Synthesis and identification of heterocyclic derivatives of fullerene C60: Unexpected reaction of anionic C60 with benzonitrile

Article abstract of DOI:10.1021/jo702678c

(Figure Presented) During the reaction of reduced C₆₀ with benzyl bromide in benzonitrile, a novel cis-1 C₆₀ adduct, 1,4-dibenzyl-2,3-cyclic phenylimidate C₆₀ (1), was obtained rather than the expected product of 1,4-dibenzyl C₆₀. The structure of compound 1 was analyzed by X-ray single-crystal diffraction, identifying the presence of a five-membered heterocycle at a [5,6] bond of C₆₀. One of the heteroatoms is assigned as a nitrogen atom; however, the identity of the other heteroatom cannot be determined unambiguously by crystallography due to similarity between the nitrogen and oxygen atoms. A related compound (2) bearing the same heterocycle was obtained from anionic C₆₀ benzonitrile solution when no benzyl bromide was added. The structure of compound 2 was determined by NMR, MALDIFT-ICR MS, and UV-vis. Results from MALDIFT-ICR MS for compound 2 show unambiguously that the second heteroatom is an oxygen atom, which is probably from traces of water in the solvent. Control experiments of the reactivity of the neutral, monoanionic, dianionic, and trianionic C ₆₀ have shown that the reactive species for the unexpected reaction is the C₆₀ trianion.

Full text of DOI:10.1021/jo702678c

Synthesis and Identification of Heterocyclic Derivatives of
Fullerene C : Unexpected Reaction of Anionic C with
6
0
60
Benzonitrile
Min Zheng, Fang-fang Li, Ling Ni, Wei-wei Yang, and Xiang Gao*
State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Graduate
School of Chinese Academy of Sciences, Chinese Academy of Sciences, Changchun, Jilin 130022, China
xgao@ciac.jl.cn
ReceiVed December 17, 2007
During the reaction of reduced C60 with benzyl bromide in benzonitrile, a novel cis-1 C60 adduct, 1,4-
dibenzyl-2,3-cyclic phenylimidate C60 (1), was obtained rather than the expected product of 1,4-dibenzyl
C60. The structure of compound 1 was analyzed by X-ray single-crystal diffraction, identifying the presence
of a five-membered heterocycle at a [5,6] bond of C60. One of the heteroatoms is assigned as a nitrogen
atom; however, the identity of the other heteroatom cannot be determined unambiguously by crys-
tallography due to similarity between the nitrogen and oxygen atoms. A related compound (2) bearing
the same heterocycle was obtained from anionic C60 benzonitrile solution when no benzyl bromide was
added. The structure of compound 2 was determined by NMR, MALDI FT-ICR MS, and UV-vis. Results
from MALDI FT-ICR MS for compound 2 show unambiguously that the second heteroatom is an oxygen
atom, which is probably from traces of water in the solvent. Control experiments of the reactivity of the
neutral, monoanionic, dianionic, and trianionic C60 have shown that the reactive species for the unexpected
reaction is the C60 trianion.
3
Introduction
areas such as organic electronics, biological and medicinal
4
5
chemistry, materials chemistry, and catalytic reagent chem-
istry. C60, the most abundant member in the fullerene family,
has received the majority of synthetic attention due to its
availability. It is now well-established that C60 is electron-
deficient and reactive toward nucleophiles. It has been shown
that the reactions of C60 mainly involve the cleavage of double
bonds at the [6,6] site and lead to the formation of 1,2- or 1,4-
adducts depending on the size of addends. Occasionally,
1
The chemistry of fullerenes has developed dramatically since
6
a method for macroscopic production of fullerenes was re-
2
ported. The unique electronic and three-dimensional structure
of fullerenes has enabled organic chemists to discover an
astonishing variety of unprecedented reactions and functional-
ized regioisomers which have shown remarkable potential in
1
1
(1) (a) Taylor, R.; Walton, D. R. M. Nature 1993, 363, 685-693. (b)
Diederich, F.; Thilgen, C. Science 1996, 271, 317-324. (c) Prato, M.;
Maggini, M. Acc. Chem. Res. 1998, 31, 519-526. (d) Hirsch, A. Top. Curr.
Chem. 1999, 199, 1-65. (e) Hirsch, A.; Brettreich, M. Fullerenes:
Chemistry and Reactions; Wiley-VCH: Weinheim, Germany, 2005. (f)
Mart ´ı n, N. Chem. Commun. 2006, 2093-2104.
derivatives with [5,6] open structure (fulleroids) involving the
cleavage of a [5,6] single bond are also observed.7
The chemistry of fullerene anions has also attracted great
attention since, unlike neutral fullerenes, they are electron-rich
(
2) Kr a¨ tschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffman, D. R.
1e,8
Nature 1990, 347, 354-358.
and reactive toward electrophiles. Due to the strong electron-
1
0.1021/jo702678c CCC: $40.75 © 2008 American Chemical Society
Published on Web 03/15/2008
J. Org. Chem. 2008, 73, 3159-3168 3159

Zheng et al.
accepting ability of the molecule,9 C60 anions are readily
of preparing multiple C60 adducts not accessible from neutral
C60. In addition, it has been shown that the anions of C60 are
8
b,e,f,h,k-m,10
8i
generated by either chemical
or electrochem-
8
a,c,d,g,i,j,n,11
8e,f
ical
reductive methods. Previous work has shown that
excellent electron transfer mediators for reaction initiation
1
7
anionic C60 exhibits a different reactivity to neutral C60. For
example, it has been shown that organofullerenes can be
and electrocatalytic reductive reactions. However, studies on
the reactivity of C60 anions are mostly limited to the reactions
of C60 with organic halides. Much less work has appeared
on the reactivity of C60 , and no apparent difference has been
recorded for the reactivity between C60 and C60 except that
1
e,12,13
2-
8
obtained from neutral C60 by cyclopropanation reaction,
while anions of organofullerenes can undergo retro-cyclopro-
3
-
1
4,15
2-
3-
panation reactions to selectively remove the addends.
The
monoanionic C60 has demonstrated a better ability to complex
with cyclodextrins, forming water-soluble supramolecular struc-
tures and constructing interfacial supramolecular self-assembled
monolayers on a gold surface.16 The dianionic C60 is also capable
(
8) (a) Caron, C.; Subramanian, R.; D’Souza, F.; Kim, J.; Kunter, W.;
Jones, M. T.; Kadish, K. M. J. Am. Chem. Soc. 1993, 115, 8505-8506. (b)
Chen, J.; Cai, R.-F.; Huang, Z.-E.; Wu, H.-M.; Jiang, S.-K.; Shao, Q.-F. J.
Chem. Soc., Chem. Commun. 1995, 1553-1554. (c) Boulas, P. L.; Zuo,
Y.; Echegoyen, L. Chem. Commun. 1996, 1547-1548. (d) Mangold, K.-
M.; Kunter, W.; Dunsch, L.; Fr o¨ hner, J. Synth. Met. 1996, 77, 73-76. (e)
Subramanian, R.; Kadish, K. M.; Vijayashree, M. N.; Gao, X.; Jones, M.
T.; Miller, D. M.; Krause, K.; Suenobu, T.; Fukuzumi, S. J. Phys. Chem.
1996, 100, 16327-16335. (f) Fukuzumi, S.; Suenobu, T.; Hirasaka, T.;
Arakawa, R.; Kadish, K. M. J. Am. Chem. Soc. 1998, 120, 9220-9227. (g)
Kadish, K. M.; Gao, X.; Van Caemelbecke, E.; Hirasaka, T.; Suenobu, T.;
Fukuzumi, S. J. Phys. Chem. A 1998, 102, 3898-3906. (h) Allard, E.;
Rivi e` re, L.; Delaunay, J.; Dubois, D.; Cousseau, J. Tetrahedron Lett. 1999,
40, 7223-7226. (i) Kadish, K. M.; Gao, X.; Van Caemelbecke, E.; Suenobu,
T.; Fukuzumi, S. J. Am. Chem. Soc. 2000, 122, 563-570. (j) Kadish, K.
M.; Gao, X.; Van Caemelbecke, E.; Suenobu, T.; Fukuzumi, S. J. Phys.
Chem. A 2000, 104, 3878-3883. (k) Allard, E.; Delaunay, J.; Cheng, F.;
Cousseau, J.; Orduna, J.; Garin, J. Org. Lett. 2001, 3, 3503-3506. (l) Cheng,
F.; Murata, Y.; Komatsu, K. Org. Lett. 2002, 4, 2541-2544. (m) Meier,
M. S.; Bergosh, R. G.; Gallagher, M. E.; Spielmann, H. P.; Wang, Z. J.
Org. Chem. 2002, 67, 5946-5952. (n) Zheng, M.; Li, F.; Shi, Z.; Gao, X.;
Kadish, K. M. J. Org. Chem. 2007, 72, 2538-2542.
(3) (a) Kim, Y.; Cook, S.; Tuladhar, S. M.; Choulis, S. A.; Nelson, J.;
Durrant, J. R.; Bradley, D. D. C.; Giles, M.; Mcculloch, I.; Ha, C.-S.; Ree,
M. Nat. Mater. 2006, 5, 197-203. (b) Hasobe, T.; Kamat, P. V.; Troiani,
V.; Solladie, N.; Ahn, T. K.; Kim, S. K.; Kim, D.; Kongkanand, A.;
Kuwabata, S.; Fukuzumi, S. J. Phys. Chem. B 2005, 109, 19-23. (c)
Konishi, T.; Ikeda, A.; Shinkai, S. Tetrahedron 2005, 61, 4881-4899. (d)
Kim, K.-S.; Kang, M.-S.; Ma, H.; Jen, A. K.-Y. Chem. Mater. 2004, 16,
5
058-5062. (e) Ikeda, A.; Hatano, T.; Shinkai, S.; Akiyama, T.; Yamada,
S. J. Am. Chem. Soc. 2001, 123, 4855-4856. (f) Yu, G.; Gao, J.; Hummelen,
J. C.; Wudl, F.; Heeger, A. J. Science 1995, 270, 1789-1791. (g) Sariciftci,
N. S.; Heeger, A. J. Int. J. Mod. Phys. B 1994, 8, 237-274. (h) Boutorine,
A. S.; Tokuyama, H.; Takasugi, M.; Isobe, H.; Nakamura, E.; H e´ l e` ne, C.
Angew. Chem., Int. Ed. Engl. 1995, 33, 2462-2465. (i) Hebard, A. F.;
Rosseinsky, M. J.; Haddon, R. C.; Murphy, D. W.; Glarum, S. H.; Palstra,
T. M. T.; Ramirez, A. P.; Kortan, A. R. Nature 1991, 350, 600-601. (j)
Holczer, K.; Klein, O.; Huang, S.-M.; Kaner, R. B.; Fu, K.-J.; Whetten, R.
L.; Diederich, F. Science 1991, 252, 1154-1157. (k) Herance, J. R.; Peris,
E.; Vidal, J.; Bourdelande, J. L.; Marquet, J.; Garc ´ı a, H. Chem. Mater. 2005,
(9) (a) Xie, Q.; Perez-Cordero, E.; Echegoyen, L. J. Am. Chem. Soc.
1992, 114, 3978-3980. (b) Echegoyen, L.; Echegoyen, L. E. Acc. Chem.
Res. 1998, 31, 593-601. (c) Reed, C. A.; Bolskar, R. D. Chem. ReV. 2000,
100, 1075-1120.
(10) (a) Subramanian, R.; Boulas, P.; Vijayashree, M. N.; D’Souza, F.;
Jones, M. T.; Kadish, K. M. J. Chem. Soc., Chem. Commun. 1994, 1847-
1848. (b) Wei, X.-W.; Wu, M.-F.; Qi, L.; Xu, Z. J. Chem. Soc., Perkin
Trans. 2 1997, 1389-1393.
1
7, 4097-4102. (l) Hoshi, H.; Yamada, T.; Ishikawa, K.; Takezoe, H.;
Fukuda, A. Phys. ReV. B 1995, 52, 12355-12365. (m) Mart ´ı n, N.; S a´ nchez,
L.; Illescas, B.; P e´ rez, I. Chem. ReV. 1998, 98, 2527-2547. (n) Lin, J.;
Zheng, M.; Chen, J.; Gao, X.; Ma, D. Inorg. Chem. 2007, 46, 341-344.
(4) (a) Friedman, S. H.; DeCamp, D. L.; Sijbesma, R. P.; Srdanov, G.;
Wudl, F.; Kenyon, G. L. J. Am. Chem. Soc. 1993, 115, 6506-6509. (b)
Nakamura, E.; Isobe, H. Acc. Chem. Res. 2003, 36, 807-815. (c) Gharbi,
N.; Pressac, M.; Hadchouel, M.; Szwarc, H.; Wilson, S. R.; Moussa, F.
Nano Lett. 2005, 5, 2578-2585. (d) Zakharian, T. Y.; Seryshev, A.;
Sitharaman, B.; Gilbert, B. E.; Knight, V.; Wilson, L. J. J. Am. Chem. Soc.
(11) (a) Lawson, D. R.; Feldheim, D. L.; Foss, C. A.; Dorhout, P. K.;
Elliott, C. M.; Martin, C. R.; Parkinson, B. J. Electrochem. Soc. 1992, 139,
L68-L71. (b) Khaled, M. M.; Carlin, R. T.; Trulove, P. C.; Eaton, G. R.;
Eaton, S. S. J. Am. Chem. Soc. 1994, 116, 3465-3474. (c) Dubois, D.;
Jones, M. T.; Kadish, K. M. J. Am. Chem. Soc. 1992, 114, 6446-6451. (d)
Stinchcombe, J.; P e´ nicaud, A.; Bhyrappa, P.; Boyd, P. D. W.; Reed, C. A.
J. Am. Chem. Soc. 1993, 115, 5212-5217. (e) Rapta, P.; Bartl, A.; Gromov,
A.; Sta sˇ ko, A.; Dunsch, L. Chem. Phys. Chem. 2002, 3, 351-356. (f) Dunch,
L.; Rapta, P.; Gromov, A.; Sta sˇ ko, A. J. Electroanal. Chem. 2003, 547,
35-43.
2
005, 127, 12508-12509. (e) Tagmatarchis, N.; Shinohara, H. Mini-ReV.
Med. Chem. 2001, 1, 339-348. (f) Jensen, A. W.; Wilson, S. R.; Schuster,
D. I. Bioorg. Med. Chem. 1996, 4, 767-779. (g) Da Ros, T.; Spalluto, G.;
Prato, M. Croat. Chem. Acta 2001, 74, 743-755.
(5) (a) Tagmatarchis, N.; Prato, M. Struct. Bonding 2004, 109, 1-39.
(b) Prato, M. J. Mater. Chem. 1997, 7, 1097-1109. (c) Wudl, F. J. Mater.
Chem. 2002, 12, 1959-1963. (d) Cravino, A.; Sariciftci, N. S. J. Mater.
Chem. 2002, 12, 1931-1943. (e) Yamada, H.; Imahori, H.; Fukuzumi, S.
J. Mater. Chem. 2002, 12, 2034-2040. (f) Brusatin, G.; Signorini, R. J.
Mater. Chem. 2002, 12, 1964-1977. (g) Guldi, D. M.; Mart ´ı n, N. J. Mater.
Chem. 2002, 12, 1978-1992. (h) Murata, Y.; Ito, M.; Komatsu, K. J. Mater.
Chem. 2002, 12, 2009-2020. (i) Barazzouk, S.; Hotchandani, S.; Kamat,
P. V. J. Mater. Chem. 2002, 12, 2021-2025. (j) Guldi, D. M.; Zerbetto,
F.; Georgakilas, V.; Prato, M. Acc. Chem. Res. 2005, 38, 38-43. (k) Hasobe,
T.; Imahori, H.; Kamat, P. V.; Ahn, T. K.; Kim, S. K.; Kim, D.; Fujimoto,
A.; Hirakawa, T.; Fukuzumi, S. J. Am. Chem. Soc. 2005, 127, 1216-1228.
(12) (a) Bingel, C. Chem. Ber. 1993, 126, 1957-1959. (b) Diederich,
F.; Isaacs, L.; Philp, D. Chem. Soc. ReV. 1994, 243-255.
(13) (a) Eiermann, M.; Haddon, R. C.; Knight, B.; Chan Li, Q.; Maggini,
M.; Mart ´ı n, N.; Ohno, T.; Prato, M.; Suzuki, T.; Wudl, F. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1591-1593. (b) Ohno, T.; Mart ´ı n, N.; Knight, B.;
Wudl, F.; Suzuki, T.; Yu, H. J. Org. Chem. 1996, 61, 1306-1309. (c)
Knight, B.; Mart ´ı n, N.; Ohno, T.; Ort ´ı , E.; Rovira, C.; Veciana, J.; Vidal-
Gancedo, J.; Viruela, P.; Viruela, R.; Wudl, F. J. Am. Chem. Soc. 1997,
109, 9871-9882. (d) Mart ´ı n, N.; S a´ nchez, L.; Illescas, B.; P e´ rez, I. Chem.
ReV. 1998, 98, 2527-2547. (e) Beulen, M. W. J.; Echegoyen, L.; Rivera,
J. A.; Herranz, M. A.; Mart ´ı n-Domenech, A.; Mart ´ı n, N. Chem. Commun.
2000, 917-918.
(6) (a) Nishibayashi, Y.; Saito, M.; Uemura, S.; Takekuma, S.; Takekuma,
H.; Yoshida, Z. Nature 2004, 428, 279-280. (b) Takekuma, S.; Takekuma,
H.; Yoshida, Z. Chem. Commun. 2005, 1628-1630. (c) Posp ´ı sˇ il, L.;
Bul ´ı cˇ kov a´ , J.; Hromadov a´ , M.; G a´ l, M.; Cici sˇ , S.; Cihelka, J.; Tar a´ bek, J.
Chem. Commun. 2007, 22, 2270-2272.
(14) (a) Knight, B.; Mart ´ı n, N.; Ohno, T.; Ort ´ı , E.; Rovira, C.; Veciana,
J.; Vidal-Gancedo, J.; Viruela, P.; Viruela, R.; Wudl, F. J. Am. Chem. Soc.
1997, 119, 9871-9882. (b) Boudon, C.; Gisselbrecht, J.-P.; Gross, M.;
Herrmann, A.; R u¨ ttimann, M.; Crassous, J.; Cardullo, F.; Echegoyen, L.;
Diederich, F. J. Am. Chem. Soc. 1998, 120, 7860-7868. (c) Kessinger, R.;
Crassous, J.; Herrmann, A.; R u¨ ttimann, M.; Echegoyen, L.; Diederich, F.
Angew. Chem. Int. Ed. 1998, 37, 13-14. (d) Crassous, J.; Rivera, J.; Fender,
N. S.; Shu, L.; Echegoyen, L.; Thilgen, C.; Herrmann, A.; Diederich, F.
Angew. Chem. Int. Ed. 1999, 38, 1613-1671. (e) Kessinger, R.; Fender,
N. S.; Echegoyen, L. E.; Thilgen, C.; Echegoyen, L.; Diederich, F. Chem.s
Eur. J. 2000, 6, 2184-2192. (f) Lukoyanova, O.; Cardona, C. M.; Altable,
M.; Filippone, S.; Domenech, AÄ . M.; Mart ´ı n, N.; Echegoyen, L. Angew.
Chem. Int. Ed. 2006, 45, 7430-7433.
(7) (a) Suzuki, T.; Li, Q.; Khemani, K. C.; Wudl, F.; Almarsson, O¨ .
Science 1991, 254, 1186-1188. (b) Suzuki, T.; Li, Q.; Khemani, K. C.;
Wudl, F. J. Am. Chem. Soc. 1992, 114, 7301-7302. (c) Prato, M.; Li, O.;
Wudl, F.; Lucchini, V. J. Am. Chem. Soc. 1993, 115, 1148-1150. (d)
Hawker, C. J.; Saville, P. M.; White, J. W. J. Org. Chem. 1994, 59, 3503-
3
505. (e) Smith, A. B., III; Strongin, R. M.; Brard, L.; Furst, G. T.;
Romanow, W. J.; Owens, K. G.; Goldschmidt, R. J.; King, R. C. J. Am.
Chem. Soc. 1995, 117, 5492-5502. (f) Hummelen, J. C.; Knight, B. W.;
LePeq, F.; Wudl, F.; Yao, J.; Wilkins, C. L. J. Org. Chem. 1995, 60, 532-
5
1
38. (g) Schick, G.; Gr o¨ sser, T.; Hirsch, A. J. Chem. Soc., Chem. Commun.
995, 2289-2290. (h) Ikeda, A.; Fukuhara, C.; Shinkai, S. Tetrahedron
Lett. 1996, 37, 7091-7094. (i) Zhu, C.-C.; Xu, Y.; Liu, Y.-Q.; Zhu, D.-B.
(15) (a) Knight, B.; Mart ´ı n, N.; Ohno, T.; Ort ´ı , E.; Rovira, C.; Veciana,
J.; Vidal-Gancedo, J.; Viruela, P.; Viruela, R.; Wudl, F. J. Am. Chem. Soc.
1997, 119, 9871-9882. (b) Beulen, M. W. J.; Rivera, J. A.; Herranz, M.
AÄ .; Illescas, B.; Mart ´ı n, N.; Echegoyen, L. J. Org. Chem. 2001, 66, 4393-
4398.
J. Org. Chem. 1997, 62, 1996-2000. (j) Schick, G.; Hirsch, A. Tetrahedron
1
998, 54, 4283-4296. (k) Weisman, R. B.; Heymann, D.; Bachilo, S. M.
J. Am. Chem. Soc. 2001, 123, 9720-9721. (l) Ulmer, L.; Mattay, J. Eur. J.
Org. Chem. 2003, 2933-2940.
3160 J. Org. Chem., Vol. 73, No. 8, 2008

Heterocyclic DeriVatiVes of Fullerene C60
C603- is more reactive than C602- under similar conditions.
11b,18
6a,c
Recent reports on dinitrogen fixation and CdC, CdO, and
+
6b
N-N bond reduction by an anionic γ-cyclodextrin-bicapped
C60 complex under mild conditions have aroused interest to
further study the reactivity of anionic C60 since it suggests that
the chemistry of fullerenes is not completely understood as
indicated by a recent review.1f
During our recent investigation into the reactivity of anionic
C60, an unexpected reaction involving trianionic C60, water, and
benzonitrile was discovered. Unlike previous studies, where the
reducing potential was switched off when the number of
coulombs reached the theoretical number required for a complete
conversion of C60 to C602 , in our case, the reducing potential
was continuously applied to the anionic C60 benzonitrile solution
after the theoretical number of coulombs had been reached. An
unexpected cis-1 adduct of C60, 1,4-dibenzyl-2,3-cyclic phen-
ylimidate C60 (1), was regioselectively formed after the addition
of benzyl bromide to the anionic C60 benzonitrile solution, while
a related compound (2) bearing the same heterocycle as
compound 1 was obtained from the anionic C60 benzonitrile
solution when no benzyl bromide was added. A control
experiment was performed where tetra-n-butylammonium hexaflu-
orophosphate (TBAPF6) was used instead of tetra-n-butylam-
monium perchlorate (TBAP) as the electrolyte, and the same
product was obtained, indicating that the oxygen atom in
compounds 1 and 2 is not from perchlorate ion but likely from
traces of water present in solvent. In addition, reactivity of the
neutral, monoanionic, dianionic, and trianionic C60 toward
benzonitrile was examined. It showed that the real reactive
species was the C60 trianion.
- 8
FIGURE 1. Cyclic voltammograms of the resulting anionic C60 with
a coulometric value of slightly more than two electrons per molecule
in benzonitrile solution containing 0.1 M TBAP under (a) Ar and (b)
2
N .
becomes stronger as the reducing potential is applied for a longer
time. However, the cyclic voltammogram of the anionic C60
solution is completely distorted and featureless of the C60 moiety
if the solution is kept under the external potential for too long,
leading to the formation of more toluene or CS2 insoluble
product than when the anionic C60 solution is kept under the
potential for less time. The potentiostat was switched off when
the irreversible anodic wave at -0.50 V versus SCE became
prominent, and at the same time, the features of C60 moiety
were retained as shown in Figure 1, then 10 equiv of PhCH2Br
was added to the solution all at once, and the reaction between
the resulting anionic C60 benzonitrile solution and benzyl
bromide was allowed to proceed for 40 min with stirring. The
solvent was removed with a rotary evaporator under vacuum,
and the residue was washed with methanol to remove TBAP
before further purification.
Herein, we report this unprecedented reaction which leads
to the formation of C60 derivatives with cyclic phenylimidate.
Although the detailed mechanism for the reaction remains to
be elucidated by further work, we wish to report the structures
of products and identify the reactive species for this novel
reaction at the current stage.
Results and Discussion
Synthesis of 1,4-Dibenzyl-2,3-Cyclic Phenylimidate C60 (1).
Fifty milligrams (69.4 µmol) of C60 was electroreduced at -1.10
V versus SCE in 50 mL of freshly distilled PhCN solution
containing 0.1 M TBAP under either a nitrogen or an argon
atmosphere. The procedures for the electrolysis are similar to
2-
Unlike previously reported results of reactions between C60
and benzyl bromide,
8e,g,j,m,n
where the only major product is 1,4-
(PhCH2)2C60, a new major product, 1,4-dibenzyl-2,3-cyclic
phenylimidate C60 (1), is found to be produced along with the
formation of 1,4-(PhCH2)2C60, trace amount of 1,2-(PhCH2)2C60
8g,i,j,n
previous work,
except the reducing potential is continuously
(3), 1,2-HPhCH2C60 (4), [6,6]methanofullerene C61HPh (5), and
applied even after the theoretical number of coulombs required
for a full conversion of C60 to C602 has been reached. Notably,
an extra irreversible anodic wave at -0.50 V versus SCE
appears in addition to the redox waves of C60 (see Figure 1) as
the reducing potential is continuously applied after the electro-
unreacted C60 with retention times at 5.01, 5.60, 5.33, 6.58, 7.14,
and 9.78 min, respectively, as shown in the HPLC trace of the
reaction mixture (Figure 2). Notably, the HPLC trace over a
Buckyprep column is nearly identical but reversed of that over
a silica column reported previously, although it does not include
-
generation of C602 is complete, indicating a reaction involving
-
19
8n
compound 1. The formation of 1,4-(PhCH2)2C60, trace amount
anionic C60 has occurred. The extra irreversible anodic wave
of 1,2-(PhCH2)2C60 (3), 1,2-HPhCH2C60 (4), and [6,6]metha-
nofullerene C61HPh (5) is typical for the reaction between the
8
n
(16) (a) Zhang, Y.; Liu, W.; Gao, X.; Zhao, Y.; Zheng, M.; Li, F.; Ye,
C60 dianion and benzyl bromide. However, the formation of
compound 1 is unprecedented and suggests a novel reaction of
anionic C60. The results show that the real reactive species in
D. Tetrahedron Lett. 2006, 47, 8571-8574. (b) Liu, W.; Zhang, Y.; Gao,
X. J. Am. Chem. Soc. 2007, 129, 4973-4980.
(
17) (a) Huang, Y.; Wayner, D. M. J. Am. Chem. Soc. 1993, 115, 367-
3-
3
68. (b) Niyazymbetov, M. E.; Evans, D. H. J. Electrochem. Soc. 1995,
this anionic C60 solution is C60 as discussed below. Compound
1
42, 2655-2658. (c) Fuchigami, T.; Kasuga, M.; Konno, A. J. Electroanal.
1
was separated by gravity column chromatography over silica
Chem. 1996, 411, 115-119. (d) D’Souza, F.; Choi, J.-P.; Hsieh, Y.-Y.;
Shriver, K.; Kutner, W. J. Phys. Chem. B 1998, 102, 212-217. (e) Sherigara,
B. S.; Kutner, W.; D’Souza, F. Electroanal. 2003, 15, 753-772.
(19) It shows that the ambient conditions have a significant influence
(18) (a) Dubois, D.; Kadish, K. M.; Flanagan, S.; Haufler, R. E.; Chibante,
on the reaction. It requires the passing of much more excessive reducing
charges in winter than in summer to have the irreversible anodic wave at
-0.50 V versus SCE appear.
L. P. F.; Wilson, L. J. J. Am. Chem. Soc. 1991, 113, 4364-4366. (b) Beulen,
M. W. J.; Echegoyen, L. Chem. Commun. 2000, 1065-1066.
J. Org. Chem, Vol. 73, No. 8, 2008 3161

Zheng et al.
to C60 via two heteroatoms at C2 and C3 positions. It is rational
to designate one of the heteroatoms as a nitrogen atom since it
actually consists of a benzonitrile molecule with the C-C6H5
group, indicating that the solvent, benzonitrile, takes part in the
reaction. As for the second heteroatom, however, due to the
similarity between nitrogen and oxygen atoms, its identity cannot
be determined unambiguously from crystallography. It shows
that it is reasonable to assign it as either a nitrogen (with a
hydrogen atom on it) or an oxygen atom. Therefore, it has to
rely on further characterizations to determine the identity of
this heteroatom. As discussed below, it shows that the second
heteroatom is an oxygen atom, and the heterocycle is therefore
an oxazoline.
The bond lengths for C1-C2 and C3-C4 are 1.581(5) and
.583(5) Å, which are within the range for a C-C single bond
1
length. For C2-C3, it is 1.624(5) Å, which is remarkably
elongated with respect to a C-C single bond. A similar bond
length elongation for this bond has also been observed in
2
0
compounds with a cis-1 pattern. The bond lengths for C4-
C5 and C1-C6 are 1.526(5) and 1.524(5) Å, respectively,
whereas it is 1.363(5) Å for C5-C6, and they are all within
the bond length range for C60. The C1-C2-C3-C4-C5-C6
ring is distorted, with C1 and C4 being uplifted by 0.128 and
0.121 Å from the mean plane, respectively. The bond angles of
C69-C68-N1/O1 and C69-C68-O2/N2 are 120.3(4) and
FIGURE 2. HPLC trace of the crude reaction mixture formed from
the anionic C60 benzonitrile solution shown in Figure 1 and benzyl
bromide. The mixture was eluted with toluene over a semipreparative
Buckyprep column at a flow rate of 4.0 mL/min with the detector
wavelength set at 380 nm. Peaks marked with 1, 3, 4, and 5 correspond
to compounds of 1, 3, 4, and 5. Peaks marked with asterisks are not
identified.
1
21.7(4)°, respectively, and the sum of bond angles of C69-
C68-N1/O1, C69-C68-O2/N2, and N1/O1-C68-O2/N2 is
2
3
60°, indicating that C68 is an sp carbon atom. The measured
bond lengths for N1/O1-C68 and O2/N2-C68 (CdN and
C-O) are almost identical with the values of 1.328(5) and
1
.319(5) Å, respectively, which is in agreement with the reported
21
values for crystal structures of oxazolines and is probably due
to the delocalization of π-electrons among NdC-O bonds.2
The bond lengths for N1/O1-C2 and O2/N2-C3 (C-N and
C-O) are also almost identical with the values of 1.469(4) and
1b
1
.478(4) Å, which are consistent with the corresponding bond
2
1
length for typical oxazolines. It is noteworthy that, since the
oxazoline is located at the [5,6] bond of C60 and there is no
symmetry plane cutting either through or perpendicular to the
plane of the imidate, compound 1 should thus have a C1
symmetry and be a racemic mixture.
UV-Visible Spectrum of Compound 1. Figure 4 shows the
UV-visible spectrum of compound 1 in hexane. It has absorp-
tion bands at 206 and 256 nm, which are also observed in the
2
2
spectrum of C60. A weak broad absorption band at around
FIGURE 3. ORTEP diagram for one 1,4-dibenzyl-2,3-cyclic phen-
ylimidate C60 molecule with 20% thermal ellipsoids. Hydrogen atoms
and CS molecule were omitted for clarity.
2
3
36 nm is observed, which is very different from the sharp and
22
strong absorption band for C60 appearing in the same region.
The compound also exhibits a spike at 429 nm, which is
characteristic for C60 derivatives with 1,2-addition pattern
8
j,23
20b,24
eluted by first using hexanes then a 1:3 v/v toluene/hexanes
mixture. Under optimum conditions for producing compound
including 1,2-
and cis-1 adducts,
and it is therefore
consistent with the cis-1 structural assignment for the compound.
1
1, the isolated yields for compound 1, 1,4-(PhCH2)2C60, and
NMR of Compound 1. Figure 5 shows the H NMR of
recovered C60 are ca. 35, 30, and 15%, respectively, while the
total amount of 1,2-(PhCH2)2C60 (3), 1,2-HPhCH2C60 (4), and
compound 1 from 4.0 to 9.0 ppm recorded in CS with C D as
2
6
6
the external lock, a full spectrum is shown in the Supporting
Information. Two AB quartets I (aa) and II (bb) centered at
4.76 and 4.72 ppm are observed, which are due to the methylene
[6,6]methanofullerene C61HPh (5) is less than 2%. The remain-
der of the reaction mixture is insoluble in toluene or CS2 and
remains to be identified.
8j
protons of benzyls, indicating the molecule has a C symmetry
1
X-ray Structure Characterization of Compound 1. Black
needle-like crystals of compound 1 were obtained by slowly
diffusing hexane into a CS2 solution of compound 1 at room
temperature. Figure 3 shows the X-ray single-crystal structure
of compound 1. It shows that two benzyl groups are bonded to
C60 at C1 and C4, respectively, while a heterocycle is bonded
(20) (a) Miller, G. P.; Tetreau, M. C.; Olmstead, M. M.; Lord, P. A.;
Balch, A. L. Chem. Commun. 2001, 1758-1759. (b) Murata, Y.; Suzuki,
M.; Rubin, Y.; Komatsu, K. Bull. Chem. Soc. Jpn. 2003, 76, 1669-1672.
(
21) (a) Braunstein, P.; Graiff, C.; Naud, F.; Pfaltz, A.; Tiripicchio, A.
Inorg. Chem. 2000, 39, 4468-4475. (b) McPherson, L. D.; Drees, M.; Khan,
S. I.; Strassner, T.; Abu-Omar, M. M. Inorg. Chem. 2004, 43, 4036-4050.
3162 J. Org. Chem., Vol. 73, No. 8, 2008

Heterocyclic DeriVatiVes of Fullerene C60
phenylimidate, and the remaining 12 resonances from 134.9 to
26.9 ppm are from the phenyl carbons of the heterocycle and
1
benzyl groups. The results are in agreement with the structural
assignment of compound 1, as displayed by the X-ray single-
crystal diffraction.
MALDI TOF Mass Spectrum of Compound 1. Compound
1
was subject to MALDI TOF MS measurement using 2,5-
dihydroxybenzoic acid (DHB) as the matrix. However, efforts
to obtain accurate molecular weight by FT-ICR MS were not
successful. Figure 7 shows the positive MALDI TOF mass
spectrum of 1. It shows the protonated molecular ion [M +
+
H] at m/z ) 1022, consistent with the structural assignment
for the compound. Ions at 930, 839, 810, 797, and 720
correspond to the fragment ions. The oxygen adduct is shown
at 736, which is likely from a fragmentation process of the
compound or a possible gas-phase reaction with the matrix as
8
i
observed in previous studies.
FIGURE 4. UV-visible spectrum of compound 1 in hexane.
Synthesis of [6,6] Cyclic Phenylimidate C60 (2). In order
to have a better understanding of the reaction, experiments were
performed where the procedures were exactly the same as those
for the generation of compound 1, except no benzyl bromide
was added to the anionic C60 benzonitrile solution shown in
Figure 1. The solution was then oxidized back to neutral
electrochemically. The procedures for purification of the
compound are similar to that for compound 1. Figure 8 shows
the HPLC trace of the crude reaction mixture. A new compound
characterized as [6,6] cyclic phenylimidate C60 (2) (Chart 1)
appeared in the HPLC at 7.51 min along with the fraction peak
of unreacted C60 at 9.74 min. Under optimum conditions, the
isolated yields are ca. 10 and 70% for 2 and recovered C60,
respectively, while the remaining product is insoluble in either
toluene or CS2 and remains to be identified.
and is consistent with the X-ray single-crystal diffraction results.
Resonances arising from the aromatic protons are also present
in the spectrum. Peaks at 8.79 (d, 2H), 8.06 (t, 1H), and 8.02
(
t, 2H) ppm are assigned to the phenyl protons of the
heterocycle, while resonances at 7.75 (d, 4H) and from 7.62 to
.51 (m, 6H) ppm are due to the phenyl protons of benzyl
7
8j
groups and partly overlap with those from the toluene residue
used for purification of the compound. Notably, no resonance
corresponding to the amine proton is observed for the compound
in the solvent system where the possibility of H-D exchange
is avoided, indicating that no amine group is present in the
molecule, which is in agreement with the assignment of the
heteroatom as an oxygen atom.
Figure 6 shows the 13C NMR of 1 recorded in a mixture of
CS2 and CDCl3. Two resonances corresponding to the methylene
carbons of the benzyl groups are shown at 46.6 and 45.9 ppm,
1
NMR of Compound 2. Figure 9 shows the H NMR of
compound 2 from 4.0 to 9.0 ppm, while a full spectrum is shown
in the Supporting Information. Similar to the spectrum of
compound 1, resonances are shown at 8.82 (d, 2H), 8.08 (t,
1H), and 8.02 (t, 2H) ppm, which correspond to the protons of
the phenyl ring of phenylimidate group, indicating the existence
of phenylimidate in the molecule. The results are further
confirmed by the observation that no resonance from the amine
proton is shown in the spectrum, consistent with the assignment
of the second heteroatom as an oxygen atom. As expected, no
resonances from methylene protons of benzyls are observed,
suggesting no benzyls are present in compound 2. Notably, a
3
while resonances corresponding to the two sp C60 carbons
bonded to benzyls are shown at 62.9 and 61.9 ppm. These values
are in agreement with a previous report on 1,4-dibenzyl C60.
8
j
The resonances at 91.5 and 98.0 ppm are assigned to the two
3
sp C60 carbons bonded to the nitrogen and oxygen atoms,
2
4,25
consistent with literature values.
A total of 62 resonances
2
are shown in the sp region, while 49 of them are assigned to
the sp carbons of C60 from 153.4 to 135.3 ppm. The resonance
2
at 162.9 ppm is due to the imine carbon of the cyclic
FIGURE 5. 1H NMR spectrum of compound 1 in CS
recorded on a 600 MHz instrument. C
D
was used as the external lock solvent.
2
6
6
J. Org. Chem, Vol. 73, No. 8, 2008 3163

Zheng et al.
FIGURE 6. 13C NMR spectrum of 1 in CS
/CDCl for (a) sp carbons and (b) sp carbons recorded on a 150 MHz instrument.
3
2
2 3
indicating that there might be some interactions between
compound 2 and toluene.
Figure 10 shows the 13C NMR of compound 2. Two weak
3
sp carbons of C60, probably due to longer relaxation time of
the carbons, are observed at 91.6 and 96.6 ppm, which are
3
attributed to the two sp carbons connected to the nitrogen and
24,25
oxygen atoms,
indicating the presence of a cyclic phenylimi-
date group as in compound 1. A total of 31 resonances for the
2
sp carbon atoms of the compound appear in the spectrum, while
2
6 of them are assigned to C60 carbons from 147.5 to 135.4
ppm, and four resonances are assigned to the phenyl carbons at
31.7 (1 C, Ph), 128.7 (2C, Ph), 128.1 (2C, Ph), and 126.5 (1C,
1
Ph) ppm, and the remaining one at 164.3 ppm is from the imine
FIGURE 7. Positive MALDI TOF MS of compound 1.
(23) (a) Isaacs, L.; Wehrsig, A.; Diederich, F. HelV. Chim. Acta 1993,
7
6, 1231-1250. (b) Linssen, T. G.; D u¨ rr, K.; Hanack, M.; Hirsch, A. J.
Chem. Soc., Chem. Commun. 1995, 103-104.
small amount of toluene is always shown in the spectrum even
though the compound was put under vacuum for several days,
(24) (a) Ulmer, L.; Siedschlag, C.; Mattay, J. Eur. J. Org. Chem. 2003,
3
811-3817. (b) Schick, G.; Hirsch, A.; Mauser, H.; Clark, T. Chem.s
Eur. J. 1996, 2, 935-943.
(
22) Hare, J. P.; Kroto, H. W.; Taylor, R. Chem. Phys. Lett. 1991, 177,
(25) Wang, G.-W.; Li, F.-B.; Xu, Y. J. Org. Chem. 2007, 72, 4774-
3
94-398.
4778.
3164 J. Org. Chem., Vol. 73, No. 8, 2008

Heterocyclic DeriVatiVes of Fullerene C60
experimental m/z values for the observed peaks. The protonated
monoisotopic molecular ion is shown at m/z ) 840.04214 as
the base peak, while ions appearing at 841.04535, 842.04878,
and 843.05378 are due to the isotopic peaks since the mass
difference between them is around 1.003. A relatively small
peak is observed at 839.03437, which has a 1.008 mass
difference with the base peak at 840.04214, indicating that the
difference is caused by a proton; it is therefore reasonable to
assign the peak at 839.03437 as the molecular ion. Fragment
ions are also observed in the spectrum, and their identities are
listed in Table 1. The observed m/z values all have an excellent
agreement with the theoretical values, confirming that the
heterocycle is a cyclic phenylimidate ring. Considering that
compounds 1 and 2 should have the same heterocycle according
to NMR results and similar synthetic methods, it is reasonable
to assign the unresolved heteroatom in the crystal structure of
compound 1 as an oxygen atom.
FIGURE 8. HPLC trace of the crude reaction mixture formed from
the anionic C60 benzonitrile solution shown in Figure 1. The mixture
was eluted by toluene over a semipreparative Buckyprep column at a
flow rate of 4.0 mL/min with the detector wavelength set at 380 nm.
UV-Visible Spectrum of Compound 2. Figure 12 shows
the UV-visible spectrum of compound 2 in hexane. It has three
major absorption bands at 207, 255, and 314 nm along with a
weak shoulder absorption band at 415 nm. The bands at 207
and 255 nm are also observed in the spectrum of C6022 and are
probably due to the electronic transitions of C60. The absorption
at 314 nm is blue-shifted with respect to the 330 nm absorption
carbon of the phenylimidate group. Resonances at 136.6, 128.3,
127.6, and 124.8 ppm are due to the toluene residue used for
2
2
purification of the compound. The appearance of only 26 C60
carbons indicates that the compound possesses a Cs symmetry,
and it indicates that, unlike the compound 1, the cyclic
phenylimidate group in compound 2 has to be on a [6,6] bond
of C60 rather than a [5,6] bond; otherwise, compound 2 would
have a C1 symmetry because there is no symmetry plane cutting
either through or perpendicular to the plane of the oxazoline.
Since compound 2 was obtained from the same anionic C60
benzonitrile solution as compound 1, except no benzyl bromide
was added, it is reasonable that the two compounds are related
by bearing the same adduct due to a similar reaction. However,
it is noteworthy that the two structures have different addition
positions for phenylimidate, which is probably due to the further
addition of 1,4-dibenzyls in compound 1.
band of C60; however, it is not usually shown in the spectra
for C60 derivatives where the adducts are bonded to C60 via
carbon atoms but observed in the spectrum of C60 derivatives
with at least one heteroatom connected directly to C60.2 The
absorption band is therefore probably related to the type of atoms
bonded to C60. Notably, no absorption around 428 nm, which
is characteristic feature for C60 derivatives with 1,2-pattern,7e
is shown in the spectrum, although the compound is a [6,6]
derivative. The absence of this absorption peak has been
previously reported for [6,6] C60 derivatives with a cyclic
5,26
2
6a
imidate group, indicating the origin of the absorption at 428
nm might also be associated with the type of atoms bonded to
C60 in addition to the addition pattern.
Reactive Species: C602 or C603-? The reactivity of the C60
-
MALDI FT-ICR Mass Spectrum of Compound 2. The
identity of the heterocycle was determined to be a cyclic
phenylimidate unambiguously by the MALDI FT-ICR mass
spectrum of compound 2 using 2,5-dihydroxybenzoic acid
dianion has been studied in benzonitrile, and no such reaction
8
a,d-g,i,j,m,n,11a,c
has been reported.
Notably, it requires more
reducing charges than the theoretical number for the conversion
2
-
of C60 to C60 to have the reaction. It is therefore possible that
the real reactive species for the reaction is not the C60 dianion,
but the trianion instead.
(DHB) as the matrix. Figure 11 shows the positive MALDI FT-
ICR mass spectrum of 2, and Table 1 lists the calculated and
FIGURE 9. 1H NMR spectrum of compound 2 in CS
marked with asterisks are from toluene residue used for purifying 2.
recorded on a 600 MHz instrument. C
D
6
was used as the external lock solvent. Peaks
2
6
J. Org. Chem, Vol. 73, No. 8, 2008 3165

Zheng et al.
FIGURE 10. 13C NMR spectrum of compound 2 in CS
a 150 MHz instrument. DMSO-d was used as the external lock solvent.
recorded on
2
6
FIGURE 12. UV-vis spectrum of compound 2 in hexane.
TABLE 1. Calculated and Observed m/z Values for Molecular and
Fragment Ions of Compound 2
deviation
species
calculated m/z
observed m/z
(ppm)
+
[
C67H5NO + H]
840.04494
839.03711
797.03913
720.00000
840.04214
839.03437
797.03642
719.99815
-3.3
-3.3
-3.4
-0.3
+
C67H5NO
C66H5
C60
+
and the reducing potential was continuously applied after C60-
was fully generated, and the total reaction time was also 3 h.
For the reaction of the dianionic C60, the reduction was stopped
once the number of electrons reached 90% of the theoretical
FIGURE 11. Positive MALDI FT-ICR HRMS of compound 2.
2
-
number required for the conversion to C60 . As for trianionic
C60, it was generated electrochemically with a reducing potential
of -1.60 V versus SCE in benzonitrile, and the electrolysis was
terminated once the theoretical number of coulombs required
CHART 1. Compound 2
3
-
for the conversion of C60 to C60 was reached. Notably, the
3
-
cyclic voltammogram of the resulting C60 is very similar to
that of the anionic C60 benzonitrile solution shown in Figure 1,
with the appearance of the irreversible anodic peak at around
-
0.50 V versus SCE. For the anionic C60 solutions, the anions
were oxidized back to neutral electrochemically by applying 0
V versus SCE at the end of the experiment. The workup for
the purification of the reaction products from the control
experiments is the same as those for the isolation of compound
2
. Figure 13 shows the HPLC traces of the crude products
obtained from neutral and different anionic species of C60.
As shown in the figure, there is no reaction product for the
neutral, monoanionic, and dianionic C60 benzonitrile solution.
While a significant amount of compound 2 (retention time )
7
.50 min, and confirmed by UV-vis measurement) was
To probe further into the reaction, control experiments were
carried out where the reactivity of neutral, monoanionic,
dianionic, and trianionic C60 was examined toward benzonitrile.
For neutral C60, the total reaction time (since C60 was put into
benzonitrile) was about 3 h, which is the typical total reaction
time for the generation of compound 2 from the anionic C60
solution. As for the reaction of monoanionic C60, it was
generated with a reducing potential of -0.70 V versus SCE,
3-
obtained from the C60 benzonitrile solution with an isolation
yield of ca. 50%, which is much higher than that obtained from
the anionic C60 solution shown in Figure 1, indicating that C60
is the real reactive species for the observed reaction in the
anionic C60 benzonitrile solution.
3-
In fact, it is not straightforward to perceive that the C60
trianion can be generated in significant amount in a solution
where the reducing potential (-1.10 V vs SCE) is about 220
2-/3-
mV less negative than the half-wave potential of C60
(-1.32
(
26) (a) Banks, M. R.; Cadogan, J. I. G.; Gosney, I.; Hodgson, P. K. G.;
Langridge-Smith, P. R. R.; Rankin, D. W. H. J. Chem. Soc., Chem. Commun.
994, 1365-1366. (b) Gr o¨ sser, T.; Parto, M.; Lucchini, V.; Hirsh, A.; Wudl,
F. Angew. Chem., Int. Ed. Engl. 1995, 34, 1343-1345.
27
3-
V vs SCE) since, according to Nernst law, the amount of C60
is only about 0.02% at equilibrium. However, the fact that the
reaction requires the passing of excessive reducing charges to
1
3166 J. Org. Chem., Vol. 73, No. 8, 2008

Heterocyclic DeriVatiVes of Fullerene C60
Origin of Oxygen Atom in the Cyclic Phenylimidate. Since
TBAP could be a possible oxygen source for the reaction, a
control experiment was performed where tetra-n-butylammo-
nium hexafluorophosphate (TBAPF6) was used as the electrolyte
instead of TBAP. The same product was obtained as shown by
1
HPLC, UV-vis, and H NMR characterizations, indicating that
it is unlikely that the oxygen atom is from the perchlorate ions
used in the experiment. Due to the high sensitivity of C60 anions
toward oxygen, the reaction was carried out under inert
atmosphere with ultrahigh purity, and it is unlikely that C603
could be generated if there were traces of oxygen in the system.
Therefore, under the experimental conditions, the most likely
source for the oxygen atom is traces of water, which may exist
in the solvent and glassware used for experiment. Benzonitrile
is known to have some common impurities such as water,
-
2
8
benzoic acid, isonitriles, and different amines. It has been
shown that benzonitrile suitable for electrochemical study can
be obtained by distillation over phosphorus pentoxide at reduced
28
pressure. An irreversible wave corresponding to water reduc-
tion is shown at around -1.90 V versus SCE in undistilled
benzonitrile, which interferes severely with the fourth redox
process of C60, and it is not suitable for electrochemical
generation of C60 anions. In contrast, when distilled benzonitrile
is used, no irreversible wave corresponding to water reduction
at around -1.90 V versus SCE is observed, and the fourth redox
process of C60 is nicely displayed in the cyclic voltammogram
along with the first three redox processes. However, traces of
water might still exist in the distilled benzonitrile even if it is
suitable for electrochemical study and generation of C60 anions.
Eaton et al. have reported that even under a much more strict
drying condition, where the solvent was kept over 4 Å molecular
FIGURE 13. HPLC traces of crude mixture from benzonitrile solution
of (a) neutral C60, (b) monoanionic C60, (c) 90% electrolyzed dianionic,
and (d) trianionic C60. The mixture was eluted by toluene over a
semipreparative Buckyprep column at a flow rate of 4.0 mL/min with
the detector wavelength set at 380 nm.
proceed suggests that C602- is unlikely the real reactive species
for the reaction. Previous studies on the electrogenerated C60
dianion have shown that it is inert toward solvent containing a
8,11a-c
nitrile functional group such as benzonitrile and acetonitrile,
and no formation of compounds 2 or 1 has been reported from
electrogenerated dianionic C60 benzonitrile solution or reaction
of the C60 dianion with benzyl bromide, confirming that the
C60 dianion is unlikely the real reactive species for the reaction.
Notably, previous work on the ESR studies of the C60 monoan-
ion suggests the possibility that a significant amount of higher
C60 anions can be generated at a less negative reducing potential.
It has been shown that there is often a narrow spike in addition
-
6
sieves for 24 h at 400 °C, 10 Torr, trace water was still
considered to be the major impurity in the solvent.1
1b
Conclusion
This report demonstrates a novel type of reaction of C603
-
to the broad line width signal present in the ESR spectrum of
with benzonitrile and water. Several C60 derivatives with cyclic
the C60 monoanion.9
c,11b-f
Recent studies have shown that the
1
phenylimidate have been obtained and characterized by H
1
3
spike is due to a trace amount of C60 derivatives formed by the
reaction of higher C60 anions present in the C60 monoanion
solution,1 indicating that higher C60 anions can be generated
at a much less negative potential. Although the intensity of the
spike is not related directly to the amount of C60 dianion present
in the solution since C60 dianion is ESR silent,9 the amount of
C60 dianion does have a significant effect on the intensity of
the spike since it affects the amount of C60 derivatives formed
NMR, C NMR, MALDI MS, UV-vis, and X-ray single-
crystal diffractions. Control experiments of neutral, monoan-
ionic, dianionic, and trianionic C60 have shown that the C60
trianion is the real reactive species for the reaction. The origin
of the oxygen atom is attributed to traces of water present in
the system. The reported reaction provides a promising method
for preparation of heterocyclic C60 derivatives and opens up
new perspectives on fullerene chemistry. Further studies on the
reaction mechanism are currently under investigation in the
laboratory.
1e,f
c
2-
in the solution. Theoretically, only a very small amount of C60
can be generated at a reducing potential set for the electrogen-
eration of the C60 monoanion;11d however, it has been observed
that the intensity of the spike in the ESR spectrum for the C60
monoanion increases significantly as the electrolysis is carried
out with a coulometric value of more than one electron per C60
Experimental Section
Spectral Characterization of 1,4-Dibenzyl-2,3-Cyclic Phen-
+
1
1c,e,f
ylimidate C60 (1): Positive MALDI TOF MS m/z 1022(MH );
molecule,
indicating that the amount of higher C60 anion
1
UV-vis (hexane) λmax ) 206, 256, 336, 429 nm; H NMR (600
can be much higher than theoretical prediction under more
reducing conditions. It is therefore possible that a significant
MHz, in CS
2
, C
6
D
6
was used as the external lock solvent) δ 8.79
(d, 2H), 8.06 (t, 1H), 8.02 (t, 2H), 7.75 (d, 4H) and from 7.62 to
3-
amount of C60 can be generated when more than a theoretical
number of reducing charges is transferred, even though the
reducing potential is less negative than the half-wave potential
7
1
.51 (m, 6H), 5.03 (d, 1H), 4.94 (d, 1H), 4.51 (d, 1H), 4.49 (d,
H); C NMR (150 MHz, CS /CDCl ) δ 162.9 (1C, CdN), 153.4
2 3
13
(1C), 153.0 (1C), 149.4 (1C), 149.3 (1C), 149.1 (1C), 148.9 (1C),
148.1 (1C), 147.3 (1C), 147.2 (2C), 147.2 (1C), 147.0 (1C), 146.6
2
-
3-
of C60 /C60
.
(
27) Dubois, D.; Moninot, G.; Kutner, W.; Jones, M. T.; Kadish, K. M.
(28) Kadish, K. M.; Anderson, J. E. Pure Appl. Chem. 1987, 59, 703-
714.
J. Phys. Chem. 1992, 96, 7137-7145.
J. Org. Chem, Vol. 73, No. 8, 2008 3167

Zheng et al.
(
1
(
1
(
1
(
1C), 146.2 (1C), 146.0 (1C), 145.5 (2C), 145.4 (2C), 145.4 (1C),
45.3 (1C), 144.9 (1C), 144.7 (1C), 144.6 (1C), 144.5 (2C), 144.3
2C), 144.3 (2C), 144.1 (2C), 144.1 (1C), 144.0 (1C), 143.9 (1C),
43.6 (1C), 143.6 (1C), 143.4 (1C), 143.2 (1C), 143.1 (1C), 143.0
1C), 142.9 (1C), 142.6 (1C), 142.3 (1C), 142.2 (1C), 141.9 (1C),
41.9 (1C), 141.7 (1C), 141.7 (1C), 139.8 (1C), 139.5 (1C), 138.1
1C), 137.5 (1C), 137.1 (1C), 136.7 (1C), 135.3 (1C), 134.9 (1C,
840.04494, found 840.04214; UV-vis (n-hexane) for 2 λmax ) 207,
1
255, 314, and 415 nm; H NMR (600 MHz, in CS
2
, C
6
D
6
was used
as the external lock solvent) δ 8.82 (d, 2H), 8.08 (t, 1H) and 8.02
(t, 2H); 13C NMR (150 MHz, CS , DMSO-d was used as external
2
6
lock solvent) δ 164.3 (1C, CdN), 147.5 (2C), 147.1 (1C), 145.7
(4C), 145.5 (2C), 145.4 (2C), 145.3 (2C), 145.1 (2C), 145.0 (2C),
145.0 (2C), 144.8 (2C), 144.5 (2C), 144.4 (2C), 143.9 (2C), 143.6
(2C), 143.0 (2C), 142.1 (3C), 142.1 (2C), 142.0 (2C), 141.7 (4C),
141.6 (2C), 141.4 (2C), 141.3 (4C), 139.8 (2C), 138.9 (2C), 137.1
Ph), 134.3 (1C, Ph), 132.4 (1C, Ph), 132.1 (2C, Ph), 132.0 (2C,
Ph), 128.9 (2C, Ph), 128.8 (2C, Ph), 128.1 (2C, Ph), 127.9 (2C,
Ph), 127.5 (1C, Ph), 127.2 (1C, Ph), 126.9 (1C, Ph), 98.0 (1C, sp ,
C-O), 91.5 (1C, sp , C-N), 62.9 (1C, sp , C-CH Ph), 61.9 (1C,
2
2 2 2
sp , C-CH Ph), 46.6 (1C, CH ), 45.9 (1C, CH ).
3
3
3
(
(
2C), 135.4 (2C), 96.6 (1C, sp , C-O), 91.6 (1C, sp , C-N), 131.7
3
3
1 C, Ph), 128.7 (2C, Ph), 128.3 (2C, Ph), 126.5 (1C, Ph).
3
X-ray Single-Crystal Diffraction of 1,4-Dibenzyl-2,3-Cyclic
Acknowledgment. We thank the reviewers of the manuscript
Phenylimidate C60 (1). Black needle-like crystals of 1 were
obtained by slowly diffusing hexane into a CS solution of 1 at
2
for their deep insights and constructive remarks. We thank
Professors Luis Echegoyen, Karl Kadish, and Shuichi Fukuzumi
for inspiring discussions. The work was partly supported by
Hundred Talents Program of the Chinese Academy of Sciences,
NSFC International Collaboration Grants 00550110367 and
00650110171.
room temperature. Single-crystal X-ray diffraction data were
collected on a Bruker SMART Apex equipped with a CCD area
detector using graphite-monochromated Mo KR radiation (λ )
0
.71073 Å) in the scan range 1.65° < θ < 25.03°. The structure
was solved with the direct method of SHELXS-97 and refined with
full-matrix least-squares techniques using the SHELXL-97 program
within WINGX. Crystal data of 1‚0.5 CS
2
: C81.50
H
19NOS, M
060.04, triclinic, space group P 1h , a ) 9.9186(8) Å, b )
2.4763(10) Å, c ) 18.4869(14) Å, R ) 93.812(2)°, â )
w
)
Supporting Information Available: General experimental
1
1
1
1
methods, the CIF file for 1,4-dibenzyl-2,3-cyclic phenylimidate C60
,
1
H NMR spectra of 1, 2, and the product isolated from the reaction
3
04.287(2)°, γ ) 95.025(2)°, V ) 2199.2(3) Å , z ) 2, Dcalcd
)
3-
mixture of C60
benzonitrile solution where 0.1 M tetra-n-
) was used as
-
3
-1
.601 Mg m , µ ) 0.139 mm , T ) 187(2) K, crystal size 0.23
0.16 × 0.10 mm; reflections collected 11 403, independent
reflections 7612; 4493 with I > 2σ(I); R ) 0.0698 [I > 2σ(I)],
wR ) 0.1258 (all data), wR ) 0.1713
all data), GOF (on F ) ) 1.026.
Spectral Characterization of [6,6] Cyclic Phenylimidate C60
butylammonium hexafluorophosphate (TBAPF
6
×
13
electrolyte, and C NMR spectra of compound 1, 2, and the
expanded region from 148 to 124 ppm for compound 2. This
material is available free of charge via the Internet at http://
pubs.acs.org.
1
2
) 0.1431 [I > 2σ(I)]; R
1
2
2
(
+
(2): MALDI FT-ICR MS m/z calcd for C67
H
5
NO [M + H]
JO702678C
3168 J. Org. Chem., Vol. 73, No. 8, 2008