Solvent-mediated generation of cobalt-cluster-stabilised propargyl cations and radicals: Allyl migration versus peroxide formation

Article abstract of DOI:10.1002/chem.200801402

The protonation of [Co₂(CO)₆,[9-[(allyldimethylsilyl) ethynyl]-9H-fluoren-9-ol}] (4), with HBF₄ in CH₂Cl ₂ led to migration of the allyl group from silicon to the cobalt-stabilised cationic site to furnish [Co₂(CO) ₆{9-allyl-9-[(dimethylfluorosilyl)ethynyl]-9H-fluorene}] (17). However, under the same conditions, [Co₂(CO)₆{9- [(benzyldimethylsilyl)ethynyl]-9H-fluoren-9-ol}] (5) underwent desilylation and rearrangement of the resulting terminal alkyne-dicobalt complex to give [Co ₃(CO),(9H-fluorenylmethylcarbynyl)] (24); moreover, dimerisation of the (benzyldimethylsilyl)ethynyl-9H-fluorenyl moiety led to the propargyl-allene 26. In contrast, protonation of 5 in THF yielded [(Co₂(CO) ₆([(benzyldimethylsilyl)ethynyl-9H-fluorenyl])}₂peroxide] (27) through a radical coupling process. Analogously, protonation of [Co ₂-(CO)₆{9-[(vinyldimethylsilyl)ethynyl]-9H-fluoren-9-ol}] (6) yields the corresponding peroxide 28. X-ray crystallographic data are reported for, among others, complexes 17, 24, 26, 27 and 28.

Full text of DOI:10.1002/chem.200801402

DOI: 10.1002/chem.200801402
Solvent-Mediated Generation of Cobalt-Cluster-Stabilised Propargyl Cations
and Radicals: Allyl Migration versus Peroxide Formation
John H. Kaldis,[b] Michael A. Brook,[b] and Michael J. McGlinchey*[a]
Abstract:
[Co2(CO)6{9-[(allyldimethylsilyl)eth
yl]-9H-fluoren-9-ol}] (4), with HBF4 in
CH2Cl2 led to migration of the allyl
group from silicon to the cobalt-stabi-
The
protonation
of
the resulting terminal alkyne–dicobalt
complex to give [Co3(CO)9(9H-fluor-
enylmethylcarbynyl)] (24); moreover,
dimerisation of the (benzyldimethylsi-
lyl)ethynyl-9H-fluorenyl moiety led to
the propargyl–allene 26. In contrast,
[{Co2(CO)6{[(benzyl
ynyl-9H-fluorenyl])}2peroxide]
through radical coupling process.
Analogously, protonation of [Co2-
(CO)6{9-[(vinyldimethylsilyl)ethynyl]-
9H-fluoren-9-ol}] (6) yields the corre-
sponding peroxide 28. X-ray crystallo-
graphic data are reported for, among
others, complexes 17, 24, 26, 27 and 28.
G
ACHTREUNG
N
ACHTREUNG
A
a
A
G
ACHTREUNG
lised
[Co2(CO)6{9-allyl-9-[(dimethylfluoro-
silyl)ethynyl]-9H-fluorene}] (17). How-
cationic
site
to
furnish
protonation of
5 in THF yielded
A
N
ACHTREUNG
ever, under the same conditions,
[Co2(CO)6{9-[(benzyldimethylsilyl)-
Keywords: allylsilanes
silanes · cluster compounds · cobalt ·
peroxides
·
benzyl-
AHCTREUNG
A
went desilylation and rearrangement of
Introduction
The ability of alkyne–dicobalt clusters to stabilise an a-cat-
ionic site is the basis of the Nicholas reaction[1]—a syntheti-
cally valuable procedure used, for example, in the prepara-
tion of enediynes,[2a] (+)-begamide E,[2b] blastinomycine,[2c]
cyclocolorenone,[2d] tetrapyrroles,[2e] insect pheromones,[2f] vi-
tamin A derivatives,[2g] carotenes,[2h] pyrones[2i] and benzofur-
ans.[2i] This stabilisation can be attributed to partial delocali-
sation of the charge onto a cobalt atom through overlap of a
filled metal d orbital with the vacant p orbital on the elec-
tron-deficient a-carbon atom,[3] and is facilitated by a struc-
tural perturbation such that the carbocationic centre leans
toward a metal vertex as illustrated in Scheme 1. This geo-
metric distortion is evident in a variety of crystallographic
determinations, whereby the a-carbon atom is clearly
bonded to a cobalt,[4] iron,[5] or molybdenum[6] cluster
vertex. Moreover, such stabilisation by a metal allows the
Scheme 1. Cobalt-stabilised propargyl cations.
convenient generation of otherwise inaccessible species such
as the non-classical bornyl or fenchyl cations,[7,8] or the anti-
aromatic indenyl or fluorenyl cations.[9]
This phenomenon, together with the known enhanced sta-
bilisation of b-silyl cations,[10] has been elegantly exploited
by Schreiber in a number of cyclisations,[11] as depicted in
Scheme 2. An additional factor in this process is the bending
of the alkyne upon complexation to a {Co2(CO)6} moiety
À ꢀ À
such that the originally linear C C C C unit now makes
angles of approximately 135–1458 at the central carbon
atoms, thus permitting the formation of cycloalkynes con-
taining as few as six ring carbon atoms.[12] More recently,
Green has used this approach to synthesise a series of cyclo-
phanes.[13]
[a] Prof. Dr. M. J. McGlinchey
School of Chemistry and Chemical Biology
University College Dublin, Belfield, Dublin 4 (Ireland)
Fax : (+353)1-716-1178
E-mail: michael.mcglinchey@ucd.ie
Although nucleophilic attack at the a-cationic site has
been extensively exploited for synthetic purposes, studies on
the development of radical character at the a-carbon are of
[b] Dr. J. H. Kaldis, Prof. Dr. M. A. Brook
Department of Chemistry
McMaster University, Hamilton, ON, L8S 4M1 (Canada)
10074
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10074 – 10084

FULL PAPER
went intramolecular nucleophil-
ic attack by the terminus of the
allyl group thus furnishing the
b-silyl cation 9a. Subsequent
formation of a thermodynami-
cally favourable silicon–fluorine
bond yielded 10a, in which the
allyl group has migrated from
Scheme 2. Metal-cluster-mediated cyclisations of allylsilanes.
silicon to carbon, thereby gen-
much more recent vintage.[14] Pioneering studies by Nicholas
and Melikyan initially focussed on the deliberate reduction
of the cations by use of zinc dust,[15] but subsequently it
became evident that in certain solvents, notably diethyl
ether or tetrahydrofuran, electron-transfer processes led di-
rectly to metal-cluster-complexed propargylic radicals with-
out the need to add an extra reducing agent.[16] For example,
erating a quaternary centre.[20]
In contrast, protonation of the cluster 7b (R=Ph) led to
the carbynyl–tricobalt cluster 11, rather than the anticipated
allyl migration product 10b. One might hypothesise that the
initially generated carbocation 12, surrounded as it is by two
phenyl groups and a cobalt cluster, is reluctant to undergo
addition to the allyl group to produce the cation 13, which
would be stabilised only by a hyperconjugative interaction
with the b-silicon. Instead, as
ꢀ À
as depicted in Scheme 3, treatment of [Co2(CO)6(HC C
shown in Scheme 5, desilyla-
tion[21] yields the terminal
alkyne complex 14, which un-
dergoes the known rearrange-
ment to a tricobalt cluster.[22]
The mechanism of rearrange-
Scheme 3. Reactions of cobalt-complexed propargyl cations and radicals.
CPh2OH)] (1) with HBF4 in SO2 yields the isolable cation
+
[Co2(CO)6(HC C CPh2)] (2);[17] however, in THF as the
ꢀ À
solvent, the reaction takes an entirely different course and
the major product is 2,5-bis(diphenylmethylene)-3-cyclopen-
tenone (3).[18] Although the mechanistic details have not yet
been fully elucidated, the formation of 3 is believed to pro-
ꢀ À
C
ceed through dimerisation of [Co2(CO)6(HC C CPh2)] rad-
icals, such that coupling occurs at the less sterically hindered
termini of the complexed alkynes to form a diallene–cobalt
complex, which, after a carbonyl migration and reductive
elimination, ultimately yields the cyclopentenone 3.
Herein we describe the reactions of [Co2(CO)6{9-[(allyldi-
methylsilyl)ethynyl]-9H-fluoren-9-ol}] (4), [Co2(CO)6{9-
[(benzyldimethylsilyl)ethynyl]-9H-fluoren-9-ol}] (5) and
[Co2(CO)6{9-[(vinyldimethylsilyl)ethynyl]-9H-fluoren-9-ol}]
(6) with fluoroboric acid in dichloromethane and in tetrahy-
drofuran, yielding products that clearly involve both cationic
and radical intermediates.
Scheme 4. A metal-cluster-mediated allyl migration.
ment of alkyne–dicobalt hexacarbonyl clusters, [Co2(CO)6-
ꢀ
A
tricobalt systems, [RCH2 CCo3(CO)9], is not yet fully un-
derstood, but a proposal for the initial stages of the process
has been advanced.[23] In the present case, one might have
expected formation of the tertiary alcohol 15, but elimina-
tion of water led directly to the observed product 11, which
has been unequivocally characterised by X-ray crystallogra-
phy.[18]
Results and Discussion
Our involvement in this area arose as the result of an at-
tempt to study the relative ease of generation of metal-clus-
ter-stabilised silylium ions versus carbocations from the bi-
functional starting material 7. The former were potentially
preparable by elimination of propene from an allyl–silane,[19]
while the latter could be generated through protonation of
an alcohol or an alkene, as shown in Scheme 4. In the case
of 7a (R=Me), the observed outcome of the experiment in-
volved initial formation of the carbocation 8a, which under-
Protonation of [Co2(CO)6{9-[(allyldimethylsilyl)ethynyl]-
9H-fluoren-9-ol}] in CH2Cl2: On the premise that allyl mi-
gration to the diphenyl-substituted cationic site in 12 is
thwarted by its enhanced stability, we chose instead to link
the two phenyls so as to incorporate a fluorenyl skeleton, as
in complex 4. The derived cation (16) not only presents a
Chem. Eur. J. 2008, 14, 10074 – 10084
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10075

M. J. McGlinchey et al.
and the di-lithio derivative of 9-
ethynylfluoren-9-ol, also yield-
ed several minor by-products.
These materials were not readi-
ly separable as free ligands but
were obtained as cobalt clus-
ters, 19, 20 and 21, during the
subsequent
chromatographic
purification procedure.
The bis-silylated complex 19
evidently arises in the course of
the reaction of the dilithio salt
Scheme 5. Rearrangement of 7b to form a tricobalt cluster.
flat, less sterically demanding target for the incoming allyl
group, but, as a 12p antiaromatic system, is also strongly
electronically disfavoured unless the charge deficiency is al-
leviated by a neighbouring cobalt vertex. Gratifyingly, in
this case, protonation results in efficient transfer of the allyl
group from silicon to the metal-stabilised carbocationic
centre of the anti-aromatic fluorenyl framework
(Scheme 6); the X-ray crystal structure of the rearranged
product [Co2(CO)6{9-allyl-9-[(dimethylfluorosilyl)ethynyl]-
9H-fluorene}] (17) is shown in Figure 1.
of 9-ethynylfluoren-9-ol with two molecules of allylchlorodi-
methylsilane. The second product, 20, which was character-
ised X-ray crystallographically (see Figure 2), presumably
results from hydration of the allylsilane 18. Although b-hy-
droxysilanes are normally rather susceptible to dehydra-
tion,[25] the excellent crystallinity of 20 appears to be en-
hanced by a combination of strong inter- and intramolecular
hydrogen-bonding interactions, evident within the solid-state
structure as illustrated in Figure 3. Two crystallographically
independent molecules are found in the unit cell, for which
Scheme 6. Cobalt-mediated allyl migration from silicon to carbon.
Figure 1. Mercury[24] representation of the molecular structure of complex
17, (30% thermal ellipsoids; aromatic hydrogen atoms have been exclud-
ed for clarity.)
Interestingly, the synthesis of 9-[(allyldimethylsilyl)-
Figure 2. Molecular structure of the dicobalt cluster diol, 20, (30% ther-
A
mal ellipsoids; hydrogen atoms have been excluded for clarity.)
10076
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10074 – 10084

Cluster Compounds
FULL PAPER
Figure 4. Mercury[24] representation of the molecular structure of 21, ex-
cluding hydrogen atoms, with 30% thermal probability ellipsoids. The in-
tramolecular hydrogen bond contact O9···O16 is 2.790 Š.
Figure 3. Inter- and intramolecular hydrogen-bonding interactions in 20;
cobalt carbonyl groups have been excluded for clarity.
that the three-carbon unit originated on the allyl substituent
attached to the silane. Nevertheless, it is not immediately
evident how the allyl functionality could have migrated to
the alkynyl terminus with concomitant cleavage of the dime-
thylsilyl fragment, but mechanistic investigations are ongo-
ing.
the four hydroxy groups adopt a slightly distorted square
planar arrangement (see Table 1), and deviate merely 0.2 Š
from planarity.[26,27] The hydrogen bonding within this spe-
Table 1. Relevant distances and angles within the hydrogen bonding
regime of 20.
Protonation of [Co2(CO)6{9-[(benzyldimethylsilyl)ethynyl]-
9H-fluoren-9-ol}] in CH2Cl2: With the aim of broadening
the scope of this intramolecular rearrangement, we chose to
prepare the analogous (benzyldimethylsilyl)ethynyldicobalt
complex 5, which contains an allyl subsystem. The objective
was to determine whether the thermodynamic driving force
provided by the formation of a silicon–fluorine bond was
great enough to induce a benzyl migration. For this to
happen, the aromatic character of the phenyl ring would
have to be lost, albeit temporarily, as illustrated in
Scheme 7. Thus, the initial result of rearrangement would be
the energetically disfavoured methylene–cyclohexadiene 22,
which should rapidly re-aromatise to generate an ortho-tolyl
substituent, as in 23.
À
À
161(4)
154
170(7)
166(10)
D H···A
D···H [Š]
H···A [Š]
D···A [Š]
D H···A [8]
À
O9 H···O19
0.87(4)
0.84
0.72(5)
0.63(7)
1.90(4)
1.91
2.03(5)
2.22(7)
2.732(6)
2.691(7)
2.740(8)
2.835(8)
À
O9A H···O19A
À
O19 H···O9A
À
O19A H···O9
cies undoubtedly contributes to its stability and permits its
isolation, whereas the corresponding hydroxylsilane may not
be favoured in the analogous diphenyl system, because of
rotational interference engendered by the aryl rings.
With respect to the other unexpected product (21), our in-
itial thought was that it may have arisen through reaction of
the fluorenyl-alkynyl anion, generated in situ, with traces of
acetone left from cleaning the
glassware prior to conducting
the experiment. However, repe-
tition of this reaction, taking
meticulous care to avoid any
possibility of contamination by
acetone, still afforded small
quantities of the product, which
was unambiguously identified
X-ray
crystallographically
(Figure 4). As with the hydroxy-
silane complex 20, hydrogen
bonding is a prominent feature
of the crystal structure; howev-
er, unlike 20, it is only manifest-
ed
intramolecularly.
One
cannot overlook the possibility Scheme 7. Products from the protonation of complex 5 with HBF4 in CH2Cl2.
Chem. Eur. J. 2008, 14, 10074 – 10084
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10077

M. J. McGlinchey et al.
However, 5 does not undergo this silicon–carbon migra-
tion process; instead, when protonated with HBF4 in di-
chloromethane at À788C, it furnished the (9-fluorenyl)me-
thylcarbynyl–tricobalt cluster 24, the molecular structure of
which is shown in Figure 5. Apparently, in the absence of an
cluster, presumably 25, that exhibited 13C NMR resonances
at 206 ppm and 200 ppm—characteristic of an allene linkage
and a cobalt carbonyl, respectively. However, it was rather
unstable, and repeated attempts to recrystallise the crude
material, led to loss of cobalt and isolation of the free ligand
that was characterised by X-ray crystallography (Figure 6)
as the unsymmetrical propargylallene dimer, 26, formed by
coupling of the alkynylfluorenyl radical with its fluorenyl-
A
thane is not normally considered to be an electron-donating
solvent, it is possible that radical formation was initiated by
single-electron transfer from a cobalt radical fragment gen-
erated during the final step in the formation of the trinu-
clear complex 24.[23] In the crystalline state, the propargyl–
allene dimer exhibits a 50:50 disorder, such that the mod-
elled overlapped molecular orientations (Figure 7) generate
pseudo Ci symmetry. This same centrosymmetric disorder
phenomenon is also found in the corresponding bis(trime-
thylsilyl)propargylallene.[32] Interestingly, a Cope rearrange-
ment of the hexa-1,2-diene-5-yne 26 would regenerate itself,
presumably through a twofold symmetric transition state.
Figure 5. Molecular structure of complex 24. Thermal parameters were
established only for the cobalt atoms and are represented as 30% proba-
bility ellipsoids.[30]
Protonation of [Co2(CO)6{9-[(benzyldimethylsilyl)ethynyl]-
9H-fluoren-9-ol}] in THF: As noted above, in contrast to
the well-established chemistry of [Co2(CO)6(propargyl alco-
hol)] complexes, which, upon protonation in dichlorome-
thane, yield metal-stabilised carbocations, it is now evident
that changing to an ether-based solvent leads to the genera-
tion of the corresponding propargyl radicals. To compare
the results of protonations in dichloromethane with ones
carried out in a solvent known to engender radical products,
the benzylsilane-dicobalt complex 5 was treated with fluo-
energetically viable migration pathway, the system suffers
desilylation to yield a terminal alkyne complex that subse-
quently undergoes the known rearrangement from
[22]
ꢀ
À
[Co2(CO)6
(RC CH)] to [Co3(CO)9
U
Intrigu-
ingly, the tricobalt cluster formed (see Scheme 7) does not
correspond to the unsaturated system 11, obtained from
À ꢀ
[Co2(CO)6{Ph2C(OH) C CH}] and HBF4 in dichlorome-
G
1
thane,[18] but instead follows the pattern observed by Hagi-
roboric acid at À788C in THF. The H and 13C NMR spectra
hara,[28] whereby the reaction of
À ꢀ
[Co2(CO)6{Ph2C(OH) C CH}]
A
with concentrated sulfuric acid
forms the saturated molecule
[29]
À
À
[Co3(CO)9{Ph2CH CH2 C}].
A possible contributory factor
is the enhanced tendency of the
anti-aromatic fluorenyl cation,
even when complexed to the
metal, to undergo reduction to
the corresponding radical (and
thereby initiate hydrogen ab-
straction) relative to that of the
much more stable diphenyl-
methyl cation. It is also worth
noting that the reaction of
[Co2(CO)6{9-[(vinyldimethylsi-
lyl)ethynyl]-9H-fluoren-9-ol}]
(6) with HBF4 in dichlorome-
thane likewise yields the (9-flu-
orenyl)methylcarbynyl–tricobalt
cluster 24.
Figure 6. Molecular structure of the alkynylallene 26 (30% thermal probability ellipsoids; hydrogen atoms
À
À
À
omitted for clarity). Selected distances (Š) and angles (8): C9 C14 1.310(15), C14 C15 1.287(9), C15 C33
The protonation of
5
in
À
À
À
1.587(18), C33 C38 1.473(12), C38 C39 1.213(10), C39 Si2 1.837(7), C9-C14-C15 177.3(10), C33-C38-C39
a second 177.4(11), C38-C39-Si2 176.7(6).
CH2Cl2 also yields
10078
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10074 – 10084

Cluster Compounds
FULL PAPER
an “open-book” type geometry, the cluster-substituted per-
oxide 27 adopts a centrosymmetric (Ci) structure with a cen-
À
tral O9 O9A distance of 1.494(3) Š and C9-O9-O9A angle
of 104.5(2)8. In this respect it resembles the very bulky
bis(9-ortho-isopropylphenyl-9-fluorenyl)peroxide, which also
À
adopts a centrosymmetric structure with a central O O dis-
tance of 1.496 Š, a C-O-O angle of 1068, and parallel fluo-
renyl planes.[33] However, as illustrated in Figure 9, the addi-
tional bulk of the dicobalt hexacarbonyl clusters in 27 re-
sults in the fluorenyl ligands being distorted such that each
six-membered ring deviates from planarity by approximately
78.
Figure 7. View showing the extensive overlap of both partially occupied
models of 26.
of the sole product (27) revealed resonances very similar to
those of the starting material, except for the absence of the
alcoholic proton signal. In the absence of a distinct molecu-
lar ion in the mass spectrum, the unequivocal identification
of 27 as either an ether or a peroxide could not be made,
since the first major bond cleavage would be that of the
oxygen–oxygen linkage, and the fragmentation patterns
would not be expected to differ significantly. However, iso-
lation of single crystals of 27 revealed the product to be the
peroxide (see Scheme 8), the molecular structure of which is
shown in Figure 8. Unlike hydrogen peroxide which exhibits
Figure 9. Perpendicular view of the benzylsilyl peroxide 27 without hy-
drogen atoms and carbonyl groups, emphasising the curved nature of the
fluorenyl ligands.
The vinylsilanedicobalt complex 6 was protonated under
identical conditions, and also afforded the analogous perox-
ide 28 as the only isolable product. Although only weakly
diffracting crystals were obtained for 28, the data were suffi-
ciently good to establish the atom connectivity of the mole-
cule, as shown in Figure 10. In this case, the fluorenyl li-
gands deviate merely 3.58 from planarity, somewhat less
than was seen in the benzyl analogue, possibly attributable
Scheme 8. Formation of peroxides 27 and 28 upon protonation of 5 and
6, respectively, in THF.
Figure 8. Mercury[24] representation of the molecular structure of the ben-
zylsilyl peroxide 27 (30% thermal probability ellipsoids; hydrogen atoms
excluded for clarity).
Figure 10. Mercury[24] representation of the molecular structure of the
vinyl-silyl peroxide 28; aromatic hydrogen atoms have been excluded for
clarity.
Chem. Eur. J. 2008, 14, 10074 – 10084
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10079

M. J. McGlinchey et al.
Syntheses of 4, 17, 18, 19, 20 and 21: 9-(Trimethylsilyl)ethynyl-9H-fluo-
ren-9-ol[32] (2.45 g, 8.8 mmol) was treated with two equivalents of nBuLi
(11.0 mL of a 1.6m hexanes solution, 17.6 mmol) at À788C and left to
react for 4 h at low temperature. A solution of allyldimethylchlorosilane
(1.1 mL, 8.8 mmol) in anhydrous diethyl ether (20 mL) was then added
dropwise and the solution allowed to warm gradually to room tempera-
ture. The mixture was quenched with aqueous NaHCO3, and the product
was extracted with diethyl ether, dried using MgSO4, and filtered to fur-
nish 9-[(allyldimethylsilyl)ethynyl]-9H-fluoren-9-ol (18) as a yellow oil
that was subjected to chromatography on a silica gel column with a 50:50
mixture of CH2Cl2/hexanes; the isolated yield was 74%. Subsequent
treatment of the alkynol 18 with an equimolar quantity of [Co2(CO)8] in
THF at room temperature for 12 h, removal of solvent and chromato-
graphic separation with hexanes gave complex 4 as a dark red oily solid
(86%). In the preparation of 18, several other oily products were also
formed, but were not separable at this stage; however, treatment of the
mixture with dicobalt octacarbonyl yielded complexes 19, 20 and 21 after
chromatographic separation. X-ray diffraction quality crystals of 17, 20
and 21 were obtained by recrystallisation from CH2Cl2/hexanes.
to a reduction in packing influences afforded by the smaller
vinyl substituent. Repetition of both protonation reactions
in diethyl ether likewise yielded the peroxides 27 and 28 as
the major and only identifiable products.
Conclusion
These observations confirm and extend earlier reports, nota-
bly by Melikyan,[15,16] that protonations of cobalt-complexed
propargylic alcohols in either THF or diethyl ether yield
radical products. The generation of peroxides is a character-
istic of reasonably long-lived radicals that react with dioxy-
C
gen to produce R-O-O intermediates. Although these are,
to our knowledge, the first metal-complexed fluorenyl per-
oxides to have been synthesised, there are several known
metal-free difluorenyl peroxides, including phenyl-, isopro-
pyl- and benzyl- substituted systems.[33,34] It should be em-
phasised that organometallic peroxides are relatively un-
common: not only are preformed peroxide ligands insuffi-
ciently thermally stable to survive the conditions required
for complexation by a metal, but also low-valent transition
metals generally bring about cleavage of the peroxy linkage.
Previously known examples include numerous mercuric per-
oxides[35] and transition-metal-containing porphyrins.[36] The
very first organometallic peroxide was prepared by passing
oxygen gas through a solution of cobaltocene in diethyl
ether at À788C.[37] Since then, several ferrocenyl systems[38]
and also a rhodocene have been reported;[39] however, these
molecules were prepared either by coordination of dioxygen
to cationic complexes of the metal in the absence of any
known radical initiating solvents, or by nucleophilic addition
of the metal complex to a preformed peroxide.
As far as we are aware, the molecules described herein
are the first examples of metal-cluster-based peroxides.
Moreover, the technique of initiating the formation of an or-
ganometallic radical by use of tetrahydrofuran as solvent,
with subsequent exposure to atmospheric oxygen, provides a
novel and efficient synthetic methodology, which has the po-
tential to be applied to the preparation of a wide range of
transition metal-containing peroxides.
Protonation of 4 in CH2Cl2: Treatment of 4 (600 mg, 1.02 mmol) with an
equimolar amount of HBF4 in dichloromethane at À788C yielded, after
hydrolysis and column chromatography, complex 17 in 60% yield. X-ray
diffraction quality crystals of 17 were obtained by recrystallisation from
CH2Cl2/hexanes.
Data for 18: 1H NMR (200 MHz, CDCl3): d=7.97–7.25 (m, 1H; fluoren-
yl), 5.79 (m, 1H; CH=C), 4.90 (m, 2H; C=CH2), 2.79 (s, 1H; OH), 1.64
(d, J=7.3 Hz, 2H; CH2), 0.17 ppm (s, 6H; Si
(CH3)2); 13C NMR
T
(50 MHz): d=146.9, 139.1 (Ar Cꢁs), 133.7 (CH=CH2), 129.5, 128.4, 124.3,
120.1 (Ar Cꢁs), 113.9 (CH=CH2), 105.8 (CCSi), 86.6 (CCSi), 74.9 (C-OH),
23.7 (CH2), À2.5 ppm (Si
C
Data for 4: 1H NMR (200 MHz, CDCl3): d=7.60–7.53 (m, 4H; fluoren-
yl), 7.35–7.24 (m, 4H; fluorenyl), 5.65 (m, 1H; CH=C), 4.84 (m, 2H; C=
CH2), 2.57 (s, 1H; OH), 1.46 (d, J=7.2 Hz, 2H; CH2), 0.00 ppm (s, 6H;
Si
134.2 (CH=CH2), 129.5, 127.6, 124.3, 120.2 (Ar Cꢁs), 117.8 (CCSi), 114.0
(CH3)2); MS (DEI):
(CH3)2); 13C NMR (50 MHz): d=199.9 (COꢁs), 150.3, 138.8 (Ar Cꢁs),
(CH=CH2), 83.3 (C-OH), 24.6 (CH2), À1.7 ppm (Si
m/z (%): 534 (21) [MÀ2CO]+, 506 (10) [MÀ3CO]+, 478 (43)
[MÀ4CO]+, 450 (42) [MÀ5CO]+, 422 (100) [MÀ6CO]+, 380 (30)
[M+HÀC3H5À6CO]+; MS (CI, NH3): m/z (%): 573 (100) [MÀOH]+,
545 (60) [MÀOHÀCO]+, 517 (48) [MÀOHÀ2CO]+, 489 (13)
[MÀOHÀ3CO]+, 478 (7) [MÀ4CO]+, 450 (6) [MÀ5CO]+, 422 (7)
[MÀ6CO]+, 304 (10) [MÀCo2(CO)6]+, 287 (13) [MÀOHÀCo2(CO)6]+,
181 (12) [C13H8OH]+; HRMS: m/z calcd for C24H20Co2O5Si
R
Data for 17: Dark red solid, m.p. 968C; H NMR (200 MHz, CDCl3): d=
7.72 (m, 2H; fluorenyl), 7.58 (m, 2H; fluorenyl), 7.42–7.29 (m, 4H; fluo-
renyl), 5.10–4.65 (m, 3H; CH=CH2), 3.17 (d, J=6.6 Hz, 2H; CH2),
0.57 ppm (d, 3JH-F =6.9 Hz, 6H; Si(CH3)2); 13C NMR (50 MHz): d=199.0
H
(COꢁs), 149.9, 140.2 (Ar Cꢁs), 132.9 (CH=CH2), 128.3, 127.3, 124.3, 119.6
(Ar Cꢁs), 118.2 (CH=CH2), 57.8 (C-9), 45.0 (CH2), 0.9 ppm (2JC-F
16.8 Hz, Si
=
Experimental Section
(CH3)2); MS (DEI): m/z (%): 536 (10) [MÀ2CO]+, 508 (18)
[MÀ3CO]+, 452 (26) [MÀ5CO]+, 424 (25) [MÀ6CO]+, 265 (32)
[MÀC3H5ÀCo2(CO)6]+; MS (DCI, NH3): m/z (%): 610 (25) [M+NH4]+,
582 (56) [M+NH4ÀCO]+, 554 (24) [M+NH4À2CO]+, 537 (34)
[M+NH4ÀOHÀ2CO]+, 509 (12) [M+NH4ÀOHÀ3CO]+, 469 (44)
All syntheses were carried out under a dry nitrogen atmosphere, and sol-
vents were dried and distilled according to standard procedures. 9-(Tri-
methylsilyl)ethynyl-9H-fluoren-9-ol was prepared as previously de-
scribed.[32] 1H and 13C NMR spectra were acquired on Bruker DRX-500
or AC-200 spectrometers; assignments were based on standard 2D NMR
[M+NH4ÀHÀ5CO]+,
424
(100)
[MÀ6CO]+,
324
(30)
[MÀC3H5ÀCoÀ6CO]+, 265 (12) [MÀC3H5ÀCo2(CO)6]+; elemental anal-
ysis calcd (%) for C26H19Co2FO6Si: C 52.72 H 3.23; found: C 52.97, H
3.46.
1
techniques (1H-1H COSY, H-13C HSQC and HMBC). Mass spectra were
acquired with a Finnigan EI/CI mass spectrometer system, using direct
electron impact and chemical ionisation methods. Chemical ionisation
was induced using NH3 as the collision gas. High-resolution mass spectra
(HRMS) were obtained using a Micromass GCT, GC Time-of-Flight
mass spectrometer. Silica gel, 230–400 mesh, was used for flash column
chromatography. Elemental analyses are from Guelph Chemical Labora-
tories, Guelph, Ontario, or from the Microanalytical Laboratory at Uni-
versity College Dublin.
Data for 19: Dark red oily solid (55 mg); 1H NMR (200 MHz, CDCl3):
d=7.66–7.53 (m, 4H; fluorenyl), 7.42–7.27 (m, 4H; fluorenyl), 5.63 (m,
2H; CH=C), 4.82 (m, 4H; C=CH2), 1.46 and 1.36 (d, J=7.8 Hz, each
2H; CH2), 0.01 and À0.25 ppm (s, each 6H; Si
C
10080
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10074 – 10084

Cluster Compounds
FULL PAPER
(CCSi), 25.6 and 24.6 (CH2), À1.1 and À1.8 ppm (Si
C
611.9850; found 611.9864; elemental analysis calcd (%) for
C30H22Co2O7Si: C 56.26, H 3.46; found: C 56.08, H 3.70.
[MÀ5CO]+,
520 (100) [MÀ6CO]+, 478 (10) [MÀC3H5ÀHÀ6CO]+; MS ( DCI, NH):
Data for 24: 1H NMR (200 MHz, CDCl3): d=7.69–7.10 (m, 8H; fluoren-
yl), 5.15 (s, 1H; CH), 3.34 ppm (2H; CH2); 13C NMR (50 MHz, CDCl3):
d=199.8 (Co-COꢁs), 146.8, 140.7, 128.6, 126.9, 125.7, 120.1, 58.7,
50.5 ppm; elemental analysis calcd (%) for C24H11Co3O9: C 46.48, H 1.79;
found: C 46.08, H 1.90.
m/z (%): 573 (100) [MÀOSi
C
(C3H5)]+, 548 (60) [MÀ5CO]+, 520
632.0296; found: 632.0284.
Data for 20: Dark red solid (30 mg), m.p. 132–1338C; 1H NMR
(200 MHz, CDCl3): d=7.72–7.65 (m, 4H; fluorenyl), 7.40–7.32 (m, 4H;
fluorenyl), 4.10 (s, 1H; SiOH), 2.63 (s, 1H; OH), 1.36 (br, 2H; CH2),
1
Data for 26: Yellow solid, m.p. 169–1708C; H NMR (200 MHz, CD2Cl2):
d=7.58–7.45 (m, 8H; Ar), 7.14–7.05 (m, 6H; Ar) 6.74–6.71 (m, 4H; Ar),
1.20 (s, 3H; CH3), 0.19 and 0.15 ppm (s, 3H; Si
(CH3)); 13C NMR
1.72 and 1.67 (s, 2H; CH2), À0.40 and À0.43 ppm (s, 6H; Si
C
(50 MHz): d=199.4 (COꢁs), 150.6, 139.2, 129.2, 127.4, 124.2, 120.0 (Ar
Cꢁs), 117.8 (CCSi), 83.4 (CCSi), 66.5 (C-OH), 37.2 28.3, 27.6, 0.9 and
À0.1 ppm (Si
C
26.1 and 25.7 (CH2ꢁs), À2.6 and À3.1 ppm (Si
C
ꢀ
[MÀH]+, 337 (100) [C13H8CCSi
A
U
[C13H8CCSi
(CH3)2CH2Ph]+, 149 (47) [Si(CH3)2CH2Ph]+; HRMS: m/z
calcd for C48H42Si2 ([M]+): 674.2825; found: 674.2836; elemental analysis
calcd (%) for C48H42Si2: C 85.41, H 6.27; found: C 85.29, H 6.60.
Data for 21: Dark red solid (86 mg), m.p. 120–1218C; 1H NMR
(200 MHz, CDCl3): d=7.84–7.71 (m, 4H; fluorenyl), 7.53–7.36 (m, 4H;
fluorenyl), 4.09 and 3.03 (s, 1H; OH), 1.58 ppm (s, 6H; C(CH3)2);
Data for 27: 1H NMR (200 MHz, CDCl3): d=7.58 (d, 3JH-H =7.4 Hz, 4H;
3
Ar), 7.42–7.08 (m, 18H; Ar), 6.85 (d, JH-H =7.0 Hz, 4H; Ar), 1.71 (s, 4H;
CH2ꢁs), À0.26 ppm (s, 12H; Si
(COꢁs), 147.3, 139.7, 139.2, 129.5, 128.3, 128.1, 127.1, 125.1, 124.2, 120.1
(Ar Cꢁs), 26.6 (CH2), À1.9 ppm (Si(CH3)2); MS (ESI, negative, CH3OH):
(CH3)2Ph]À; elemental anal-
A
13C NMR (50 MHz): d=199.1 (COꢁs), 150.1, 138.8, 129.4, 127.6, 124.5,
120.1 (Ar Cꢁs), 106.9 and 102.6 (C=C), 83.5 and 73.4 (C-OH), 32.9 ppm
(CH2); MS (DEI): m/z (%): 522 (4) [MÀCO]+, 494 (3) [MÀ2CO]+, 466
(12) [MÀ3CO]+, 438 (11) [MÀ4CO]+, 410 (8) [MÀ5CO]+, 382 (48)
[MÀ6CO]+, 364 (50) [MÀHÀOHÀ6CO]+, 324 (12) [M+HÀCoÀ6CO]+;
MS (DCI, NH3): m/z (%): 533 (35) [MÀOH]+, 505 (34) [MÀHÀ2OH]+,
449 (12) [MÀOHÀ3CO]+, 438 (32) [MÀ4CO]+, 410 (16) [MÀ5CO]+,
382 (24) [MÀ6CO]+, 364 (16) [MÀHÀOHÀ6CO]+, 247 (13)
[MÀOHÀCo2(CO)6]+, 230 (11) [MÀ2OHÀCo2(CO)6]+, 181 (100)
[C13H8OH]+; elemental analysis calcd (%) for C24H16Co2O8: C 52.39, H
2.93; found: C 52.08, H 2.90.
AHCTREUNG
m/z (%): 639 (35) [MÀC13H8OC2Co2(CO)6Si
ACHTREUNG
ysis calcd (%) for C60H42Co4O14Si2: C 56.35, H 3.31; found: C 56.57, H
2.98.
Syntheses of 6 and 28: As for 18, treatment of 9-(trimethylsilyl)ethynyl-
9H-fluoren-9-ol with nBuLi and vinyldimethylchlorosilane furnished 9-
(vinyldimethylsilyl)ethynyl-9H-fluoren-9-ol as a yellow oil in 74% yield.
1H NMR (200 MHz, CDCl3): d=7.68–7.53 (m, 4H; fluorenyl), 7.36–7.29
(m, 4H; fluorenyl), 6.15–5.76 (m, 3H; CH=CH2), 3.42 (s, 1H; OH), 2.08
(s, 2 H; CH2), 0.19 ppm (s, 6H; Si
(CH3)2); 13C NMR (50 MHz): d=146.9,
U
Syntheses of 5, 24, 26 and 27: As for 18, treatment of 9-(trimethylsilyl)-
139.0 (Ar Cꢁs), 136.1 (CH=C), 133.0 (C=CH2), 129.4, 128.3, 124.2, 119.9,
(Ar Cꢁs), 106.3 (CCSi), 85.6 (CCSi), 74.7 (C-OH), 30.7 (CH2), À1.8 ppm
A
nished 9-(benzyldimethylsilyl)ethynyl-9H-fluoren-9-ol as a white solid,
m.p. 87–888C, in 68% yield. 1H NMR (200 MHz, CDCl3): d=7.74–7.62
(m, 4H; fluorenyl), 7.44–7.39 (m, 4H; fluorenyl), 7.21–7.03 (m, 5H; Ph),
(Si
A
(DCI, NH3): m/z (%): 290 (10) [M]+, 273 (100) [MÀOH]+. Subsequent
treatment of the alkynol with an equimolar quantity of [Co2(CO)8] in
THF at room temperature for 12 h, removal of solvent and chromato-
graphic separation with hexanes gave complex 6 in 71% isolated yield.
2.69 (s, 1H; OH), 2.21 (s, 2H; CH2), 0.17 ppm (s, 6H; Si(CH3)2);
13C NMR (50 MHz): d=146.7, 139.0, 138.7, 129.6, 128.4, 128.3, 128.0,
124.3, 120.1 (Ar Cꢁs), 106.2 (CCSi), 86.5 (CCSi), 74.9 (C-OH), 25.9 (CH2),
Treatment of 6 (500 mg, 0.87 mmol) with HBF4 (0.12 mL, 0.9 mmol) in
THF (or in diethyl ether) afforded the peroxide (28) as a dark red solid
in 60% (or 52%), respectively, after chromatographic separation on a
silica column using a 50:50 dichloromethane/hexanes solution as eluent.
À2.3 ppm (Si
C
Data for 6: Dark red solid, m.p. 71–728C; 1H NMR (200 MHz, CDCl3):
Treatment of 5 (1.0 g, 1.56 mmol) with 1.5 equivalents of HBF4 (0.33 mL,
2.3 mmol) in CH2Cl2 at À788C, and subsequent chromatographic separa-
tion on a silica column, using 1:1 dichloromethane/hexanes as eluent, af-
forded the tricobalt cluster 24 as a dark red solid, and the propargylallene
26 in 32 and 42% yields, respectively.
d=7.63–7.58 (m, 4H; fluorenyl), 7.38–7.26 (m, 4H; fluorenyl), 6.06–5.70
(m, 3H; CH=CH2), 2.58 (s, 1H; OH), 0.16 ppm (s, 6H; Si(CH3)2);
A
13C NMR (50 MHz): d=199.8 (COꢁs), 150.3, 138.9 (Ar Cꢁs), 137.4 (CH=
C), 132.8 (C=CH2), 129.5, 127.6, 124.3, 117.4, (Ar Cꢁs), 118.2 (CCSi), 83.4
(CCSi), 78.2 (COH), À1.3 ppm (Si
C
Treatment of 5 (500 mg, 0.78 mmol) with an equimolar amount of HBF4
(0.11 mL, 0.8 mmol) in THF (or in diethyl ether) afforded the peroxide
27 as a dark red solid in 63% (or 56%), respectively, after chromato-
graphic separation on a silica column using 50:50 dichloromethane/hex-
anes as eluent.
Data for 5: Dark red solid, m.p. 998C; 1H NMR (200 MHz, CDCl3): d=
7.69–7.55 (m, 4H; fluorenyl), 7.44–6.90 (m, 9H; fluorenyl and Ph), 2.59
Data for 28: Dark red solid, m.p. 144–1458C, 1H NMR (200 MHz,
CDCl3): 1H NMR (200 MHz, CDCl3): d=7.59 (d, J=6.9 Hz 4H; Ar),
7.39–7.20 (m, 12H; Ar), 5.82–5.50 (m, 6H; CH=CH2), À0.09 ppm (s,
(s, 1H; OH), 1.90 (s, 2H; CH2), À0.09 ppm (s, 6H; Si
C
12H; Si
(CH3)ꢁs); 13C NMR (50 MHz): d=198.8 (COꢁs), 147.4, 139.7 (Ar
A
124.3, 120.2 (Ar Cꢁs), 83.3 (C-OH), 26.6 (CH2), À1.9 ppm (Si
C
Cꢁs), 137.6 (CH=C), 132.4 (C=CH2), 129.3, 127.0, 125.2, 119.9 (Ar Cꢁs),
À1.5 ppm (Si(CH3)2); MS (ESI, negative, CH3OH): m/z (%): 559 (10)
[C13H8C2Co2(CO)6Si
(CH3)2C3H5]À, 273 (20) [C13H8CCSi(CH3)2C3H5]À; el-
AHCTREUNG
A
ACHTREUNG
[C13H8OH]+, 149 (90) [Si
A
emental analysis calcd (%) for C50H34Co4O14Si2: C 52.19, H 2.98; found:
C 51.97, H 2.90.
3
(100) [C13H8OH]+. HRMS: m/z calcd for C29H22O6SiCo2 ([MÀCO]+):
Chem. Eur. J. 2008, 14, 10074 – 10084
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10081

M. J. McGlinchey et al.
Table 2. Crystallographic Data.
17
20
21
26
27
28
empirical formula
Mr
C26H19Co2FO6Si
592.36
C26H22Co2O8Si
608.39
C24H16Co2O8
550.23
C48H42Si2
675.00
C60H42Co4O14Si2
1278.84
C50H34Co4O14Si2
1150.67
crystal system
space group
a [Š]
b [Š]
c [Š]
monoclinic
P21/n (#14)
8.624(2)
20.573(5)
14.556(4)
90
95.017(4)
90
2572.6(11)
4
monoclinic
P21/c (#14)
20.445(7)
10.233(3)
27.202(9)
90
105.938(5)
90
5472(3)
tetragonal
P421/c (#114)
22.489(5)
22.489(5)
9.846(3)
90
90
90
4980(2)
8
monoclinic
P21/c (#14)
13.467(6)
16.297(7)
8.923(4)
90
95.692(8)
90
1948.7(14)
2
triclinic
P1 (#2)
orthorhombic
Pbca (#62)
22.17(4)
13.24(3)
16.93(3)
90
90
90
4969(17)
4
1.538
¯
10.516(5)
12.0516)
12.382(6)
86.312(8)
76.860(8)
66.034(8)
1395.6(12)
1
a [8]
b [8]
g [8]
V [Š3]
Z
8
1calcd [gcmÀ3
T [K]
]
1.529
173(2)
1.382
1.477
173(2)
1.302
1.468
197(2)
1.376
1.150
173(2)
0.123
1.522
173(2)
1.278
173(2)
1.426
m [mmÀ1
]
F
A
1200
1952
2224
716
650
2328
1.72–27.00
À11ꢁhꢁ11
À23ꢁkꢁ26
À18ꢁlꢁ17
22551
1.04–23.75
À23ꢁhꢁ21
À11ꢁkꢁ11
À30ꢁlꢁ30
35492
1.28–24.99
À26ꢁhꢁ26
À26ꢁkꢁ26
À11ꢁlꢁ11
37626
1.97–25.00
À16ꢁhꢁ15
À19ꢁkꢁ19
À9ꢁlꢁ10
14547
1.69–27.50
À13ꢁhꢁ12
À15ꢁkꢁ14
À16ꢁlꢁ15
12858
1.84–22.50
À23ꢁhꢁ23
À13ꢁkꢁ14
À18ꢁlꢁ17
26564
index ranges
reflections measured
reflections used (Rint
)
5615 (0.0608)
5615/0/402
0.0389/0.0658
0.0718/0.0712
1.010
8346 (0.1443)
8346/243/834
0.0607/0.0762
0.1486/0.0945
0.975
4385 (0.1088)
4385/0/372
0.0394/0.0686
0.0856/0.0835
1.077
3424
(0.0522)
6305 (0.0315)
6305/0/446
0.0323/0.0673
0.0536/0.0759
1.009
3244 (0.1512)
3244/0/323
0.0847/0.1172
0.1315/0.1306
1.129
data/restraints/parameters
final R values [I>2s(I)] R1/wR2
R values (all data) R1/wR2
3424/36/296
0.0651/0.1168
0.1189/0.1388
1.020
goodness-of-fit on F2
À3
largest diff peak/hole [eŠ
]
0.368/À0.278
0.425/À0.377
0.328/À0.311
0.343/À0.305
0.357/À0.313
0.383/À0.344
X-ray measurements for 17, 20, 21, 26, 27 and 28: Crystal data were col-
lected using P4 Bruker diffractometer, equipped with Bruker
Likewise, for molecules 27 and 28 only half of each molecule was refined,
and the other half was generated by the appropriate symmetry element.
a
a
SMART 1 K CCD area detector, using the program SMART,[40] and a ro-
tating anode, using graphite-monochromated MoKa radiation (l=
0.71073 Š), and are listed in Table 2. A full sphere of the reciprocal
space was scanned by phi-omega scans. Data reduction was carried out
by using the program SAINT,[41] which applied Lorentz and polarisation
corrections to the three-dimensionally integrated diffraction spots.
Pseudo-empirical absorption correction based on redundant reflections
was performed by the program SADABS.[42] The structures were solved
by using the direct-methods procedure in the Bruker SHELXTL program
library,[43] and refined by full-matrix least-squares methods on F2 with
anisotropic thermal parameters for all non-hydrogen atoms. Hydrogen
atom treatment varied from compound to compound, depending on the
crystal quality. In 17, 21 and 27, all hydrogen atoms were located in the
difference fourier map and allowed to refine freely with isotropic thermal
displacement parameters. In 26 and 28 hydrogen atoms were added at
calculated positions and refined using a riding model. Molecule 20 crys-
tallised with two crystallographically unrelated molecules in the unit cell,
while the remaining structures contained only one molecule in the asym-
metric unit. A small slip-disorder of one of the hydrogen-bonded fluoren-
yl ligands in 20 containing 73% occupancy of the major fragment, re-
quired a partial restraining of these ligands to optimise the model. The
hydrogen atoms attached to disordered carbon atoms in 20 were added
at calculated positions and refined by using a riding model; the hydrogen
attached to the disordered oxygen was set to ride on the oxygen of the
major component, but all other hydrogen atoms were refined freely. The
carbonyl carbon atoms of 24 could not be refined anisotropically as a
result of the weak intensity of the data, and so the fluorenyl moieties
were restrained to simulate isotropic behaviour to provide a better
model. A partial twist-disorder of the silyl groups in 28, modelled so that
the major component consisted of 57%, did not allow for anisotropic re-
finement of the silicon methyls and vinyl group. As previously men-
tioned, the allene 26 exhibited a centrosymmetric disorder, whereby the
fluorenyl ligands and the silyl substituents of the alkyne and allene in the
two equally occupied molecular orientations, overlap with one another.
CCDC-693776 (17), CCDC-693777 (20), CCDC-693778 (21), CCDC-
693779 (26), CCDC-693780 (27) and CCDC-693781 (28) contain the sup-
plementary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgements
We are grateful to the Natural Sciences and Engineering Research Coun-
cil of Canada (NSERC), Science Foundation Ireland and University Col-
lege Dublin for generous financial support. J.H.K. thanks NSERC and
the Province of Ontario for postgraduate scholarships. Mass spectra were
acquired courtesy of Dr. Kirk Green of the McMaster Regional Mass
Spectrometry Centre. We thank Dr. Helge Müller-Bunz, Director of the
UCD X-ray Diffraction Facility, for crystallographic advice.
[1] a) K. M. Nicholas, Acc. Chem. Res. 1987, 20, 207–214; b) J. E.
OꢁBoyle, K. M. Nicholas, Tetrahedron Lett. 1980, 21, 1595–1598;
c) A. J. M. Caffyn, K. M. Nicholas, in Comprehensive Organometallic
Chemistry II, Vol. 12 (Eds.: G. Wilkinson, F. G. A. Stone, E. W.
Abel) Pergamon, Oxford (UK), 1995, Chapter 7.1, pp. 685–702;
d) M. J. McGlinchey, L. Girard, R. Ruffolo, S. Kainz, H. K. Gupta,
M. A. Brook, M. J. McGlinchey, J. Organomet. Chem. 1997, 547,
217–226, Coord. Chem. Rev. 1995, 143, 331–381; e) H. El Amouri,
M. Gruselle, Chem. Rev. 1996, 96, 1077–1103.
[2] a) P. Magnus, R. T. Lewis, J. C. Huffman, J. Am. Chem. Soc. 1988,
110, 6921–6923; b) C. Mukai, S. M. Moharram, O. Kataoka, M. Ha-
naoka, J. Chem. Soc. Perkin Trans. 1 1995, 2849–2854; c) C. Mukai,
O. Katuoka, M. Hanaoka, J. Org. Chem. 1993, 58, 2946–2952; d) M.
Saha, B. Bagby, K. M. Nicholas, Tetrahedron Lett. 1986, 27, 915–
918; e) P. A. Jacobi, S. Rajeswari, Tetrahedron Lett. 1992, 33, 6231–
10082
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10074 – 10084

Cluster Compounds
FULL PAPER
6234; f) C. A. Henrick, Tetrahedron 1977, 33, 1845–1889; g) F. Kien-
zle, Pure Appl. Chem. 1976, 47, 183–190; h) S. M. Makin, Pure
Appl. Chem. 1976, 47, 173–181; i) M. Soleilhavoup, L. Maurette, C.
Lamirand, B. Donnadieu, M. J. McGlinchey, R. Chauvin, Eur. J.
Org. Chem. 2003, 1652–1660.
[22] a) R. Markby, I. Wender, R. A. Friedel, F. A. Cotton, H. W. Stern-
berg, J. Am. Chem. Soc. 1958, 80, 6529–6533; b) U. Kruerke, W.
Hübel, Chem. Ind. 1960, 1264–1264; c) P. W. Sutton, L. F. Dahl, J.
Am. Chem. Soc. 1967, 89, 261–268.
[23] J. H. Kaldis, P. Morawietz, M. J. McGlinchey, Organometallics 2003,
22, 1293–1301.
[24] Mercury 1.4.2, available from http://www.ccdc.cam.ac.uk/mercury/
[25] P. F. Hudrlik, D. Peterson, J. Am. Chem. Soc. 1975, 97, 1464–1468.
[26] For a comprehensive discussion of square-planar and other hydro-
[3] a) B. E. R. Schilling, R. Hoffmann, J. Am. Chem. Soc. 1978, 100,
6274–6274; b) M. F. DꢁAgostino, C. S. Frampton, M. J. McGlinchey,
J. Organomet. Chem. 1990, 394, 145–166.
[4] G. G. Melikyan, S. Bright, T. Monroe, K. I. Hardcastle, J. Ciurash,
Angew. Chem. 1998, 110, 170–172; Angew. Chem. Int. Ed. 1998, 37,
161–164.
[5] a) D. Osella, G. Dutto, G. Jaouen, A. Vessi›res, P. R. Raithby, L. De
Benedetto, M. J. McGlinchey, Organometallics 1993, 12, 4545–4552;
b) W. Rohde, G. Fendersak, Z. Naturforsch. B 1991, 46, 1169–1176.
[6] L. Girard, P. E. Lock, H. El Amouri, M. J. McGlinchey, J. Organo-
met. Chem. 1994, 478, 189–196, and references therein.
[7] M. Gruselle, H. El Hafa, M. Nikolski, G. Jaouen, J. Vaissermann, L.
Li, M. J. McGlinchey, Organometallics 1993, 12, 4917–4925.
[8] M. Kondratenko, H. El Hafa, M. Gruselle, G. Jaouen, J. Vaisser-
mann, M. J. McGlinchey, J. Am. Chem. Soc. 1995, 117, 6907–6913.
[9] J. A. Dunn, W. J. Hunks, R. Ruffolo, S. S. Rigby, M. A. Brook, M. J.
McGlinchey, Organometallics 1999, 18, 3372–3382.
[10] a) J. B. Lambert, Y. Zhao, J. Am. Chem. Soc. 1996, 118, 7867–7868;
b) H.-U. Siehl, F.-P. Kaufmann, Y. Apeloig, V. Braude, A. Danovich,
N. Berndt, N. Stamatis, Angew. Chem. 1991, 103, 1546–1549;
Angew. Chem. Int. Ed. Engl. 1991, 30, 1479–1482.
[11] a) S. L. Schreiber, T. Sammakia, W. E. Crowe, J. Am. Chem. Soc.
1986, 108, 3128–3130; b) S. L. Schreiber, M. T. Klimas, T. Samma-
kia, J. Am. Chem. Soc. 1987, 109, 5749–5759.
[12] a) P. Magnus, T. Pitterna, J. Chem. Soc. Chem. Commun. 1991, 110,
541–543; b) P. Magnus, R. Carter, M. Davies, J. Elliott, T. Pitterna,
Tetrahedron 1996, 52, 6283–6306; c) M. H. Chisholm, K. Folting,
J. C. Huffman, E. A. Lucas, Organometallics 1991, 10, 535–537.
[13] a) J. DiMartino, J. R. Green, Tetrahedron 2006, 62, 1402–1409; b) R.
Gibe, J. R. Green, G. Davidson, Org. Lett. 2003, 5, 1003–1005; c) R.
Gibe, J. R. Green, Chem. Commun. 2002, 1550–1552; d) J. R.
Green, Curr. Org. Chem. 2001, 5, 809–826; e) J. R. Green, Synlett
2001, 353–356; f) R. Guo, J. R. Green, Chem. Commun. 1999,
2503–2504; g) J. R. Green, Chem. Commun. 1998, 1751–1752.
[14] G. G. Melikyan, G. Vostrowsky, W. Bauer, H. J. Bestmann, M. Khan,
K. M. Nicholas, J. Org. Chem. 1994, 59, 222–229, and references
therein.
gen-bonded packing motifs found in [Co2(CO)6(alkynol)] clusters,
U
see: D. Braga, F. Grepioni, D. Walther, K. Heubach, A. Schmidt, W.
Imhof, H. Gçrls, T. Klettke, Organometallics 1997, 16, 4910–4919.
[27] Other examples, involving hydrogen-bonded ferrocenylalkynols,
have been reported in: a) H. Schottenberger, K. Wurst, M. R. Buch-
meiser, J. Organomet. Chem. 1999, 584, 301–309; b) E. V. Banide, Y.
Ortin, B. Chamiot, A. Cassidy, J. Niehaus, A. Moore, C. M. Seward,
H. Müller-Bunz, M. J. McGlinchey, Organometallics 2008, 27, 4173–
4182.
[28] P.-J. Kim, N. Hagihara, Bull. Chem. Soc. Jpn. 1968, 41, 1184–1187.
[29] One could also envisage that the formation of the saturated cluster
24, rather than the fluorenyl analogue of the alcohol 15, may have
occurred by dehydration, as from 15 to 11, and subsequent hydroge-
nation through reaction of the acid with cobalt-containing decompo-
sition products.
[30] Crystal data for 24: C24H11Co3O9, Mr =620.12, monoclinic, space
group P21/n (#14), a=8.702(6), b=30.56(2), c=26.752(19) Š, b=
98.679(12)8, V=7032(9) Š3, Z=12, 1calcd =1.757 gcmÀ3
2.151 mmÀ1, T=173 K, R1 =0.1098.
,
m=
[31] Head-to-tail propargyl-allene dimers, which are conventionally pre-
pared by copper-mediated coupling of bromoallenes, exhibit photo-
chromism in the solid state: J. L. Scott, K. Tanaka, Cryst. Growth
Des. 2005, 5, 1209–1213, and references therein.
[32] E. V. Banide, B. C. Molloy, Y. Ortin, H. Müller-Bunz, M. J. McGlin-
chey, Eur. J. Org. Chem. 2007, 2611–2622.
[33] P. D. Robinson, Y. Hou, C. Y. Myers, Acta Crystallogr. Sect. C 1999,
55, 9900147.
[34] a) T. Kitagawa, A. Miyabo, H. Fujii, T. Okazaki, T. Mori, M. Matsu-
dou, T. Sugie, K. Takeuchi, J. Org. Chem. 1997, 62, 888–892; b) T.
Pilati, M. Simonetta, S. Quici, Cryst. Struct. Commun. 1982, 11,
1027–1032; c) G. Baum, H. Shechter, J. Org. Chem. 1976, 41, 2120–
2124; d) E. E. Van Tamelen, T. M. Cole Jr. , J. Am. Chem. Soc.
1971, 93, 6158–6166; e) J. I. G. Cadogan, D. H. Hey, W. A. Sander-
son, J. Chem. Soc. 1960, 82, 3203–3210.
[15] a) G. G. Melikyan, R. C. Combs, J. Lamirand, M. Khan, K. M. Nich-
olas, Tetrahedron Lett. 1994, 35, 363–366; b) G. G. Melikyan, A. De-
ravakian, J. Organomet. Chem. 1997, 544, 143–145; c) G. G. Meliky-
an, A. Deravakian, S. Myer, S. Yadegar, K. I. Hardcastle, J. Ciurash,
P. Toure, J. Organomet. Chem. 1999, 578, 68–75; d) G. G. Melikyan,
C. Wild, P. Toure, Organometallics 2008, 27, 1569–1581.
[16] a) G. G. Melikyan, F. Villena, A. Florut, S. Sepanian, H. Sarkissian,
A. Rowe, P. Toure, D. Mehta, N. Christian, S. Myer, D. Miller, S.
Scanlon, M. Porazik, M. Gruselle, Organometallics 2006, 25, 4680–
4690; b) G. G. Melikyan, S. Sepanian, F. Villena, J. Jerome, B.
Ahrens, R. McClain, J. Matchett, S. Scanlon, A. Abrenica, K. Paul-
sen, K. I. Hardcastle, J. Organomet. Chem. 2003, 683, 324–330; c) P.
Toure, S. Myer, G. G. Melikyan, J. Phys. Chem. A 2001, 105, 4579–
4584.
[17] R. E. Connor, K. M. Nicholas, J. Organomet. Chem. 1977, 125, C45–
48.
[18] J. H. Kaldis, M. J. McGlinchey, Tetrahedron Lett. 2002, 43, 4049–
4053.
[19] a) R. Ruffolo, A. Decken, L. Girard, H. K. Gupta, M. A. Brook,
M. J. McGlinchey, Organometallics 1994, 13, 4328–4335; b) R. Ruf-
folo; S. Kainz; H. K. Gupta; M. A. Brook, M. J. McGlinchey, J. Or-
ganomet. Chem. 1997, 547, 217–226.
[20] R. Ruffolo, M. A. Brook, M. J. McGlinchey, Organometallics 1998,
17, 4992–4996.
[35] a) J. R. Nixon, M. A. Cudd, N. A. Porter, J. Org. Chem. 1978, 43,
4048–4052; b) J. Halfpenny, R. W. H. Small, J. Chem. Soc. Chem.
Commun. 1979, 879–880; c) P. H. Dussault, D. R. Davies, Tetrahe-
dron Lett. 1996, 37, 463–466; d) A. J. Bloodworth, K. A. Johnson,
Tetrahedron Lett. 1994, 35, 8057–8060; e) A. J. Bloodworth, R. J.
Curtis, M. Spencer, N. A. Tallant, Tetrahedron 1993, 49, 2729–2750;
f) A. J. Bloodworth, A. Shah, J. Chem. Soc. Chem. Commun. 1991,
947–948.
[36] a) C. O. Dietrich-Buchecker, J. P. Sauvage, J. Weiss, Tetrahedron
Lett. 1986, 27, 2257–2260; b) J. Besancon, S. Top, J. Tirouflet, B.
Gautheron, Y. Dusausoy, J. Organomet. Chem. 1975, 94, 35–46;
c) M. Kohno, Bull. Chem. Soc. Jpn. 1988, 61, 1509–1515; d) D. A.
Buckingham, J. Dekkers, A. M. Sargeson, J. Am. Chem. Soc. 1973,
95, 4173–4179; e) H. Kalish, H. M. Lee, M. M. Olmstead, L. Latos-
Grazynski, S. P. Rath, A. L. Balch, J. Am. Chem. Soc. 2003, 125,
4674–4675; f) G. G. Haraldsson, J. E. Baldwin, J. G. Jones, J. Deber-
nardis, Polyhedron 1993, 12, 2453–2455; g) A. Gold, W. Ivey, G. E.
Toney, R. Sangaiah, Inorg. Chem. 1984, 23, 2932–2935; h) W.
Freyer, S. Flatau, Tetrahedron Lett. 1996, 37, 5083–5086; i) W.
Freyer, H. Stiel, M. Hild, K. Teuchner, D. Leupold, Photochem.
Photobiol. 1997, 66, 596–604.
[37] H. Kojima, S. Takahashi, N. Hagihara, J. Chem. Soc. Chem.
Commun. 1973, 230–231.
[38] a) E. A. Dikusar, V. L. Shirokii, A. P. Yuvchenko, A. V. Bazhanov,
K. L. Moiseichuk, V. N. Khrustalev, M. Yu. Antipin, Russ. J. Gen.
Chem. 1999, 69, 1266–1272; b) S. G. Stepin, O. S. Stepina, E. A. Di-
[21] J. A. Beck, S. A. R. Knox, R. F. D. Stansfield, F. G. A. Stone, M. J.
Winter, P. Woodward, J. Chem. Soc. Dalton Trans. 1982, 195–200.
Chem. Eur. J. 2008, 14, 10074 – 10084
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10083

M. J. McGlinchey et al.
kusar, V. L. Shirokii, Russ. J. Appl. Chem. 2000, 73, 1596–1599;
c) J.-R. Hamon, D. Astruc, Organometallics 1988, 7, 1036–1046.
[39] J. Jeffery, R. J. Mawby, M. B. Hursthouse, N. P. C. Walker, J. Chem.
Soc. Chem. Commun. 1982, 1411–1412.
[42] G. M. Sheldrick, SADABS, Bruker AXS Inc., Madison, WI 53711,
2000.
[43] G. M. Sheldrick, SHELXTL, Version 5.1, Bruker AXS Inc., Madi-
son, WI 53711, 1998.
[40] SMART, Release 5.611, Bruker AXS Inc., Madison, WI 53711,
2000.
Received: July 10, 2008
[41] SAINT, Release 6.02a, Bruker AXS Inc., Madison, WI 53711, 2000.
Published online: September 26, 2008
10084
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10074 – 10084