On the coordination chemistry of corannulene, the smallest "buckybowl"

Article abstract of DOI:10.1016/j.jorganchem.2005.04.044

A variety of methods, conventional and non-conventional, are used in attempts to prepare the compounds (η⁶-corannulene)M(CO) ₃ (M = Cr, Mo, W), all unsuccessful. Conventional methods are also utilized in attempts to prepare the compound [CpFe(η⁶- corannulene)]PF₆, but these result in mixtures of cationic CpFe(arene) complexes containing partially hydrogenated corannulene; similar results have been reported for other polyaromatic hydrocarbons. DFT calculations on the compound (η⁶-corannulene)Cr(CO)₃ suggest that the (η⁶-corannulene)-Cr linkage is only a few kcal/mol weaker than the corresponding bond in (η⁶-benzene)Cr(CO)₃, implying that failures in syntheses arise from kinetic, not thermodynamic problems.

Full text of DOI:10.1016/j.jorganchem.2005.04.044

Journal of Organometallic Chemistry 690 (2005) 3440–3450
www.elsevier.com/locate/jorganchem
On the coordination chemistry of corannulene, the smallest
‘‘buckybowl’’
a
a
a,
b
Michael W. Stoddart , John H. Brownie , Michael C. Baird *, Hartmut L. Schmider
a
Department of Chemistry, Queenꢀs University, 90 Bader Lane, Kingston, Ont., Canada K7L 3N6
b
High Performance Computing Virtual Laboratory, Queenꢀs University, Kingston, Ont., Canada K7L 3N6
Received 3 March 2005; revised 18 April 2005; accepted 18 April 2005
Available online 15 June 2005
Abstract
A variety of methods, conventional and non-conventional, are used in attempts to prepare the compounds (g6-corannu-
lene)M(CO)3 (M = Cr, Mo, W), all unsuccessful. Conventional methods are also utilized in attempts to prepare the compound
[CpFe(g6-corannulene)]PF6, but these result in mixtures of cationic CpFe(arene) complexes containing partially hydrogenated
corannulene; similar results have been reported for other polyaromatic hydrocarbons. DFT calculations on the compound (g6-
corannulene)Cr(CO)3 suggest that the (g6-corannulene)-Cr linkage is only a few kcal/mol weaker than the corresponding bond
in (g6-benzene)Cr(CO)3, implying that failures in syntheses arise from kinetic, not thermodynamic problems.
ꢀ 2005 Elsevier B.V. All rights reserved.
Keywords: Corannulene; Chromium; Molybdenum; Tungsten; Iron; Arene complexes
1. Introduction
compounds which might be prepared by, e.g., corannu-
lene (Fig. 1, B), which may be thought of as a C20 frag-
ment cap of C60.
Because of their unusual electronic structures and
high degree of unsaturation, fullerenes such as C60
(Fig. 1, A) have been much investigated for their abilities
to form metal-containing compounds [1–3]. Metal–C60
compounds generally fall into one of three main catego-
ries, endohedral compounds in which the metal atom is
encapsulated within the fullerene sphere [1], alkali metal
Corannulene is a polycyclic aromatic hydrocarbon
composed of five six-membered rings fused to each other
and to a five-membered central ring [4,5]; the molecule is
bowl shaped because of the need of the C6 rings to
accommodate the steric requirements of the central C5
ring, but bowl inversion is facile [5c–f]. As is indicated
in the structure (B), corannulene contains four different
types of carbon–carbon bonds which may be designated
as hub, flank, spoke and rim bonds. Structural studies
indicate that the bond orders are localized essentially
as shown [5], with the hub and flank C–C bond lengths
fulleride compounds containing discrete anionic species
nꢀ
C
ðn ¼ 1–6Þ [2a–e], and g2-transition metal complexes
60
in which the fullerene behaves essentially as an electron
deficient alkene ligand [3]. More recent attention has
also focused on the possible types of metal-containing
˚
being 1.413(3) and 1.440(2) A, respectively, the spoke
˚
and rim bond lengths being 1.391(4) and 1.402(5) A,
respectively [5a]. The three types of carbon atoms in cor-
annulene are numbered in order to facilitate later
discussion.
*
Corresponding author. Tel.: +6135332614; fax: +6135336669.
E-mail address: bairdmc@chem.queensu.ca (M.C. Baird).
0022-328X/$ - see front matter ꢀ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.04.044

M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
3441
hub
spoke
flank
M
1
3
C
rim
O
C
C
O
2
O
C
A
B
Fig. 1. Structures of C60 (A) and corannulene (B).
There are many routes to stable g6-arene compounds
of these types [7] and, as we show below, we have tried
virtually all without success. In an attempt to possibly
assess whether we were in fact dealing with a thermody-
namic problem, i.e., very low stability of the target mol-
ecules, we have also carried out DFT calculations on the
compound (g6-corannulene)Cr(CO)3 using the B3LYP
[8a–c] functional on a self-consistent DFT level of calcu-
lation. The calculations strongly suggest that the g6-cor-
annulene–Cr linkage in (g6-corannulene)Cr(CO)3 is
only a few kcal/mol weaker than the g6-benzene linkage
in (g6-C6H6)Cr(CO)3, and thus our failures may better
be attributed to kinetic problems. We have also ex-
tended our synthetic investigations to the attempted syn-
theses of [CpFe(g6-corannulene)](PF6), a potential
member of a well known class of compounds of the type
[CpFe(g6-arene)](PF6) [9]. In this case we have enjoyed
marginal success.
The potential for interesting coordination chemistry
of corannulene is obvious, as the compound might be
expected to bind to a wide variety of metal–ligand moi-
eties in g2-fashion via the rim C@C bonds, in g4-fashion
to, e.g., an Fe(CO)3 moiety via a pair of 1,3-spoke C@C
bonds, and in g6-fashion to, e.g., a Cr(CO)3 moiety via
one of the aromatic C6 rings. Furthermore, of course,
coordination to the convex (exo) or the concave (endo)
surface of the bowl seems possible. To date, however,
very few examples of metal corannulene complexes are
known and even fewer have been characterized crystal-
lographically. The first example of a transition metal
complex of corannulene was [Cp*Ru(g6-C20H10)]-
(O3SCF3), prepared by reaction of [Cp*Ru(NC-
Me)3](O3SCF3) with corannulene in CD2Cl2 in 1997
1
[6a]. Although the H and 13C{1H} NMR spectra of
the product showed clearly that the C5 axis of the coran-
nulene had been broken, the compound was not charac-
terized crystallographically and the g6 structure was
inferred largely because of the generally high stability
of compounds of the type [Cp*Ru(g6-arene)]. More re-
cently, however, an analogous diruthenium compound,
[(Cp*Ru)2(l2-g6, g6-C20H10)](PF6)2, has been prepared
and characterized crystallographically, confirming the
expected g6 mode of coordination [6b].
The second corannulene complex reported was the
very similar [Cp*Ir(g6-C20H10)](BF4)2 [6c], but again
the compound was too unstable to be fully character-
ized. A somewhat different series of corannulene com-
pounds is that of Petrukhina et al. [6d], who through
metal–vapour deposition reactions obtained the com-
pounds {[Rh2(O2CCF3)]4}m{[g2-C20H10]n} (m:n = 1:1,
3:2); coordination via both endo and exo faces was dem-
onstrated crystallographically, but the corannulene li-
gands dissociated when the compounds were dissolved
in organic solvents.
2. Experimental
All 1D and 2D NMR spectra were obtained on Bru-
ker AC 200, Avance 400 and Avance 600 NMR spec-
trometers, chemical shifts being referenced with respect
to TMS using the residual proton or carbon resonances
of the solvent. IR spectra were acquired on a Perkin El-
mer Spectrum One FT-IR spectrometer in cells with
NaCl windows at a spectral resolution of 4 cmꢀ1
.
ESMS, EIMS and LDMS were obtained on Applied
Biosystems/MDS Sciex QSTAR XL, Micromass/VG
Quattro Triple Quadrupole and Applied Biosystems
Voyager instruments, respectively.
All manipulations were carried out in an Mbraun
Labmaster glove box or using standard Schlenk line
techniques under argon purified by passage through a
˚
column of BASF copper catalyst and a column of 5 A
molecular sieves. Solvents were dried by passage
through a column of activated alumina and were de-
gassed prior to use. All chemicals, except those noted be-
low, were purchased from Aldrich and used as received.
Norbornadiene and dehydroacetic acid were purchased
from Lancaster and used as received.
Although these results imply a general lack of stabil-
ity for corannulene–metal complexes, the number of
types of metal–ligand moieties investigated is actually
very small. We decided, therefore, to attempt to synthe-
size the series of arene compounds (g6-corannu-
lene)M(CO)3 (M = Cr, Mo, W), of the type C where
exo coordination is illustrated.
The compounds M(CO)3(MeCN)3 (M = Cr, Mo, W)
were prepared in 80–90% yields by refluxing the

3442
M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
hexacarbonyls in acetonitrile while a constant stream of
argon was bubbled directly through the solution to pro-
mote the loss of CO [10a,b]. The compounds (g6-cyclo-
heating to 70 ꢁC. Norbornadiene (250 mL, 2.32 mol)
was added down the condenser and the mixture was re-
fluxed for 4 days. The solution was allowed to cool to
room temperature and left to crystallize in the freezer
overnight. The liquid was decanted off and the crude
was recrystallized from toluene to give 108 g (69%)
7,10-diacetylfluoranthene. 1H NMR (200 MHz, CDCl3):
d 2.78 (s, 6H), 7.58 (s, 2H), 7.64 (d, J = 7.6 Hz, 2H), 7.91
(d, J = 8.2 Hz, 2H), 8.33 (d, J = 7.6 Hz, 2H). Lit. 2.77
(6H), 7.59 (2H), 7.64 (2H), 7.93 (2H), 8.34 (2H) [4f].
heptatriene)M(CO)3 (M = Cr [10c,d],
W
[10e]),
M(CO)3(pyridine)3 (M = Cr [10f], Mo [10g]), (g6-ben-
zene)Cr(CO)3 [10h,i] were prepared as in the literature.
2.1. Synthesis of corannulene
The synthesis of corannulene followed largely the
procedures of Scott et al. [4f] albeit with modifications
as described below.
2.1.3. 7,10-Bis(1-chlorovinyl)fluoranthene
The literature method [4f] for this step resulted in low
yields and many impurities due to over chlorination.
The preparation was modified as follows, 7,10-diacetyl-
fluoranthene (51 g, 0.178 mol) and PCl5 (74 g, 0.356
mol) were added to a 3 L round bottom flask. Methylene
chloride (1.5 L) was added and the contents were stirred
under argon at room temperature for 6 days, after which
the flask was opened to air, cooled in an ice bath and ex-
cess PCl5 hydrolyzed by the addition of crushed ice. The
mixture was washed with water (3 · 1 L) and sat. NaCl
(aq) (2 · 1 L). The organic layer was separated, dried
over MgSO4 and evaporated to dryness under reduced
pressure. The crude product was chromatographed on
silica gel using a hexanes/dichloromethane mixture
(7:3). The entire yellow band was collected from the col-
umn as a single fraction while a brown residue was left
on the column. The collected fraction was evaporated
to dryness leaving yellow crystals. The yellow crystals
were sublimed in two fractions at 140 ꢁC overnight to
give pure 7,10-bis(1-chlorovinyl)fluoranthene; the yield
(reported as a combined average of two complete runs)
2.1.1. 2,4,6-Heptanetrione
In a procedure modified from the literature [4h,i],
dehydroacetic acid (100 g, 0.59 mol) and concentrated
HCl (400 mL) were placed in a 1 L round bottom flask,
fitted with a reflux condenser vented to an oil bubbler to
monitor the evolution of CO2 gas. The contents of the
flask were refluxed until CO2 evolution ceased (ꢁ4 h).
The flask was then allowed to cool to room temperature
and further cooled in an ice bath while NaOH pellets
(160 g, 4.0 mol) were added and dissolved until the solu-
tion was basic to litmus. The contents were transferred
to a 2 L flask containing a solution of Ba(OH)2 Æ 8H2O
(250 g, 0.84 mol) in 1 L of boiling water. The mixture
was cooled to room temperature and the resulting pre-
cipitate was filtered. The precipitate was placed in a 3
L flask followed by 1 L of water and 700 mL of diethyl
ether. Concentrated HCl (150 mL) was added slowly un-
til all solids dissolved and the aqueous layer was acidic
to litmus. The dark orange ether layer was separated,
dried over MgSO4, and evaporated to dryness under re-
duced pressure to yield 27.8 g (33%) 2,4,6-heptanetrione
that was stored in the freezer as previous work had
found the product to decompose at room temperature
1
was 49%. H NMR (200 MHz, CDCl3): d 5.71 (d, J = 1
Hz, 2H), 5.87 (d, J = 1.2 Hz, 2H), 7.30 (s, 2H), 7.64 (dd,
2H), 7.88 (d, J = 8.2 Hz, 2H), 8.30 (d, J = 7 Hz, 2H). Lit.
5.72 (2H), 5.88 (2H), 7.30 (2H), 7.64 (2H), 7.89 (2H),
8.32 (2H).
1
[4j]. H NMR (200 MHz, CDCl3): d 1.97 (s, 1.51H),
2.07 (s, 2.12H), 2.23 (s, 2.59H), 3.38 (s, 1.30H), 3.69
(s, 0.19H), 5.12 (s, 0.42H), 5.54 (s, 0.57H). Lit. d 1.95
(1.51H), 2.05 (2.12H), 2.22 (2.54H), 3.40 (1.37H), 3.68
(0.29H), 5.16 (0.53H), 5.60 (0.65H) [4k].
2.1.4. Corannulene
The apparatus used for the pyrolysis reactions was
developed based on that reported by Scott et al. [4f].
In a typical pyrolysis, 7,10-bis(1-chlorovinyl)fluoranth-
ene (1.5 g, 4.6 mmol) was placed in a porcelain sample
boat which rested inside a quartz tube (25 mm OD, 85
cm long with O-Ring joints at both ends). The tube
rested in an electric tube furnace (30 cm heating zone),
and was connected upstream to a cylinder of prepurified
argon. Downstream from the furnace, the tube was con-
nected to a liquid nitrogen cooled trap which was in turn
connected to a Schlenk line. The system was sealed and
evacuated, and the furnace was heated to 1100 ꢁC. A
slow leak of argon was initiated and slow sublimation
of the 7,10-bis(1-chlorovinyl)fluoranthene in the argon
flow was induced by a heating tape wrapped around
the upstream portion of the quartz tube. The pyrolysis
2.1.2. 7,10-Diacetylfluoranthene
Scott has made significant modifications to his origi-
nal 3-step synthesis of C20H10 [4l]. Following the new
procedure, acenaphthenequinone (100 g, 0.55 mol),
2,4,6-heptanetrione (78 g, 0.55 mol), methanol (1.6 L)
and triethylamine (137 mL) were added to a 6 L flask.
The contents were then stirred for 3 h after which 2 L
of water were added. Concentrated HCl (ꢁ125 mL)
was added until the mixture was acidic to litmus and
then a further 1 L of water was added. The mixture
was then stirred for 20 minutes and the resulting brown
precipitate was filtered and allowed to dry in the fume
hood for several days. The dry precipitate (135 g) was
dissolved in 1.4 L toluene in a 3 L round bottom by

M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
3443
was allowed to run until no starting material remained
in the sample boat (ꢁ4–6 h). The furnace was then
turned off and the system was allowed to cool to room
temperature. Once at room temperature, the apparatus
was opened to air and the tube and nitrogen trap were
washed with methylene chloride. The combined wash-
ings were filtered to remove graphitic material and evap-
orated to dryness. The residue was chromatographed
using cyclohexane on silica gel and the yellow product
was collected as one fraction while a dark brown band
was left on the column. Cyclohexane was removed on
the rotary evaporator to give pure corannulene. The
yield, reported as a combined average of 14 runs of
the pyrolysis (20.4 g, 63 mmol of starting material),
was 11.4% (1.8 g, 7.2 mmol). 1H NMR (400 MHz,
CDCl3): d 7.80; 13C{1H} NMR (150 MHz, CD2Cl2)
127.04, 130.96, 135.69. Lit. 1H: 7.81 (s), 13C{1H}:
127.18, 130.84, 135.81 [4f].
power setting and run time, but all experiments involv-
ing C20H10 and the above reactants failed. In all cases,
1H NMR spectra of the products exhibited only the sin-
gle resonance of free C20H10.
2.4. Reaction of K2[C20H10] with Cr(CO)6
A mixture of 0.10 g C20H10 (0.40 mmol) and 0.08 g
potassium metal (2 mmol) in 60 mL THF was sonicated
for 2 h, by which time the color had changed from yel-
low to green and finally the persistent deep red of
(C20H10)2ꢀ [11]. The solution was stirred overnight to
ensure complete conversion, filtered to remove excess
potassium and added to 0.10 g Cr(CO)6 (0.45 mmol).
The resulting solution was stirred for 1 h, after which
the yellow precipitate was allowed to settle and the solu-
tion was cannulated off and evaporated to dryness. The
resulting residue was washed with hexanes (3 · 20 mL)
and the washings were also evaporated to dryness. At
this point the three separation fractions (yellow precipi-
tate, hexanes soluble residue and hexanes insoluble res-
idue) were examined.
The yellow precipitate exhibited no resonances in the
1H NMR spectrum and was therefore characterized by
IR spectroscopy and electrospray mass spectrometry
(ESMS). Fragments due to the loss of CO groups con-
firm the complex to be K[H(Cr(CO)5)2]. ESMS, m/z
(amu) Relative intensities are given in parentheses and
are referenced individually to the largest peak of each
respective isotope pattern; [HCr2C10O10]: 385.96 (34),
384.96 (100), 382.97 (10); Calc.: 389 (0.1), 388 (2.1),
387 (11.8), 386 (33.7), 385 (100), 384 (2.2), 383 (10.3),
381 (0.2). [HCr2C9O9]: 359.97 (2.0), 358.96 (12), 357.96
(37.5), 356.96 (100), 354.97 (12) Calc.: 361 (0.1), 360
(1.8), 359 (11.3), 358 (32.6), 357 (100), 356 (2.2), 355
(10.3), 353 (0.2). [HCr2C8O8]: 331.96 (2.4), 330.96
(13.4), 329.96 (37.8), 328.95 (100), 327.97 (3.0), 326.96
(37.8) Calc.: 332 (1.6), 331 (10.8), 330 (31.5), 329,
(100), 328 (2.1), 327 (10.3), 325 (0.2). [HCr2C7O7]:
302.96 (10.2), 301.96 (33.3), 300.96 (100), 298.96 (10.2)
Calc.: 304 (1.5), 303 (10.1), 302 (30.3), 301, (100), 300
(2.0), 299 (10.3), 297 (0.2). IR m(CO) (cmꢀ1, THF)
1942(s), 1884(s) Lit. (sodium salt): 2032(w), 1942(s),
1879(s) [12a,b].
2.2. Thermal reactions of C20H10 and M(CO)6
(M = Cr, Mo)
A solution of 0.1 g C20H10 (0.4 mmol) and 4 mmol of
the appropriate hexacarbonyl was refluxed in 50 mL of
hexanes for two weeks (Mo) or four weeks (Cr), after
which IR spectra of the solutions exhibited only the car-
bonyl absorptions of the starting materials.
2.3. Attempted reactions of C20H10 with alternative
sources of the M(CO)3 moieties (Cr, Mo and W)
A mixture of 0.1 g C20H10 (0.4 mmol) and 0.1 g
Mo(CO)3(MeCN)3 (0.4 mmol) were ground in a mor-
tar to a uniform mixture and transferred to a Schlenk
flask. The flask was sealed, put under static vacuum
and submerged in an 80 ꢁC oil bath for 4 h. After this
time, the material had changed from yellow to dark
brown and the mixture was cooled and washed with
hexanes to give a black residue, the IR spectrum of
1
which exhibited no m(CO). A H NMR spectrum exhib-
ited only the singlet of C20H10. Similarly unproductive
were reactions of M(CO)3(MeCN)3 (M = Cr, Mo, W)
in THF at 45 ꢁC over several days, of Cr(CO)3-
(MeCN)3 in refluxing hexanes or diglyme for 24 h, of
(g6-cycloheptatriene)M(CO)3 (M = Cr, W) in hexanes
for a month, of M(CO)3(pyridine)3 (M = Cr, Mo) in
a mixture of ethyl ether and boron trifluoride diethyl
The material that was insoluble in hexanes was deter-
mined to be K2[Cr2(CO)10]; IR: m(CO) cmꢀ1 (Nujol)
1919(s), 1893(vs), 1869(2), 1825(m), 1805(m), 1762(w).
The pattern was identical to that reported in the litera-
ture for Na2[Cr2(CO)10]: 1903(s), 1880(vs), 1851(s),
1
etherate for 2 h. In all cases, H NMR spectra of the
products exhibited only the single resonance of free
C20H10.
1
A number of experiments were also preformed utiliz-
ing a CEM Discover microwave reactor. Experiments
were carried out in 10 mL glass tubes pressure rated to
200 psi and fitted with a magnetic stir bar and a pressure
rated septum cap. A variety of experimental parameters
were varied, including solvent, temperature, pressure,
1801(m), 1776(m), 1756(w) [12c]. The H NMR spec-
trum of the material did not exhibit any resonances.
The IR spectrum of the hexanes soluble material did
not exhibit any absorptions in the CO region, while
1
the H and 13C{1H} NMR spectra exhibited strong res-
onances of free corannulene.

3444
M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
2.5. Reaction of K2[C20H10] with (g6-benzene)Cr(CO)3
(6.2), 390 (1.6), 391(100), 392 (29.6), 393 (4.5), 394
(0.2). ESMS (m/z) Neg. Ions, Rel. Int.: 144.9 (100). Calc.
for PF6: 145 (100).
A solution of K2[C20H10] was prepared as above,
cooled to ꢀ78 ꢁC and transferred to a solution contain-
ing a 10-fold molar excess of (g6-benzene)tricarbonyl
chromium. Upon mixing, the deep red color of the dian-
ion turned dark brown, and the mixture was brought to
room temperature, stirred for 30 min, treated with 47 lL
of air, stirred for a further 10 min and then taken to dry-
ness under reduced pressure. The resulting residue was a
dark brown to black solid. An IR spectrum of the dark
residue exhibited no absorptions in the carbonyl region,
although a1H NMR spectrum exhibited a strong reso-
nance of free C20H10.
2.7. Attempts to prepare [CpFe(g6-corannulene)](PF6)
in decalin
A mixture of 0.29 g C20H10 (1.16 mmol), 0.03 g Al
powder (1.16 mmol), 1.55 g AlCl3 (1.16 mmol), 2.16 g
ferrocene (1.16 mmol), 0.02 mL water (1.16 mmol) and
35 mL of decalin was heated at 140 ꢁC while stirring
for 16 h. After heating, the reaction mixture was cooled
to 0 ꢁC and hydrolyzed. The aqueous layer was sepa-
rated before NH4OH was added, and the resulting solu-
tion was filtered. Aqueous HPF6 (60% wt.) (0.2 mL,
1.16 · 10ꢀ3 mol) was then added to the filtrate to precip-
itate a brown-yellow solid (0.06g) which was collected
via filtration. This solid was dissolved in acetone and
precipitated by addition of Et2O. Repeated attempts at
recrystallization by layering an acetone solution with
Et2O failed to give a pure material. This material was
passed through alumina using acetone as an eluant to
no avail. During all phases of the reaction and the reac-
tion workup, great care was taken to minimize the expo-
2.6. Attempt to prepare [CpFe(g6-corannulene)](PF6)
in the absence of solvent
A mixture of 0.65 g C20H10 (2.60 mmol), 0.14 g Al
powder (5.2 mmol), 3.46 g AlCl3 (2.60 mmol) and 4.83
g ferrocene (2.60 mmol) was heated at 140 ꢁC while stir-
ring for 22 h. After this period, the reaction mixture was
cooled to 0 ꢁC and hydrolyzed with NH4OH, and the
resulting solution was filtered. A solution of HPF6
(60% by wt) (1.16 · 10ꢀ3 mol) was then added to the fil-
trate to precipitate a brown-yellow solid (0.08 g) which
was collected via filtration. This solid was dissolved in
acetone and precipitated by addition of ethyl ether. Re-
peated attempts at recrystallization failed to give pure
material. This isolated material was passed through alu-
1
sure of the reaction mixture and products to light. H
NMR ((CD3)2CO, 25 ꢁC, 400 MHz): d 3.46 (s), 3.5–
3.7 (br peaks), 3.96 (s), 4–4.3 (br peaks), 6.22 (br),
6.60 (br), 7.06 (br), 7.8–7.9 (two br), 8.1–8.5 (br peaks).
ESMS (m/z) Pos. Ions, Rel. Int.: 371.3 (6.4), 372.3 (2.2),
373.2 (100), 374.3 (30), 375.3 (78.6), 376.2 (22.5), 377.3
(9.5), 378 (2.3), 379.4 (4.9), 380.3 (1.5), 381.3 (8.2),
382.3 (2.6), 383.4 (2.2), 384.3 (0.8), 385.4 (5.4), 386.4
(1.7). ESMS (m/z) Neg. Ions, Rel. Int.: 144.9 (100).
For calculated mass, see above reaction.
A second attempt using this method gave material
with the following NMR and ESMS properties:1H
NMR ((CD3)2CO, 25 ꢁC, 400 MHz): d 1.45–1.95 (series
of singlets), 3.7–3.9 (two br s), 3.95 (br s), 4.2 (br s), 4.8–
5.2 (br) 6.2-6.3 (br), 6.8–6.85 (br), 7.05 (br), 7.6–7.8 (two
br). Exact mass: (C25H29Fe+) measured: 385.16186.
Calc.: 385.1629, error: 2.7 ppm. (C25H31Fe+) measured:
387.17751. Calc.: 387.1770, error: 1.3 ppm.
1
mina using acetone as an eluant to no avail. H NMR
((CD3)2CO, 25 ꢁC, 400 MHz): d 3.46 (s), 3.5–3.7 (br
peaks), 3.96 (s), 4.04, (s), 4.14 (s), 4.2–4.5 (br peaks),
6.22 (br), 6.6–6.7 (br), 7.06 (br), 7.46 (d), 7.84 (d), 7.90
(d), 8.1-8.5 (series of br peaks). ESMS: (m/z) Pos. Ions,
Rel. Int.: 369.2 (6.0), 370.2 (1.5), 371.2 (9.1), 373.1 (100),
374.1 (30.4), 375.1 (86.1), 376.1 (26.2), 377.1 (70.5),
378.1 (22.0), 379.2 (53.4), 380.1 (16.1), 381.1 (74.3),
382.1 (21.8), 383.1 (14.2), 384.1 (4.2), 385.2 (30.2),
386.2 (9.8), 387.2 (7.1), 389.1 (4.6). Calc. for
C25H15Fe: 369 (6.2), 370 (1.6), 371 (100), 372 (29.6),
373 (4.5), 374 (0.2). Calc. for C25H17Fe: 371 (6.2), 372
(1.6), 373 (100), 374 (29.6), 375 (4.5), 376 (0.2). Calc.
for C25H19Fe: 373 (6.2), 374 (1.6), 375 (100), 376
(29.6), 377 (4.5), 378 (0.2). Calc. for C25H21Fe: 375
(6.2), 376 (1.6), 377 (100), 378 (29.6), 379 (4.5), 380
(0.2). Calc. for C25H23Fe: 377 (6.2), 378 (1.6), 379
(100), 380 (29.6), 381 (4.5), 382 (0.2). Calc. for
C25H25Fe: 379 (6.2), 380 (1.6), 381 (100), 382 (29.6),
383 (4.5), 384 (0.2). Calc. for C25H27Fe: 381 (6.2),
382(1.6), 383 (100), 384 (29.6), 385 (4.5), 386 (0.2). Calc.
for C25H29Fe: 383 (6.2), 384 (1.6), 385 (100), 386 (29.6),
387 (4.5), 388 (0.2). Calc. for C25H31Fe: 385 (6.2), 386
(1.6), 387 (100), 388 (29.6), 389 (4.5), 391 (0.2). Calc.
for C25H33Fe: 387 (6.2), 388 (1.6), 389 (100), 390
(29.6), 391 (4.5), 392 (0.2). Calc. for C25H35Fe: 389
2.8. DFT Calculations
To elucidate the structure of (g6-corannu-
lene)Cr(CO)3 and obtain an estimate of the Cr–corannu-
lene bond strength, we performed a series of ab initio
calculations. These were done with the B3LYP [8a–c]
functional on a self-consistent DFT level of calculation
utilizing the High Performance Computing Virtual Lab-
oratory (HPCVL) at Queenꢀs University. We obtained
results in three basis sets: the standard ‘‘split valence’’
6-31G* basis [8d–i], the same basis set augmented with
the Los Alamos Effective Core Potential on Cr [8j–l]
and in Dunningꢀs correlation consistent double-zeta

M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
3445
basis [8m]. All computations were performed using the
GAUSSIAN 98 suite of programs [8n].
2-
2-
+ C10H8
+ 2K[C10H8]
Cr
Cr
2-
C
O
C
C
O
C
Cr
C
O
C
O
O
C
3. Results and discussion
O
O
C
C
O
O
3.1. Attempts to synthesize g6-corannulene complexes of
Cr, Mo, W
O2
Cr
C
C
Considerable time was spent developing and optimiz-
ing a procedure for the synthesis of corannulene, de-
scribed in Section 2. Carbonyl compounds of the
Group 6 metals, Cr, Mo and W, are well known to form
p-arene complexes of the type (g6-arene)M(CO)3, and a
wide range of synthetic routes has been reported, rang-
ing from direct reactions of arenes with the metal hexa-
O
Cr
C
O
C
O
O
C
C
O
O
Scheme 1. Conversion of (g6-beneze)Cr(CO)3 to (g6)Cr(CO)3 [13].
ion, K2[C20H10] [11], with both Cr(CO)6 and (g6-ben-
zene)Cr(CO)3. As described in Section 2, reactions did
occur but did not yield the desired product. Instead,
electron transfer processes occurred with Cr(CO)6 to
give corannulene and the reduced chromium complexes,
K[H(Cr(CO)5)2] and K2[Cr2(CO)10] [12], identified by
IR spectroscopy and mass spectrometry, while the reac-
tion with (g6-benzene)Cr(CO)3 gave corannulene but no
carbonyl-containing products.
The inability of any of these traditional and non-
traditional routes to form any of the compounds
(g6-C20H10)M(CO)3 (M = Cr, W) was something of a
surprise. As noted above, all of the routes used have
been widely cited in the literature and several have been
successfully used to give coordination complexes with
polycyclic arenes [15]. We discuss below some of the
possible reasons for failure.
In the case of g6-arene complexes of planar polyaro-
matic hydrocarbons (PAHs), we note that the existing li-
brary of g6-arene complexes of the group 6 metals
contains many examples that illustrate the preference
of the metal for less substituted rings [15b–d]. In studies
involving coordinated 1,4-disubstituted naphthalenes to
chromium, for example, substituents such as MeO,
Me2N, MeO2C and F all result in coordination of a
group 6 metal tricarbonyl moiety to the unsubstituted
ring as in D [15e].
carbonyls,
through
reactions
involving
labile
trisubstituted metal precursors of the type M(CO)3L,
(L = 3CH3CN, 3pyridine, g6-cycloheptatriene, g6-
arene) [7], to multi-step redox exchange reactions [13]
which we have attempted to exploit. While there are
examples of coordination of the metals to mono and
polycyclic planar arenes, no complexes have been re-
ported involving coordination of these metals to curved
aromatic molecules although the possibility of coordina-
tion of corannulene to Cr and W has been suggested
[6c].
Our attempts to coordinate corannulene to the three
M(CO)3 moieties have involved lengthy (up to several
weeks) refluxing of solutions of C20H10 with all three
hexacarbonyls in various solvents, in addition to the
use of a microwave reactor, an approach which was re-
ported for effecting substitution of the hexacarbonyls
during the course of our study [14]. We have also uti-
lized as possibly more labile precursors [7], the com-
pounds M(CO)3(MeCN)3 (M = Cr, Mo, W) at various
temperatures in THF, hexanes and diglyme as well as
in the absence of solvent. We have refluxed hexanes
solutions of the compounds (g6-cycloheptatriene)-
M(CO)3 (M = Cr, W) with C20H10 for a month, and
we have reacted M(CO)3(pyridine)3 (M = Cr,W) with
C20H10 in the presence of boron trifluoride diethyl
etherate as a pyridine trap. None of these more or less
conventional approaches have produced any material
which contained coordinated C20H10.
R
R
We have also attempted to utilize a non-conven-
tional, very ingenious sequence of reactions by which
Cooper and Leong succeeded in converting (g6-ben-
zene)Cr(CO)3 to the naphthalene analogue, (g6-naph-
thalene)Cr(CO)3 [13]. In this novel, indirect route,
(g6-benzene)Cr(CO)3 is initially reduced to [(g4-ben-
zene)Cr(CO)3]2ꢀ by potassium naphthalenide as in
Scheme 1. The reaction is unusual in that it allows sub-
stitution of benzene by a polycyclic arene.
Cr
OC
CO
CO
D: R = MeO, Me2N, MeO2C, F
In the case of larger fused or condensed ring PAHs,
in which two or more atoms are shared between rings,
the metals often coordinate preferentially to the least
substituted ring [7] although distinctions have been
Attempting to exploit this type of unusual approach,
we reacted the potassium salt of the corannulene dian-

3446
M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
made in cases where thermodynamic and kinetic prod-
ucts have been isolated by different routes. Oprunenko
et al. [15b] reported this difference with the isomers of
(g6-fluoranthene) tricarbonyl chromium; kinetic and
thermodynamic isomers are shown as E and F, respec-
tively.
ity of six rings for coordination, there have been no
reports in which coronene has been coordinated to a
group 6 metal. Indeed, the only reported transition me-
tal complexes of C24H12 are of iron [16a–c] and ruthe-
nium [16d,e] (see below).
It is interesting to note that no examples exist in the
literature in which a group 6 metal has been coordinated
to any six-membered ring that is attached to three other
rings. This void in the literature and the above examples
cannot be used to conclude that coordination of C20H10
to Cr, Mo and W cannot occur. However, they do sug-
gest that C20H10 is not an isolated example in which
coordination does not take place by the same protocols
used for simpler planar arenes. Indeed, in view of our
theoretical finding that compounds of the type (g6-cor-
annulene)Cr(CO)3 should be thermodynamically stable
(see below), it would seem that the reason for synthetic
failures is kinetic in origin.
Cr
CO
Cr
CO
E. Kinetic Product
OC
CO
OC
CO
F. Thermodynamic Product
Despite the formation of the kinetic product, isomeri-
sation to the thermodynamic product in which the metal
is coordinated to the less substituted ring is evident. In-
deed, the preference of the Cr(CO)3 unit for less substi-
tuted rings is even more pronounced as the number of
condensed rings is increased. In the case of pyrene,
two different types of aromatic rings exist, those that
share three or four carbon atoms with adjacent rings.
In this case, the preference for the less substituted ring
(G) is so great that coordination to the ring with four
shared carbon atoms (H) has not been observed [15d].
The degree of aromatic stabilization has been suggested
as a reason for this preference [15d].
3.2. Attempts to synthesize g6-corannulene complexes
of Fe
The [CpFe]+ and [Cp*Fe]+ (Cp* = pentamethylcyclo-
pentadienyl) cationic moieties are known to coordinate
to a variety of aromatic and polyaromatic molecules
in g6 fashion [9,16b,17]. Numerous examples of
[CpFe(g6-polycyclic aromatic)]+ and [Cp*Fe(g6-polycy-
clic aromatic)]+ complexes have been reported, including
those of coronene, naphthalene, phenanthrene, pyrene,
triphenylene and perylene [16b]. We therefore antici-
pated that use of the [CpFe]+ system might provide a
route to corannulene complexes.
To this end we have tried two different methods, re-
ported by Lacoste et al,. to generate the desired
[CpFe(corannulene)]+ complex [16b]. In the first a mix-
ture of Cp2Fe, AlCl3, Al powder, and corannulene were
heated at 140 ꢁC, while in the second method Cp2Fe,
AlCl3, Al powder, H2O and corannulene were combined
in decalin and heated to 140 ꢁC. The solvent free method
is normally used in the melt of the aromatic substrate,
but, in view of the high melting temperature of corannu-
lene, we carried out the reaction as a heterogeneous mix-
ture. The experiments resulted in the formation of only
very small amounts of products which were character-
Cr
Cr
OC
CO
OC
CO
CO
CO
H. Not formed
G. Exclusive Product
In an effort to gain insight into the problems
involved, one might consider coronene (I), the clos-
est example of a planar condensed arene to corannu-
lene.
1
ized by H NMR spectroscopy and electrospray mass
spectrometry (ESMS).
While ESMS on the materials obtained all showed
that none of the desired [CpFe(corannulene)]+ was pres-
ent (m/z = 371 Da) (Fig. 2), there were apparent in the
spectra a series of peaks which indicated the presence
of partially hydrogenated species of the type [CpFe(cor-
annulene + nH2)]+ (n = 1–7 for the materials obtained in
the solvent free reactions, 1–9 for the materials obtained
in the reactions in decalin), albeit in varying relative
amounts from one reaction to the next. In all cases, the
isotopic distributions were as expected; in one example
I
Coronene is the simplest planar PAH in which all of
the peripheral rings are equivalent and share four car-
bon atoms with neighboring rings. Despite the availabil-

M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
3447
373.1
373.2
100
100
b
a
375.1
375.3
377.1 381.1
379.2
%
%
383.1
385.2
376.4
377.3
387.2
371.2
369.2
371.3
370
Da/e
390
Da/e
390
0
0
365
370
380
375
385
360
365
375
380
385
Fig. 2. Partial electrospray mass spectra of mixtures of complexes of the type [CpFe(partially hydrogenated corannulenes)]+ (a) from the reaction in
decalin, and (b) from the solvent free reaction.
a high resolution mass spectrum was obtained and indi-
cated unequivocally that the species [CpFe(C20H24)]+
and [CpFe(C20H26)]+ had formed, i.e., [CpFe(corannu-
lene + 7H2)]+ and [CpFe(corannulene + 8H2)]+.
tial hydrogenation. Resonances were also observed in
the region d 3.5–3.7, which is expected for the CH2
groups of coordinated partially hydrogenated corannu-
lene. A COSY spectrum exhibited a number of correla-
tions between different peaks in the aromatic region.
However because of the number of overlapping peaks
and the presence of paramagnetic broadening in all sam-
ples, other inferences are difficult to make due to the
number of complexes and the several different isomers
possible for each hydrogenated complex.
The hydrogenation of polycyclic arenes during this
type of reaction has precedence in a number of reported
examples [16b,17], for instance in reactions involving
naphthalene, anthracene, phenanthrene and pyrene.
This provides further evidence that the products formed
during the course of our reactions with corannulene, are
in fact complexes of the type [CpFe(arene)]+. Hydroge-
nation would lead to a reduction in the ring strain expe-
rienced by corannulene, providing a driving force for
hydrogenation, and we note that complexes of partially
hydrogenated corannulene are known [18].
3.3. Theoretical results
Calculations were done with the B3LYP [8a–c] func-
tional on a self-consistent DFT level of calculation uti-
lizing three basis sets, the standard ‘‘split valence’’
6-31G* basis [8d–i], the same basis set augmented with
the Los Alamos Effective Core Potential on Cr [8j–l],
and in Dunningꢀs correlation consistent double-zeta ba-
sis [8m]. We obtained two distinct low-energy structures
for (g6-corannulene)Cr(CO)3, corresponding to a posi-
tion of the Cr(CO)3 group inside (‘‘endo’’) or outside
(‘‘exo’’) of the corannulene ligand. In both cases, a stag-
gered conformation seems to be slightly favoured,
although the barrier to rotation is very low. In all basis
sets used, the exo isomer depicted as C and as in Fig. 4 is
more stable than the endo isomer, possible because of
greater steric hindrance in the latter.
1
The H NMR spectra (Fig. 3) of the materials pro-
vided further evidence for the types of complexes sug-
gested by the ESMS data. Thus the spectra exhibited
several singlets in the Cp region, d 3.9–4.5, consistent
with previously published data for cationic complexes
of the type [CpFe(g6-polycyclic aromatic)]+ [9]. Numer-
ous resonances were also observed in the aromatic re-
gion of the spectra, consistent with there being present
a mixture of compounds in which the symmetry of the
corannulene had been reduced by coordination and par-
To estimate the relative coordination strengths of
(g6-corannulene)Cr(CO)3, we also performed vibra-
tional analyses in the same basis sets and on the same le-
vel of calculation, for both structures, as well as for the
Cr(CO)3 and the corannulene fragments. The results
were used to obtain zero-point corrected bond energies.
For comparison, the same calculations were performed
for the well-known (g6-benzene)Cr(CO)3 for which the
experimentally determined Cr–benzene bond dissocia-
tion energy is 53 kcal/mol [19]. We imposed D6h
8.5
8.0
7.5
7.0
6.5
δ
Fig. 3. The aromatic region of the 1H NMR spectrum of a mixture of
complexes of the type [CpFe(partially hydrogenated corannulenes)]+.

3448
M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
kcal/mol stronger than that of the endo isomer. More-
over, the corresponding bond strength is approximately
7 kcal/mol weaker for the exo corannulene isomer than
for the corresponding benzene complex. Although the
values of the bond strengths differ somewhat for differ-
ent basis sets, the relative differences are remarkably
consistent.
Important calculated bond lengths are given in Table
2, where we compare calculated data for free corannu-
lene with experimental X-ray and electron diffraction
data. As can be seen, the calculated and experimental
X-ray diffraction data differ only marginally, and the
trends in the experimental C–C bond lengths are re-
flected very closely in the calculations. Also given in Ta-
ble 2 are the corresponding C–C bond lengths for the
arene ring coordinated to the Cr(CO)3 fragment. As
can be seen, the hub, spoke and flank C–C bonds all
lengthen slightly on coordination but the rim C–C bond
lengthens considerably. That the latter observation re-
flects a stronger interaction with the metal is consistent
with our finding that the Cr–C(3) distances are notably
shorter than the others.
The significant Cr–corannulene bond strength in
(g6-corannulene)Cr(CO)3, although less than that of
the corresponding benzene–Cr bond strength, suggests
that (g6-corannulene)Cr(CO)3 should indeed be ther-
mally stable under the experimental conditions described
here. Perhaps a kinetic explanation may be found for our
inability to prepare (g6-corannulene)Cr(CO)3, since the
thermodynamic stability of the compound suggests that
Fig. 4. The optimized structure of (g6-corannulene)Cr(CO)3 in its
‘‘exo’’ form.
symmetry on benzene, C5v symmetry on corannulene,
C3v symmetry on the (g6-benzene)Cr(CO)3 and the
Cr(CO)3 fragment. The exo and endo isomers of (g6-cor-
annulene)Cr(CO)3 have approximate Cs symmetry, but
no restrictions were imposed. The results are summa-
rized in Table 1.
As can be seen, while only two of the basis sets give
satisfactory values for the Cr–benzene bond dissociation
energy, all basis sets suggest that the chromium–coran-
nulene bond strength of the exo isomer is about 5
Table 1
Calculated Cr–arene bond energies (kcal/mol, zero-point corrected)
Compound
6-31G* [8d–i]
6-31G*/LanL2DZ [8d–l]
cc-pVDZ [8m]
Exp.
endo (g6-corannulene)Cr(CO)3
exo (g6-corannulene)Cr(CO)3
(g6-benzene)Cr(CO)3
51.9
57.4
64.5
42.3
47.3
54.0
44.9
50.5
57.7
53 [19]
Table 2
˚
6
Bond distances (A) in corannulene and of the g bound ring of (g6-corannulene)Cr(CO)3 as determined using the cc-pVDZ [8m] basis set. The C
atom numbering is as in Fig. 1 B
C–C Bond in corannulene
This work
X-ray [5a]
Electron diffraction [5b]
Hub
1.420
1.387
1.450
1.392
1.413(2)
1.391(4)
1.440(2)
1.402(5)
1.414(6)
Spoke
Flank
Rim
1.414(20)
1.447(16)
1.380(16)
C–C Bond in exo-(g6-corannulene)Cr(CO)3
Hub
1.428
1.394
1.453
1.431
Spoke
Flank
Rim
Cr–C Bond in exo-(g6-corannulene)Cr(CO)3
Cr–C(1)
Cr–C(2)
Cr–C(3)
2.296
2.413
2.200

M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
3449
[3] (a) W. Sliwa, Transit. Met. Chem. 21 (1996) 583;
(b) A.H.H. Stephens, M.L.H. Green, Adv. Inorg. Chem. 44
(1997) 1;
it should be isolable if prepared. Mechanistic studies of,
e.g., arene exchange of compounds of the type (g6-arene)-
Cr(CO)3 indicate a major role for associate processes
[20], and it is possible that this is more difficult for large,
polycyclic aromatic hydrocarbons.
(c) A.L. Balch, M.M. Olmstead, Chem. Rev. 98 (1998) 2123;
(d) V.I. Sokolov, Pure Appl. Chem. 70 (1998) 789;
(e) K. Lee, H. Song, J.T. Park, Acc. Chem. Res. 36 (2003) 78.
[4] (a) W.E. Barth, R.G. Lawton, J. Am. Chem. Soc. 88 (1966) 380;
(b) W.E. Barth, R.G. Lawton, J. Am. Chem. Soc. 93 (1971) 1730;
(c) A. Sygula, G. Xu, Z. Marcinow, P.W. Rabideau, Tetrahedron
57 (2001) 3637;
4. Summary
(d) H.-J. Kno¨lker, A. Braier, D.J. Bro¨cher, P.G. Jones, H.
Piotrowski, Terahedron Lett. 40 (1999) 8075;
In an extensive series of attempts to synthesize the
compounds (g6-corannulene)M(CO)3 (M = Cr, Mo,
W), we have utilized virtually every standard and some
non-standard of the published procedures for the syn-
thesis of compounds of the type (g6-arene)M(CO)3.
All approaches have failed miserably, none producing
even traces of the desired products, and it would seem
that the target molecules are for some reason much less
stable than are other gn-corannulene complexes which
have been reported. However, DFT calculations suggest
that the g6-corannulene–Cr bond of (g6-corannu-
lene)Cr(CO)3 is only marginally weaker than the
correspond arene–metal bond in the very stable
(g6-benzene)Cr(CO)3, and it seems that the reasons for
synthetic failures are kinetic in nature.
(e) T.J Seiders, E.L. Elliott, G.H. Grube, J.S. Siegel, J. Am.
Chem. Soc. 121 (1999) 7804;
(f) L.T. Scott, P.-C. Cheng, M.M. Hashemi, M.S. Bratcher, D.T.
Meyer, H.B. Warren, J. Am. Chem. Soc. 119 (1997) 10963;
(g) G. Mehta, G. Panda, Terahedron Lett. 38 (1997) 2145;
(h) G. Schwarzenbach, K. Lutz, Liebigs Ann. 23 (1940) 1186;
(i) A. El-Qisairi, P.M. Henry, J. Organomet. Chem. 603 (2000) 50;
(j) A. Kumar, M.M. Agrawal, B.P. Buchlas, Acta Cienc. Indica,
Chem. 9 (1983) 73;
(k) E. Marcus, J.K. Chan, C.B. Strow, J. Org. Chem. 31 (1966)
1369;
(l) M. Bancu, L. T. Scott, Private communication, 2002.
[5] (a) J.C. Hanson, C.E. Nordman, Acta Crystallogr., Sect. B 32
(1976) 1147;
(b) L. Hedberg, K. Hedberg, P.C. Cheng, L.T. Scott, J. Phys.
Chem. A 104 (2000) 7689;
(c) L.T. Scott, M.M. Hashemi, M.S. Bratcher, J. Am. Chem. Soc.
114 (1992) 1920;
(d) A. Sygula, A.H. Abdourazak, P.W. Rabideau, J. Am. Chem.
Soc. 118 (1996) 339;
Acknowledgements
(e) T.J. Seiders, K.K. Baldridge, J.S. Siegel, J. Am. Chem. Soc.
118 (1996) 2754;
We thank the Government of Ontario (Ontario
Graduate Scholarship to J.H.B.), Queenꢀs University
(Queenꢀs Graduate Award to M.W.S) and the Natural
Sciences and Engineering Research Council (Research
Grant to M.C.B.) for funding this research. We also
thank Professor T. Baer, who supplied us with data
from his computations, and Professor L.T. Scott for
helpful comments.
(f) T.J. Seiders, K.K. Baldridge, G.H. Grube, J.S. Siegel, J. Am.
Chem. Soc. 123 (2001) 517.
[6] (a) T.J. Seiders, K.K. Baldrige, J.M. OꢀConnor, J.S. Siegel, J.
Am. Chem. Soc. 119 (1997) 4781;
(b) P.A. Vecchi, C.M. Alvarez, A. Ellern, R.J. Angelici, A.
Sygula, R. Sygula, P.W. Rabideau, Angew. Chem. Int. Ed. 43
(2004) 4497;
(c) C.M. Alvarez, R.J. Angelici, A. Sygula, R. Sygula, P.W.
Rabideau, Organometallics 22 (2003) 624;
(d) M.A. Petrukhina, K.W. Andreini, J. Mack, L.T. Scott, Angew
Chem. Int. Ed. 42 (2003) 3375.
[7] (a) R. Davis, L.P.A. Kane-Maguire, in: G. Wilkinson, F.G.A.
Stone, E.W. Abel (Eds.), Comprehensive Organometallic Chem-
istry, vol. 3, Pergamon Press, Oxford, UK, 1982, p. 953;
(b) R. Davis, L.P.A. Kane-Maguire, in: G. Wilkinson, F.G.A.
Stone, E.W. Abel (Eds.), Comprehensive Organometallic Chem-
istry, vol. 3, Pergamon Press, Oxford, UK, 1982, p. 1149;
(c) R. Davis, L.P.A. Kane-Maguire, in: G. Wilkinson, F.G.A.
Stone, E.W. Abel (Eds.), Comprehensive Organometallic Chem-
istry, vol. 3, Pergamon Press, Oxford, UK, 1982, p. 1321;
(d) M.J. Morris, in: E.W. Abel, F.G.A. Stone, G. Wilkinson
(Eds.), Comprehensive Organometallic Chemistry II, vol. 5,
Pergamon Press, Oxford, UK, 1995, p. 471.
References
[1] (a) D.S. Bethune, R.D. Johnson, J.R. Salem, M.S. de Vries, C.S.
Yannoni, Nature (London) 366 (1993) 123;
(b) I. Holleman, M.G.H. Boogaarts, G. Meijer, Rec. Trav. Chim.
Pays-Bas 113 (1994) 543;
(c) F.T. Edelmann, Angew. Chem., Int. Ed. Engl. 34 (1995) 981;
(d) S. Nagase, K. Kobayashi, T. Akasaka, Bull. Chem. Soc. Jpn.
69 (1996) 2131;
(e) S. Guha, K. Nakamoto, Coord. Chem. Rev. 249 (2005) 1111.
[2] (a) M.S. Dresselhaus, G. Dresselhaus, P.C. Eklund, Science of
fullerenes and Carbon Nanotubes, Academic Press, 1996;
(b) M.J. Rosseinsky, Chem. Mat. 10 (1998) 2665;
(c) K. Prassides, in: W. Andreoni (Ed.), The Physics of Fullerene-
based and Fullerene-related Materials, Kluwer Academic Pub-
lishers, Netherlands, 2000, p. 175;
[8] (a) A.D. Becke, J. Chem. Phys. 98 (1993) 5648;
(b) C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785;
(c) B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett.
157 (1989) 200;
(d) R. Ditchfield, W.J. Hehre, J.A. Pople, J. Chem. Phys. 54
(1971) 724;
(d) C.A. Reed, R.D. Bolskar, Chem. Rev. 100 (2000) 1075;
(e) S. Liu, S. Sun, J. Organomet. Chem. 599 (2000) 74;
(f) For theoretical discussions, see N. Goldberg, R. Hoffmann, J.
Am. Chem. Soc. 118 (1996) 3315;
(e) W.J. Hehre, R. Ditchfield, J.A. Pople, J. Chem. Phys. 56
(1972) 2257;
(f) P.C. Hariharan, J.A. Pople, Mol. Phys. 27 (1974) 209;
(g) M.S. Gordon, Chem. Phys. Lett. 76 (1980) 163;
(g) N. Goldberg, R. Hoffmann, Inorg. Chem. 35 (1996) 4369.

3450
M.W. Stoddart et al. / Journal of Organometallic Chemistry 690 (2005) 3440–3450
[14] M. Ardon, G. Hogarth, D.T.W. Oscroft, J. Organomet. Chem.
(h) P.C. Hariharan, J.A. Pople, Theor. Chim. Acta 28 (1973) 213;
(i) R.C. Binning, L.A. Curtiss, J. Comp. Chem. 11 (1990) 1206;
(j) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270;
(k) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 284;
(l) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299;
(m) D.E. Woon, T.H. Dunning, J. Chem. Phys. 98 (1993) 1358;
(n) M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A.
Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery, R.E.
Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels,
K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M.
Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford,
J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma,
D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.
Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin,
D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara,
M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W.
Wong, J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Repl-
ogle, J.A. Pople, GAUSSIAN 98, Revision A.9, Gaussian, Inc.,
Pittsburgh, PA, 1998.
689 (2004) 2429.
[15] (a) S. Lin, P. Boudjouk, J. Organomet. Chem. 187 (1980) C11;
(b) Y. Oprunenko, S. Malyugina, A. Vasilꢀko, K. Lyssenko, C.
Elschenbroich, K. Harms, J. Organomet. Chem. 641 (2002) 208;
(c) R.B. King, F.G.A. Stone, J. Am. Chem. Soc. 82 (1960) 4557;
(d) A. Arrais, E. Diana, G. Gervasio, R. Gobetto, D. Marabello,
P.L. Stranghellini, Eur. J. Inorg. Chem. (2004) 1505;
¨
(e) B. Deubzer, H.P. Fritz, C.G. Kreiter, K. Ofele, J. Organomet.
Chem. 7 (1967) 289.
[16] (a) G. Schmitt, W. Keim, J. Fleischhauer, U. Walbergs, J.
Organomet. Chem. 152 (1978) 315;
(b) M. Lacoste, H. Rabaa, D. Astruc, A. Le Beuze, J.-Y. Saillard,
´
G. Precigoux, C. Courseille, N. Ardoin, W. Bowyer, Organomet-
allics 8 (1989) 2233;
(c) H. Rabaa, M. Lacoste, M.-H. Delville-Desboise, J. Ruiz, B.
Gloaguen, N. Ardoin, D. Astruc, A. Le Beuze, J.-Y. Saillard, J.
Linares, F. Varret, J.-M. Dance, E. Marquestaut, Organometallics
14 (1995) 5078;
(d) I. Chaves, A. Cisternas, M. Otero, E. Roman, U. Mueller, Z.
¨
Naturforsch. 45b (1990) 658;
[9] (a) A.J. Deeming, in: G. Wilkinson, F.G.A. Stone, E.W. Abel
(Eds.), Comprehensive Organometallic Chemistry, vol. 4, Perg-
amon Press, Oxford, UK, 1982, p. 377;
(e) S. Suravajjala, J.R. Polam, L.C. Porter, Organometallics 13
(1994) 37.
(b) R.C. Kerber, in: E.W. Abel, F.G.A. Stone, G. Wilkinson
(Eds.), Comprehensive organometallic chemistry II, vol. 7, Perg-
amon Press, Oxford, UK, 1995, p. 101;
[17] (a) A.N. Nesmeyanov, N.A. Volꢀkenau, I.N. Bolesova, Dokl.
Akad. Nauk. SSSR 166 (1966) 607;
(b) A.N. Nesmeyanov, L.I. Denisovich, S.P. Gubin, N.A.
Volꢀkenau, F.I. Sirotkina, I.N. Bolesova, J. Organomet. Chem.
20 (1969) 169;
´
(c) R.G. Sutherland, M. Iqbal, A. Piorko, J. Organomet. Chem.
302 (1986) 307.
[10] (a) D.P. Tate, W.R. Knipple, J.M. Augl, Inorg. Chem. 1 (1962)
433;
(c) W.H. Morrison, E.Y. Ho, D.N. Hendrickson, J. Am. Chem.
Soc. 96 (1974) 3603;
(b) H. Werner, K. Deckelmann, U. Scho¨nenberger, Helv. Chim.
Acta 53 (1970) 2002;
(d) W.H. Morrison, E.Y. Ho, D.N. Hendrickson, Inorg. Chem.
14 (1975) 500;
(c) F.A. Cotton, J.A. McCleverty, J.E. White, Inorg. Synth. 28
(1990) 45;
(e) C.C. Lee, K.J. Demchuk, W.J. Pannekoek, R.G. Sutherland,
J. Organomet. Chem. 162 (1978) 253;
(d) E.W. Abel, M.A. Bennett, R. Burton, G. Wilkinson, J. Chem.
Soc., Abstr. (1958) 4559;
(f) C.C. Lee, K.J. Demchuk, R.G. Sutherland, Can. J. Chem. 57
(1979) 933;
(e) G.J. Kubas, Inorg. Synth. 27 (1990) 1;
(f) P. Hannesschlager, P. Brun, Appl. Organomet. Chem. 14
(2000) 371;
(g) R.G. Sutherland, W.J. Pannekoek, C.C. Lee, Ann. N.Y.
Acad. Sci. 295 (1977) 192;
(h) M. Lacoste, D. Astruc, J. Chem. Soc., Chem. Commun.
(1987) 667;
(g) A.J. Pearson, E. Schoffers, Organometallics 16 (1997) 5365;
(h) C.A.L. Mahaffy, P.L. Pauson, Inorg. Synth. 28 (1990) 136;
(i) R.D. Fischer, Chem. Ber. 93 (1960) 165.
[11] M. Baumgarten, L. Gherghel, M. Wagner, A. Weitz, M. Rabi-
novitz, P.-C. Cheng, L.T. Scott, J. Am. Chem. Soc. 117 (1995)
6254.
(i) E.P. Ku¨ndig, P. Jeger, G. Bernardinelli, Inorg. Chim. Acta 357
(2004) 1909.
[18] (a) R.M. Chin, M.S. Jarosh, J.D. Russell, R.J. Lachicotte,
Organometallics 21 (2002) 2027;
(b) R.M. Chin, B. Baird, M. Jarosh, S. Rassman, B. Barry, W.D.
Jones, Organometallics 22 (2003) 4829.
[12] (a) R.G. Hayter, J. Am. Chem. Soc. 88 (1966) 4376;
(b) R.M. Wing, D.C. Crocker, Inorg. Chem. 6 (1967) 289;
(c) W.C. Kaska, J. Am. Chem. Soc. 90 (1968) 6340.
[13] (a) V.S. Leong, N.J. Cooper, Organometallics 7 (1988) 2058;
(b) V.S. Leong, N.J. Cooper, J. Am. Chem. Soc. 110 (1988)
2644.
[19] S.L. Mukerjee, R.F. Lang, T. Ju, G. Kiss, C.D. Hoff, Inorg.
Chem. 31 (1992) 4885.
[20] (a) T.G. Traylor, K.J. Stewart, M.J. Goldberg, J. Am. Chem.
Soc. 106 (1984) 4445;
(b) T.G. Traylor, M.J. Goldberg, Organometallics 6 (1987) 2413.