Functionalization of C60 via organometallic reagents

Article abstract of DOI:10.1016/j.tet.2008.08.017

The reaction of [60]fullerene with organolithium and Grignard reagents carrying orthoester, acetal or other end groups yielded adducts 3-5 at the 6-6 bond of C60 after quenching with trifluoroacetic acid. The adducts could be easily methylated or benzylated with methyl iodide or benzyl bromide in the presence of potassium tert-butoxide to yield exclusively the 1,4-disubstituted C60 6 and 7a,b. Cleavage of the orthoester, acetal and silylether groups gave the corresponding carboxylic acid 9, aldehydes 10a,b and 11 and alcohols 12 and 13a,b. The carboxylic acid 9 readily reacted with alanine ethyl ester under standard peptide coupling conditions to give 14 in 55% yield. Attempts to generate a silyl enol ether from the reaction of aldehyde 10b with TIPSOTf and triethylamine failed. Instead the reaction led to a cyclized ether 16a (or alcohol 16b in the absence of silylating agent) resulting from the addition of an initially formed fulleride anion to the aldehyde group. The corresponding acetal 4b reacted similarly. The reaction of aldehyde 10b with aniline also gave a cyclized product 19. Surprisingly, aldehyde 11, which no longer carried an acidic fullerene proton reacted with aniline to give a product 20 resulting from an intramolecular Diels-Alder reaction followed by aromatization of a primarily formed N-phenylimine. Alcohol 13b could be readily converted to the corresponding bromide using tetramethyl-α-bromoenamine. The bromide was reacted with the carbanion derived from the protected glycine derivative to yield the diastereomeric fullerene amino acid derivatives 1-benzyl-4-[α-propyl-tert-butylglycinate benzophenone imine] 1,4-dihydro[60]fullerenes 24a and 24b.

Full text of DOI:10.1016/j.tet.2008.08.017

Tetrahedron 64 (2008) 10319–10330
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Functionalization of C60 via organometallic reagents
,
b
a
a
c
b
a,
*
Elise Champeil a , Conor Crean , Carlos Larraya , Gennaro Pescitelli , Gloria Proni , Leon Ghosez
´
a Institut Europe´en de Chimie et Biologie-IECB, Univ. Bordeaux 1-CNRS UMR 5248, 2, rue Robert Escarpit, F-33607 Pessac, France
b Department of Science, John Jay College of Criminal Justice, 445 west 59th street, New York, NY 10019, United States
c Dipartimento di Chimica e Chimica Industriale, Universita di Pisa, Pisa, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 April 2008
Received in revised form 22 July 2008
Accepted 7 August 2008
Available online 9 August 2008
The reaction of [60]fullerene with organolithium and Grignard reagents carrying orthoester, acetal or
other end groups yielded adducts 3–5 at the 6–6 bond of C60 after quenching with trifluoroacetic acid.
The adducts could be easily methylated or benzylated with methyl iodide or benzyl bromide in the
presence of potassium tert-butoxide to yield exclusively the 1,4-disubstituted C60 6 and 7a,b. Cleavage of
the orthoester, acetal and silylether groups gave the corresponding carboxylic acid 9, aldehydes 10a,b
and 11 and alcohols 12 and 13a,b. The carboxylic acid 9 readily reacted with alanine ethyl ester under
standard peptide coupling conditions to give 14 in 55% yield. Attempts to generate a silyl enol ether from
the reaction of aldehyde 10b with TIPSOTf and triethylamine failed. Instead the reaction led to a cyclized
ether 16a (or alcohol 16b in the absence of silylating agent) resulting from the addition of an initially
formed fulleride anion to the aldehyde group. The corresponding acetal 4b reacted similarly. The reaction
of aldehyde 10b with aniline also gave a cyclized product 19. Surprisingly, aldehyde 11, which no longer
carried an acidic fullerene proton reacted with aniline to give a product 20 resulting from an intra-
molecular Diels–Alder reaction followed by aromatization of a primarily formed N-phenylimine. Alcohol
Keywords:
C60
Nucleophilic addition
Alkylation
Amino acid
Peptide
13b could be readily converted to the corresponding bromide using tetramethyl-
a-bromoenamine. The
bromide was reacted with the carbanion derived from the protected glycine derivative to yield the
diastereomeric fullerene amino acid derivatives 1-benzyl-4-[
imine] 1,4-dihydro[60]fullerenes 24a and 24b.
a
-propyl-tert-butylglycinate benzophenone
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
acidic proton allowing the generation of a carbanion centre on the
C60 backbone, which can be further reacted with electrophiles.5f
In the context of a search of new fullerene-containing chemo-
therapeutic agents, we needed a series of C60-derived building
blocks characterized by the presence of functional groups linked to
the C60 backbone by a flexible tether of variable length. We selected
organometallic reagents carrying orthoester, acetal and silylether
functional groups as precursors for carboxylic acids, aldehydes and
alcohols, respectively. In this paper, we describe our results on the
representative additions of functionalized organometallics to C60
as well as some transformations of the resulting adducts.
As a result of its unique properties, the all-carbon molecule C60
is a potential source of new materials or chemotherapeutic agents.1
This has been a stimulus for the study of chemical modifications,
which would not significantly modify the properties of the C60
backbone while allowing the introduction of substituents required
for a specific application.2 Some of the most commonly studied
reactions of C60 include cycloadditions,3 cyclopropanations,4 ad-
dition of organometallic reagents,5 and photo-induced electron
transfer reactions.6 Initial studies on the addition of alkyl lithium
and Grignard reagents to C60 have setup the best conditions for the
formation of a monoadduct. Thus Hirsch et al. have shown that
12 equiv of the Grignard reagent derived from 2-(2-bromoethyl)-
1,3-dioxolane was required to give an optimum yield of 52% of the
1,2-monoadduct 1-[2-(1,3-dioxanyl)ethyl]-2-hydrofullerene[60].5b
Monoadducts have also been obtained from the reaction of lithium
acetylides to C60.5d These dihydrofullerene adducts contain an
2. Addition of functionalized organometallics to C60
2.1. Synthesis of an organodihydrofullerene orthoester
The reaction of [60]fullerene with organolithium and Grignard
reagents5 had never been exploited to incorporate a protected
carboxylic acid functionality onto the C60 sphere. An orthoester
was chosen as an equivalent for a carboxylic acid with no acidic
a
-proton. Since we expected some unstability of the classical
* Corresponding author. Tel.: þ33 5 4000 2217/2223; fax: þ33 5 4000 2215.
E-mail address: l.ghosez@iecb.u-bordeaux.fr (L. Ghosez).
methyl or ethyl orthoesters under the conditions of formation of
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.08.017

10320
E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
the organometallic reagents, we decided to use the more stable
bicyclic orthoester prepared by following the general procedure
described by Corey (Scheme 1).7a,b
associated sp3 fullerene core carbons at 65.4 and 65.6 ppm. The
UV–vis spectra showed an absorption band at 432 nm confirming
1,2-additions of the Grignard reagents. Further characterization
was provided by mass spectral data showing molecular ions (MþH)
at 824.1 for 4a and 838.1 for 4b.
O
i
Br
Br
HO
O
O
O
2.3. Synthesis of an organodihydrofullerene tert-
butyldimethylsilylether
ii
O
O
iii
O
O
The Grignard reagent9 obtained from 1-bromo-3-(tert-butyldi-
methylsilyloxy)propane was added to a toluene solution of C60
immediately generating a dark green solution indicative of the
formation a the fulleride anion.10 Quenching with excess tri-
fluoroacetic acid and work-up gave 1-hydro-2-[30-(tert-butyldi-
methylsiloxy)propyl]-1,2-dihydro[60]fullerene 5 in 67% yield based
on consumed C60 (Scheme 4). The separation of adduct 5 from C60
could be effected by flash chromatography using a 10:1 mixture of
cyclohexane/toluene. The 1H NMR spectrum of 5 revealed a singlet
at 6.51 ppm characteristic of the proton attached to the fullerene
core of a monoadduct. The other NMR signals unambiguously
showed the attachment of a tert-butyl dimethylsiloxypropyl group.
The 13C NMR spectrum showed signals for two sp3 carbons of the
fullerene core at 63 and 64 ppm together with 28 signals in
the region 135–156 ppm. The number of signals suggested that the
product also resulted from a 1,2-addition. The mass spectrum
showed the correct molecular ion (MþH) at 896.1.
O
I
O
1
Scheme 1. Reagents and conditions: (i) BrCH2CH2CH2COCl, pyridine/THF, 0 ꢀC, 1 h; (ii)
BF3$Et2O, CH2Cl2, 0 ꢀC, 3 h; (iii) NaI, DMF, 100 ꢀC, 2 h.
Compound 1 was transformed into the corresponding organo-
lithium reagent by reaction with n-BuLi in THF. Treatment of
a toluene solution of [60]fullerene with 1.6 equiv of 1-(2-pro-
pyllithium)-4-methyl-2,6,7-trioxabicyclo[2,2,2]-octane 1 gave a black
precipitate, which, upon quenching with trifluoroacetic acid, gave
1-hydro-2-[30-(400-methyl-200,600,700-trioxabicyclo[200,200,200]-octyl)-
propyl]-1,2-dihydro[60]fullerene 3 as a black solid in 20–30% yield
(Scheme 2).
O
O
H
i
ii
O
+
Li
O
O
O
H
2
3 (22%)
i, ii
BrMg
O Si
+
Si
O
Scheme 2. Reagents and conditions: (i) addition of 2 (prepared from the reaction of
n-butyl lithium with 1 in 50:50 pentane/diethylether) to C60 in toluene at ꢁ78 ꢀC, then
rt; (ii) excess CF3COOH.
5 (67%)
Scheme 4. Reagents and conditions: (i) Grignard reagent in THF added to C60 in
toluene, rt; (ii) excess CF3COOH.
2.2. Synthesis of organodihydrofullerene acetals
Two commercially available 2-(bromoalkyl)-1,3-dioxolanes
were used as precursors of the Grignard reagents for the addition
reactions to C60 following the general procedure described by
Hirsch et al.5b Thus, a molar solution of C60 in toluene was treated
with 12 equiv of the Grignard reagents derived from 2-(2-bro-
moethyl)-1,3-dioxolane and 2-(3-bromopropyl)-1,3-dioxolane. A
colour change (purple to dark green/black) immediately occurred
suggesting the formation of a fulleride anion. Protonation with
trifluoroacetic acid yielded the fullerene acetals 4a (55%) and 4b
(42%) (Scheme 3).
The UV–vis and NMR spectra of adducts 4a and 4b were in
agreement with C60 derivatives bearing two organic groups
attached at the 1,2-position of the 6,6-junction bond. A singlet at
6.49 ppm was observed in the 1H NMR spectrum, which is
characteristic of a proton attached to the C60 core in a fullerene
monoadduct. The 13C NMR showed 27 of the 30 expected sp2 car-
bon resonances for a product with a 1,2-addition pattern, with the
3. Alkylation of the 1,2-organodihydrofullerenes
The proton attached to the fullerene core in 1,2-adducts is acidic
and its pKa has been estimated at 5.7 for 1-hydro-2-tert-butyl 1,2-
dihydro[60]fullerene in dimethylsulfoxide.5a This acidity limits the
chemistry that could be carried out on such fullerene derivatives and
thus we chose to synthesize derivatives where the proton was
replaced by a methyl or benzyl group. This substitution should give
fullerene derivatives with increased solubility in organic solvents,
which would then ease purification by chromatography and also help
the investigation of further reactions of the substituted fullerenes.
The addition of organolithium or organomagnesium reagents to
C60 followed by protonation of the resulting carbanions has been
shown to give 1,2-adducts.5e This is an energetically favourable
situation (>8 kcal/mol more stable) because the corresponding 1,4-
adduct contains a double bond in one of the pentagons. However,
the replacement of hydrogen by bulkier substituents could reverse
this thermodynamic preference as shown previously in a number of
cases.11
C60 adducts 4b and 5 were deprotonated by treatment with
1 equiv of potassium tert-butoxide in THF to form green solutions
indicative of the formation of the fulleride anion (Scheme 5).
Compounds 6 and 7a were formed after reaction of the corre-
sponding fulleride anion with 100 equiv of methyl iodide at room
temperature for 24 h. Compound 7b was formed upon treatment
with 6 equiv of benzyl bromide for 2 h in refluxing THF.
O
H
( )n
O
+
i, ii
BrMg
O
( )n
O
4a: n=1 (55%)
b: n=2 (42%)
The 1,4-relationship between the two [60]fullerene substituents
was demonstrated by the replacement of the absorption at 436 nm
Scheme 3. Reagents and conditions: (i) addition of the Grignard reagent (prepared
from the corresponding bromide in THF) to C60 in toluene, rt; (ii) excess CF3COOH.

E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
10321
OH
CH3
H
O
O
O
i, ii
O
H
H
O
O
i
HO
O
O
O
ii
4b
6 (59%)
8 (55 %)
3
complex mixture
iii
CH3
H
i, ii
OTBDMS
OTBDMS
OH
O
OTBDMS
OTBDMS
H
5
7a (77%)
Ph
9 (45%)
H
i, iii
Scheme 6. Reagents and conditions: (i) DCM, trifluoroacetic acid, H2O, rt; (ii) NaOH or
K2CO3 or NaH or LiOH in toluene, rt; (iii) trifluoroacetic acid, H2O, 80 ꢀC.
5
7b (65%)
the fullerene core at
region 136–157 ppm. A singlet at
d
¼59.9 and 65.5 together with 29 signals in the
Scheme 5. Reagents and conditions: (i) potassium tert-butoxide, THF, 15 min, rt; (ii)
CH3I, 24 h, rt; (iii) benzyl bromide, 2 h, reflux.
d
¼6.65 was observed in the 1H
NMR spectrum, which is characteristic of a proton attached to the
C60 core in a fullerene monoadduct. UV–vis spectra showed an
absorption band at 432 nm confirming the 1,2-addition pattern.
Compound 9 is the first fullerene derivative with a carboxylic acid
functionality linked to the C60 sphere via a flexible alkyl chain.
with a broad shoulder at 450 nm.12 13C NMR spectra confirmed this
assignment: products 6, 7a and 7b, which have no symmetry ele-
ments showed the expected 60 signals. The spectra also show the
presence of the expected sp3 hybridized carbon atoms in the C60
4.2. Synthesis of organodihydrofullerene aldehydes
core (signals appeared at
for 7a, and at
d
¼58.4 and 65.2 for 6, at ¼59.5 and 63.34
d
d
¼60.9 and 63.1 for 7b). The high resolution mass
Despite the fact that a fullerene acetal had been previously
synthesized, its transformation into aldehyde had not been repor-
ted.5b Hydrolysis of acetals 4a, 4b and 6 were carried out in
refluxing toluene using a trifluoroacetic acid/water mixture to give
the organodihydrofullerene aldehydes 10a, 10b and 11 in 40–63%
yield (Scheme 7).
spectra (HRMS) showed the correct molecular ions for compounds
6, 7a and 7b (837.0897 (MþH) for 6, 909.1675 (MþH) for 7a and
985.1994 (MþH) for 7b). The 1H NMR spectra of 6 and 7a revealed
singlets at
protons in the
d
¼2.86 for 6 and
d¼2.83 for 7a characteristic of methyl
b
position of the fullerene core.
Since the alkylation reaction is irreversible, the preference for
substitution at position 4 versus 2 organodihydrofullerene proba-
bly resulted from less steric interactions in the linear transition
state when the alkyl chain and the incoming electrophile are much
further apart.
The resulting aldehydes were insoluble in toluene and CS2 had
to be used both for work-up and for NMR analysis. 1H NMR spectra
showed fullerene protons at
d
¼6.43 (10a) or 6.59 (10b) and alde-
hyde protons at
d
¼10.26 (10a), 10.10 (10b) and 10.14 (11). The
methyl group attached to the fullerene sphere (11) appeared at
d
¼2.90. 13C NMR spectra of 10b and 11 showed two fullerene sp3
4. Cleavage of the protecting groups
signals at
d
¼60.1 and 60.3 and ¼54.1 and 62.1, respectively; 26 of
d
the 30 expected core fullerene sp2 carbons for 10b and 58 signals
for the sp2 carbon atoms of 11 were also observed.
4.1. Synthesis of organodihydrofullerenecarboxylic acid
Treatment of the fullerene orthoester 3 with trifluoroacetic acid
in toluene at room temperature partially cleaved the orthoester to
give compound 8 in 55% yield (Scheme 6). This type of ester is
known to give the corresponding carboxylic acid upon treatment
with base.12a However, reactions of 8 with various bases (NaOH,
LiOH, K2CO3) at room temperature resulted in the formation of an
insoluble black powder, which could not be characterized. This was
probably due to the high electrophilicity of the fullerene sphere,
which makes it more susceptible to nucleophilic attack than the
ester group.2b
H
H
O
O
i
H
( )n
( )n
O
10a (53%)
b (40%)
4a : n=1
b : n=2
The ester group was eventually cleaved by treatment of 8 in
a mixture of trifluoroacetic acid and water in toluene at 80 ꢀC for
16 h. The crude carboxylic acid 9 was washed with methanol, ether
and CHCl3. The resulting black powder was taken up in pyridine
and insoluble material was filtered off. The product could not be
purified by column chromatography. The 13C NMR of the carboxy-
late salt in deuterated pyridine showed the presence of a carboxy-
O
O
O
H
i
CH3
CH3
11 (60%)
6
late group with a signal at
d
¼176.0 for the carbonyl carbon. The 13
C
NMR spectrum showed signals for two sp3 hybridized carbons of
Scheme 7. Reagents and conditions: (i) toluene, trifluoroacetic acid, H2O, reflux.

10322
E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
4.3. Synthesis of organodihydrofullerene alcohols
amide proton appeared at
showed signals at
This result showed that standard peptide coupling conditions
are compatible with the presence of an acidic proton on the ful-
lerene core.
d
¼6.29 and the 13C NMR spectrum
d
¼171.7 and 173.2 for the two carbonyl groups.
Removal of the tert-butyldimethylsilyl protecting group of 5, 7a
and 7b was easily performed upon addition of 6 N HCl.5g The cor-
responding alcohols 12,13a and 13b were obtained in 53–72% yield.
Alcohol 12 was insoluble in toluene, and CS2 was again necessary
for purification by chromatography (Scheme 8).
5.2. Attempt at the formation of a silyl enol ether from
aldehyde 4b
Trialkylsilyl enol ethers are useful synthetic intermediates,
which allow the introduction of a wide variety of functional groups
at the a-position of an aldehyde or a ketone. These reactions are
electrophilic additions, which do not require the presence of a base
and should therefore be compatible with the presence of an acidic
proton on the C60 core.
The formation of the silyl enol ether 15 derived from 10b was
attempted by reaction of the aldehyde with a highly electrophilic
silylating agent (triisopropylsilyl-trifluoromethanesulfonate) in the
H
H
i
O Si
OH
OH
5
12 (75%)
O Si
i
presence of a weak base in order to favour reaction at the a-carbon
atom of the carbon chain and avoid deprotonation of the fullerene
core (Scheme 10). Aldehyde 10b was stirred at 80 ꢀC in a solution
of TIPSOTf and triethylamine in DMF. TLC showed complete
disappearance of aldehyde 10b after 12 h.
R
R
7a : R=Me
b : R=PhCH2
13a (42%)
b (72%)
Scheme 8. Reagents and conditions: (i) 6 N HCl, THF, rt.
+OSi(iPr)3
O
O
ii
i
OSi(iPr)3
H
H
Compound 12 was not sufficiently soluble to obtain a good
H
H
H
quality 13C NMR spectrum but the 1H NMR spectrum confirmed the
loss of the silyl group and showed a signal at
d
¼6.51 for the full-
15
10b
ii
erenic proton. The mass spectral data confirmed the structural
assignment (molecular ion (MþH) at 782.1).
The fullerene alcohols 13a and 13b could be purified by chro-
matography. They showed higher solubility in toluene and CS2 as
compared to alcohol 12. The 13C NMR spectrum of 13b showed two
16 a = Si(iPr)3
b = H
H
fullerene sp3 signals at
d
¼61.0 and 63.1; 52 of the expected 58
-
-
OR
H
fullerene sp2 signals are also present in the 13C NMR spectrum in
the range 136–159 ppm. The 1H NMR spectrum of 13b showed
Scheme 10. Reagents: (i) Et3N; (ii) TIPSOTf.
signals at
d
¼7.47, 7.55 and 7.67 for the benzyl protons. Compound
13b also showed characteristic CH2/CH3 signals in the range
2–4.5 ppm similar to those of 13a. (cf. section 7). Compounds 13a
and 13b were also characterized by mass spectrometry.
A black powder was isolated after filtration and evaporation of
the solvent. This powder had an enhanced solubility compared to
aldehyde 10b and was very soluble in most common organic sol-
vents. The powder was purified by flash chromatography to give
compound 16a in 75% yield. We expected this powder to be the
desired silyl enol ether 15. However, the 1H NMR spectrum revealed
that there was no longer a fullerenic proton in this product as no
5. Selected transformations of 1,2- and 1,4-[60]fullerene
adducts
signal around
d
¼6.5 was observed. Furthermore, signals corre-
5.1. Coupling of hydro[60]fullerenyl acid 9 with amino acids
sponding to vinyl protons were also absent. The 1H NMR data
showed that all alkyl protons were no longer equivalent. Also,
a sharp absorption at 431 nm in the UV–vis spectrum was in-
dicative of a C60 derivative with two organic groups attached at the
6,6-junction.8 The 13C NMR spectrum exhibited signals for two sp3
C60 carboxylic acid 9 was subjected to standard peptide cou-
pling conditions in the presence of alanine ethyl ester hydrochlo-
ride (Scheme 9). In contrast to the acid precursor 9, the coupling
product 14 was easily purified by chromatography and isolated as
a brown powder soluble in most common organic solvents. The
structure of 14 was established by standard spectroscopic methods.
carbons in the C60 core (d 68.4, 69.4) and 58 signals (partially
overlapping) for the sp2 carbons of the C60 core. On the basis of
these data, we concluded that a stereogenic centre was attached to
the fullerene sphere and that the product was deprived of Cs
symmetry. These data were indicative of a cyclic ether probably
resulting from deprotonation of 10b followed by addition of the
fullerenyl anion to the aldehyde group. Indeed treatment of 10b
with triethylamine in DMF yielded the simple cyclized product 16b.
Cyclization products had already been observed by Cousseau et al.
when they reacted C602ꢁ with diiodo derivatives I–(CH2)n–I (n¼3
and 4).13 We also decided to examine if an enol ether could be di-
rectly formed from acetal 4b: the oxygen atom of an acetal is indeed
more basic than a carbonyl group and silylation should occur more
The 1H NMR signal for the fullerenic proton (
downfield compared to that of the acid precursor (
d
¼6.56) shifted
d
¼6.65). The
H
CH3
N
OH
+
H2N
CH3
i
H
O
H
O
H
H
COOEt
COOEt
14 (66%)
Scheme 9. Reagents and conditions: (i) EDC$HCl, HOAt, DCM.
9

E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
10323
readily with the net result of increasing the acidity of the
a
-proton
examine the reaction of aniline on the corresponding methylated
of 4b, which could then be preferentially removed by the tertiary
base. Triisopropylsilyl-trifluoromethanesulfonate was slowly added
to a CS2 solution of 4b and then N,N0-diisopropylethylamine was
added slowly to the resulting mixture (Scheme 11).
aldehyde 11 (Scheme 13).
O
N
H
N
H3C
H3C
6
5
TIPS
+
7
1
O
4
HO
3 2
H3C
Diels-Alder
8
i
O
i
O
condensation
H
-
O
17
11
O
H
N
H
TIPS
TIPS
H
+
N
O
O
O
O
4b
Ha
H
H3C
H
H3C
aromatisation
Base
18
Scheme 11. Reagents and conditions: (i) TIPSOTf in CS2, 30 ꢀC, 12 h then Hu¨nig’s base.
20
Scheme 13. Reagents and conditions: (i) aniline (1 equiv), benzene, 4 Å MS, rt, 3 days.
However, these conditions also led to the formation of a cyclized
product 17 instead of the expected enol ether 18. The structure of 17
was determined from analysis of its spectroscopic properties,
which were similar to those of compounds 16a and 16b. These
results suggested that the formation of a silyl enol ether from an
aldehyde carried by a side chain attached to C60 would require
the replacement of the acidic hydrogen of the fullerene core by
a base-insensitive substituent if the length of the tether allows for
a cyclization reaction.
The reaction of 11 with an equimolar amount of aniline was
carried out in benzene in the presence of 4 Å MS. The condensation
was expected to proceed smoothly as high yields of a full-
eroaldimine were previously obtained by Saigo et al. under the
same conditions.14 The reaction was stopped after 16 h and the
crude 1H NMR spectrum showed no imine proton.
TLC revealed one major spot and compound 20 was isolated by
chromatography. The 1H NMR spectrum showed four distinct aro-
5.3. Reaction of 10b and 11 with aniline
matic protons at
d
¼7.84, 7.02, 6.69 and 6.57. The proton at the bi-
cyclic ring junction (He) appeared as an octuplet with coupling
constants of 2.8, 4.4 and 7.8. The four alkyl protons on the five-
membered ring (Ha, Hb, Hc, Hd) were all diastereotopic and
The reaction of 10b with amines was expected to lead to the
corresponding imines, which are themselves precursors of amines
and amino acids. We selected the reaction of 10b with aniline as
model for this transformation (Scheme 12). A mixture of 10b, ani-
line and molecular sieves in benzene yielded a black powder 19,
which could be purified by chromatography on silica gel (yield:
65%). Compound 19 showed enhanced solubility in most common
organic solvents compared to aldehyde 10b. The 1H NMR spectrum
appeared between
d
¼2.30 and 3.01. The protons of the methyl
group at the C60 surface gave a sharp singlet at
d
¼2.35. In the 13C
NMR spectrum 53 fullerene sp2 carbon signals and 1 phenyl ring
signal (assigned by HMBC and HMQC experiments) were observed
between
phenyl ring appeared at
d
¼133 and 159. The remaining five sp2 carbons of the
d
¼129.6, 129.3, 124.0, 120.6 and 116.8. The
of 19 showed no signal in the
proton of the imine group. The absence of a signal at around
d
¼7.5 region as expected for the
carbon at the bicyclic junction appeared at
of the alkyl groups of the five-membered ring appeared at
and 34.1. The signal at
d
¼65.6 while the proton
d¼160
d¼40.7
typical for a carbon of an imine bond further confirmed the absence
of the expected condensation product. Also no signal was observed
d
¼27.3 was characteristic of the sp3 carbon
for the protons of the methyl group on the C60 surface. The tetra-
at around
d
¼6.5 indicative that the fullerenic proton was lost dur-
addition pattern on C60 was confirmed by the 13C NMR spectrum,
ing the transformation. As a matter of fact, NMR and UV spectra of
19 were quite similar to those of cyclic ether 16a or alcohol 16b. The
alkyl protons were now nonhomotopic, the UV–vis spectrum
showed a sharp absorption at 430 nm in agreement with a C60
derivative with two organic groups attached at the 6,6-junction
bond and the 13C NMR spectrum exhibited signals for two sp3
which exhibited four sp3 carbons of the C60 framework at
d
¼72.1,
60.8, 58.0 and 53.5. Table 1 shows the 1H NMR chemical shifts of
compound 20.
Compound 20 probably resulted from the intermediate imine,
which then could have undergone a Diels–Alder reaction between
the newly formed C]N bond, the phenyl ring and a C]C bond on
the C60 surface. Fullerene derivatives are good dienophiles and
readily undergo Diels–Alder reactions. However, in our case, this
mechanism implies a rather unusual localization of the electrons in
the diene region. Alternatively, the bicyclic compound could be the
carbons in the C60 core (
d
¼67.2, 70.0) and 58 signals (partially
overlapping) for the sp2 carbons of the C60 core. We concluded that
the reaction product was structure 19 resulting from a nucleophilic
attack of the fullerenide anion on the imine group. Thus access to an
imine from aldehyde 10b was not possible as a result of the pres-
ence of the acidic fullerene proton. We therefore decided to
Table 1
1H NMR chemical shifts for compound 20
O
N
Protons
Chemical shifts (ppm)
H
i
H
CH3
2.35
H
NHPh
H
Ha, Hb
Hc, Hd
He
2.30, 2.38
2.78, 3.01
4.27
10b
19
Hf, Hg, Hh, Hi
7.84, 7.02, 6.69, 6.57
Scheme 12. Reagents and conditions: (i) C6H5NH2 (1 equiv), benzene, rt, 72 h.

10324
E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
result of an electron transfer reaction mechanism between the
phenylimine donor moiety and the fullerene acceptor moiety. The
third step would involve the rearomatization of the six-membered
ring (Scheme 14). Three double bonds on the surface of C60 (C2–C3,
C5–C6, C7–C8) were in close enough proximity to react with the
newly formed imine bond and the phenyl ring. Thus three different
adducts could in principle be obtained (Fig. 1). The structural
assignment of the product was difficult because structures 20, 21a
and 21b would be expected to show similar 1H and 13C NMR
spectra. HMBC and HMQC experiments were performed but did not
provide sufficient indication regarding the addition pattern be-
cause of the absence of a fullerenyl proton directly attached to the
C60 sphere. Hence, the connectivities of the sp3 and sp2 hybridized
fullerenyl carbon atoms could not be determined via such NMR
experiments. The connectivities of the carbon atoms can be
assigned with 2D INADEQUATE or C–C HOHAHA technique but this
would have required long experimental time and was not available
at the time of the realization of this work. However, a 2D T-ROESY
NMR experiment provided information in elucidating the addition
pattern of the C60 addend. No NOE correlation between the methyl
CH3 and any of the protons from the bicyclic structure was
observed, which was indicative of a 1–4–7–8 addition product
(compound 20) rather than a 1–2–3–4 (compound 21b) or 1–4–5–6
(compound 21a) addition. Indeed, in the case of a 1–2–3–4 or
a 1–4–5–6 addition, the phenyl protons (Hf, Hg, Hh, Hi) would have
been expected to show some NOE correlation with the methyl
group on the C60 sphere, as they are located in close proximity.
Pri
H
H
O
N
H
H
H
H
Me
22
Figure 2. Compound 22 synthesized by Rubin et al.15
Table 2
UV–vis absorption bands for compound 20
Compound
Addition pattern
Absorption bands
20
1–4–7–8
1–4–7–8
1–2–3–4
252, 314, 406, 446, 526, 682
252, 312, 408, 446, 510, 682
256, 328, 404, 432, 648, 676, 712.
Rubin’s compound
Rubin’s compound
compound 20 retained a C]C double bond at the 5,6-ring
junction.
5.4. Synthesis of a protected glycine derivative
The strategy adopted to generate a protected fullerene amino
acid derivative from the fullerene alcohol 13b involved conversion
of the alcohol into a fullerene alkyl bromide followed by sub-
stitution of the halogen with an anion derived from N-(diphe-
nylmethylene)glycine tert-butyl ester.16
The conversion of the fullerene alcohol to alkyl bromide 23 in
44% yield was rendered possible by the use of (1-bromo-2-methy-
lprop-1-enyl)-N,N-dimethylamine,17 a smooth brominating agent,
which reacts under neutral conditions (Scheme 14).
Ph
N
O
O
Ph
Br
OH
Br
i
O
Ph
Ph
Ph
Ph
N
N
O
Ph
The diastereomeric fullerene amino acid derivatives 1-benzyl-4-
[a-propyl-tert-butylglycinate benzophenone imine] 1,4-dihy-
ii, iii
dro[60]fullerenes (24a and 24b) were generated via the addition of
N-(diphenylmethylene)glycine tert-butyl ester to 1-benzyl-4-[30-
bromopropyl] 1,4-dihydro[60]fullerene (23).
24a and b
23
13b
Alkylation of N-(diphenylmethylene)glycine tert-butyl ester has
been extensively studied.16,18,19,20 In general, the tert-butyl and
diphenylimine protecting groups can be removed under mild
conditions. The active methylene group of N-(diphenylmethyle-
ne)glycine tert-butyl ester has been exploited in the syntheses of
protected fullerene glycine derivatives obtained via Bingel cyclo-
propanation reactions. However, deprotection of those protected
fullerene glycine derivatives was not successful and was assigned to
the electron-withdrawing character of the fullerene sphere making
the imine group less susceptible to acidic or basic hydrolysis.21–24 In
our case, we expect molecules 24a and 24b to be more easily
deprotected due to the presence of the alkyl spacer group between
the fullerene and the amino acid group since electron deficiency in
Scheme 14. Reagents and conditions: (i) DCM, rt, 45 min; (ii) THF, ꢁ78 ꢀC, 60 min,
warming to rt, then 60 min at rt; (iii) NH4Cl.
Further evidence for the formation of 20 was obtained from its
UV–vis absorption spectrum. It showed remarkable similarity with
another compound 22 (Fig. 2) synthesized by Rubin et al., which
had an identical addition pattern (1–4–7–8).15
Both spectra show considerable reduction in band structure
compared to 1–2–3–4 or 1–4–5–6 addition products. Furthermore,
the number, position and intensity of the absorption bands of both
compounds were extremely similar. Table 2 shows the absorption
bands observed by Rubin et al.15 and by us (compound 20), and
those observed for a 1–2–3–4 addition product. Noteworthy were
the absorptions at 406 and 446 nm. They reflected the fact that
the fullerene will not interfere through five
s bonds. Initial alkyl-
ation reactions using 50% KOH (aq) as base to generate the active
anion were unsuccessful leading to decomposition. It was then
decided to generate the anion with LDA and use an excess of
N-(diphenylmethylene)glycine tert-butyl ester anion. The fullerenic
anion was formed as shown by the appearance of a green colour but
no alkylation product could be identified after quenching. NMR
analysis of the black powder suggested the presence of products
resulting from multiple additions. When the addition was carried
out using exactly 1 equiv of the anion or using a slight excess of the
fullerene alkyl bromide a pair of diastereomeric racemic products
24a and 24b were isolated and separation of the diastereomers was
possible by flash chromatography. The 1H NMR spectra of the two
products were practically identical except in the region of benzyl
Hc/d
Ha/b
Ha/b
Hc/d
He
Hc/d
Hf
Hc/d
Ha/b
Hc/d
He
Hf
Hf
Hg
Hh
Hg
Hh
H
H
N
Hc/d
Ha/b
Ha/b
H
N
Hg
Hh
He
N
5
3
6
6
Hi Me
Hi
Hi
1
7
Me
4
Ha/b
2
5
3
8
1
4
1
2
2
Me
6
5
3
4
20
21a
Figure 1. Possible Diels–Alder adducts.
21b

E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
10325
CH2 resonances. These two CH2 protons are diastereotopic and
therefore give rise to a pattern of double-doublets typical of an AB
system (Fig. 3).
Ph2
Ph1
N
COOtBu
4
In 24a, the two signals are almost isochronous and appear as
very distorted doublets with J¼12.6 Hz. In 24b, the doublets with
J¼12.8 Hz are separated by Dd¼250 Hz (0.50 ppm at 500 MHz). The
chemical shift difference in 24b could be explained by restricted
motion of the benzyl group as a result of an interaction with the
diphenylimine group. A similar phenomenon was observed in a
series of disubstituted fullerenes, 25a:(C6H5CH2)2C60, 25b:
(2-BrC6H5CH2)2C60, 25c: (3-BrC6H5CH2)2C60 and 25d: (4-BrC6H5-
CH2)2C60, studied by Kadish et al.25 Their 1H NMR spectra showed
two similar AB doublets for the benzyl CH2 protons of 25a, 25c and
25d; 25b gave a different pattern, similar to that of 24b with a much
larger Dd value (0.22 ppm) than the other three compounds (<0.1).
This was suggested to be due to a hindered rotation of the phenyl
ring in the 2-bromo compound in analogy with studies by White-
sides et al.,26,27 who showed that the conformation of the phenyl
ring with respect to the benzyl CH2 protons has an effect on the
NMR signal. This led us to undertake a molecular modelling study
to verify if a similar phenomenon was plausible for compounds 24a
and 24b.
3'
(Ph
)
2'
Bn
1
9
1'
Ha
Hb
Figure 4.
24b may explain the large anisochrony between Ha (pro-S) and Hb
(pro-R) protons for this diastereomer.
In addition, CS (conformational searches) calculations were run
using the procedure described in Section 7.26 to further highlight
structural discrepancies between 24a and 24b. The two sets of 50
low-energy structures showed in fact significant differences. We
were especially interested in structures featuring clear
p–p in-
teractions28 between the benzyl and the diphenylmethylene moi-
eties. For 24b, several low-energy structures showed interacting
phenyl rings, including the absolute minimum (shown as 24b[1] in
Fig. 5, right). It exhibited a clear face-to-face
p-stacking between
the benzyl phenyl PhBn and one imino phenyl (Ph1), which was
especially close to proton Ha (distancez3.3 Å from Ph1 centre)
with respect to Hb (z4.5 Å). In such a situation, Ha is more sub-
jected to the ring current exerted by Ph1, which may help
explaining the large chemical shift difference with Hb. For 24a, on
the contrary, only a few low-energy structures showed interacting
We deemed that phenyl–phenyl interactions29 between the
benzyl and the diphenylmethylene moieties could play a decisive
role in defining the preferential conformations adopted by 24a and
24b. Therefore, as a force field for the molecular modelling in-
vestigation, we choose OPLS as a result of the correct treatment of
intermolecular forces in benzene oligomers29 and other aromatic
phenyl rings, including the third most stable conformation (24a[3]
,
p
stackings30 reported with this force field. We built two di-
astereomeric structures assuming (fC)–(
S) configuration for 24a
and (fC)–(
R) for 24b, as shown in Figure 3 and ran molecular dy-
Fig. 5, left). This exhibited an edge-to-face interaction with much
larger distances Ph1–Hbz5.4 Å and Ph1–Haz7.0 Å.
a
In conclusion, the observed 1H NMR patterns for 24a and 24b
must be related to a different conformational freedom, possibly
associated with the ring current shift exerted by the diphenyl-
a
namics (MD) and conformational searches (CF) calculations with
OPLS.
methylene moiety. On this basis, (fC)–(
a
S) or (fA)–(
a
R) and (fC)–(
a
R)
MD (molecular dynamics) calculations at 300 K showed no
substantial difference in the phenyl–CH2 torsions between the two
diastereomers, as the C30–C20–C10–Ha and C30–C20–C10–Hb di-
hedral angles (see Fig. 4) spanned comparable ranges of values in
the simulations. On the contrary, the conformation of the benzyl
or (fA)–(
respectively (Fig. 3).
a
S) relative configurations were assigned to 24a and 24b,
6. Conclusions
group as
a whole varied somewhat between the two di-
astereomers: the C20–C10–C1–C9 dihedral was almost unrestricted
in 24a, while for 24b it tended to assume values around ꢁ60ꢀ and
180ꢀ. The restricted motion experienced by the benzyl moiety in
This study has confirmed that functional groups carried by
a tether can be readily introduced to C60 by addition of function-
alized organolithium or Grignard reagents. The resulting
orthoesters, acetals or silyl enol ethers can be readily transformed
into the parent carboxyl, carbonyl or alcohol functional groups.
Preliminary experiments on functional manipulations of these
O
(S)
4
1
O
N
H
Ph
(fC)
Ph
Ph
24a
O
(R)
4
1
O
N
H
(fC)
Ph
Ph
Ph
24b
Figure 5. Lowest-energy OPLS structures for 24a and 24b exhibiting phenyl–phenyl
interactions.
Figure 3. 1H NMR spectra of fullerene amino acid derivatives 24a and 24b.

10326
E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
tethered fullerene derivatives showed some limitations in the use
of these functionalized [60]fullerenes: thus, while peptide-type
coupling could be readily effected with a carboxylic acid derivative,
the corresponding aldehyde could not be transformed into an enol
ether, cyclization initiated by the formation of a fulleride anion
mixture was filtered off and extracted with water and brine
(2ꢂ20 mL). The solvent was evaporated and the resulting brown
residue was purified by flash chromatography (SiO2, toluene) to
yield
1-hydro-2-[30-(400-methyl-200,600,700-trioxabicyclo[200,200,200]-
octyl)propyl]-1,2-dihydro[60]fullerene 2 (27 mg, 22% yield) as
a black solid: mp >300 ꢀC; 1H NMR (5:1, CS2:CDCl3) 0.84 (s, 3H),
2.22 (t, J¼7.8, 2H), 2.72 (m, 2H), 3.43 (m, 2H), 4.04 (s, 6H), 6.50 (s,
1H); 13C NMR (5:1, CS2:CDCl3) 14.6, 20.9, 30.4, 37.0, 46.9, 59.6, 65.0,
72.7, 109.0, 135.9, 136.6, 140.0, 141.6, 141.6, 142.0, 142.0, 142.0, 142.0,
142.2, 142.5, 142.5, 143.2, 144.6, 144.7, 145.4, 145.4, 145.9, 146.2,
146.2, 146.3, 146.4, 146.5, 147.2, 147.3, 147.5, 153.9, 156.2; UV–vis
(CHCl3) lmax 255 nm, 325, 432. HRMS calculated 893.1177, found
893.1184.
occurred faster than the deprotonation
a to the aldehyde. The same
was observed with the corresponding acetals or when the aldehyde
was reacted with aniline. Replacement of hydrogen by methyl in
aldehyde 4b was expected to allow the formation of an imine by
reaction with aniline. This was the case but the imine spontane-
ously underwent an intramolecular Diels–Alder reaction involving
a 6–6 double bond of the fullerene core.
Clearly, further work would be needed to establish the impor-
tance of the length of the tether on the functional group manipu-
lation of these tethered fullerenes. It is expected that some
reactions will only be possible on derivatives where the acidic
fullerene proton is replaced by an inert alkyl group. This should not
be a major problem since the properties of the C60 core will not be
significantly modified after alkylation.
We have also demonstrated the synthetic use of nucleophilic
additions to generate a fullerenyl protected amino acid with
a flexible alkyl linker between the fullerene sphere and the amino
acid functional group. The two diastereomeric products thus pre-
pared showed different 1H NMR spectra, which were analyzed with
the help of molecular modelling. This opens a route to synthetic
analogs of phenylalanine carrying a [60]fullerene ring.
7.4. .1-Hydro-2-[20-(100,300dioxolanyl)ethyl]-1,2-
dihydro[60]fullerene 4a
2-(3-Ethylmagnesium bromide)-1,3-dioxolane was prepared
from the reaction between 2-(3-bromoethyl)-1,3-dioxolane
(1.206 g, 6.66 mmol) and magnesium turnings (24.3 mg, 6.66 mmol)
in THF (40 mL) at room temperature using a water bath to control
the temperature. The solution of the Grignard reagent was added
dropwise via a cannula to a rapidly stirred solution of C60 (400 mg,
0.55 mmol) in toluene (250 mL). An immediate colour change
(purple to dark green) was observed. The reaction was quenched
with excess trifluoroacetic acid (0.5 mL) and the solvents removed
in vacuo. Purification by flash chromatography (SiO2, toluene) gave
1-hydro-2-[200-(100,300-dioxanyl)ethyl]-1,2-dihydro[60]fullerene 4a
(250 mg, 55%) as a brown solid: mp >300 ꢀC; 1H NMR (5:1, CS2/
CDCl3) 2.95 (m, 2H), 3.58 (m, 2H), 4.08 (m, 2H), 4.21 (m, 2H), 5.40 (t,
J¼4.2, 1H), 6.48 (s, 1H); 13C NMR (5:1, CS2/CDCl3) 31.2, 40.4,
59.4, 63.9, 65.2, 103.6, 135.7, 136.2, 139.9, 140.0, 141.3, 141.6, 141.7,
141.7, 141.7, 141.9, 142.3, 142.7, 142.9, 144.3, 144.4, 145.1, 145.2, 145.2,
145.5, 145.8, 145.9, 146.1, 146.7, 147.1, 153.2, 155.2; UV–vis (CHCl3)
lmax 254 nm, 325, 405, 433. HRMS calculated 823.0759, found
823.0803.
7. Experimental
7.1. General
NMR spectra were recorded using a Bruker Avance 400NB
spectrometer. UV–vis spectra were recorded using a Varian Cary
300 UV–vis spectrometer. High resolution mass spectrometry was
carried out by Cesamo mass spectrometry service, University of
Bordeaux 1. C60 was purchased from Aldrich Chemical Company
or M.E.R. (Materials and Electrochemical Research company).
Solvents were dried by filtration over activated alumina. All re-
actions were carried out in standard glassware under an inert
atmosphere of N2 or Ar. Organometallic reactions were performed
in flame dried Schlenk tubes flushed three times with argon prior
to use. Solvents for organometallic reactions were transferred via
cannula and deoxygenated twice prior to use. 1-Bromo-3-(tert-
butyldimethylsiloxy)propane,9 2-(3-bromopropyl)-1,3-dioxolane31
and 1-(2-iodopropyl)-4-methyl-2,6,7-trioxabicyclo[2,2,2]-octane7b
were synthesized following the published procedures.
7.5. 1-Hydro-2-[30-(100,300-dioxolanyl)propyl]-
1,2-dihydro[60]fullerene 4b
2-(3-Propylmagnesium bromide)-1,3-dioxolane was prepared
from the reaction between 2-(3-bromopropyl)-1,3-dioxolane
(151 mg, 0.833 mmol) and magnesium turnings (31 mg, 1.25 mmol)
in THF (10 mL) at room temperature using a water bath to prevent
excessive reflux. After 2 h, the solution of the Grignard reagent was
added dropwise via a cannula to a rapidly stirred solution of C60
(50 mg, 0.069 mmol) in toluene (50 mL). An immediate colour
change (purple to dark green) was observed. The reaction was
quenched with excess trifluoroacetic acid (0.5 mL) and the solvents
were removed in vacuo. Purification by flash chromatography (SiO2,
toluene) gave 1-hydro-2-[30-(100,300-dioxolanyl)propyl]-1,2-dihy-
dro[60]fullerene 4b (67 mg, 42% yield based on consumed C60) as
a brown solid: mp >300 ꢀC; 1H NMR (5:1, CS2/CDCl3) 2.19 (m, 2H),
2.72 (m, 2H), 3.50 (m, 2H), 3.96 (m, 4H), 5.17 (t, J¼4.4, 1H), 6.49 (s,
1H); 13C NMR (5:1, CS2/CDCl3) 22.1, 34.9, 47.7, 60.2, 65.4, 65.6, 104.7,
136.3, 136.4, 137.0, 140.6, 140.7, 142.1, 142.4, 142.5, 142.5, 142.7,
143.0, 143.7, 145.1, 145.1, 145.8, 145.9, 145.9, 145.9, 146.3, 146.6,
146.7, 146.8, 146.9, 147.5, 154.1, 156.2. HRMS calculated 837.0916,
found 837.0923.
7.2. 1-(2-Iodopropyl)-4-methyl-2,6,7-trioxabicyclo[2,2,2]-
octane 1
The orthoester was prepared according to the literature
procedures.7b
1H NMR (CDCl3) 0.78 (s, 6H), 1.75 (t, J¼7.4, 2H), 1.98 (m, 2H), 3.19
(t, J¼7, 2H), 3.87 (s, 6H); 13C NMR (CDCl3) 6.78,14.4, 27.6, 30.12, 37.2.
HRMS calculated 299.0144, found 299.0148.
7.3. 1-Hydro-2-[30-(400-methyl-200,600,700-trioxabicyclo[200,200,200]-
octyl)propyl]-1,2-dihydro[60]fullerene 3
2.28 mL (1.6 equiv) of a 0.1 mol/L solution (ether/pentane, 1:1)
of
1-(2-iodopropyl)-4-methyl-2,6,7-trioxabicyclo[2,2,2]-octanyl
7.6. 1-Hydro-2-[30-(tert-butyldimethylsiloxy)propyl]-
lithium (prepared from the reaction of 1 equivalent of n-butyl-
lithium with 1 in 50:50 pentane:diethylether) was added to C60 in
toluene at ꢁ78 0C were syringed into toluene solution (100 mL) of
fullerene[60] (100 mg, 0.138 mmol). The reaction was quenched
with trifluoroacetic acid after 5 min at room temperature. The
1,2-dihydro[60]fullerene 5
3-(tert-Butyldimethylsiloxy)propyl magnesium bromide was
produced from the reaction of magnesium turnings (538 mg,
0.021 mol) and 1-bromo-3-(tert-butyldimethylsiloxy)propane

E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
10327
(2.11 g, 8.33 mmol) in THF (10 mL) at room temperature. After 1.5 h,
the solution of the Grignard reagent was added dropwise via
a cannula to a solution of C60 (0.5 g, 0.69 mmol) in toluene
(500 mL). The addition resulted in the immediate formation of
a dark green/black solution. The reaction was quenched with 1 mL
trifluoroacetic acid and the solvents were removed in vacuo. The
solution of the brown residue in 250 mL CHCl3 was washed with
50 mL of brine and water (2ꢂ50 mL), and dried over anhydrous
MgSO4. Evaporation of the solvents left a brown residue, which was
purified by flash chromatography (SiO2, cyclohexane/toluene, 10:1)
to give C60 (25 mg) and 1-hydro-2-[30-(tert-butyldimethylsiloxy)-
propyl]-1,2-dihydro[60]fullerene 5 (393 mg, 67% yield based on
consumed C60) as a brown solid: mp >300 ꢀC; 1H NMR (CDCl3)
0.23 (s, 6H), 1.04 (s, 9H), 2.80 (m, 2H), 3.50 (m, 2H), 4.15 (t, J¼6, 2H),
6.51 (s, 1H); 13C NMR (CDCl3) 5.1, 18.5, 26.1, 30.1, 43.6, 59.7, 63.0,
64.8, 135.9, 136.5, 140.1, 140.2, 141.6, 141.7, 141.9, 142.0, 142.1, 142.2,
142.6, 143.1, 143.3, 144.6, 144.7, 145.4, 145.4, 145.9, 146.2, 146.3,
146.4, 146.5, 146.6, 147.1, 147.5, 147.9, 153.4, 156.1. HRMS calculated
895.1518, found 895.1520.
144.5, 144.8, 145.1, 145.2, 145.2, 145.4, 145.4, 145.5, 145.8, 146.6,
147.3, 147.3, 147.4, 147.4, 147.4, 147.5, 147.6, 147.9, 148.7, 149.0, 149.1,
152.3, 153.1, 158.02 159.4. HRMS calculated 909.1675, found
909.1671.
7.9. 1-Benzyl-4-[30-(tert-butyldimethylsiloxy)propyl]
1,4-dihydro[60]fullerene 7b
To
1-hydro-2-[30-(tert-butyldimethylsiloxy)propyl]-1,2-dihy-
dro[60]fullerene 5 (200 mg, 0.223 mmol) in THF (100 mL) was
added potassium tert-butoxide (50 mg, 0.446 mmol) as a solution
in THF (10 mL) via a cannula. After 15 min, benzyl bromide (229 mg,
1.34 mmol) was added dropwise and the reaction mixture was
refluxed for 2 h. The solvent was removed in vacuo. The resulting
brown residue was purified by flash chromatography (SiO2, cyclo-
hexane/toluene, 10:1) to give 1-benzyl-4-[30-(tert-butyldime-
thylsiloxy)propyl] 1,4-dihydro[60]fullerene 7b (150 mg, 65% yield)
as a brown solid: mp >300 ꢀC; 1H NMR (CDCl3) 0.16 (s, 6H), 0.98 (s,
9H), 2.12 (m, 4H), 3.79 (m, 2H), 4.31 (m, 2H), 7.43 (t, J¼6.8, 1H), 7.51
(t, J¼7.2, 2H), 7.65 (d, J¼7.6, 2H); 13C NMR (CDCl3) ꢁ4.7, 18.6, 30.5,
38.8, 49.6, 59.5, 60.9, 63.1, 128.0, 128.9, 131.5, 136.8, 138.3, 138.6,
139.1, 139.3, 141.0, 142.4, 142.4, 142.8, 142.9, 142.9, 143.0, 143.0,
143.3, 143.3, 143.5, 143.5, 143.5, 144.1, 144.2, 144.2, 144.3, 144.4,
144.6, 144.6, 144.6, 144.7, 144.8, 145.1, 145.1, 145.2, 145.3, 145.4,
145.4, 145.8, 145.8, 147.3, 147.3, 147.4, 147.6, 147.8, 147.8, 149.0, 149.0,
149.0, 151.1, 153.3, 158.2, 158.3. HRMS calculated 985.1988, found
985.1994.
7.7. 1-Methyl-4-[20-(100,300-dioxolanyl)propyl]-
1,2-dihydro[60]fullerene 6
A solution of potassium tert-butoxide (31 mg, 0.28 mmol) in THF
(5 mL) was added to 1-hydro-2-[20-(100,300-dioxanyl)propyl]-1,2-
dihydro[60]fullerene 4b (200 mg, 0.243 mmol) in THF (200 mL) via
a cannula. After 15 min, iodomethane (3.4 g, 1.49 mL, 24 mmol) was
added dropwise. The reaction was then stirred for 24 h at room
temperature. The solvent was removed in vacuo and the residue
dissolved in CS2 (50 mL). The solution was washed with brine
(2ꢂ25 mL) and water (2ꢂ25 mL), and then dried over anhydrous
MgSO4. Evaporation of the solvents left a brown residue, which was
purified by flash chromatography (SiO2, CS2/toluene, 1:1) to give
1-methyl-4-[20-(100,300-dioxanyl)propyl]-1,2-dihydro[60]fullerene 6
(120 mg, 59% yield) as a brown solid: mp >300 ꢀC; 1H NMR (5:1,
CS2/CDCl3) 2.72 (m, 2H), 2.86 (s, 3H), 3.18 (m, 2H), 4.00 (m, 2H), 4.12
(m, 2H), 5.25 (t, J¼4.35, 1H); 13C NMR (5:1, CS2/CDCl3) 29.1, 31.5,
36.8, 54.3, 58.4, 65.2, 103.9, 137.4, 137.5, 138.4, 138.7, 138.8, 140.2,
140.8, 141.9, 142.0, 142.2, 142.3, 142.4, 142.5, 142.6, 142.9, 142.9,
143.0, 143.0, 143.1, 143.1, 143.6, 143.7, 143.8, 144.00, 144.1, 144.2,
144.2, 144.2, 144.3, 144.6, 144.6, 144.7, 144.8, 145.0, 145.0, 145.1,
145.4, 145.6, 146.8, 146.8, 146.8, 146.8, 146.9, 147.1, 147.1, 147.2, 147.3,
148.1, 148.5, 148.5, 148.6, 151.4, 152.7, 157.0, 158.7; UV–vis (CHCl3)
lmax 257 nm, 364, 450 (br). HRMS calculated 837.0915, found
837.0896.
7.10. 1-Hydro-2-[30-(200,200-bis(hydroxymethyl)propyloxy-
carbonyl)propyl]-1,2-dihydro[60]fullerene 8
1-Hydro-2-[30-(400-methyl-200,600,700-trioxabicyclo[200,200,200]-octyl)-
propyl]-1,2-dihydro[60]fullerene 3 (100 mg, 0.112 mmol) was dis-
solved in a mixture of dichloromethane (6 mL) and trifluoroacetic
acid (6 mL). Two drops of water were added to the solution, which
was stirred overnight at room temperature. The solvents were
evaporated and the resulting brown residue was washed with
methanol and then purified by flash chromatography (SiO2, tolu-
ene) to yield 1-hydro-2-[30-(200,200-bis(hydroxymethyl)propyloxy-
carbonyl)propyl]-1,2-dihydro[60]fullerene 8 (56 mg, 55% yield) as
a black solid: mp >300 ꢀC; 1H NMR (CDCl3) 1.22 (s, 3H), 2.89 (m,
2H), 2.91 (m, 2H), 3.46 (m, 2H), 4.25 (s, 2H), 4.43 (d, J¼3.2, 4H), 6.53
(s, 1H); 13C NMR (CDCl3) 16.9, 22.0, 33.7, 38.9, 46.1, 59.6, 64.6, 65.3,
68.6, 135.9, 136.4, 140.1, 140.2, 141.6, 141.6, 141.9, 142.0, 142.1, 142.2,
142.6, 143.2, 144.5, 144.7, 145.4, 145.4, 145.4, 145.86, 146.2, 146.2,
146.4, 146.4, 147.0, 147.3, 147.5, 153.5, 155.3, 156.8, 157.2, 172.6; UV–
vis (CHCl3) lmax 256 nm, 325, 433. HRMS calculated 912.1361, found
912.1305.
7.8. 1-Methyl-4-[30-(tert-butyldimethylsiloxy)propyl]
1,4-dihydro[60]fullerene 7a
To
1-hydro-2-[30-(tert-butyldimethylsiloxy)propyl]-1,2-dihy-
7.11. 1-Hydro-2-(butanoic acid)-1,2-dihydro[60]fullerene 9
dro[60]fullerene 5 (320 mg, 0.36 mmol) in THF (200 mL) was added
potassium tert-butoxide (44 mg, 0.39 mmol) in THF (15 mL) via
a cannula. After 15 min, iodomethane (5.07 g, 0.036 mol) was
added dropwise. The reaction was stirred for 24 h at room tem-
perature. The solvent was then removed in vacuo and the residue
dissolved in dichloromethane (200 mL). The solution was washed
with brine (2ꢂ25 mL) and water (2ꢂ25 mL), and then dried over
anhydrous MgSO4. Evaporation of the solvent left a brown residue,
which was purified by flash chromatography (SiO2, cyclohexane/
toluene, 10:1) to give 1-methyl-4-[30-(tert-butyldimethylsiloxy)-
propyl] 1,2-dihydro[60]fullerene 7a (250 mg, 77% yield) as a brown
solid: mp >300 ꢀC; 1H NMR (CDCl3) 0.17 (s, 6H), 0.99 (s, 9H), 2.56
(m, 2H), 2.83 (s, 3H), 3.12 (m, 2H), 4.01 (t, J¼6, 2H); 13C NMR (CDCl3)
4.7, 18.8, 26.4, 29.6, 30.7, 40.03, 54.8, 59.4, 63.3, 138.0, 138.9, 139.1,
139.4, 141.2, 142.3, 142.4, 142.6, 142.9, 142.9, 143.0, 143.0, 143.4,
143.5, 143.6, 144.2, 144.2, 144.3, 144.4, 144.5, 144.6, 144.7, 144.7,
1-Hydro-2-[30-(200,200-bis(hydroxymethyl)propyloxycarbonyl)-
propyl]-1,2-dihydro[60]fullerene
8 (100 mg, 0.138 mmol) were
dissolved in a mixture of toluene (6 mL) and trifluoroacetic acid
(6 mL). Two drops of water were added to the solution, which was
stirred overnight at 80 ꢀC. The solvents were evaporated and the
resulting brown residue was washed with methanol (3ꢂ20 mL) and
ether (3ꢂ20 mL), dissolved in pyridine, and filtered to afford 1-
hydro-2-(butanoic acid)-1,2-dihydro[60]fullerene 9 (40 mg, 45%
yield) as a black solid: mp >300 ꢀC; 1H NMR (pyridine-d5) 2.92 (m,
2H), 3.00 (t, J¼6.4, 2H), 3.40 (m, 2H), 6.65 (s, 1H); 13C NMR (pyri-
dine-d5) 23.2, 35.0, 46.4, 59.8, 65.4, 136.3, 136.7, 140.4, 140.4, 140.4,
141.9, 142.3, 142.3, 142.6, 142.8, 143.2, 144.8, 144.9, 145.0, 145.6,
145.6, 145.6, 145.6, 146.2, 146.4, 146.4, 146.6, 146.6, 146.9, 147.5,
147.7, 147.7, 154.9, 156.8, 176.0; UV–vis (CHCl3) lmax 254 nm, 328,
432. HRMS calculated 809.0603, found 809.0612.

10328
E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
7.12. 1-Hydro-2-(20-ethanal)-1,2-dihydro[60]fullerene 10a
chromatography (SiO2 toluene/ethanol, 4:1) to give 1-hydro-2-(30-
propanol) 1,2-dihydro[60]fullerene 12 (30 mg, 75% yield) as a brown
solid: mp >300 ꢀC; 1H NMR (5:2 CS2/CDCl3) 2.82 (m, 2H), 3.56 (m,
2H), 4.19 (m, 2H), 6.51 (s, 1H). HRMS calculated 781.0653, found
781.0634.
Trifluoroacetic acid (2 mL) and water (0.5 mL) were added to
1-hydro-2-[200-(100,300-dioxolanyl) ethyl]-1,2-dihydro[60]fullerene
4a (100 mg, 0.121 mmol) dissolved in toluene (50 mL). The reaction
mixture was refluxed for 16 h and the solvents were then removed
in vacuo. The residue was dissolved in CS2 (30 mL) and the resulting
solution was washed with brine (2ꢂ5 mL) and water (2ꢂ5 mL), and
dried over anhydrous MgSO4. The solvent was removed in vacuo
and the residue was purified by flash chromatography (SiO2, tolu-
ene) and washed with ether (3ꢂ4 mL) to give 1-hydro-2-(20-etha-
nal)-1,2-dihydro[60]fullerene 10a (50 mg, 53% yield) as a black
solid: mp >300 ꢀC; 1H NMR (5:1, CS2/CDCl3) 2.89 (m, 2H), 3.75 (m,
2H), 3.83 (m, 2H), 6.43 (s, 1H), 10.26 (s, 1H). HRMS calculated
779.0497, found 779.0520.
7.16. 1-Methyl 4-(30-propanol) 1,4-dihydro[60]fullerene 13a
To 1-methyl-4-[30-(tert-butyldimethylsiloxy)propyl] 1,2-dihy-
dro[60]fullerene 7a (302 mg, 0.33 mmol) in THF (80 mL) was added
6 N HCl (5 mL) dropwise. After stirring the reaction for 2 h at room
temperature, the solvent was removed in vacuo and the residue
was dissolved in CS2 (30 mL). The solution was washed with water
(2ꢂ15 mL), dried over anhydrous MgSO4 and the solvent evapo-
rated to give a brown residue, which was purified by flash chro-
matography (SiO2 toluene/ethanol, 10:1) and washed with pentane
(3ꢂ2 mL) to give 1-methyl 4-(30-propanol) 1,2-dihydro[60]fuller-
ene 13a (100 mg, 41.6% yield) as a black solid: mp >300 ꢀC; 1H NMR
(5:2 CS2/CDCl3) 2.62 (m, 2H), 2.87 (s, 3H), 3.13 (m, 2H), 4.25 (m, 2H).
HRMS calculated 795.0810, found 795.0811.
7.13. 1-Hydro-2-(30-propanal)-1,2-dihydro[60]fullerene 10b
To 1-hydro-2-[30-(100,300-dioxolanyl) propyl]-1,2-dihydro[60]-
fullerene 4b (40 mg, 0.048 mmol) dissolved in toluene (30 mL) were
added trifluoroacetic acid (1 mL) and water (0.5 mL). The reaction
mixturewas refluxed for 16 hand the solvents werethen removed in
vacuo. The residue was dissolved in CS2 (10 mL) and the resulting
solution was washed with brine (2ꢂ5 mL) and water (2ꢂ5 mL), and
dried over anhydrous MgSO4. The solvent was removed in vacuo and
the residue was purified by flash chromatography (SiO2, toluene)
and washed with pentane (2ꢂ2 mL) to give 1-hydro-2-(30-propa-
nal)-1,2-dihydro[60]fullerene 10b (15 mg, 40% yield) as a black
solid: mp >300 ꢀC; 1H NMR (5:1, CS2/CDCl3) 2.89 (m, 2H), 3.04 (m,
2H), 3.46 (m, 2H), 6.59 (s,1H),10.10 (s,1H); 13C NMR (5:1, CS2/CDCl3)
23.4, 32.6, 44.5, 60.1, 60.2,136.4,136.8,140.6,140.7,141.1,142.1,142.2,
142.3,142.4,142.6,143.0,143.7,145.0,145.8,145.8,145.9,145.9,146.0,
146.2, 146.6, 146.7, 146.8, 146.8, 147.4, 153.9, 155.8, 200.9. HRMS
calculated 793.0653, found 793.0635.
7.17. 1-Benzyl-4-(30-propanol) 1,4-dihydro[60]fullerene 13b
To 1-benzyl-4-[30-(tert-butyldimethylsiloxy)propyl] 1,4-dihy-
dro[60]fullerene 7b (244 mg, 0.25 mmol) in THF (50 mL) was added
6 N HCl (5 mL). After stirring the reaction for 2 h at room temper-
ature, the solvent was removed in vacuo and the residue was dis-
solved in CHCl3 (50 mL). The solution was washed successively with
brine (2ꢂ25 mL) and water (2ꢂ25 mL), and dried over anhydrous
MgSO4. Following purification by flash chromatography (SiO2, tol-
uene/ethanol,
4:1),
1-benzyl-4-(30-propanol)
1,4-dihydro-
[60]fullerene 13b (155 mg, 72% yield) was obtained as a black solid:
mp >300 ꢀC; 1H NMR (CDCl3) 2.04 (m, 2H), 2.38 (m, 2H), 3.81 (m,
2H), 4.31 (m, 2H), 7.47 (m, 1H), 7.55 (t, J¼6.8, 2H), 7.67 (d, J¼7.2,
2H).13C NMR (CDCl3) 30.6, 38.4, 49.5, 59.3, 61.0, 63.1, 128.0, 129.0,
131.6, 136.9, 138.1, 138.6, 139.2, 139.2, 141.0, 142.3, 142.4, 142.5,
142.8, 142.9, 143.0, 143.0, 143.4, 143.4, 143.5, 143.5, 144.1, 144.2,
144.3, 144.3, 144.4, 144.6, 144.6, 144.7, 144.7, 145.0, 145.1, 145.2,
145.2, 145.3, 145.4, 145.8, 145.9, 147.2, 147.3, 147.3, 147.3, 147.6, 147.6,
147.7, 148.9, 149.0, 149.0, 149.1, 151.0, 153.2, 158.0, 158.5. HRMS
calculated 871.1123, found 871.1114.
7.14. 1-Methyl-4-(30-propanal)-1,4-dihydro[60]fullerene 11
Trifluoroacetic acid (2 mL) and water (0.5 mL) were added to 1-
methyl-4-[10-(10,30-dioxolanyl) ethyl]-1,4-dihydro[60]fullerene
6
(100 mg, 0.119 mmol) in toluene (50 mL). The reaction mixture was
refluxed for 2 h and the solvents were then removed in vacuo. The
residue was dissolved in CS2 (30 mL) and the resulting solution was
washed with brine (2ꢂ5 mL) and water (2ꢂ5 mL), and dried over
anhydrous MgSO4. The solvents were removed in vacuo and the
residue was purified by flash chromatography (SiO2, toluene) and
washed with ether (3ꢂ4 mL) to give 1-methyl-4-(30-propanal)-1,4-
dihydro[60]fullerene 11 (60 mg, 63% yield) as a black solid: mp
>300 ꢀC; 1H NMR (5:1, CS2/CDCl3) 2.90 (s, 3H), 3.40 (dt, J¼7.8, 14.5,
2H), 3.52 (t, J¼7.6, 2H), 10.14 (s, 1H); 13C NMR (5:1, CS2/CDCl3) 29.5,
34.8, 43.8, 54.0, 62.1, 138.3, 138.3, 139.1, 140.1, 140.7, 141.1, 141.2,
141.5, 141.6, 141.7, 141.8, 142.0, 142.1, 142.3, 142.3, 142.7, 142.8, 142.9,
143.2, 143.6, 143.7, 143.9, 144.0, 144.0, 144.0, 144.0, 144.0, 144.3,
144.3, 144.4, 144.5, 144.6, 144.6, 144.8, 144.9, 145.0, 145.1, 145.2,
145.3, 145.6, 145.9, 146.1, 146.2, 146.5, 146.6, 146.8, 146.9, 147.3,
147.3, 147.4, 147.7, 148.10, 148.3, 148.4, 152.3, 154.9, 158.2, 158.48,
197.9. HRMS calculated 793.0653, found 793.0659.
7.18. 1-Hydro-2-[30-(
L-ALA-OEt) propyl]-
1,2-dihydro[60]fullerene 14
1-Hydroxy-7-azabenzotriazole (HOAt) (10.13 mg, 0.0748 mmol)
and 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochlo-
ride (EDC$HCl, 14 mg, 0.07 mmol) were added to a suspension of
1-hydro-2-(butanoic
0.0498 mmol) 9 in dichloromethane (10 mL). The mixture was
stirred at room temperature for 15 min. A solution of HCl$H- -Ala-
OEt (8.36 mg, 0.0548 mmol) and N-methyl morpholine (5.8 L,
acid)-1,2-dihydro[60]fullerene
(40 mg,
L
m
5.54 mg, 0.0548 mmol) in 1 mL of dichloromethane was added to
the suspension and the resulting mixture was stirred 30 min at
room temperature. The solvents were evaporated and the resulting
brown residue was purified by flash chromatography (SiO2, tolu-
ene/ethylacetate, 4:1) to yield 1-hydro-2-[30-(
L-Ala-OEt) propyl]-
7.15. 1-Hydro-2-(30-propanol) 1,2-dihydro[60]fullerene 12
1,2-dihydro[60]fullerene 14 (75 mg, 66% yield) as a black solid: mp
>300 ꢀC; 1H NMR (CDCl3) 1.34 (t, J¼7.2, 3H), 1.54 (d, J¼6.8, 3H), 2.77
(t, J¼7, 1H), 2.95 (m, 2H), 3.47 (m, 2H),4.28 (q, J¼7.2, 2H), 4.75 (m,
1H), 6.29 (d, J¼6.8,1H), 6.56 (s, 1H); 13C NMR (CDCl3) 14.1, 18.7, 22.6,
36.0, 46.2, 48.2, 59.5, 61.7, 64.7, 135.9,136.5,140.1,140.2,141.6,141.6,
141.9, 142.0, 142.1, 142.2, 142.5, 143.2, 144.6, 144.7, 145.4, 145.4,
145.8, 146.1, 146.2, 146.3, 146.3, 146.4, 147.1, 147.3, 147.4, 153.8, 155.7,
171.6, 173.2; UV–vis (CHCl3) lmax 253 nm, 385, 432. HRMS calcu-
lated 908.1287, found 908.1279.
To
1-hydro-2-[30-(tert-butyldimethylsiloxy)propyl]-1,2-dihy-
dro[60]fullerene 5 (46 mg, 0.051 mmol) dissolved in THF (5 mL) was
added 6 N HCl (1 mL). After stirring the reaction for 2 h at room
temperature, the solvent was removed in vacuo and the residue was
dissolved in CS2 (15 mL). The solution was washed with water
(2ꢂ5 mL), dried over anhydrous MgSO4 and the solvent evaporated
to give
a brown residue, which was purified by flash

E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
10329
7.19. Cyclopenta-[1,2]-[60]fullerene-30-triisopropylsilyloxy 16a
7.22. Cyclopenta-[1,2]-[60]fullerene-30-N-phenylamine 19
Triisopropylsilyl-trifluoromethanesulfonate (0.0476 mL, 54 mg,
0.177 mmol) was added to a suspension of 1-hydro-2-(propanal)-
1,2-dihydro[60]fullerene 10b (106 mg, 0.136 mmol) in DMF
(40 mL). After stirring for 30 min at 0 ꢀC, 0.0415 mL (30 mg,
0.297 mmol) of triethylamine was added. The solution was allowed
to warm to room temperature and subsequently stirred for 12 h at
80 ꢀC. The mixture was dissolved in CS2 and washed with ice-cold
dilute acid followed by a solution of NaHCO3 and brine. The solution
was then dried over anhydrous MgSO4 and the solvent evaporated.
The resulting brown residue was purified by flash chromatography
(SiO2, toluene) to yield cyclopentanyl-1,2-[60]fullerene-30-(triiso-
propylsilyl)oxy 16a (95 mg, 75% yield) as a black solid: mp >300 ꢀC;
1H NMR (5:1, CS2/CDCl3) 1.07 (s, 3H), 1.21 (s, 6H), 1.22 (s, 6H), 1.24 (s,
6H), 2.92 (m, 1H), 3.09 (m, 1H), 3.58 (m, 1H), 3.93 (m, 1H), 5.96 (dd,
J¼4, 4, 1H); 13C NMR (5:1, CS2/CDCl3) 12.8, 18.0, 18.2, 18.2, 34.1, 39.5,
68.3, 69.4, 83.8, 134.9, 135.6, 135.9, 136.9, 139.3, 139.8, 140.0, 140.1,
141.5, 141.5, 141.6, 141.7, 141.8, 141.9, 141.9, 141.9, 142.0, 142.0, 142.1,
142.3, 142.3, 142.4, 142.4, 142.8, 142.9, 142.9, 144.2, 144.3, 144.3,
144.4, 145.0, 145.0, 145.0, 145.1, 145.1, 145.4, 145.5, 145.5, 145.7,
145.8, 145.8, 145.8, 146.0, 146.3, 146.0, 146.4, 147.1, 147.1, 147.1, 154.5,
155.1, 156.6, 156.9; UV–vis (CHCl3) lmax 253 nm, 327, 404, 431.
HRMS calculated 935.1831, found 935.1817.
To
[60]fullerene 10b (56 mg, 0.072 mmol) in benzene (40 mL) were
added 6.55 L (6.69 mg, 0.072 mmol) of aniline and 800 mg of
a
suspension of 1-hydro-2-(30-propanal)-1,2-dihydro-
m
molecular sieves (4 Å). After stirring at room temperature for 3
days, the solvent was evaporated. The resulting brown residue was
purified by flash chromatography (SiO2, toluene) to yield cyclo-
penta-[1,2]-[60]fullerene-30-N-phenylamine 19 (40 mg, 65% yield)
as a black solid: mp >300 ꢀC; 1H NMR (5:1, CS2/CDCl3) 2.95 (m, 1H),
3.11 (m, 1H), 3.69 (m, 1H), 3.74 (m, 1H), 4.50 (d, J¼8.8, 1H), 5.66 (dt,
J¼5.2, 7.2, 1H), 6.76 (t, J¼7.4, 1H), 6.94 (d, J¼8, 2H), 7.22 (t, J¼8, 2H);
13C NMR (5:1, CS2/CDCl3) 31.7, 40.0, 67.2, 70.0, 74.6, 113.9, 118.4,
129.4, 134.7, 135.0, 135.1, 135.1, 135.1, 135.2, 136.8, 139.7, 139.7, 140.1,
140.2, 141.4, 141.6, 141.7, 141.8, 141.9, 141.9, 141.9, 142.1, 142.2, 142.4,
142.4, 142.5, 142.8, 144.2, 144.3, 144.3, 144.9, 145.0, 145.0, 145.1,
145.1, 145.1, 145.3, 145.3, 145.4, 145.4, 145.8, 145.8, 145.9, 145.9,
146.0, 146.1, 146.1, 146.3, 146.5, 146.6, 147.0, 147.3, 151.8, 155.7, 155.7,
157.1; UV–vis (CHCl3) lmax 256 nm, 325, 430. HRMS calculated
854.0969, found 854.0963.
7.23. 1,4,7,8-Functionalised fullerene 20
Aniline (16.78 mL, 17.15 mg, 0.184 mmol) and 4 g of molecular
7.20. Cyclopenta-[1,2]-[60]fullerene-30-hydroxy 16b
sieves (4 Å) were added to a suspension of 1-methyl-4-(30-
propanal)-1,4-dihydro[60]fullerene 10b (146 mg, 0.184 mmol) in
benzene (100 mL). After stirring at room temperature for 3 days,
the solvent was evaporated. The resulting brown residue was
purified by flash chromatography (SiO2, toluene) to yield com-
pound 20 (60 mg, 38% yield) as a black solid: mp >300 ꢀC; 1H
NMR (5:1, CS2/CDCl3) 2.30 (m, 1H), 2.35 (s, 3H), 2.48 (m, 1H),
2.78 (m, 1H), 3.01 (m, 1H), 3.81 (d, J¼2.8, 1H), 4.27 (ddd, J¼2.8,
4.4, 7.8, 1H), 6.57 (d, J¼8, 1H), 6.69 (m, 1H), 7.02 (m, 1H), 7.84 (d,
J¼8, 1H); 13C NMR (5:1, CS2/CDCl3) 27.3, 34.0, 40.6, 53.5, 57.9,
60.8, 65.6, 72.0, 116.8, 120.5, 123.9, 129.3, 129.5, 133.4, 137.0,
137.3, 138.2, 138.3, 140.0, 140.1, 141.2, 141.4, 141.8, 141.8, 142.1,
142.3, 142.4, 142.7, 143.2, 143.5, 143.7, 143.9, 144.3, 144.5, 144.7,
144.8, 144.9, 145.0, 145.1, 145.2, 145.2, 145.3, 146.0, 146.1, 146.3,
146.4, 146.6, 146.8, 146.9, 147.0, 147.0, 147.2, 147.6, 147.6, 147.9,
148.1, 148.3, 148.5, 148.8, 148.9, 149.5, 149.8, 151.6, 151.9, 155.8,
155.9, 158.1; UV–vis (CHCl3) lmax 252 nm, 314, 406, 446, 526,
682. HRMS calculated 868.1126, found 868.1142.
Triethylamine (0.0457 mL, 33 mg, 0.327 mmol) were added to
a suspension of 1-hydro-2-(30-propanal)-1,2-dihydro[60]fullerene
10b (30 mg, 0.136 mmol) in DMF (40 mL). The solution was stirred
for 12 h at 80 ꢀC. The mixture was dissolved in CS2 and washed with
water and brine. The solution was then dried over anhydrous
MgSO4 and the solvent evaporated. The resulting brown residue
was purified by flash chromatography (SiO2, CS2/CH2Cl2, 4:1) to
afford cyclopentan-30-ol-1,2-[60]fullerene 16b (25 mg, 83% yield)
as a black solid: mp >300 ꢀC; 1H NMR (CDCl3) 2.72 (d, J¼4, 1H), 2.93
(m, 1H), 3.15 (m, 1H), 3.63 (m, 1H), 3.97 (m, 1H), 5.83 (m, 1H); 13C
NMR (CDCl3) 33.1, 39.7, 69.6, 76.7, 82.7, 134.8, 135.5, 135.7, 138.0,
139.7, 140.0, 140.1, 141.5, 141.6, 141.7, 141.7, 141.8, 141.9, 141.9, 141.9,
142.1, 142.2, 142.4, 142.4, 142.4, 142.9, 144.2, 144.2, 144.3, 144.3,
145.0,145.0,145.1,145.1,145.1,145.2,145.2,145.3,145.3,145.7,145.8,
145.8, 146.0, 146.0, 146.1, 147.0, 147.1, 152.0, 154.6, 156.4, 156.7;
UV–vis (CHCl3) lmax 255 nm, 325, 404, 431. HRMS calculated
779.0497, found 779.0513.
7.21. Cyclopenta-[1,2]-[60]fullerene-30-oxyethanol 17
7.24. 1-Benzyl-4-[30-bromopropyl] 1,4-dihydro[60]fullerene 23
Triispropylsilyl-trifluoromethanesulfonate (21.54
m
L, 26.44 mg,
N-(1-Bromo-2-methylprop-1-enyl)-N,N-dimethylamine (15 mg,
0.084 mmol) was added dropwise to 1-benzyl-4-(30-propanol)
1,4-dihydro[60]fullerene 13b (60 mg, 0.069 mmol) in dichloro-
methane (30 mL). The reaction mixture was then stirred for 45 min
at room temperature. Addition of 20 mL of dichloromethane, work-
up with brine (2ꢂ10 mL) and water (2ꢂ10 mL) and drying of the
organic layer over anhydrous MgSO4 gave a brown solid after re-
moval of the solvent. Purification by flash chromatography (SiO2,
toluene/ethanol, 4:1) followed by washing with pentane (3ꢂ2 mL)
gave 1-benzyl-4-[30-bromopropyl] 1,4-dihydro[60]fullerene 23
(26 mg, 40% yield) as a brown solid: mp >300 ꢀC; 1H NMR (CDCl3)
1.96 (m, 2H), 2.62 (m, 2H), 3.51 (m, 2H), 4.29 (m, 2H), 7.49 (m, 1H),
7.56 (t, J¼6.8, 2H), 7.67 (d, J¼7.2, 2H); 13C NMR (CDCl3) 30.4, 33.2,
40.6, 48.9, 49.4, 58.7, 60.9, 127.9, 128.8, 131.5, 136.6, 137.0, 138.1,
139.1, 142.3, 142.8, 142.8, 142.9, 143.0, 143.0, 144.1, 144.3, 144.3,
144.5, 144.6, 144.6, 144.6, 144.7, 144.7, 145.0, 145.1, 145.1, 145.2,
145.3, 145.4, 145.8, 146.5, 146.6, 147.2, 147.3, 147.3, 147.3, 147.5, 147.6,
149.0, 149.0, 149.1, 151.2, 152.8, 152.9, 158.3. HRMS calculated
933.0279, found 933.0250.
0.119 mmol) was added to a suspension of 1-hydro-2-[20-(100,300-
dioxolanyl)ethyl]-1,2-dihydro[60]fullerene 4b (82 mg, 0.099 mmol)
in CS2 (40 mL). After stirring for 30 min at 0 ꢀC, 23.42
mL (17.38 mg,
0.134 mmol) of N,N0-diisopropylethylamine was added. The solu-
tion was allowed to warm to room temperature and subsequently
stirred for 12 h at 30 ꢀC. The solvent was evaporated and the
resulting brown residue was purified by flash chromatography
(SiO2, CS2/CH2Cl2, 9:1) to afford cyclopentanyl-1,2-[60]fullerene-30-
oxyethanol 17 (72 mg, 88% yield) as a black solid: mp >300 ꢀC; 1H
NMR (5:1, CS2/CDCl3) 2.11 (t, J¼6.2, 1H), 2.97 (m, 1H), 3.07 (m, 1H),
3.57 (m, 1H), 3.83 (m, 1H), 3.94 (m, 2H), 4.15 (m, 2H), 5.03 (dd,
J¼4.4, 5.6, 1H); 13C NMR (5:1, CS2/CDCl3) 30.4, 39.3, 62.1, 69.7, 72.4,
73.3, 90.7, 135.0, 135.7, 135.9, 136.9, 139.9, 140.1, 140.2, 140.2, 141.7,
141.9, 142.0, 142.0, 142.0, 142.1, 142.2, 142.2, 142.2, 142.3, 142.5,
142.6, 142.6, 143.0, 143.1, 144.4, 144.5, 144.5, 145.2, 145.3, 145.3,
145.3, 145.6, 145.7, 145.7, 145.8, 146.0, 146.1, 146.2, 146.2, 146.3,
146.7, 147.3, 153.7, 155.2, 156.5, 156.7; UV–vis (CHCl3) lmax 257 nm,
356, 403, 431. HRMS calculated 823.0759, found 823.0733.

10330
E. Champeil et al. / Tetrahedron 64 (2008) 10319–10330
7.25. 1-Benzyl-4-[
a
-propyl-tert-butylglycinate benzophenone
References and notes
imine] 1,4-dihydro[60]fullerene 24
1. (a) Nakamura, E.; Isobe, H. Acc. Chem. Res. 2003, 36, 807–815; (b) Yang, J.;
Alemany, L. B.; Driver, J.; Hartgerink, J. D.; Barron, A. R. Chem.dEur. J. 2007, 13,
2530–2545; (c) Yang, J.; Wang, K.; Driver, J.; Yang, J.; Barron, A. R. Org. Biomol.
Chem. 2007, 5, 260–266; (d) Sofou, P.; Elemes, Y.; Panou-Pomonis, E.; Stavra-
koudis, A.; Tsikaris, V.; Sakarellos, C.; Sakarellos-Daitsiotis, M.; Maggini, M.;
Formaggio, F.; Toniolo, C. Tetrahedron 2004, 60, 2823–2828.
2. (a) Taylor, R.; Walton, D. R. M. Nature 1993, 363, 685–693; (b) Hirsch, A. The
Chemistry of Fullerenes; Thieme: Stuttgart, New York, NY, 1994; (c) Diederich, F.;
Isaacs, L.; Phillip, D. Chem. Soc. Rev. 1994, 243–255; (d) Hirsch, A. Synthesis 1995,
895–913; (e) Hirsch, A. Top. Curr. Chem. 1999, 199, 1–65; (f) Da Ros, T.; Prato, M.
Chem. Commun. 1999, 663–669.
3. (a) Wudl, F. Acc. Chem. Res. 1992, 25, 157–161; (b) Rubin, Y.; Khan, S.;
Freedberg, D. I.; Yeretzian, C. J. Am. Chem. Soc. 1992, 115, 344–345; (c) Prato,
M.; Suzuki, T.; Foroudian, H.; Li, Q.; Khemani, K.; Wudl, F.; Leonetti, J.; Little,
R. D.; White, T.; Rickborn, B.; Yamago, S.; Nakamura, E. J. Am. Chem. Soc.
1993, 115, 1594–1595; (d) An, Y.-Z.; Anderson, H. L.; Rubin, Y. J. Org. Chem.
1993, 58, 4799–4801; (e) Maggini, M.; Scorrano, G.; Prato, M. J. Am. Chem.
Soc. 1993, 115, 9798–9799.
4. (a) Bingel, C. Chem. Ber. 1993, 126, 1957–1959; (b) Hirsch, A.; Lamparth, I.;
Karfunkel, H. R. Angew. Chem., Int. Ed. Engl. 1994, 33, 437–438.
5. (a) Fagan, P. J.; Krusic, P. J.; Evans, D. H.; Lerke, S. A.; Johnston, E. J. Am.
Chem. Soc. 1992, 114, 9697–9699; (b) Hirsch, A.; Gro¨sser, T.; Skiebe, A.; Soi, A.
Chem. Ber. 1993, 126, 1061–1067; (c) Hirsch, A.; Soi, A.; Karfunkel, H. R.
Angew. Chem., Int. Ed. Engl. 1992, 31, 766–768; (d) Komatsu, K.; Murata, Y.;
Takimoto, N.; Mori, S.; Sugita, N.; Wan, T. S. M. J. Org. Chem. 1994, 59, 6101–
6102; (e) Keshavarz, M. K.; Knight, B.; Srdanov, G.; Wudl, F. J. Am. Chem.
Soc. 1995, 117, 11371–11372; (f) Timmerman, P.; Anderson, H. L.; Faust, R.;
Nierengarten, J.-F.; Habicher, T.; Seiler, P.; Diederich, F. Tetrahedron 1996, 52,
4925–4927; (g) Fujiwara, F.; Murata, Y.; Wan, T. S. M.; Komatsu, K. Tetra-
hedron 1998, 54, 2049–2058.
6. Mikami, K.; Matsumoto, S.; Ishida, A.; Takamuku, S.; Suenobu, T.; Ukuzumi, S.
J. Am. Chem. Soc. 1995, 117, 11134–11141.
7. (a) Corey, E. J.; Raju, N. Tetrahedron Lett. 1983, 24, 5571–5574; (b) Keinan, E.;
Sinha, S. C.; Singh, S. P. Tetrahedron 1991, 47, 4631–4638.
8. Kadish, K. M.; Gao, X.; Van Caemelbecke, E.; Suenobu, T.; Fukuzumi, S. J. Phys.
Chem. A 2000, 104, 3878–3883.
9. Maurer, K. W.; Armstrong, R. W. J. Org. Chem. 1996, 61, 3106–3116.
10. Kitagawa, T.; Takeuchi, K. I. Bull. Chem. Soc. Jpn. 2001, 74, 785–800.
11. (a) Gonzalez, R.; Wudl, F.; Pole, D. L.; Sharma, P. K.; Warkentin, J. J. Org.
Chem. 1996, 61, 5837–5839; (b) Kitagawa, T.; Tanaka, T.; Murakita, H.;
Nishikawa, A.; Takeuchi, K. Tetrahedron 2001, 57, 3537–3547; (c) Meier, M. S.;
Bergosh, R. G.; Gallagher, M. E.; Spielmann, H. P.; Wang, Z. J. Org. Chem. 2002,
5946–5952.
12. (a) See Ref. 8. (b) Murata, Y.; Komatsu, K.; Wan, T. S. M. Tetrahedron Lett. 1996,
37, 7061–7064; (c) Fukuzumi, S.; Suenobu, T.; Hirasaka, T.; Arakawa, R.; Kadish,
K. M. J. Am. Chem. Soc. 1998, 120, 9220–9227.
Lithium diisopropylamide (0.071 mmol, 35 mL of 2 M stock)
was added to THF (2 mL) and cooled to –78 ꢀC. After 15 min,
N-(diphenylmethylene)glycine tert-butyl ester (21 mg, 0.071 mmol)
in THF (2 mL) was added via cannula and the solution was main-
tained at ꢁ78 ꢀC for 45 min. After this time, the solution was added
dropwise via cannula to 1-benzyl-4-[30-bromopropyl] 1,4-dihy-
dro[60]fullerene 23 (66 mg, 0.071 mmol) in THF (10 mL) at ꢁ78 ꢀC.
The solution was maintained at ꢁ78 ꢀC for 1 h, and then warmed to
room temperature and held at this temperature for a further hour.
The reaction was then quenched with saturated NH4Cl (5 mL),
dichloromethane (25 mL) was added and the isolated organic layer
was washed successively with brine (2ꢂ10 mL) and water
(2ꢂ10 mL). The solution was then dried over anhydrous MgSO4 and
the solvent evaporated. The resulting brown residue was purified
by flash chromatography (SiO2, toluene) to give small amounts
(<8 mg) of two products with very similar Rf values (0.51, 0.52).
Compound 24a: 1H NMR (CDCl3) 1.44 (s, 9H), 2.6 (m, 2H), 2.9 (m,
1H), 3.2 (m, 2H), 3.3 (m, 1H), 3.53–4.02 (A, B, m J1¼12.4, J2¼12.8,
2H), 5.24 (s, 1H), 7.1–7.9 (m, 15H); 13C NMR (CDCl3) 28.5, 42.8, 43.8,
47.1, 58.4, 59.9, 65.3, 70.8, 75.2, 127.4, 128.4, 128.6, 129.1, 129.7,
129.9, 130.2, 131.2, 135.2, 135.2, 136.1, 136.6, 139.2, 139.5, 140.6,
141.5, 141.7, 141.8, 142.5, 142.7, 142.9, 143.3, 143.5, 144.1, 144.8, 145.2,
145.3, 145.4, 145.5, 145.6, 145.7, 145.7, 145.9, 145.9, 146.1, 146.8,
146.8, 147.0, 147.1, 147.1, 147.2, 147.8, 148.1, 148.3, 148.6, 148.9, 149.5,
149.9, 150.4, 151.4, 152.6, 154.9, 156.2, 156.3, 161.2, 163.9, 168.3,
172.4. HRMS calculated 1148.2589, found 1148.2637.
Compound 24b: 1H NMR (CDCl3) 1.52 (s, 9H), 2.5 (m, 2H), 2.9 (m,
1H), 3.01 (m, 2H), 3.4 (m, 1H), 3.87–3.94 (A, B, m, J1¼12.8, J2¼12.8,
2H), 5.19 (s, 1H), 7.2–7.7 (m, 15H). HRMS calculated 1148.2589,
found 1148.2633.
7.26. Computational studies
Molecular modelling calculations were run using the native
OPLS force field in Hyperchem 7.52, Hypercube. Molecular
dynamics (MD) simulations were run with the following conditions:
simulation run time 10,000 ps, simulation step 1 fs, heat time
100 ps, starting temperature 0 K, simulation temperature 300 K
and temperature step 10 K. Simulations were run in vacuo and
repeated in duplicate. Conformational searches (CS) on 24a and 24b
were run with the following procedure: (1) all possible dihedral
angles were chosen (eight overall) responsible for different con-
formations of the benzyl and the propylglycinate moieties; (2) in
each step, 1–8 dihedrals were varied randomly; (3) the geometry
thus obtained was optimized and retained if passing some post-
convergence tests (chirality inversion, duplicate structures); (4)
after 1000 steps, all structures within 6 kcal/mol were retained
(each of them may be found multiple times in the CS), and the
lowest energy 50 structures analyzed.
13. Allard, E.; Rivie`re, L.; Delaunay, J.; Dubois, D.; Cousseau, J. Tetrahedron Lett. 1999,
40, 7223–7226.
14. Hamada, M.; Hino, T.; Kinbara, K.; Saigo, K. Tetrahedron Lett. 2001, 42, 5069–
5071.
15. Rubin, Y.; Ganapathi, P. S.; Franz, A.; An, Y.-Z.; Qian, W.; Neier, R. Chem.dEur. J.
1999, 5, 3162–3184.
16. O’Donnell, M. J.; Bennett, W. D.; Wu, S. J. Am. Chem. Soc. 1989, 111, 2353–2355.
17. Ghosez, L.; Georges-Koch, I.; Patiny, L.; Houtekie, M.; Bovy, P.; Nshimyumukiza,
P.; Phan, T. Tetrahedron 1998, 54, 9207–9222.
18. Lipkowitz, K. B.; Cavanaugh, M. W.; Baker, B.; O’Donnell, M. J. J. Org. Chem. 1991,
56, 5181–5192.
19. Corey, E. J.; Xu, F.; Noe, M. C. J. Am. Chem. Soc. 1997, 119, 12414–12415.
20. Lygo, B.; Crosby, J.; Lowdon, T. R.; Peterson, J. A.; Wainwright, P. G. Tetrahedron
2001, 57, 2403–2409.
21. Burley, G. A.; Keller, P. A.; Pyne, S. G.; Ball, G. E. Chem. Commun. 1998, 2539–
2540.
22. Burley, G. A.; Keller, P. A.; Pyne, S. G. Chem. Commun. 2000, 1717–1718.
23. Burley, G. A.; Keller, P. A.; Pyne, S. G.; Ball, G. E. Chem. Commun. 2001, 563–564.
24. Burley, G. A.; Keller, P. A.; Pyne, S. G.; Ball, G. E. J. Org. Chem. 2002, 67, 8316–
8330.
25. Kadish, K. M.; Gao, X.; Caemelbecke, E. V.; Hirasaka, T.; Suenobu, T.; Fukusumi,
S. J. Phys. Chem. A. 1998, 102, 3898–3906.
26. Whitesides, G. M.; Grocki, J. J.; Holtz, J. D.; Steinberg, H.; Roberts, J. D. J. Am.
Chem. Soc. 1965, 87, 1058–1064.
Ring current shifts were estimated with the classical formula
Ds(ppm)¼Dc[(1–3cos2
j)/12p a nucleus lying at
r3] valid for
distance r (in Å) from the centre of the aromatic ring, with the centre-
27. Whitesides, G. M.; Holtz, D.; Roberts, J. D. J. Am. Chem. Soc. 1964, 86, 2628–2634.
28. Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic Chemistry; University
Science Books: Sausalito, CA, 2006; p. 184 ff.
nucleus direction forming an angle
j with respect to the ring. For the
benzene diamagnetic anisotropy, a value Dc¼1200 Å3 was used.
29. Rusyniak, M. J.; Ibrahim, Y. M.; Wright, D. L.; Khanna, S. N.; El-Shall, M. S. J. Am.
Chem. Soc. 2003, 125, 12001–12013.
Acknowledgements
ˇ
ˇ
30. Hobza, P.; Kabela´c, M.; Sponer, J.; Mejzlı´k, P.; Vondra´sek, J. J. Comput. Chem.
1997, 18, 1136–1150.
Financial support by the European Union (Plasmabarb: Contract
# G5RD-CT-1999-00173) and IECB is gratefully acknowledged.
31. Wenkert, D.; Ferguson, S. B.; Porter, B.; Qvarnstrom, A.; McPhail, A. T. J. Org.
Chem. 1985, 50, 4114–4119.