Alkylation of dihydrofullerenes

Article abstract of DOI:10.1021/jo020216e

The fulleride dianions C₆₀^2- and C₇₀^2- were generated by deprotonation of the corresponding hydrogenated fullerenes, 1,2-C₆₀H₂ and 1,2-C₇₀H₂. These anions were prepare

Full text of DOI:10.1021/jo020216e

Alk yla tion of Dih yd r ofu ller en es
Mark S. Meier,*,† Robert G. Bergosh,† Megan E. Gallagher,† H. Peter Spielmann,‡ and
Zhongwen Wang†
Center for Advanced Carbon Materials and Department of Chemistry, and Department of Molecular and
Cellular Biochemistry, University of Kentucky, Lexington, Kentucky 40506-0055
meier@uky.edu
Received March 29, 2002
2-
2-
The fulleride dianions C60 and C70 were generated by deprotonation of the corresponding
hydrogenated fullerenes, 1,2-C60H2 and 1,2-C70H2. These anions were prepared in the presence of
a variety of alkylating agents, and mono- or dialkylated products were obtained. Alkylation was
not successful with sulfonate ester alkylating agents. Deprotonation of monoalkylated compounds,
followed by second alkylation with a different alkylating agent, produced heterodialkylated
compounds. The monoalkyated material was invariably the 1,2-isomers, while the dialkylated
materials were generally 1,4-isomers, although some 1,2-isomer was observed in the C70 context.
The major product from alkylation of C702- was the 7,23-isomer 13a , a structure where the alkylation
took place near the equator of the fullerene cage, rather than at the more strained carbons near
the poles.
5
In tr od u ction
and C70H10
,
additional C70Hn species,10-12 as well as
other, more highly hydrogenated fullerene species.13,14
These hydrogenated fullerenes offer a versatile entry
point for further fullerene derivatization. Fullerene C-H
bonds are quite acidic: the pKa values of C60H2 are
One of the pillars of organic synthesis is the regiospe-
cific deprotonation of carbon and subsequent alkylation
of the resulting anions. The alkylation of fullerene anions
presents the possibility of reversing the normal reactivity
of a fullerene from electrophile to nucleophile. This
reversal in reactivity could in principle provide an
entrance to alkylated products with regiochemistry not
directly obtainable by current methods. Currently,
fullerene anions are most often provided by the reduction
of fullerenes, but deprotonation of hydrogenated fullerenes
is another potentially useful route. Different isomeric
hydrogenated fullerenes offer the possibility of routes to
different isomeric alkylated fullerene derivatives, some-
thing that is not possible through simple reduction
followed by alkylation.
remarkably low for a hydrocarbon (pKa1 ) 4.7, pKa2
)
16).15 These data suggest that fullerene anions (fullerides)
should be readily available by deprotonation of hydro-
genated fullerenes. In principle, it should be possible to
6-
generate C60 hexaanion by completely deprotonating
C60H6 or to make any of a series of C60Hn
the appropriate choice of starting compound and then
controlling the amount of base added. C60Hn
m-
species by
m-
species
differ from C60m- ions in that sp3 centers are present, and
these tetrahedral centers may lead to an increase in
charge and in nucleophilicity at a limited set of carbon
atoms. If this occurs, it would offer a degree of control
over the range of regioisomers formed by alkylation.
Successful implementation of this deprotonation-
alkylation strategy in fullerenes offers the possibility to
selectively control the functionalization and synthetic
elaboration of these molecules. The potential exists for
We have been investigating the synthesis and reactiv-
ity of hydrogenated fullerenes.1-6 We and others have
developed synthetic methods for the preparation of single
isomers of C60H2, C60H4, C60H6,1,7-9 C70H2, C70H4, C70H8,6
† Center for Advanced Carbon Materials and Department of
Chemistry.
‡ Department of Molecular and Cellular Biochemistry.
(1) Bergosh, R. G.; Laske Cooke, J . A.; Meier, M. S.; Spielmann, H.
P.; Weedon, B. R. J . Org. Chem. 1997, 62, 7667-7672.
(2) Guarr, T. F.; Meier, M. S.; Vance, V. K.; Clayton, M. J . Am.
Chem. Soc. 1993, 115, 9862-9863.
(8) Meier, M. S.; Weedon, B. R.; Spielmann, H. P. J . Am. Chem. Soc.
1996, 118, 11682-11683.
(9) Meier, M. S.; Laske Cooke, J . A.; Weedon, B. R.; Spielmann, H.
P. Recent Adv. Chem. Phys. Fullerenes Relat. Mater. 1996, 1193-1199.
(10) Henderson, C. C.; Cahill, P. A. Science 1993, 259, 1885-1887.
(11) Henderson, C. C.; Rohlfing, C. M.; Assink, R. A.; Cahill, P. A.
Angew. Chem., Int. Ed. Engl. 1994, 33, 786-788.
(3) Meier, M. S.; Vance, V. K.; Corbin, P. K.; Clayton, M.; Mollman,
M.; Poplawska, M. Tetrahedron Lett. 1994, 35, 5789-5792.
(4) Weedon, B. R.; Haddon, R. C.; Meier, M. S.; Spielmann, H. P. J .
Am. Chem. Soc. 1999, 335, 5-340.
(12) Henderson, C. C.; Rohlfing, C. M.; Gillen, K. T.; Cahill, P. A.
Science 1994, 264, 397-399.
(5) Spielmann, H. P.; Weedon, B. R.; Meier, M. S. J . Org. Chem.
2000, 65, 2755-2758.
(13) Nossal, J .; Saini, R. K.; Alemany, L. B.; Meier, M. S.; Billups,
W. E. Eur. J . Chem. 2001, 4167-4180.
(6) Spielmann, H. P.; Wang, G.-W.; Meier, M. S.; Weedon, B. R. J .
Org. Chem. 1998, 9865, 5-9871.
(14) For other methods for the preparation of reduced fullerenes,
see: Goldshleger, N. F.; Moravshii, A. P. Russ. Chem. Rev. 1997, 66,
323-342.
(7) Bergosh, R. G.; Meier, M. S.; Spielmann, H. P.; Wang, G.-W.;
Weedon, B. R. Recent Adv. Chem. Phys. Fullerenes Relat. Mater. 1997,
4, 240-245.
(15) Niyazymbetov, M. E.; Evans, D. H.; Lerke, S. A.; Cahill, P. A.;
Henderson, C. C. J . Phys. Chem. 1994, 98, 13093-13098.
10.1021/jo020216e CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/13/2002
5946
J . Org. Chem. 2002, 67, 5946-5952

Alkylation of Dihydrofullerenes
SCHEME 1
sequential deprotonation-alkylation steps in which the
regiochemistry of the resulting alkylated species reflects
the regiochemistry of the starting reduced fullerene. For
example, if the starting regiochemistry can be manifest
in the product, then regiochemically pure R1R2R3R4R5R6C60
species could be made from 1,2,33,41,42,50-C60H6. Our
first work toward this goal is reported here.
Preparation of fulleride anions by deprotonation of
fullerene C-H bonds is a complementary route to forma-
tion of these anions by direct reduction of C60 or C70.
Solutions of fulleride anions have been produced tradi-
tionally by electrochemical or by dissolving-metal reduc-
tion of the parent fullerene.16-20 Alternatively, reduction
of C60 in THF solution with thiolates21,22 or with quinone
dianions can achieve the same goal23 or by a number of
different alkali metal reductions.24-27 Alkylation of
fullerene ions prepared electrochemically28-30 and by
dissolving-metal reduction has been achieved, although
a mixture of products was obtained.31 Alkylation of
fullerene anions produced by quinone dianion and by
thiolates has been successful as well.22,23
SCHEME 2
ating agents should then provide the desired mono- and/
or dialkyl fullerenes (Scheme 1).
Resu lts a n d Discu ssion
Treatment of 1 with base (BuLi, LiHMDS, NaH, DBU,
K2CO3, or KOtBu) in THF, toluene, benzene, methanol-
THF, or o-dichlorobenzene solution containing a signifi-
cant excess of benzyl bromide produced little if any mono-
or dialkyl fullerene. The majority of the fullerene-derived
product was an uncharacterized brown material that was
insoluble in toluene but soluble in THF.
Mon oa lk yla tion . Given the acidity of hydrogenated
fullerenes, double deprotonation of 1,2-C60H2 (1) 1,10,32-35
to produce C602- should be possible with very mild bases.
Treatment of the resulting dianion with suitable alkyl-
(16) Dubois, D.; Kadish, K. M.; Flanagan, S.; Wilson, L. J . J . Am.
Chem. Soc. 1991, 113, 7773-7774.
(17) Greaney, M. A.; Gorun, S. M. J . Phys. Chem. 1991, 95, 7142-
A survey of additional common bases did not provide
a solution to this problem. However, we found that the
7144.
(18) Wei, X.; Suo, Z.; Zhou, K.; Xu, Z.; Zhang, W.; Wang, P.; Shen,
H.; X., L. J . Chem. Soc., Perkin Trans. 2 1999, 121-126.
(19) Wei, X.; Suo, Z.; Yin, G.; Xu, Z. J . Organomet. Chem. 2000, 599,
69-73.
2-
C60 dianion can be generated in benzonitrile solution
by treating 1 with tetrabutylammonium hydroxide. The
characteristic deep red color of the dianion forms very
rapidly, but in the presence of excess diphenylmethyl
bromide, the dark red color is rapidly replaced by a deep
green color, typical of an RC60 anion. Immediately
quenching the reaction mixture with acetic acid produces
a mixture of Ph2CHC60H (2) and C60 (Scheme 2). No
unreacted 1 was detected under these conditions.
HPLC purification of this crude reaction mixture on a
Cosmosil Buckyprep column (100% toluene mobile phase)
provided clean samples of 2 in 60% yield. The absorption
spectrum of 2 is virtually superimposable on the spec-
trum of 1 (Figure 1), indicating that the two sp3 carbons
of 2 are arranged in a 1,2 fashion, an addition pattern
common among fullerene derivatives.
Scrupulous deoxygenation is required for successful
alkylation. The best yields (typically 60% isolated yield)
were obtained in reactions where freeze-pump-thaw
(FPT) deoxygenation was used to remove oxygen from
the reaction mixture. Even under these conditions,
roughly 30% yields (by HPLC) of C60 were obtained in
every example. This undesired oxidation could not be
suppressed, even with FPT deoxygenation of the reaction
mixture and the base solution and carrying out the
reaction in an inert-atmosphere glovebox.
(20) For a review of fulleride anions, see: Reed, C. A.; Bolskar, R.
D. Chem. Rev. 2000, 100, 1075-1119.
(21) Subramanian, R.; Boulas, P.; Vijayashree, M. N.; D’Souza, F.;
J ones, M. T.; Kadish, K. M. J . Chem. Soc. Chem. Commun. 1994,
1847-1848.
(22) Allard, E.; Riviere, L.; Delaunay, J .; DuBois, D.; Cousseau, J .
Tetrahedron Lett. 1999, 40, 7223-7226.
(23) (a) Subramanian, R.; Kadish, K. M.; Vijayashree, M. N.; Gao,
X.; J ones, M. T.; Miller, M. D.; Krause, K. L.; Suenobu, T.; Fukuzumi,
S. J . Phys. Chem. 1996, 100, 16327-16335. (c) Kusukawa, T.; Ando,
W. Angew. Chem., Int. Ed. Engl. 1996, 35, 1315-1317. (d) Sawamura,
M.; Toganoh, M.; Suzuki, K.; Hirai, A.; Iikura, H.; Nakamura, E.
Organic Lett. 2000, 2, 1919-1921. (d) Allard, E.; Delaunay, J .; Cheng,
F.; Cousseau, J .; Ordu´na, J .; Gar´ın, J . Org. Lett. 2001, 3, 3503-3506.
(24) Boyd, P. D. W.; Bhyrappa, P.; Paul, P.; Stinchcombe, J .; Bolshar,
R. D.; Sun, Y.; Reed, C. A. J . Am. Chem. Soc. 1995, 117, 2907-2914.
(25) Chen, J .; Cai, R.-F.; Huang, Z.-E.; Wu, H.-M.; J iang, S.-K.; Shao,
Q.-F. J . Chem. Soc. Chem. Commun. 1995, 1553-1554.
(26) Sun, Y.; Reed, C. A. J . Chem. Soc. Chem. Commun. 1997, 747-
748.
(27) Paul, P.; Xie, A.; Bau, R.; Boyd, P. D. W.; Reed, C. A. J . Am.
Chem. Soc. 1994, 116, 4145-4146.
(28) Kadish, K. M.; Gao, X.; Caemelbecke, E. V.; Hirasaka, T.;
Suenobbu, T.; Fukuzumi, S. J . Phys. Chem. A 1998, 102, 3898-3906.
(29) Caron, C.; Subramanian, R.; D’Souza, F.; Kim, J .; Kutner, W.;
J ones, M. T.; Kadish, K. M. J . Am. Chem. Soc. 1993, 115, 8505-8506.
(30) Fukuzumi, S.; Suenobu, T.; Hirasaka, T.; Arakawa, R.; Kadish,
K. M. J . Am. Chem. Soc. 1998, 120, 9220-9227.
(31) Bausch, J . W.; Surya Prakash, G. K.; Olah, G.; Tse, D. S.;
Lorents, D. C.; Bae, Y. K.; Malhotra, R. J . Am. Chem. Soc. 1991, 113,
3205-3206.
(32) Becker, L.; Evans, T. P.; Bada, J . L. J . Org. Chem. 1993, 58,
7630-7631.
Deoxygenation of the base solution by FTP did not
result in better yields than did deoxygenation by simple
sparging. Since increasingly stringent deoxygenation
methods do not affect the amount of C60 formed, we
(33) Ballenweg, S.; Gleiter, R.; Kra¨tschmer, W. Tetrahedron Lett.
1993, 34, 3737-3740.
(34) Fukuzumi, S.; Suenobu, T.; Kawamura, S.; Shida, A.; Mikami,
K. J . Chem. Soc. Chem. Commun. 1997, 291-292.
(35) Morosin, B.; Henderson, C.; Schirber, J . E. Appl. Phys. A 1994,
59, 178-180.
2-
believe that the roughly 30% of the C60 dianion is lost
J . Org. Chem, Vol. 67, No. 17, 2002 5947

Meier et al.
(producing 6), occurred so rapidly that dialkyl products
dominated the reaction mixture within minutes of the
addition of base.
Heter od ia lk yla tion . Much of the potential of the
deprotonation-alkylation route to fullerene derivatives
lies in the ability to sequentially replace several protons
with different alkyl groups. To explore this possibility,
we treated 2 with TBAOH in THF solution containing
excess MeOTs, leading to a deep green solution charac-
-
teristic of the RC60 anion, but after 24 h there was no
detectable formation of CH3C60(CHPh2). The anion slowly
decomposed to a brownish material that was nearly
insoluble in toluene. These results are consistent with
2-
our experience attempting to alkylate C60 with sul-
fonate esters. However, treatment of 2 in PhCN solution
with TBAOH and allyl bromide produces CH2dCHCH2C60-
(CHPh2) (8). The UV-vis spectrum of this compound is
1
that of a 1,4-dialkylated product, and the H NMR and
13C NMR are consistent with the expected heterodialkyl-
ated product. Similarly, we have successfully produced
the heterodialkylation product with benzyl bromide as
the alkylating agent, producing 1,4-PhCH2C60(CHPh2)
(9). 1,2-Dialkylated products were not observed in these
reactions.
F IGURE 1. Absorption spectra of 2 and 1,2-C60H2.
to oxidation through electron transfer to the alkyl
halide,36 followed by radical reactions that do not lead to
alkylated fullerenes. We have attempted to detect 1,2-
diphenylethane from the alkylation of C60H2 with benzyl
bromide, but were not able to do so definitively.
Using the same conditions for alkylation of 1 with Ph2-
CHBr, we have been able to monoalkylate C60H2 with
several other alkyl halides. However, not all alkyl halides
will alkylate the C602- anion. Specifically, alkylation was
achieved with Ph2CHBr (2, 60%), propargyl bromide (3,
21%), allyl bromide (4, 54%), and PhCH2Br (5, 38% by
Alk yla tion of C70 An ion s Der ived fr om 1,2-C70H2.
The alkylation of C70 anions derived from 1,2-C70H2 (10),
as well as other C70Hn species, poses an interesting
regiochemical question. In C702- the charge is delocalized
over five symmetrically unique positions,38 any one of
which may react in the first alkylation step.
HPLC). In the latter case, dialkylation was very rapid
(see below) and made isolation of 5 very difficult. No
alkylation was detected when octyl bromide, octyl iodide,
or cyclopropylmethyl bromide was used. All attempts to
2-
alkylate C60 with sulfonate esters were unsuccessful.
These results are consistent with a radical mechanism.23
Hom od ia lk yla tion . If these reactions were allowed
to proceed for several hours, dialkylated products were
isolated. For example, stirring a deoxygenated mixture
of 1, TBAOH, and excess benzyl bromide produced a 45%
2-
isolated yield of 6. Dibenzylation of C60 generated by
p-benzoquinone dianion reduction of C60 is known to
produce 1,4-disubstitution.23 The 1,4-regiochemistry is
established by comparison with literature spectra.37 In
addition, we have successfully produced 1,4-dialkylated
derivative 7 in 49% yield, using allyl bromide. Interest-
ingly, reaction with PhCH2Br, the second alkylation step
In practice, deprotonation of 10 (prepared by Zn(Cu)
reduction of C706) in the presence of benzyl bromide, leads
to similar color changes as seen with 1, although com-
plete alkylation requires significantly more reaction time
than in the C60 case.
(36) D’Souza, F.; Choi, J .-P.; Kutner, W. J . Phys. Chem. B 1998,
102, 4247-4252.
(37) Kadish, K. M.; Gao, X.; Caemelbecke, E. V.; Suenobu, T.;
Fukuzumi, S. J . Phys. Chem. A 2000, 104, 3878-3883.
(38) Roduner, E.; Reid, I. D. Chem. Phys. Lett. 1994, 223, 149-154.
5948 J . Org. Chem., Vol. 67, No. 17, 2002

Alkylation of Dihydrofullerenes
HPLC analysis of the crude reaction mixture reveals
a set of three bands containing dialkylated products (by
mass spectrometry). The first and third bands contained
single compounds 13 (10% isolated yield) and 12 (1%
isolated yield), while the middle band contained several
species.
The middle band resolved into two fractions upon
passage through a Cosmosil Buckyprep column (100%
toluene mobile phase). Examination of the 1H NMR
spectrum of the first fraction (10% isolated yield) revealed
an AB pattern and a singlet in the benzyl methylene
region. The 13C NMR spectrum showed two fullerene sp3
resonances, two benzylic resonances, and more than 60
sp2 resonances. This combination of data indicates that
this band is a mixture of two compounds, both of which
have significant symmetry. We were unable to resolve
this fraction into its individual components.
The second fraction of the middle band was composed
of 11. The 1H NMR spectrum of 11 consisted of two
singlets for the benzyl methylene resonances, indicating
two inequivalent benzyl groups. The absence of an AB
pattern suggests that the two benzyl groups are on a
plane of symmetry. The 13C NMR spectrum showed two
fullerene sp3 resonances and 31 fullerene sp3 resonances,
several of which having double intensity (indicative of
overlapping resonances). These data are consistent with
a plane of symmetry containing two different benzyl
groups. In addition, the absorption spectrum of this
compound is very similar to that of 10 (see Figure 2).
These data suggest that the structure of this minor
isomer (1% yield) is 1,2-(PhCH2)2C70 (11). While this is a
sterically crowded arrangement, 1,2-(PhCH2)2C60 has
been previously prepared in low yield.37
F IGURE 2. Comparison of absorption spectra of two different
1,2-C70R2 species: 10 and 11.
A number of structures with C2-symmetry are possible.
Two representative examples are shown below. Structure
13a was proposed earlier for the Bn2C70 species prepared
by alkylation39 of electrochemically generated C702-. This
assignment was based, in turn, on calculations by Cahill
on the thermodynamically most stable isomers of C70H2
with nonadjacent hydrogens.40 Given the fact that this
reaction is probably kinetically controlled rather than
thermodynamically controlled, we used 13C INADEQUATE
NMR methods to definitively answer this structural
question.
The 1H NMR spectrum of 12 exhibited a single AB
pattern for the benzyl methylene protons. The 13C NMR
spectrum possessed one fullerene sp3 resonance, one
benzylic resonance, four phenyl resonances, and at least
32 fullerene sp2 resonances (some overlapping). These
data suggest that 12 has either a plane of symmetry or
a C2 axis of symmetry. The limited amount of material
available made it impossible to definitively assign the
structure of this minor product.
Examination of the 1H NMR spectrum of the first
fraction, containing 13, revealed a single AB pattern for
the benzyl methylene resonances, indicating a structure
with either a plane of symmetry or a C2 axis of symmetry
and hindered rotation of the benzyl groups. The 13C NMR
spectrum exhibited a single benzyl resonance and a single
fullerene sp3 resonance. The sp2 region included four
resonances from the phenyl CH carbons (three with
NOEs from attached hydrogens) and 34 single intensity
fullerene resonances. This confirms the C2 symmetry of
this isomer.
A sample of 13% 13C-enriched 10 was prepared, puri-
fied, and subjected to alkylation with PhCH2Br. INAD-
EQUATE NMR experiments revealed 46 of 52 expected
correlations. The connectivity exposed by these correla-
tions allowed us to make a definitive structural assign-
ment. The key observation from the INADEQUATE
spectrum is that there are two sp2 carbons that are
correlated (adjacent) to each other, and both are cor-
related to the sp3 carbon (Figure 3). Only one C2 structure
meets these requirements: 13a . The connectivity re-
vealed by the rest of the correlations is also consistent
with this structure.
(39) Kadish, K. M.; Gao, X.; Gorelik, O.; Van Caemelbecke, E.;
Suenobu, T.; Fukuzumi, S. J . Phys. Chem. A 2000, 104, 2902-2907.
(40) Henderson, C. C.; Rohlfing, C. M.; Cahill, R. A. Chem. Phys.
Lett. 1993, 213, 383-388.
J . Org. Chem, Vol. 67, No. 17, 2002 5949

Meier et al.
needlestock techniques. Unless otherwise noted, yields refer
to isolated, HPLC-purified material.
Gen er a l P r oced u r e for Alk yla tion of Dih yd r ofu ller -
en es: 1-Diph en ylm eth yl-1,2-dih ydr o[60]fu ller en e (2).C60H2
(63.4 mg, 0.088 mmol), diphenylmethyl bromide (2.36 g, 9.57
mmol), and PhCN (31 mL) were combined in a 100 mL Schlenk
flask. This mixture was subjected to 10 freeze-pump-thaw
(FPT) cycles. The solution was warmed to room temperature
and put under an Ar atmosphere. 3 mL of TBAOH (1 M in
MeOH, deoxygenated by sparging with Ar) was quickly added
and the reaction mixture turned deep red then quickly turned
deep green. The reaction was quenched immediately with 2
mL of acetic acid, resulting in a brown solution. This solution
was filtered and passed through a short silica plug to remove
tetrabutylammonium salts. The PhCN was removed by short-
path vacuum distillation. The resulting brown solid was
dissolved in CS2 and passed through another silica plug to
remove the diphenylmethyl bromide starting material. This
solution was diluted with toluene and concentrated in vacuo.
The product was purified by HPLC (Cosmosil semipreparative
Buckyprep column, toluene mobile phase, 5 mL/min flow rate,
monitored at 400 nm), producing 46.6 mg (0.052 mmol, 60%
yield): MS (negative ion FABS) m/z 888; 1H NMR (200 MHz,
CS2/acetone-d6) δ ) 6.04 (s, 1 H), 6.78 (s, 1 H), 7.37 (m, 2 H),
7.49 (m, 4 H), 8.08 (m, 4 H); 13C NMR (100 MHz, CS2/acetone-
d6) δ ) 60.25, 66.81, 70.68, 128.31, 129.47, 131.07, 136.75,
136.9, 139.55, 140.66, 140.76, 141.97, 142.05, 142.19, 142.51,
142.57, 142.63, 143.03, 143.08, 143.69, 144.96, 145.21, 145.82,
145.83, 145.97, 146.25, 146.58, 146.70, 146.85, 147.38, 147.65,
147.87, 155.00, 155.13.
F IGURE 3. A portion of the structure of 13a as determined
by INADEQUATE NMR spectroscopy. Numbers in bold on the
left side are chemical shifts, numbers in plain font on the right
are 13C-13C coupling constants, and the numbers on the
vertexes denote equivalent carbons.
The preference for alkylation of C702- near the equator
indicates high reactivity in carbons that are typically the
least reactive in the neutral fullerene. While the charge
is significantly delocalized in C702-, AM1 calculations41
of the SOMO (which is likely to be involved in the
irreversible, C-C bond forming step) suggest that this
key orbital has the highest density at the very positions
(C7/C23, which are equivalent by symmetry) where we
find the benzyl groups attached. After the first alkylation
step occurs at C7, the resulting PhCH2C70- species shows
significant charge concentration at C23, the other carbon
to which we find benzyls attached. In contrast to the
1-P r op a r gyl-1,2-d ih yd r o[60]fu ller en e (3): MS (positive
ion MALDI-MS), m/z 760; 1H NMR (200 MHz, CS2/acetone) δ
) 2.86 (t, 1H), 4.41 (d, 2H), 6.81 (s, 1H); 13C NMR (75 MHz,
CS2/acetone) δ ) 40.50, 62.73, 66.98, 77.74, 83.34, 139.95,
139.97, 143.47, 143.60, 145.03, 145.10, 145.38, 145.50, 145.71,
145.90, 145.94, 146.57, 147.93, 148.01, 148.69, 148.75, 148.80,
149.10, 149.19, 149.50, 149.58, 149.70, 149.78, 150.42, 150.60,
150.80, 157.31, 157.66; HR MALDI (9-nitroanthracene matrix)
760.027, calcd for C63H4 760.031.
-
situation found for RC60 anions, the charge density is
not as high at the carbons adjacent to the alkylated
carbon.
Con clu sion s
1-Allyl-1,2-d ih yd r o[60]fu ller en e (4): MS (negative ion
FABS-MS), m/z 762; 1H NMR (200 MHz, CS2/benzene) δ ) 3.99
(d, 2h), 5.67 (m, 2H), 6.33 (s, 1H), 6.68 (m, 1H); 13C NMR (100
MHz, CS2/benzene) δ ) 51.36, 58.90, 64.55, 121.84, 133.64,
136.40, 136.64, 140.56, 140.59, 141.96, 142.23, 142.27, 142.34,
142.51, 142.89, 143.57, 144.88, 145.02, 145.68, 145.70, 145.77,
145.85, 146.08, 146.41, 146.49, 146.54, 146.70, 146.72, 147.27,
147.58, 147.73, 154.03, 155.61.
1-Ben zyl-1,2-d ih yd r o[60]fu ller en e (5): MS (negative ion
FABS-MS), m/z 812; 1H NMR (400 MHz, CS2/CDCl3) δ ) 4.76
(s, 2H), 6.62 (s, 1H), 7.41 (m, 1H), 7.50 (m, 2H), 7.90 (m, 2H);
13C NMR (100 MHz, CS2/CDCl3) δ ) 52.50, 59.09, 66.0, 127.80,
127.72, 128.75, 131.50, 132.05, 135.80, 135.95, 136.25, 139.90,
140.05, 141.50, 142.62, 141.81, 141.66, 141.92, 142.18, 142.40,
143.13, 144.42, 144.60, 145.33, 145.40, 145.81, 146.05, 146.08,
146.20, 146.95, 147.20, 153.90, 155.30.
We have demonstrated that it is possible to use reduced
fullerenes as precursors to fulleride anions under syn-
thetically useful conditions. Yields for monoalkylation are
typically near 60% with some loss of dianion to oxidation
to the parent fullerene. Alkyl halides that can easily
produce radicals upon electron transfer are the only
halides that are useful in this reaction. Aliphatic chlo-
2-
rides and all sulfonates tested do not react with C60
produced under these conditions. A second alkylation step
does occur within a few minutes to a few hours of reaction
time. Both the first and the second alkylation steps
appear to be much more efficient with alkyl halides than
with alkyl sulfonates. C70H2 was deprotonated and alkyl-
ated under similar conditions, and while a mixture of
products was obtained, it appears that the most nucleo-
philic site is C7. The resulting anion (PhCH2C70-) alkyl-
ates at C23, producing a 1,4-dialkylated compound whose
structure was assigned by 13C INADEQUATE experi-
ments. Rather than addition to the strained carbons at
the poles of the molecule, addition occurs to positions
near the equator, offering a route to novel C70 derivatives.
Gen er a l P r oced u r e for Hom od ia lk yla tion of Dih yd r o-
fu ller en es: 1,4-Diben zyl-1,4-dih ydr o[60]fu ller en e (6). C60H2
(1) (50.1 mg, 0.069 mmol), benzyl bromide (1.0 mL, 4.07 mmol),
and 25 mL of PhCN were combined in a 100 mL Schlenk flask.
The mixture was subjected to 10 FPT cycles, warmed to room
temperature, and put under an Ar atmosphere. Then 3.5 mL
of TBAOH (1M in MeOH, sparged with Ar) was quickly added
and the reaction mixture turned deep red and immediately
deep green. The reaction was allowed to stir for 2 h and slowly
turned to a brown solution. HCl (10%, 2 mL) was added, and
the PhCN was removed using a short-path vacuum distillation.
The resulting brown solid was dissolved in toluene and washed
with 10% HCl. The product was purified by HPLC [Cosmosil
semipreparative Buckyprep column, toluene/hexane (70:30)
mobile phase, 5 mL/min flow rate, and monitored at 400 nm],
producing 28.1 mg (0.0311 mmol, 45%): MS (negative ion
Exp er im en ta l Section
General experimental details are available in previous
publications.1 Representative procedures for the preparation
of mono- and dialkylfullerenes are given below. All reactions
were carried out under Ar or N2 atmospheres using standard
(41) The SOMO was calculated using MacSpartan Pro, Wave-
function, Inc.
1
FABS), m/z 903 (MH); H NMR (200 MHz, CDCl3) δ ) 3.74
5950 J . Org. Chem., Vol. 67, No. 17, 2002

Alkylation of Dihydrofullerenes
(qAB, 4H), 7.34 (m, 2H), 7.46 (m, 4H), 7.56 (m, 4H); 13C NMR
(100 MHz, CDCl3) δ ) 48.42, 50.32, 127.38, 128.24, 130.93,
136.09, 137.60, 138.58, 140.31, 141.79, 141.81, 142.27, 142.34,
142.44, 142.77, 142.89, 142.98, 143.52, 143.73, 143.92, 144.04,
144.08, 144.16, 144.47, 144.52, 144.58, 144.80, 145.29, 145.97,
146.69, 146.71, 146.77, 146.98, 148.43, 151.51, 157.74; HR
MALDI (9-nitroanthracene matrix) 902.103, calcd for C74H14
902.110.
1,4-Dia llyl-1,4-d ih yd r o[60]fu ller en e (7): MS (negative ion
FABS-MS), m/z 802; 1H NMR (200 MHz, CS2/acetone) δ ) 3.90
(m, 4H), 5.61 (m, 4H), 6.68 (m, 2H); 13C NMR (100 MHz, CS2/
acetone) δ ) 47.82, 59.75, 120.94, 133.83, 139.02, 139.63,
141.39, 142.67, 142.73, 143.16, 143.25, 143.33, 143.65, 143.83,
144.44, 144.62, 144.78, 144.93, 144.98, 145.08, 145.44, 145.55,
145.80, 146.13, 146.69, 147.54, 147.59, 147.85, 148.06, 149.25,
149.38, 152.26, 158.12.
toluene. The toluene was then removed in vacuo to minimize
solution quantity. Short-path vacuum distillation was then
used to remove the benzonitrile. The resulting brown solid was
dissolved in toluene and concentrated. This solution was then
purified by two different HPLC columns: first by preparative
HPLC [4.6 × 250 mm Cosmosil 5PBB column, toluene:hexane
(70:30) as mobile phase, monitored at 310 nm] to obtain three
fractions. Fraction 1 contained 13a and fraction 3 contained
the minor product 12. Fraction 2 was further purified by
preparative HPLC (4.6 × 250 mm Buckyprep column, toluene
as mobile phase, monitored at 310 nm) and was resolved into
two chromatographic bands. The first band was determined
to be an inseparable mixture of two isomers, and the second
band was composed of a single compound (11). The yield of
the four resolved components (13a , 11, and 12) was 10%:1%:
1%, with an additional 10% yield of the unresolved pair of
isomers.
Gen er al P r ocedu r e for Heter odialkylation of Dih ydr o-
fu ller en es: (Diph en ylm eth yl)(allyl)-1,4-dih ydr o[60]fu ller -
en e (8). HC60(CHPh2) (46.6 mg, 0.053 mmol), allyl bromide (3
mL), and PhCN (23 mL) were combined in a 100 mL Schlenk
flask. This mixture was subjected to 10 freeze-pump-thaw
(FPT) cycles. The solution was warmed to room temperature
and put under an Ar atmosphere. Then 4.5 mL of TBAOH (1M
in MeOH, sparged with Ar) was quickly added and the reaction
mixture turned deep green and slowly became brown. The
reaction was stirred overnight and quenched with 4.5 mL of
AcOH. This solution was filtered and passed through a short
silica plug to remove tetrabutylammonium salts. The PhCN
was removed by short-path vacuum distillation. This solution
was diluted with toluene and concentrated in vacuo. The
product was purified by HPLC [Cosmosil semipreparative
Buckyprep column, toluene/hexane (70:30) mobile phase, 5 mL/
min flow rate, monitored at 400 nm], producing 14.2 mg (0.015
1,2-Diben zyld ih yd r o[70]fu ller en e (11): MS (negative ion
FABS-MS, ca. 13% 13C-enriched), m/z 1032; 1H NMR (400
MHz, CS2/CDCl3) δ ) 4.15 (s, 2H), 4.34 (s, 2H), 7.10-7.64 (m,
10H); 13C NMR (150 MHz, CS2/CDCl3) δ ) 60.22, 60.97, 130.85,
130.89, 130.93, 131.69, 133.54, 135.81, 139.03, 139.58, 142.63,
145.48, 145.86, 145.95, 146.00, 146.52, 146.59, 146.64, 147.13,
148.38, 148.48, 148.61, 148.85, 148.93, 149.26, 149.40, 149.48,
149.95, 150.28, 150.92, 151.02, 154.50, 159.99.
Diben zyld ih yd r o[70]fu ller en e (12): MS (negative ion
FABS-MS, ca. 13% 13C-enriched), m/z 1033; 1H NMR (400
MHz, CDCl3) δ ) 3.04 (qAB, 4H), 7.15-7.45 (m, 10H); 13C NMR
(100 MHz, CDCl3) δ ) 46.29, 57.83, 127.44, 128.03, 128.32,
130.54, 130.77, 131.89, 132.50, 132.85, 134.06, 136.26, 138.81,
140.16, 141.63, 143.91, 144.13, 144.41, 145.05, 145.39, 145.61,
145.77, 146.72, 146.88, 148.29, 148.61, 148.73, 148.94, 149.45,
149.74, 150.02, 150.21, 150.54, 151.54, 152.30, 157.02, 158.21.
7,23-Diben zyld ih yd r o[70]fu ller en e (13a ): MS (MALDI-
MS), m/z 1033; 1H NMR (600 MHz, CDCl3) δ ) 3.66 (qAB, 4H),
7.13-7.37 (m, 10H); 13C NMR (150 MHz, CDCl3) δ ) 49.81,
57.43, 127.52, 128.29, 130.63, 132.24, 133.69, 134.11, 134.22,
135.78, 139.35, 140.12, 141.11, 142.06, 142.50, 142.82, 143.61,
144.33, 144.58, 144.87, 145.21, 145.30, 145.38, 145.91, 146.11,
146.88, 146.99, 147.09, 147.34, 147.44, 148.28, 148.48, 148.65,
148.76, 148.91, 149.71, 150.53, 150.63, 151.18, 158.50.
INADEQUATE NMR Exp er im en ts. Two 13C INADE-
QUATE experiments were performed. Both were performed
on Varian Inova spectrometers at 10 °C, one was performed
at 100 MHz (exp1) and a second at 150 MHz (exp 2).
Experiment 1 (exp 1) used a spectral width of 11 056 Hz and
16 384 real points in ω2, and experiment 2 (exp 2) used a
spectral width of 5422 Hz and 16 384 real points in ω2. Digital
oversampling by a factor of 36 (for exp 1) and 68 (for exp 2)
was employed to reduce baseline distortions associated with
the spectrometer’s low-pass filters.42 All experiments were
recorded using the hypercomplex method of phase incremen-
tation to obtain quadrature phase detection in ω1.43 The
spectral width of ω1 was the same as in ω2, which led to folding
of the spectrum, and 1024 (for exp 1) and 2048 (for exp 2)
complex points were collected, and the total number of
transients recorded were 16 and 32 (for exp 1) and 8 and 16
(for exp 2) for real and complex data, respectively. For all
experiments, the value of ∆ for the double quantum polariza-
tion transfer was set to 4.3 ms, corresponding to a maximum
1
mmol, 29% yield): MS (negative ion FABS-MS), m/z 929; H
NMR (200 MHz, CDCl3) δ ) 2.84 (m, 2H), 5.40 (m, 2H), 5.56
(s, 1H), 6.42 (m, 1H), 7.42 (m, 6H), 7.90 (m, 4H); 13C NMR
(100 MHz, CDCl3) δ ) 59.08, 60.52, 64.49, 65.50, 119.91,
127.71, 127.74, 128.76, 128.77, 130.15, 130.27, 133.23, 134.39,
137.88, 138.70, 138.81, 138.95, 139.92, 139.97, 140.38, 141.39,
141.94, 141.97, 142.21, 142.37, 142.52, 142.56, 142.76, 142.94,
143.05, 143.10, 143.11, 143.15, 143.34, 143.46, 143.70, 143.75,
143.97, 143.99, 144.09, 144.10, 144.19, 144.22, 144.26, 144.33,
144.47, 144.55, 144.58, 144.61, 144.63, 144.88, 144.92, 145.04,
145.42, 145.60, 146.77, 146.84, 146.86, 147.06, 147.07, 147.30,
147.59, 148.44, 148.47, 148.58, 148.72, 151.24, 151.55, 156.48,
157.35.
(Diph en ylm eth yl)(ben zyl)-1,4-dih ydr o[60]fu ller en e (9):
1
54% isolated yield; MS (negative ion FABS-MS), m/z 978; H
NMR (400 MHz, CDCl3) δ ) 3.43 (qAB, 2H), 5.48 (s, 1H), 7.20-
7.55 (m, 14H, some residual solvent present), 7.98 (d, 2H), 8.05
(d, 2H); 13C NMR (100 MHz, CDCl3) δ ) 47.97, 59.94, 64.15,
65.34, 127.05, 127.64, 127.68, 128.02, 128.49, 128.68, 130.00,
130.14, 130.64, 135.46, 137.74, 138.49, 138.60, 138.67, 139.65,
140.20, 141.28, 141.61, 141.75, 142.03, 142.10, 142.33, 142.39,
142.63, 142.70, 142.88, 142.94, 142.96, 143.09, 143.34, 143.53,
143.74, 143.88, 143.93, 143.95, 143.98, 144.06, 144.15, 144.22,
144.35, 144.69, 144.76, 144.79, 145.25, 145.35, 146.58, 146.63,
146.66, 146.70, 146.84, 146.86, 147.29, 147.30, 148.26, 148.37,
148.46, 151.11, 151.40, 156.20, 156.72.
Diben zyld ih yd r o[70]fu ller en es. 13C-enriched C70H2 (59.1
mg, 0.07 mmol, prepared from ca. 13% 13C C70), benzyl bromide
(250 µL, 2 mmol), and benzonitrile (60 mL) were combined in
a 250 mL Schlenk flask and deoxygenated through 10-12 FPT
cycles. TBAOH (1 M in methanol), present in a separate 25
mL Schlenk flask, was also deoxygenated by 15 FPT cycles.
After degassing, the two flasks were placed under an Ar
atmosphere, and 500 µL (1.6 mmol) of TBAOH in MeOH
(sparged with Ar) was then added to the reaction mixture
using a gastight syringe. The solution turned a darker brown
immediately and stirred for 18 h. At 18 h, 1 mL of acetic acid
was added for workup. Ammonium salts were removed by
passing the solution through a silica plug and eluting with
1
magnetization transfer for J CC ) 57 Hz that was the average
of the sp2-sp2 one bond coupling constants. Decoupling of 1H
during the entire experiment was performed using the WALTZ
scheme built into the spectrometer hardware. The total recycle
delay (acquisition time plus relaxation delay) was set to 12.7
s (for exp 1) and 12.6 (for exp 2), which are approximately 0.64
(for exp 1) and 0.63 (for exp 2) multiplied by an average sp2
13C spin-lattice relaxation time of 20 s (measured in C60H2).
(42) Delsuc, M. A.; Lallemand, J . Y. J . Magn. Reson. 1986, 69, 504-
507.
(43) States, D. J .; Haberkorn, R. A.; Ruben, D. J . J . Magn Reson.
1982, 48, 286-292.
J . Org. Chem, Vol. 67, No. 17, 2002 5951

Meier et al.
The 13C T1 relaxation times were measured by recording 11
1D experiments with relaxation delays T of 0.1, 0.5, 1, 2, 3, 4,
6, 11, 18, 28, and 45 s with proton decoupling. Intensities of
the resonance peaks in the NMR spectra were determined as
peak heights. The relaxation rate constants were obtained
from nonlinear fits to I(T) ) I∞ - (I∞ - I0) exp(-R1T), in which
I(T) is the intensity at time T, I∞ is the intensity at time )
infinity, and I0 is the intensity at time ) zero.
part by the MRSEC Program of the National Science
Foundation under Award Number DMR-9809686.
Su p p or tin g In for m a tion Ava ila ble: Characterization
data (HPLC and 13C NMR spectra) for 2-9 and 11, absorption
spectrum, and a figure showing the complete chemical shift
assignments and J CC values from the INADEQUATE experi-
ments on 13a . This material is available free of charge via
the Internet at http://pubs.acs.org.
Ack n ow led gm en t. This work was supported by the
J O020216E
National Science Foundation (CHE-9816339) and in
5952 J . Org. Chem., Vol. 67, No. 17, 2002