 can be accomplished in reasonable yield
even by the very crude method of exposing an ap-
propriately prepared sample to the atmosphere. Unfor-
tunately, while the same reaction takes place under
pure O2, the use of O2 at various pressures does not offer
improved results. Oxidation with other reagents proved
fruitless, as oxygen atom donors are totally ineffective,
and oxidations through electron transfer or using ni-
trosobenzene do not yield clean reactions. The specific
need for dioxygen suggests a mechanism involving
oxygen insertion into the Rh-Rh bond, which has been
observed for a similar Rh(II) dimer by Sharp and co-
workers.37,38 As shown in Scheme 7, the peroxo ligand
produced by oxygen insertion could be replaced by
pinacol via protonolysis to give two molecules of 5 after
dimer cleavage. The hydrogen peroxide produced could
then react with another 1 equiv of 11 by addition across
the Rh-Rh bond, and subsequent alcohol exchange
would yield 5 and H2O. The reaction of 11 with tert-
butyl hydroperoxide provides support for the viability
of the second phase of the mechanism described above.
Given that the mononuclear pinacolate 5 decomposes
slowly in the presence of either oxygen or peroxide, we
suspect that this decomposition pathway (or similar
reaction by an intermediate rhodium dihydroxo species)
limits the yield of the stoichiometric oxidation reaction.
The fact that one of the photoproducts can incorporate
additional pinacol to generate 5 suggested that the
catalytic conversion of pinacol to acetone should be
possible in this system. This has been found to be true,
and several turnovers are consistently observed in NMR
tube scale reactions using either 5 or 11 as a catalyst
and oxygen as the oxidant. Results are slightly better
when 5 is used (Table 2, entries 4 and 5), likely due to
the modest yield of the initial oxidation of 11. Under
these conditions, the number of turnovers observed is
clearly dependent upon the pinacol concentration (Table
2, entry 1), which is consistent with the notion that
rapid conversion of the proposed dihydroxo species is
important in preventing decomposition.
Monitoring by 1H NMR provides evidence that di-
nuclear pinacolate 11, while it was the subject of our
initial oxidation experiments, is not produced in the
course of the catalysis in C6D6. Furthermore, when
conditions are optimized for the production of 11 (by the
use of C6D12 in place of C6D6), no improvement in the
yield of the catalytic reaction is observed (entries 3 and
4). This suggests that the stoichiometric oxidation
mechanism discussed above is not operative in the
catalytic system (Scheme 8). One possible alternative
is that other photoproducts 9 and 10 are oxidized as
well, but efforts to model this step stoichiometrically
produced 5 in very low yield, and a mixture of photo-
products showed diminished catalytic activity relative
Sch em e 7
(37) Hoard, D. W.; Sharp, P. R. Inorg. Chem. 1993, 32, 612.
(38) Sharp, P. R.; Hoard, D. W.; Barnes, C. L. J . Am. Chem. Soc.
1990, 112, 2024.

2258 Organometallics, Vol. 20, No. 11, 2001
Holland et al.
otherwise noted. Cp*Co(pin) and solutions containing Cp*Co-
(pin) or Cp*Rh(pin) were protected from light by aluminum
foil except where otherwise noted. Pinacol was distilled under
vacuum from 45 to 0 °C, dried over 4 Å sieves in pentane, and
recrystallized from pentane at -30 °C. Catechol was dried over
4 Å sieves in pentane and recrystallized from pentane at -30
°C. The complexes [Cp*RhCl2]2,40 [Cp*IrCl2]2,40 [Cp*RhOMe]2
(12),21 [Cp*CoCl2]2,41 [Cp*CoCl]2,42 and Cp*Rh(PPh3)(o-O2C6H4)
(15)27 were prepared according to published literature proce-
dures. All other reagents were purchased commercially and
used as received.
Sch em e 8
Except where otherwise noted, yields of NMR tube reactions
were determined by integration against a 1,3,5-trimethoxy-
benzene internal standard (added with the solvent) using
single pulse spectra and an initial delay of at least 30 s. The
internal standard (δ 6.25 (s, 3H), 3.30 (s, 9H) ppm in C6D6)
was first calibrated with respect to starting materials by
means of an initial spectrum acquired before irradiation,
heating, or the addition of reactive species.
“Mercury-lamp” irradiation was performed using a 500 W
medium-pressure Hanovia mercury vapor lamp housed in a
water-cooled quartz immersion well. During irradiation, samples
were immersed in spectral grade methanol maintained by a
Neslab Endocal circulating refrigeration bath at 20 °C. “Ambi-
ent light” photolyses were conducted under typical fluorescent
room lights in the glovebox or on the laboratory bench.
Cp *Co(p in ) (4). A 20 mL vial wrapped with aluminum foil
was charged with a stir bar, [Cp*CoCl2]2 (507 mg, 0.959 mmol),
pinacol (227 mg, 1.92 mmol), and THF (10 mL). To the
resulting green slurry was added, dropwise with stirring, a
THF slurry (5 mL) of KN(TMS)2 (764 mg, 3.83 mmol). Over
the course of the addition the mixture became homogeneous
and took on a brownish tint. The vial was capped, and the
solution was stirred for 2 h at room temperature in the dark.
The solution was then reduced to dryness under vacuum to
yield a brownish green solid, which was extracted with benzene
(20 mL) to yield a green extract and brown-yellow residue. The
extract was reduced to dryness under vacuum, and recrystal-
lization of the resulting green solid from ether (6 mL) at -30
to 5 (entries 4 and 6). Rather, we suspect that the
[Cp*Rh] fragment is oxidized directly to a peroxo or
dihydroxo species by oxygen or hydroperoxide, respec-
tively.
This conjecture is supported by the observation that
nitrous oxide, failing to oxidize photoproducts 9-11
under either thermal or photochemical conditions, is
effective as a replacement for oxygen in the catalytic
system. This modification allows for catalysis at much
lower concentrations of pinacol (and thus much better
conversion), presumably because the side products of
the oxidation are only water and nitrogen gas rather
than hydrogen peroxide. The mechanism of this oxida-
tion pathway is still under investigation but presumably
involves either trapping of [Cp*Rh] by pinacol and
subsequent hydrogen atom abstraction by N2O, or
oxygen atom transfer from N2O to [Cp*Rh] (to form a
short-lived rhodium-oxo species) followed by protonoly-
sis.
1
°C yielded 4 as blocky green crystals (442 mg, 75% yield). H
NMR (C6D6): δ 1.68 (s, 15H), 0.97 (s, 12H) ppm. 13C{1H} NMR
(C6D6): δ 90.53 (s, C5Me5), 80.88 (s, OCMe2), 29.39 (s, OCMe2),
8.43 (s, C5Me5) ppm. IR (KBr): 2959 (s), 1457 (m), 1348 (m),
1140 (s), 953 (s), 891 (s) cm-1. UV-vis (nm (M-1 cm-1)): 227
(11 000), 286 (11 000), 365 (3900), 662 (200), 922 (330). Anal.
Calcd for C16H27O2Co: C, 61.93; H, 8.77. Found: C, 61.93; H,
8.91.
Con clu sion s
Cr ysta llogr a p h ic Stu d y of Cp *Co(p in ) (4). A green
plate-shaped crystal having approximate dimensions 0.20 ×
0.22 × 0.18 mm obtained by recrystallization from ether at
-30 °C was mounted on a glass fiber using Paratone N
hydrocarbon oil. Data were collected using a SMART CCD area
detector with graphite monochromated Mo KR radiation.
Hydrogen atoms were included but not refined. The structure
obtained is disordered about the mirror plane at y ) 1/4,
because the pinacolate is coordinated almost perpendicular to
the mirror plane and apparently adopts both “left-handed” and
“right-handed” conformations in the solid state. Because the
two conformations have very similar steric bulk, neither
conformation is favored upon crystallization, resulting in the
two superimposed five-membered rings. The two conformations
also superimpose the pinacolate methyl groups upon each
other. Two of these methyl groups, C10 and C11, are located
near the mirror plane at y ) 1/4 and have been refined as full-
occupancy carbon atoms lying on the mirror plane. The large
“thermal motion” parameters for these two atoms result from
Both cobalt and rhodium pinacolate complexes have
been observed to produce acetone upon irradiation. The
[Cp*M] fragments produced by this reaction have been
trapped cleanly by 2,3-dimethylbutadiene and have also
been observed to react with their respective starting
materials to form M(II) dimers. The rhodium dimer 11
reacts with pinacol and oxygen to regenerate the
original Cp*Rh(pin) species (5), likely by a mechanism
involving dioxygen insertion into the Rh-Rh bond and
subsequent exchange of diolate for the peroxo ligand.
Photolysis of the rhodium complex in the presence of
pinacol and either oxygen or nitrous oxide results in
catalytic oxidative cleavage of pinacol to acetone. Ap-
plication of this chemistry to less highly substituted
diols will require the use of metal centers less prone to
â-hydride elimination.
Exp er im en ta l Section
Gen er a l P r oced u r es. The general experimental proce-
(40) Kang, J . W.; Moseley, K.; Maitlis, P. M. J . Am. Chem. Soc. 1969,
91, 5970.
dures previously outlined by our group39 were followed unless
(41) Plitzko, K.-D.; Boekelheide, V. Organometallics 1988, 7, 1573.
(42) Koelle, U.; Fuss, B.; Belting, M.; Raabe, E. Organometallics
1986, 5, 980.
(39) Alaimo, P. J .; Arndtsen, B. A.; Bergman, R. G. Organometallics
2000, 19, 2130.

Group 9 Metal Pinacolate Complexes
Organometallics, Vol. 20, No. 11, 2001 2259
the assumption that they are imposed on the mirror plane.
The clarified structure given in Figure 1 is based on modeling
of the central carbon and oxygen atoms of the ligand as half-
occupancy pairs of atoms with isotropic thermal parameters.
Selected bond lengths are given in Figure 1. Crystallographic
data are as follows: space group Pnma, a ) 16.3777(5) Å, b )
12.3570(4) Å, c ) 7.9567(3) Å, V ) 1610.27(8) Å3, Z ) 4, Fcalcd
) 1.280 g/cm3, µ(Mo KR) ) 10.62 cm-1, 1733 unique reflections,
1317 reflections with I > 3.00σ(I), R ) 3.12%.
13C{1H} NMR: δ 85.54 (s, OCMe2), 81.61 (s, C5Me5), 27.44 (s,
OCMe2), 10.10 (s, C5Me5) ppm. IR (KBr): 1139 (s), 951 (m),
893 (m) cm-1. Anal. Calcd for C16H27O2Ir: C, 43.32; H, 6.16.
Found: C, 43.24; H, 6.14.
Cr ysta llogr a p h ic Stu d y of Cp *Ir (p in ) (6). A red prism
crystal having approximate dimensions 0.20 × 0.12 × 0.10 mm
was mounted on a glass fiber using Paratone N hydrocarbon
oil. Data were collected using a SMART CCD area detector
with graphite-monochromated Mo KR radiation. Hydrogen
atoms were included but not refined. The structure showed
disorder similar to that observed for 4 (see above), and the
structure given in Figure 3 is based on modeling of the central
carbon atoms of the ligand as half-occupancy pairs of atoms
with isotropic thermal parameters. Selected bond lengths are
given in Figure 3. Crystallographic data are as follows: space
group Pnma, a ) 17.028(1) Å, b ) 12.3059(9) Å, c ) 7.9491(6)
Å, V ) 1665.6(2) Å3, Z ) 4, Fcalcd ) 1.77 g/cm3, µ(Mo KR) )
80.35 cm-1, 1645 unique reflections, 1225 reflections with I >
3.00σ(I), R ) 1.8%.
(Cp *Co)2(p in ) (7). A 20 mL vial was charged with a stir
bar, [Cp*CoCl]2 (110 mg, 239 mmol), pinacol (29.0 mg, 246
mmol), and 5 mL of THF. To the resulting brown solution was
added, dropwise with stirring, 5 mL of a THF solution of KN-
(TMS)2 (95.9 mg, 482 mmol). Over the course of the addition,
the mixture took on a greenish tint. The vial was capped, and
the solution was stirred for 2 h at room temperature. The
solution was then reduced under vacuum to yield a brownish
green solid, which was extracted with 20 mL of benzene to
yield a brownish green extract and a brown residue. The
extract was reduced to dryness under vacuum, and recrystal-
lization of the resulting brownish green solids from 2 mL of
diethyl ether at -30 °C yielded (Cp*Co)2(pin) as small brown-
green crystals (85.1 mg, 70% yield). 1H NMR (C6D6): δ 1.76
(s, 30H), 1.51 (s, 12H) ppm. 13C{1H} NMR (C6D6): δ 87.82 (s,
C5Me5), 84.26 (s, OCMe2), 30.21 (s, OCMe2), 10.36 (s, C5Me5)
ppm. IR (KBr): 2975 (s), 1452 (s), 1378 (s), 1140 (m), 953 (m)
cm-1. EI-MS: m/z 504 ([(Cp*Co)2(pin)]+), 194 ([Cp*Co]+). Anal.
Calcd for C26H42O2Co2: C, 61.90; H, 8.39. Found: C, 61.74; H,
8.56.
Cp *Rh (p in ) (5). A 20 mL vial wrapped with aluminum foil
was charged with a stir bar, [Cp*RhCl2]2 (519 mg, 0.840 mmol),
pinacol (199 mg, 1.68 mmol), and THF (10 mL). To the
resulting brick red slurry was added, dropwise with stirring,
a THF slurry (5 mL) of KN(TMS)2 (669 mg, 3.36 mmol). Over
the course of the addition, the mixture became homogeneous
and turned purple-brown. The vial was capped, and the
solution was stirred for 1 h at room temperature in the dark.
The solution was then reduced to dryness under vacuum to
yield a purple solid, which was extracted with benzene (30 mL)
to yield a purple extract and brown-yellow residue. The extract
was reduced to dryness under vacuum, and recrystallization
of the resulting purple solids from ether (7 mL) at -30 °C
yielded 5 as blocky purple crystals (430 mg, 72% yield). 1H
NMR (C6D6): δ 1.42 (s, 12H), 1.31 (s, 15H) ppm. In the
presence of pinacol, the δ 1.42 ppm resonance is shifted upfield
and slightly broadened due to facile exchange with free diol.
For all cases where pinacol is present, 5 was identified and
quantified only on the basis of the δ 1.31 ppm signal or the
sum of the two signals when they overlapped. Removal of
pinacol by sublimation returned the pinacolate resonance to
δ 1.42 ppm. 13C{1H} NMR (C6D6): δ 89.35 (d, J Rh-C ) 8.7 Hz,
C5Me5), 84.67 (s, OCMe2), 28.07 (s, OCMe2), 9.40 (s, C5Me5)
ppm. IR (KBr): 1140 (s), 1028 (s), 951 (s) cm-1. UV-vis (nm
(M-1 cm-1)): 234 (12 000), 335 (3900), 585 (570). EI-MS: m/z
296 ([Cp*Rh(CMe2O)]+), 238 ([Cp*Rh]+). Anal. Calcd for
C
16H27O2Rh: C, 54.23; H, 7.68. Found: C, 54.00; H, 7.98.
Cr ysta llogr a p h ic Stu d y of Cp *Rh (p in ) (5). Small red-
brown crystals of the compound were obtained by slow crystal-
lization from ether at -30 °C. A crystal was mounted on a
glass fiber using polycyanoacrylate cement. Data were col-
lected at -126 °C using an Enraf CAD-4 diffractometer. The
2438 raw intensity data were converted to structure factor
amplitudes and their esd’s by correction for scan speed,
background, and Lorentz and polarization effects. The struc-
ture was refined in the space group Pnam, and removal of
systematically absent data and averaging of Friedel planes
gave 1137 unique data. The structure was solved by Patterson
methods and refined via standard least-squares and Fourier
techniques. The structure showed disorder similar to that
observed for 4 (see above), and the structure given in Figure
2 is based on modeling of the central carbon and oxygen atoms
of the ligand as half-occupancy pairs of atoms with isotropic
thermal parameters. Selected bond lengths are given in Figure
2. Crystallographic data are as follows: space group Pnma, a
) 16.9561(4) Å, b ) 7.8747(5) Å, c ) 12.2873(4) Å, V ) 1640.6-
(8) Å3, Z ) 4, Fcalcd ) 1.434 g/cm3, 1137 unique reflections, 1062
reflections with I > 3.00σ(I), R ) 2.19%.
Cp *Ir (p in ) (6). A 20 mL vial was charged with a stir bar,
[Cp*IrCl2]2 (340 mg, 0.427 mmol), pinacol (102 mg, 0.863
mmol), and THF (10 mL). To the resulting orange solution was
added, dropwise with stirring, a THF slurry (5 mL) of KN-
(TMS)2 (343 mg, 1.72 mmol). Over the course of the addition,
the mixture became homogeneous and turned dark red-purple.
The vial was capped, and the solution was stirred for 2 h at
room temperature. The solution was then reduced to dryness
under vacuum to yield a red powder, which was extracted with
benzene (20 mL) to yield a red-purple extract and brown-yellow
residue. The extract was reduced to dryness under vacuum,
and recrystallization of the resulting red residue from ether
(4 mL) at -30 °C yielded 6 as 295 mg of dark red needles (80%
yield). 1H NMR (C6D6): δ 1.48 (s, 12H), 1.41 (s, 15H) ppm.
Tr ea tm en t of Cp *Co(p in ) P h otolysis In ter m ed ia te
w ith Cp *Co(p in ) (4). In the glovebox, 4 (2.0 mg, 6.5 µmol)
was dissolved in C6D6 to yield a bright green solution which
was transferred to a J . Young tube. The tube was irradiated
for 60 min, after which time the solution was purple, and
monitoring of the reaction by 1H NMR spectroscopy showed
that acetone (δ 1.54 (s, 2 × 6H) ppm) was the only NMR-active
product. (When such purple, irradiated samples were allowed
to stand in the dark at room temperature, they turned brown
1
over the course of 24 h and yielded only broad, noisy H NMR
spectra.) In the box, more Cp*Co(pin) (2.0 mg, 6.5 µmol) was
added to the purple solution, and it was left in the dark at 20
°C for 2 days. The reaction was monitored by 1H NMR
1
spectroscopy, which indicated formation of 7 in 80% yield. H
NMR (C6D6): δ 1.76 (s, 30H), 1.51 (s, 12H) ppm.
P h otolysis of Cp *Co(p in ) (4) in C6D12 w ith 2,3-Dim -
eth ylbu ta d ien e. In the glovebox, 4 (5.8 mg, 18.8 µmol) and
2,3-dimethylbutadiene (10 µL, 120 µmol) were dissolved in
C6D12 and the solution was transferred to a J . Young NMR
1
tube. The tube was irradiated for 40 min, after which time H
NMR spectroscopy indicated complete conversion to acetone
and the product 8. Removal of volatile materials under vacuum
yielded 5.0 mg of spectroscopically pure 8 as a brown solid (93%
1
yield). H NMR (C6D12): δ 1.96 (s, 2 × 6H, acetone), 1.81 (s,
6H), 1.74 (s, 15H), 0.89 (br, 2H), -0.70 (br, 2H) ppm. 1H NMR
(C6D6): δ 1.80 (s, 6H), 1.69 (s, 15H), 1.06 (br, 2H), -0.41 (br,
2H) ppm. 13C{1H} NMR (C6D6): δ 88.62 (s, C5Me5), 87.64 (s,
CH2CMe), 38.63 (s, CH2CMe), 17.29 (CH2CMe), 9.98 (C5Me5)
ppm. (Lit.20 13C{1H} NMR (C6D6): δ 88.6, 87.6, 38.6, 17.3, 10.0
1
ppm. No H NMR data were included in the reference.)
P h otolysis of Cp *Co(p in ) (4) w ith 2,3-Dim eth ylbu ta -
d ien e in C6D6. In the glovebox, 5 (6.9 mg, 22.2 µmol) and 2,3-

2260 Organometallics, Vol. 20, No. 11, 2001
Holland et al.
13C{1H} NMR (C6D6): δ 184.59 (m,38 aryl), 96.13 (t, J Rh-C
1.8 Hz, aryl), 94.50 (d, J Rh-C ) 4.8 Hz, C5Me5), 75.85 (d, J Rh-C
) 5.5 Hz, aryl), 10.82 (s, C5Me5) ppm.
dimethylbutadiene (15 µL, 180 µmol) were dissolved in C6D6,
and the solution was transferred to a J . Young NMR tube. The
tube was irradiated for 40 min, and 1H NMR spectroscopy
indicated acetone (δ 1.54 ppm) was the only NMR-active
product. Although the starting material was completely con-
sumed, no diene had been consumed. The reaction was allowed
to continue at 20 °C in the dark for 2 days, after which time
)
P h otolysis of Cp *Rh (p in ) (5) in C6D6. (a ) Va r iou s Ligh t
Sou r ces. In the glovebox, a stock solution of 5 (18.1 mg, 51.0
µmol) in C6D6 (2.5 mL, 20 mM) was prepared and divided
roughly evenly among four Pyrex J . Young NMR tubes and a
quartz Schlenk cuvette. The quartz cuvette (1) and one tube
(2) were irradiated directly with the mercury lamp, another
tube (3) was irradiated through a uranium glass filter, and
another (4) was irradiated through an amber filter. After 20
min of irradiation, the contents of the cuvette were transferred
to an NMR tube in the glovebox, and 1H NMR spectra of all
the tubes were acquired. At this point, no starting material
remained in the solution irradiated through quartz. The other
tubes 2-4 (showing approximately 50, 40, and 30% conversion,
respectively) were irradiated as before and monitored every
10 min by 1H NMR spectroscopy until no starting material
remained. The fourth tube was exposed to ambient light in
1
monitoring by H NMR spectroscopy showed that the product
8 had been formed in 80% yield. Complex 8 was isolated as a
brown solid by recrystallization from pentane at -30 °C (3.9
mg, 64% yield).
(Cp *Rh )2(p in ) (11). In the glovebox, 5 (50.7 mg, 143 µmol)
was dissolved in pentane (50 mL) in a 100 mL glass bomb.
This was exposed to ambient light for 1 week, over which time
a color change from purple to maroon occurred. The solution
was then reduced under vacuum to a dark oily material, which
was recrystallized repeatedly from diethyl ether to yield 11
1
as a black solid (23.1 mg, 54% yield). H NMR (C6D6): δ 1.75
(s, 30H), 1.51 (s, 12H) ppm. 13C{1H} NMR (C6D6): δ 87.86 (s,
OCMe2), 86.91 (d, J Rh-C ) 8.6 Hz, C5Me5), 29.95 (s, OCMe2),
11.42 (s, C5Me5) ppm. IR (KBr): 2957 (s), 1452 (m), 1376 (s),
1139 (s), 950 (m), 890 (m) cm-1. EI-MS: m/z 592 ([(Cp*Rh)2-
(pin)]+), 373 ([Cp*2Rh]+), 238 ([Cp*Rh]+). Anal. Calcd for
1
the laboratory and monitored by H NMR spectroscopy peri-
odically until no starting material remained. All were subse-
quently irradiated for an additional 24 h, and 1H NMR
spectroscopy after this time revealed no significant change in
any of the product mixtures. Relative product concentrations
were determined by 1H NMR spectroscopy and reported in
Table 1 as proportions of the total Cp* resonances. Only the
Cp* resonances reported for 9 and 10 were observed. Com-
pound 10 was isolated from the combined product mixture by
recrystallization from ether (2.0 mg black solid, 11% yield),
and a mass spectrum was obtained. EI-MS: m/z 798 ([(Cp*Rh)3-
(C6D6)]+), 373 ([Cp*2Rh]+).
(b) Ir r a d ia tion Ca r r ied Ou t u sin g Low Con cen tr a tion
of 5. In the glovebox, a 20 mL vial was charged with 5 (0.4
mg, 1.1 µmol) and C6H6 (10 mL, 0.1 mM). The faint purple
solution was left exposed to ambient light and observed every
20 min. Within 60 min, the color had changed to a pale yellow
that did not change visibly over the next 1 h. The solution was
reduced to dryness under vacuum, and the brown residue was
dissolved in C6D6 and transferred to an NMR tube. The product
ratio was determined and reported as described above.
(c) Ea r ly Con ver sion w ith Mer cu r y-La m p Ir r a d ia tion .
In the glovebox, 5 (2.5 mg, 7.1 µmol) was dissolved in C6D6,
and the solution was transferred to a J . Young NMR tube. The
tube was irradiated for 2 min, and a 1H NMR spectrum was
acquired which showed < 5% conversion of 5 to only products
11 (40%), 9 (60%), and acetone. The tube was then irradiated
for 40 min, and 1H NMR spectroscopy showed complete
conversion of 5 to predominantly 10 and 9.
C
26H42O2Rh2: C, 52.71; H, 7.15. Found: C, 52.63; H, 7.06.
Tr ea tm en t of (Cp *Rh OMe)2 (12) w ith P in a col. In the
glovebox, (Cp*RhOMe)2 (4.2 mg, 7.8 µmol) and pinacol (2.0 mg,
17.0 µmol) were dissolved in C6D6. The resulting purple
solution was transferred to a J . Young NMR tube, which was
subsequently heated at 45 °C for 1 h. After this time, the
1
solution was maroon. Monitoring of the reaction by H NMR
spectroscopy indicated quantitative conversion to product 11
and methanol (δ 3.05 (s, 2 × 3H) ppm). Recrystallization of
the product from ether yielded 11 as a black solid (1.9 mg,
39% yield).
Tr ea tm en t of (Cp *Rh )2(p in ) (11) w ith MeOH. In the
glovebox, a glass bomb was charged with a stir bar, 11 (3.1
mg, 5.2 µmol), and MeOH (1 mL). The maroon solution was
stirred at room temperature for 3 days, during which time a
slight darkening was observed. The solution was then reduced
to dryness under vacuum, and C6D6 was added. The products
1
were identified by H NMR spectroscopy, which showed that
11 had been completelyconsumed and converted to(Cp*RhOMe)2
(12) and pinacol, both in 60% yield. The product mixture was
not clean, and the products were not isolated. Partial 1H NMR
(C6D6): δ 3.56 (s, 6H), 1.66 (s, 30H), 1.03 (s, 12H, pinacol
C-CH3) ppm. (Lit.21 1H NMR (C6D6): δ 3.54 (s, 6H), 1.66 (s,
30H) ppm.)
(Cp *Rh )3C6H6 (10) a n d (Cp *Rh )2C6H6 (9). In the glove-
box, a glass bomb was charged with 5 (150 mg, 0.423 mmol)
and C6H6 (200 mL, 2 mM). The sample was irradiated for 16
h and reduced to dryness under vacuum to yield a brown solid.
This was recrystallized from diethyl ether at -30 °C, and
compound 10 was isolated as black, poorly formed crystals
(30.1 mg, 27% yield). Complex 9 could not be cleanly separated
from residual 10; therefore, the spectroscopic assignments for
9 that follow are based upon spectra which included 10 as a
∼20% impurity.
(d ) Ea r ly Con ver sion w ith Am bien t Ligh t. In the
glovebox, 5 (2.5 mg, 6.8 µmol) was dissolved in C6D6, and the
solution was transferred to a J . Young NMR tube. The tube
1
was exposed to ambient light for 2 h, and a H NMR spectrum
was acquired which showed <5% conversion of 5 to only
products 11 (60%), 9 (40%), and acetone. The tube was exposed
to ambient light for 2 more days, after which 1H NMR
spectroscopy showed complete conversion of 5 to predomi-
nantly 11 and 10.
1
(Cp *Rh )3C6H6 (10). H NMR (C6D6): δ 7.64 (m, 2H, aryl),
P h otolysis of Cp *Rh (p in ) (5) in C6D12. In the glovebox,
5 (2.4 mg, 6.8 µmol) was dissolved in C6D12 and the purple
solution was transferred to a J . Young NMR tube. The tube
was irradiated for 40 min to yield a maroon solution, and 1H
NMR spectroscopy indicated complete conversion to the prod-
ucts. The solution was then reduced to dryness under vacuum,
and C6D6 was added. Subsequent 1H NMR spectroscopy
4.36 (m, 2H, aryl), 2.23 (m, 2H, aryl), 2.06 (s, 15H, Cp*), 1.78
(s, 30H, Cp*) ppm. 13C{1H} NMR (C6D6): δ 140.74 (m,43 aryl),
97.61 (d, J Rh-C ) 4.2 Hz, C5Me5), 97.40 (d, J Rh-C ) 4.3 Hz,
C5Me5), 75.90 (d, J Rh-C ) 5.7 Hz, aryl), 71.48 (d, J Rh-C ) 17.0
Hz, aryl), 12.30 (s, C5Me5), 11.17 (s, C5Me5) ppm. IR (KBr):
2904 (s), 1376 (s), 1021 (m), 718.2 (m) cm-1. EI-MS: m/z 792
([(Cp*Rh)3(C6H6)]+), 373 ([Cp*2Rh]+). Anal. Calcd for C36H51
Rh3: C, 54.56; H, 6.49. Found: C, 54.95; H, 6.39.
-
1
confirmed the product to be 11. H NMR (C6D12): δ 1.96 (s, 2
× 6H, acetone), 1.75 (s, 30H), 1.21 (s, 12H) ppm. 1H NMR
(Cp *Rh )2C6H6 (9). 1H NMR (C6D6): δ 8.73 (m, 2H, aryl),
(C6D6): δ 1.75 (s, 30H), 1.51 (s, 12H) ppm.
5.81 (m, 2H, aryl), 4.58 (m, 2H, aryl), 1.75 (s, 30H, Cp*) ppm.
P h otolysis of Cp *Rh (p in ) (5) w ith 2,3-Dim eth ylbu ta -
d ien e. In the glovebox, 5 (5.2 mg, 14.6 µmol) and 2,3-
dimethylbutadiene (10 µL, 120 µmol) were dissolved in C6D6
and the solution was transferred to a J . Young NMR tube. The
(43) This multiplet has a complex pattern that indicates coupling
to multiple rhodium nuclei but is otherwise unhelpful in assigning the
structure.

Group 9 Metal Pinacolate Complexes
Organometallics, Vol. 20, No. 11, 2001 2261
tube was irradiated for 40 min, and 1H NMR spectroscopy
indicated complete conversion to acetone and the brown diene
adduct 13. The diene adduct was subsequently isolated by
recrystallization from pentane as red-brown crystals (3.1 mg,
66% yield). 1H NMR (C6D6): δ 1.83 (s, 15H), 1.75 (s, 6H), 1.74
(br, 2H), 0.53 (br, 2H) ppm. 1H NMR (CD2Cl2): δ 1.83 (s, 15H),
1.72 (s, 6H), 1.51 (br, 2H), 0.06 (br, 2H) ppm. (Lit.44 1H NMR
(CD2Cl2): δ 1.83 (s, 15H), 1.72 (s, 6H), 1.51 (br, 2H), 0.06 (br,
2H) ppm.)
Tr ea t m en t of (Cp *R h )2(p in ) (11) w it h P in a col a n d
Oxygen . In the glovebox, 11 (1.3 mg, 2.2 µmol) and pinacol
(4.4 mg, 37 µmol) were dissolved in C6D6 to yield a maroon
solution which was transferred to a J . Young tube. The tube
was then freeze-pump-thawed twice on the vacuum line and
pressurized with O2 (223 Torr). The reaction was monitored
by 1H NMR spectroscopy. After 36 h, all of the starting
material was gone and the product 5 was evident in 72% yield.
Only a small amount of pinacol (approximately 0.5 equiv for
every 1 equiv of 5 produced) was consumed.
Tr ea tm en t of (Cp *Rh )2(p in ) (11) w ith P in a col a n d ter t-
Bu tyl Hyd r op er oxid e. In the glovebox, 11 (1.5 mg, 2.5 µmol)
and pinacol (3.6 mg, 31 µmol) were dissolved in C6D6 to yield
a maroon solution which was transferred to a J . Young tube.
tert-Butyl hydroperoxide (0.40 µL, 2.1 µmol) was later added
as a C6D6/nonane solution, and the tube was shaken and left
at room temperature. The reaction was monitored by 1H NMR
spectroscopy and found to be complete within 10 min. A total
of 70% of the starting material was consumed, and 5 was
produced in 70% yield based on the amount of 11 consumed.
tert-Butyl alcohol (δ 1.05 (s, 9H) ppm) was also produced. Only
a small amount of pinacol (approximately 0.5 equiv for every
1 equiv of 5 produced) was consumed.
in the pinacol methyl resonance (δ 1.03 (s, 12H) ppm). The
values thus obtained were always in good agreement with the
amount of acetone (δ 1.54 (s, 6H) ppm) produced. Results are
reported in Table 2, and turnover values for experiments in
which 11 was used as a catalyst are corrected for the fact that
it includes two rhodium centers.
Ca ta lytic Con ver sion of P in a col to Aceton e Usin g
Nitr ou s Oxid e. In the glovebox, the catalyst (3.5 mg, 10 µmol)
and pinacol (24.0 mg, 203 µmol) were dissolved in C6D6 (0.25
mL) and the solution was transferred to a J . Young tube. The
tube was then moved to the vacuum line where the solution
was twice freeze-pump-thawed. The line was then charged
with 761 Torr of nitrous oxide, and the NMR tube (at room
temperature) was opened to the line and then closed. The tube
was shaken vigorously for several seconds and irradiated for
12 h with additional shaking every 2 h. After this time, 1H
NMR spectroscopy showed that >95% of the starting pinacol
(δ 1.03 (s, 12H) ppm) had been consumed relative to internal
standard, and acetone (δ 1.54 (s, 6H) ppm) accounting for 85%
of the pinacol had been produced.
Tr ea tm en t of Cp *Rh (p in ) (5) w ith Ca tech ol. In the
glovebox, 5 (11.8 mg, 33.3 µmol) was dissolved in THF-d8. No
internal standard was used. To the purple solution was added
catechol (3.7 mg, 34 µmol), resulting in an immediate color
change to intense blue. The reaction was monitored by 1H
NMR spectroscopy, and product was found to have formed in
quantitative yield. The solution was then reduced to dryness
under vacuum and recrystallization from ether at -30 °C
1
afforded product 14 as dark crystals (8.0 mg, 70% yield). H
NMR (THF-d8): δ 6.50 (m, 2H), 6.71 (m, 2H), 1.92 (s, 15H),
ppm. 1H NMR (CDCl3): δ 6.82 (m, 4H), 1.95 (s, 15H), ppm.
(Lit.27 1H NMR (CDCl3): δ 6.80 (m, 4H), 1.95 (s, 15H) ppm.)
Ca t a lyt ic Con ver sion of P in a col t o Acet on e Usin g
Oxygen . Specific quantities for this reaction are given in Table
2. In the glovebox, the catalyst (see Table 2) and pinacol were
dissolved in 300 µL of the appropriate solvent, and the solution
was transferred to a J . Young tube. The tube was then moved
to the vacuum line where the solution was twice freeze-
pump-thawed. The line was then charged with oxygen, the
NMR tube (at room temperature) was opened to the line, and
the pressure in the system was noted. The tube was then
closed, shaken vigorously for several seconds, and irradiated
for 16-20 h. After this time, the number of turnovers was
determined using 1H NMR spectroscopy by noting the change
Ack n ow led gm en t. We are grateful for support of
this research from the National Science Foundation
(Grant No. CHE-0094349). The Center for New Direc-
tions in Organic Synthesis is supported by Bristol-Myers
Squibb as Sponsoring Member.
Su p p or tin g In for m a tion Ava ila ble: Data collection and
refinement details and listings of atomic coordinates, thermal
parameters, and bond lengths and bond angles of all three
crystallographically characterized complexes. This material is
available free of charge via the Internet at http://pubs.acs.org.
(44) Lee, H.-B.; Maitlis, P. M. J . Chem. Soc., Dalton Trans. 1975,
2315.
OM0010870