Study on the thermal reactions of [60]fullerene with amino acids and amino acid esters

Article abstract of DOI:10.1039/c2ob26066b

Thermal reactions of [60]fullerene with a series of amino acids and amino acid esters under aerobic and dark conditions have been investigated. Fulleropyrrolidines can be obtained from these reactions although an aldehyde is not added purposely. Possible reaction mechanisms involving uncommon C-N bond cleavages have been proposed to generate aldehydes, which then react with amino acids and amino acid esters to provide azomethine ylides, followed by 1,3-dipolar cycloaddition to [60]fullerene affording fulleropyrrolidines. Control experiments support our proposed mechanisms, and elucidate the innate nature of C-N bond cleavages of amino acids and amino acid esters.

Full text of DOI:10.1039/c2ob26066b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 8720
www.rsc.org/obc
PAPER
Study on the thermal reactions of [60]fullerene with amino acids and
amino acid esters†
San-E Zhu,a Xin Cheng,a Yu-Jin Li,a Cheng-Kang Mai,a Yong-Shun Huang,a Guan-Wu Wang,*a,b
Ru-Fang Peng,b Bo Jinb and Shi-Jin Chub
Received 3rd June 2012, Accepted 11th September 2012
DOI: 10.1039/c2ob26066b
Thermal reactions of [60]fullerene with a series of amino acids and amino acid esters under aerobic and
dark conditions have been investigated. Fulleropyrrolidines can be obtained from these reactions although
an aldehyde is not added purposely. Possible reaction mechanisms involving uncommon C–N bond
cleavages have been proposed to generate aldehydes, which then react with amino acids and amino acid
esters to provide azomethine ylides, followed by 1,3-dipolar cycloaddition to [60]fullerene affording
fulleropyrrolidines. Control experiments support our proposed mechanisms, and elucidate the innate
nature of C–N bond cleavages of amino acids and amino acid esters.
disclosed a few photochemical reactions between amino acid
Introduction
esters and C60.4 Gan and co-workers have also investigated the
Many years have elapsed since the discovery of fullerenes, but
their chemical modifications continue to be of great appeal due
to the potential applications of fullerene derivatives in many
fields of life and materials sciences.1 Amino acids and their
derivatives are important biological molecules. They have been
utilized to functionalize [60]fullerene (C60) via the Prato reac-
tion, i.e., 1,3-dipolar cycloaddition of azomethine ylides to C60.2
Azomethine ylides are generated in situ from the decarboxylation
of iminium salts derived from the condensation of amino acids
with aldehydes/ketones or from the formal 1,2-H shift of imines
of amino acid esters. The formed fulleropyrrolidine derivatives
have recently been employed as the acceptors of photovoltaic
solar cells with a power conversion efficiency (PCE) of up to
3.44% and are superior to methyl [6,6]-phenyl-C61-butylate
([C60]-PCBM) under the same conditions.3 Fulleropyrrolidines
can also be further transformed to ionic fullerene derivatives and
applied to supramolecular assemblies.1d
reactions of amino acid esters with C60 in the presence of iodo
reagents under ultrasonification.5 We have explored the reaction
of C60 with amino acid ester hydrochlorides by using triethyl-
amine (Et3N) as a base to remove HCl.6,7 We found that the reac-
tion in CS2 at room temperature proceeded well and afforded
fullerene products containing amino acid ester, thioamide and
thiourea units.6 In contrast, the reaction in refluxing o-dichloro-
benzene (ODCB) gave fulleropyrrolidine derivatives containing
the CH3CH moiety originated from an unusual C–N bond clea-
vage of Et3N. The fulleropyrrolidines were formed through the
reaction of C60 with azomethine ylides from amino acids and
acetaldehyde, which in turn were generated by the hydrolysis of
the amino acid ester hydrochlorides in the presence of Et3N and
by the fragmentation/oxidization of Et3N, respectively.7 On the
other hand, studies on the direct thermal reaction of C60 with
amino acids themselves are scarce. Gan and co-workers investi-
gated the reaction of C60 with glycine or alanine in the presence
of sodium hydroxide and obtained complex mixtures of multi-
adducts.8 Herein, we report the direct thermal reactions of C60
with a series of amino acids and amino acid esters in the absence
of an aldehyde and under aerobic and dark conditions affording
isolable monoadducts.9,10 More importantly, we have gained a
mechanistic insight into these reactions deduced from the struc-
tural motifs of isolated products and control experiments.
Compared to the extensive application of amino acids and
their derivatives in the Prato reaction, much fewer reports on the
functionalization of C60 using amino acids or amino acid esters
alone have been described. Gan’s group and others have
aHefei National Laboratory for Physical Sciences at Microscale, CAS
Key Laboratory of Soft Matter Chemistry, and Department of Chemistry,
University of Science and Technology of China, Hefei, Anhui 230026,
P. R. China. E-mail: gwang@ustc.edu.cn; Fax: (+86) 551-360-7864
bState Key Laboratory Cultivation Base for Non-metal Composites and
Functional Materials, Southwest University of Science and Technology,
Mianyang, Sichuan 621010, P. R. China
†Electronic supplementary information (ESI) available: NMR spectra of
products 2a–2d, 4a–4d, 5b, 5c, 7a–7d, 8a–8c, 9 and 10c. See DOI:
10.1039/c2ob26066b
Results and discussion
In our preliminary communication,10 we reported the thermal
reaction of C60 with a few amino acids. In order to extend the
substrate scope and better understand the reaction mechanism,
8720 | Org. Biomol. Chem., 2012, 10, 8720–8729
This journal is © The Royal Society of Chemistry 2012

View Article Online
we have extended the thermal reaction of C60 with more amino
acids as well as that with amino acid esters. In contrast to the
previous experimental procedure,10 and for rational comparison
amongst the reactions of C60 with different amino acids and
amino acid esters, the concentration of C60 was kept constant; a
sealed tube instead of a conventional flask was used for the reac-
tion to prevent any loss of the volatile starting materials or
in situ generated intermediates; the reaction temperature was
180 °C and 10 equiv. of amino acid/amino acid ester were used
in most cases.
We first systematically explored the thermal reaction of C60
with representative N-unsubstituted amino acids, that is, glycine
(1a), alanine (1b), phenylalanine (1c) and phenylglycine (1d).
Among the employed N-unsubstituted amino acids, 1a–1c are
naturally occurring amino acids. To avoid the potential interfer-
ence of a photochemical reaction,4 the thermal reaction of C60
was conducted under dark conditions. Intriguingly, the products
from the thermal reaction of C60 with 1a–1d in refluxing ODCB
in the presence of air proved to be N-unsubstituted fulleropyrroli-
dines 2a–2d, instead of aminated fullerene derivatives.8 It was
found that the presence of oxygen was crucial for the thermal
reaction, as strict exclusion of oxygen from the reaction mixture
by a freeze–pump–thaw procedure failed to provide 2a–2d.11
The reaction conditions and product yields for the reaction of
C60 with 1a–1d are listed in Table 1. It should be pointed out
that the reactions with both glycine and alanine were much more
sluggish and required higher temperature than those with phenyl-
alanine and phenylglycine, probably due to the lower solubility
of the former two amino acids.
Products 2a–2d are known compounds, and their identities
were confirmed by comparison of their spectra with those
reported in the literature.12–14 The structures of adducts 2a–2d
are the same as those of products formed by the Prato reaction of
C60, 1a–1d and the aldehydes generated in situ from 1a–1d via
the apparent elimination of the COOH and NH2 groups. It
should be noted that there are trans and cis isomers depending
on the different orientation of the two R substituents when
R ≠ H. Products 2b and 2c existed as a mixture of trans and cis
isomers while product 2d consisted of only the cis isomer. The
exact same phenomenon was also observed for the correspond-
ing Prato reactions.13
Secondly, typical N-substituted amino acids, i.e., N-methyl-
glycine (sarcosine) (3a), N-ethylglycine (3b), N-benzylglycine
(3c) and N-phenylglycine (3d), were chosen to react with C60 in
refluxing ODCB under aerobic and dark conditions. The reaction
of C60 with 3b and 3c afforded two types (4b, 5b and 4c, 5c) of
monoadducts, whereas that with 3a and 3d provided only one
type (4a and 4d) of monoadducts. The reaction conditions and
product yields for the reaction of C60 with 3a–3d are listed in
Table 2.
Adducts 4a–4d, 5b and 5c also prove to be fulleropyrroli-
dines. Among them, products 4a,12 4b,15 4c16 and 5c17 are
known compounds, and their structures are confirmed by com-
parison of their spectral data with those reported previously. The
HRMS spectra of 4d and 5b showed correct molecular ions. The
1H NMR spectrum of 4d exhibited the signals for a phenyl
Table 1 Reaction conditions and product yields for the reaction of C60
with 1a–1da
Tb
(°C)
Reaction
time (h)
Yieldc
(%)
Table 2 Reaction conditions and product yields for the reaction of C60
with 3a–3da
Substrate 1
Product 2
200
24
19 (51)
d
200
48
35 (56)e
Substrate 3
Reaction time (h) Yield of 4b (%) Yield of 5b (%)
10
30 (81)
33 (58)
22 (45)
36 (71)
—
180
180
12
10
19 (66)f
37 (66)g
2.5
28
11 (19)
23 (47)
—
1.5
c
a Unless indicated otherwise, all reactions were performed with C60
(0.05 mmol) and 1 (0.5 mmol) in ODCB (6 mL). b Refers to the
temperature of oil bath. c Isolated yield; those in parentheses were based
a Unless indicated otherwise, all reactions were performed with C60
(0.05 mmol) and 1 (0.5 mmol) in ODCB (6 mL) at 180 °C. b Isolated
on consumed C60
.
d 1 mmol of 1a was used. e Trans/cis ratio = 1 : 3.6.
yield; those in parentheses were based on consumed C60.
c 0.1 mmol of
3d was used.
f Trans/cis ratio = 1.1 : 1. g Cis isomer only.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 8720–8729 | 8721

View Article Online
Table 3 Reaction conditions and product yields for the reaction of C60
with 6a–6da
group and a singlet at 5.14 ppm for the two methylene groups.
The 13C NMR spectrum of 4d displayed 16 lines for the 58 sp2-
carbons of the C60 skeleton along with a peak at 69.23 ppm for
the two sp3-carbons of the C60 core, consistent with its C2v
molecular symmetry. The 1H NMR spectrum of 5b gave two
doublets at 4.88 and 3.98 ppm with JAB = 9.2 Hz for the CH2
group attached to the fullerene cage, two double quartets at 3.50
(J = 12.0, 7.4 Hz) and 2.73 (J = 12.0, 7.0 Hz) ppm, a triplet at
1.53 ppm (J = 7.2 Hz) for the ethyl group, a quartet at 4.02 ppm
(J = 6.3 Hz), and a doublet at 1.95 ppm (J = 6.3 Hz) for the
CHCH3 moiety. The 13C NMR spectrum of 5b revealed
49 peaks in the 136–157 ppm range for the 58 sp2-carbons of
the C60 cage and two peaks at 75.89 and 69.20 ppm for the two
sp3-carbons of the C60 skeleton, consistent with its C1 molecular
symmetry. Adducts 4a–4d appeared to be the Prato reaction pro-
ducts of C60, 3 and CH2O generated in situ from 3 through the
apparent extrusion of the COOH and RNH moieties. Meanwhile,
5b and 5c might be similarly formed by the reaction of C60 with
3 and R′CHO (when R = R′CH2), which was produced via an
oxidative cleavage of the R′CH2–N bond of 3.
Finally, the corresponding N-substituted amino acid esters
including sarcosine ethyl ester (6a), N-ethylglycine ethyl ester
(6b), N-benzylglycine ethyl ester (6c) and N-phenylglycine ethyl
ester (6d) were allowed to react with C60 under the same con-
ditions as for N-substituted amino acids. Although 6c and 6d
were commercially available, 6a and 6b were supplied as hydro-
chlorides. The in situ generation of 6a and 6b from their hydro-
chlorides in the presence of Et3N during the reaction with C60
may cause hydrolysis of the amino acid ester hydrochlorides.7
Therefore, 6a and 6b were prepared by the neutralization of
sarcosine ethyl ester hydrochloride and N-ethylglycine ethyl
ester hydrochloride with dilute aqueous solution of K2CO3 prior
to use. The reaction conditions and product yields for the
reaction of C60 with 6a–d are listed in Table 3.
All adducts 7–10 were identified as fulleropyrrolidines.
Among them, products 7a,17 8a,4a 94a,17 and 10c18 are known
compounds, and their structures are substantiated by comparison
of their spectral data with those reported in the literature. Com-
pounds 7b, 7c, 7d, 8b and 8c exhibited correct molecular
weights in their high-resolution mass spectra and the expected
chemical shifts as well as the splitting patterns for all protons in
their 1H NMR spectra. In the 13C NMR spectrum of 7b–7d,
there were no more than 28 peaks in the 135–154 ppm range for
the 58 sp2-carbons of the C60 cage and one peak at ∼71 ppm for
the two sp3-carbons of the C60 skeleton, consistent with their C2
molecular symmetry. In comparison, 51 lines in the range of
136–157 ppm for the 58 sp2-carbons of the C60 cage along with
two lines at ca. 71 and 75/76 ppm for the two sp3-carbons of the
C60 skeleton were observed in the 13C NMR spectrum of 8b and
8c, agreeing with their C1 molecular symmetry.
Reaction
time (h)
Yield of Yield of Yield of Yield of
Substrate 6
7b,c (%) 8b,c (%) 9b (%)
10b (%)
4
2
24 (44)
31 (56)
9 (16)
3 (5)e
—
13 (24)
4 (7)f
—
0.7
75
16 (30)
11 (22)
10 (19)
13
0.7 (1)
(25)g
—
—
—
d
a Unless indicated otherwise, all reactions were performed with C60
(0.05 mmol) and 6 (0.5 mmol) in ODCB (6 mL) at 180 °C. b Isolated
yield; those in parentheses were based on consumed C60.
c Trans isomer
only. d 1.0 mmol of 6d was used. e Trans/cis ratio = 1 : 3.6. f Trans/cis
ratio = 1 : 3.5. g Trans/cis ratio = 1 : 3.3.
Primary and secondary amines can react with C60 thermally to
give aminated products.8,19 At first glance, it was surprising that
fulleropyrrolidines were obtained from the direct thermal reac-
tion of C60 with amino acids or amino acid esters in the absence
of a purposely added aldehyde under aerobic and dark con-
ditions. Although the exact pathways for the observed products
from the thermal reaction of C60 with amino acids 1 and 3 as
well as amino acid esters 6 remain to be clarified, structural
motifs of the isolated products were sufficient to allow us to
deduce the plausible reaction mechanisms (Schemes 1–3). The
mechanisms for the formation of all potential intermediates
involving C–N bond cleavages of amino acids and amino acid
esters are generalized in Schemes 1 and 2, and the subsequent
1,3-dipolar cycloaddition reactions with C60 affording fullero-
pyrrolidines are shown in Scheme 3.
Electron transfer from amino acids/amino acid esters to C60
gives cation radical A and anion radical C60−˙ (step i),7,19 the
latter transfers an electron to O2 to afford O2−˙ and regenerate
C60 (step ii).20 Hydrogen abstraction at the α-carbon adjacent to
the carbonyl group of A by O2−˙ produces iminium cation B and
The reaction of C60 with amino acid esters was much more
complex than that with the corresponding amino acids. Adducts
7 and 8 could be envisioned as the products from the reaction of
−
HO2 (step iii),4,7 subsequent proton transfer provides imine C
and H2O2 (step iv). Under heating conditions, H2O2 can decom-
pose to H2O and O2. Hydrolysis of C generates amine D and
carbonyl compound E (step v).7,21 When R3 = H in E, it elimin-
ates CO2 to give aldehyde F upon heating (step vi).22 Mean-
while, hydrogen abstraction at the α-carbon adjacent to R4
(when R1 = R4CH2) of A by O2−˙ attains iminium cation G and
HO2− (step vii). Proton transfer from G to HO2− furnishes imine
C
60 with 6 and aldehydes (EtO2CCHO and R′CHO when R = R′
CH2), which were produced from apparent C–N oxidative
cleavages of RNH–CH2CO2Et and R′CH2–NHCH2CO2Et (when
R = R′CH2), respectively. Meanwhile, 9 and 10 might be formed
from the reaction of C60 with the in situ generated aldehydes and
NH2CH2CO2Et.
8722 | Org. Biomol. Chem., 2012, 10, 8720–8729
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 1 Possible mechanisms for C–N bond cleavages of amino
acids and amino acid esters.
Scheme 3 1,3-Dipolar cycloaddition reactions with C60 affording
fulleropyrrolidines.
transfer from a zwittionic amino acid to C60 affords the cationic
radical K and C60−˙ (step x),4b followed by the same step ii to
form O2−˙ and regenerate C60 (step ii). Proton transfer from K to
O2−˙ furnishes radical L and HO2˙ (step xi). Loss of CO2 from L
gives radical M (step xii),23 from which a hydrogen abstraction
by HO2˙ provides imine N (step xiii). Hydrolysis of imine N pro-
duces amine R1NH2 (D) and R2CHO (F) (step xiv).7,21
For amino acids 1a–1d, R1 = R3 = H, R2 = R, aldehyde
RCHO (F′) is generated through steps i–vi and/or steps x–xiv.
The 1,3-dipolar reaction of C60 with 1a–1d and aldehyde F′
leads to fulleropyrrolidines 2a–2d (step xv). We previously
found that the cis/trans isomeric ratio of 2b and 2c was heavily
dependent on the reaction conditions.13 The reaction of C60 with
1b and acetaldehyde in refluxing toluene (110 °C) for 27 h led
to 21% of trans-2b and 5% of cis-2b, and the isolated isomeric
mixture of 2b was then heated in refluxing ODCB (180 °C) for
12 h to afford a mixture dominated by the cis isomer.13 The
Scheme 2 Alternative pathway to generate aldehydes from amino
acids.
H and H2O2 (step viii), subsequent hydrolysis of H produces
aldehyde I and amino acid/amino acid ester J (step ix).7,21
An alternative pathway leading to the formation of aldehydes
R2CHO (F) from amino acids is shown in Scheme 2. Electron
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 8720–8729 | 8723

View Article Online
direct reaction of C60 with 1b at 200 °C for 2 d gave cis-2b as
the major product (Table 1, entry 2). When the same reaction
was conducted at lower temperature (180 °C for 4 d), trans-2b
and cis-2b were obtained in 8% and 6% yields, respectively. A
similar behavior was also observed for the reaction with 1c.
Thus, a higher reaction temperature tended to give a higher cis/
trans isomeric ratio in our current cases. These experimental
results support the above proposed mechanisms for the formation
of aldehydes from amino acids, followed by 1,3-dipolar cyclo-
addition with C60 to produce fulleropyrrolidines.13
Scheme 4 Thermal reaction of C60 with 1a, 3a and CH2O.
For amino acids 3a–3d, R1 = R, R2 = R3 = H, formaldehyde
CH2O is formed via steps i–vi and/or steps x–xiv, while alde-
hyde R′CHO (CH2O, CH3CHO and PhCHO for 3a, 3b and 3c,
respectively) can be produced through steps vii–ix when R = R′
CH2. The cycloaddition reaction of C60 with 3a–3d and CH2O
gives 4a–4d (step xvi), while the reaction of C60 with 3a–3c and
R′CHO provides 5a–5c (step xvii). In the case of 3a, product 5a
was the same as adduct 4a. Therefore, more pathways leading to
4a may operate simultaneously. Similar steps vii–ix are impossi-
ble for 3d, of which the phenyl group is directly attached to the
nitrogen atom. As a result, the thermal reaction of C60 with 3d
afforded 4d only. It should be noted that it was possible for
H2NCH2COOH to react with CH2O/R′CHO and C60 to give the
corresponding N-unsubstituted fulleropyrrolidines. However, no
more than a trace amount of, if any, N-unsubstituted fulleropyrro-
lidines could be isolated from the thermal reaction of C60 with
3a–3d because the in situ generated CH2O and R′CHO (I′) pre-
ferred to react with a large excess of amino acids 3a–3d rather
than the relatively small amount of the in situ formed
H2NCH2COOH to furnish an azomethine ylide, which then
underwent [2 + 3] cycloaddition reaction with C60 to provide
4a–4d and 5b–5c, respectively.
Scheme 5 Thermal reaction of C60 with 1a, 3c and CH2O.
Scheme 6 Thermal reaction of C60 with 1a, 3c and PhCHO.
For amino acid esters 6a–6d, R1 = R, R2 = H, R3 = Et,
EtO2CCHO (E′) and R′CHO (I′, when R = R′CH2) were
obtained via steps i–v and vii–ix, respectively. Substrate 6d
bearing the Ph–N moiety cannot undergo steps vii–ix, and thus
is unable to produce the corresponding R′CHO (I′) and
NH2CH2CO2Et (J′). The reaction of C60 with 6a–6d and E′
affords 7a–7d (step xviii), and that with 6a–6c and I′ furnishes
8a–8c (step xix). For 6a–6c, C60 can react with the in situ gener-
ated J′ and EtO2CCHO or R′CHO to provide 9 or 10 (steps xx
and xxi), respectively. As expected, 9 was indeed obtained in all
cases of 6a–6c. However, only 10c could be isolated because
much more benzaldehyde and J′ from 6c might be produced
than aldehydes (formaldehyde and acetaldehyde) and J′ from 6a
and 6b (vide infra).
To better understand the product distribution for the thermal
reaction of C60 with amino acids or amino acid esters, we per-
formed additional experiments. Control experiments showed
that the reaction of C60 with 1 equiv. of 3a, 1 equiv. of
NH2CH2CO2H and 1 equiv. of CH2O at 180 °C for 7 min gave
4a in 18% yield along with only a trace amount of 2a
(Scheme 4). In comparison, the reaction of C60 with 1 equiv. of
3c, 1 equiv. of NH2CH2CO2H and 1 equiv. of CH2O at 180 °C
for 7 min afforded 4c in 16% yield together with a very small
amount of 2a (Scheme 5); meanwhile the reaction of C60 with 1
equiv. of 3c, 1 equiv. of NH2CH2CO2H and 1 equiv. of PhCHO
at 180 °C for 7 min afforded 5c in 27% yield along with a trace
amount of the corresponding N-unsubstituted fulleropyrrolidine
(Scheme 6). These experiments indicated that the reactivity of
N-substituted glycine was higher than that of glycine. Therefore,
it is understandable that we could not identify the N-unsubsti-
tuted fulleropyrrolidines from the reaction of C60 with 3a–3d
after taking into account the very small amount of the in situ
generated glycine relative to that of 3a–3d.
It was reported that fulleropyrrolidine 4a was unexpectedly
formed from the reaction of C60 with sarcosine and gossypol.24
Product 4a was also obtained from the reaction of C60 with sar-
cosine and 4,6-dimethoxysalicylaldehyde25a or hydroxyacetalde-
hyde.25b However, the authors did not provide any explanation
of how 4a was produced in their cases. Our current work shows
that 4a can be formed directly from the reaction of C60 with sar-
cosine via the aforementioned mechanisms, thus elucidating the
origin for the formation of 4a reported in the literature.24,25
Control experiments showed that the reaction of C60 with 2
equiv. of 3a under typical Prato reaction conditions (refluxing
toluene, 2 h) afforded only a trace amount of 4a. The same reac-
tion in refluxing chlorobenzene for 2 h still gave a very small
amount of 4a. However, when the reaction time was prolonged
to 48 h, as in the reaction of C60 with sarcosine and hydroxyace-
taldehyde,25b 4a was obtained in 20% yield along with 78% of
unreacted C60. These results suggest that although the formation
of fulleropyrrolidines from the direct reaction of C60 with amino
acids under typical Prato reaction conditions is negligible, a
8724 | Org. Biomol. Chem., 2012, 10, 8720–8729
This journal is © The Royal Society of Chemistry 2012

View Article Online
significant amount of fulleropyrrolidines can be generated at
high reaction temperature and long reaction time.
It is obvious from the above results and discussion that both
N-substituted glycines (R′CH2NHCH2CO2H) and N-substituted
glycine esters (R′CH2NHCH2CO2Et) could undergo C–N bond
cleavage at either side of the NH group. To further explore the
innate nature of the C–N bond cleavage and product distribution,
we have conducted additional control experiments. We found
that the reaction of C60 with 1 equiv. of CH3CH2NHCH2CO2H
(3b), 1 equiv. of CH2O and 1 equiv. of CH3CHO at 180 °C for
6 min produced 4b and 5b in 23% and 8% yields, respectively
(Scheme 7). The product ratio of 4b/5b was 3 : 1 and happened
to be the same as that from the direct thermal reaction of C60
with 3b (Table 2, entry 2), suggesting that the same molar
amount of CH2O and CH3CHO was generated from 3b. In com-
parison, the reaction of C60 with 1 equiv. of CH3CH2-
NHCH2CO2Et (6b), 1 equiv. of EtO2CCHO and 1 equiv. of
CH3CHO at 180 °C for 10 min produced 7b and 8b in 38% and
3% yields, respectively (Scheme 8). Combined with the product
yields (7b and 8b in 31% and 13% yields, respectively) for the
thermal reaction of C60 with 6b (Table 3, entry 2), it could be
deduced that EtO2CCHO and CH3CHO were generated in situ
from 6b in a molar ratio of 1 : 5. Thus, the CH3CH2–NH bond
was more easily to break via steps vii–ix in Scheme 1 than the
NH–CH2CO2Et bond via steps i–v in Scheme 1. If 3b in its
thermal reaction with C60 proceeded only via the same routes as
6b, it was expected that CH2O and CH3CHO should be formed
also in a molar ratio of 1 : 5, in sharp contrast to the observed
ratio of 1 : 1. Consequently, other routes, i.e., steps x–xiv, must
operate and contribute to the formation of CH2O. The same pro-
cesses shown in Scheme 2 should also occur for 1a–1d, 3a, 3c
and 3d to provide the corresponding aldehydes.
Scheme 9 Thermal reaction of C60 with 3c, CH2O and PhCHO.
Scheme 10 Thermal reaction of C60 with 6c, EtO2CCHO and PhCHO.
with 1 equiv. of PhCH2NHCH2CO2Et (6c), 1 equiv. of
EtO2CCHO and 1 equiv. of PhCHO at 180 °C for 3 min afforded
7c in 31% yield along with a trace amount of 8c (Scheme 10).
The observation of 5c (23%) and 8c (10%) in the direct reaction
of C60 with 3c and 6c (entry 3, Tables 2 and 3), respectively,
suggested that the PhCH2–NH bond was much more facile to
cleave than the NH–CH2CO2H and NH–CH2CO2Et bonds in
substrates 3c and 6c, respectively, and thus provided significantly
more PhCHO than CH2O (for 3c)/EtO2CCHO (for 6c). The
reactivity for the hydrogen abstraction in step vii should follow
the order of CH3–NH < CH3CH2–NH < PhCH2–NH. Therefore,
more benzaldehyde is expected to form from 3c than acet-
aldehyde from 3b due to the more reactive benzylic position of
the former. It is reasonable to assume that the amount of the
in situ generated CH2O via steps i–vi is not strongly affected by
the N-alkyl group of 3b and 3c. Consequently, the product ratio
of 5c/4c is expected to be higher than that of 5b/4b, fully con-
sistent with our experimental results (Table 2, entries 2 and 3).
Similar to amino acids 3, the formation of EtO2CCHO via steps
i–v is little affected by the N-alkyl group of 6, while the pro-
duction of R′CHO via steps vii–ix should be more favorable for
6c than that for 6b, which in turn is more facile than for 6a, cor-
roborating the observed product distribution of 8c/7c > 8b/7b >
8a/7a (see Table 3). In addition, product 10 was observed only
in the case of 6c, further supporting that much more PhCHO
from 6c was produced than CH2O from 6a and CH3CHO from
6b, respectively.
On the other hand, the reaction of C60 with 1 equiv. of
PhCH2NHCH2CO2H (3c), 1 equiv. of CH2O and 1 equiv. of
PhCHO at 180 °C for 4 min afforded 4c in 22% yield together
with a trace amount of 5c (Scheme 9), while the reaction of C60
Scheme 7 Thermal reaction of C60 with 3b, CH2O and CH3CHO.
Conclusions
Fulleropyrrolidines are commonly obtained by the Prato reaction,
i.e., the reaction of C60 with an amino acid or amino acid ester
with an aldehyde. In the present work, we have demonstrated
that the direct thermal reaction of C60 with amino acids or amino
acid esters in refluxing ODCB without the purposely added
aldehydes under aerobic and dark conditions can also afford
Scheme 8 Thermal reaction of C60 with 6b, EtO2CCHO and
CH3CHO.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 8720–8729 | 8725

View Article Online
fulleropyrrolidines. Plausible reaction mechanisms involving
unusual C–N bond cleavages are proposed to explain the for-
mation of intermediates from the employed amino acids and
amino acid esters, and subsequent addition of the azomethine
ylides to C60 furnishes fulleropyrrolidines. All isolated products
from the reaction mixtures can be well explained by the proposed
reaction mechanisms. Control experiments shed light on the
innate nature of the C–N bond cleavages of amino acids and
amino acid esters and support the proposed reaction mechan-
isms. Our current work also clarifies the origin of the unexpected
formation of N-methylfulleropyrrolidine from the reaction of C60
with sarcosine and gossypol, 4,6-dimethoxysalicylaldehyde or
hydroxyacetaldehyde reported in the literature.
for 10 h gave unreacted C60 (16.1 mg, 44%) and fulleropyrroli-
1
dine 2d13 (17.2 mg, 37%); H NMR (400 MHz, CS2/CDCl3): δ
= 7.95 (d, J = 7.2 Hz, 4H), 7.39 (t, J = 7.5 Hz, 4H), 7.30 (t, J =
7.3 Hz, 2H), 5.95 (s, 2H), 3.11 (br s, 1H).
Thermal reaction of C60 with sarcosine
By the same procedure as above, the reaction of C60 (36.0 mg,
0.05 mmol) with sarcosine (44.3 mg, 0.50 mmol) at 180 °C for
10 h gave unreacted C60 (22.6 mg, 63%) and fulleropyrrolidine
4a12 (11.7 mg, 30%). 1H NMR (300 MHz, CS2/CDCl3): δ =
4.37 (s, 4H), 2.97 (s, 3H).
Thermal reaction of C60 with N-ethylglycine
Experimental
By the same procedure as above, the reaction of C60 (36.1 mg,
0.05 mmol) with N-ethylglycine (52.2 mg, 0.51 mmol) at
180 °C for 2.5 h gave unreacted C60 (15.4 mg, 43%) and a
mixture of 4b and 5b, which was further separated by recycling
HPLC (10 × 250 mm Cosmosil Buckyprep column; flow rate
3 mL min−1; injection volume 5 mL; toluene as eluent) to afford
fulleropyrrolidines 4b15 (13.1 mg, 33%) and 5b (4.5 mg, 11%).
4b: 1H NMR (400 MHz, CS2/CDCl3): δ = 4.40 (s, 4H), 3.15 (q,
J = 7.2 Hz, 2H), 1.55 (t, J = 7.2 Hz, 3H). 5b: 1H NMR
(400 MHz, CS2/CDCl3): δ = 4.88 (d, J = 9.2 Hz, 1H), 4.02 (q,
J = 6.3 Hz, 1H), 3.98 (d, J = 9.2 Hz, 1H), 3.50 (dq, J = 12.0, 7.4 Hz,
1H), 2.73 (dq, J = 12.0, 7.0 Hz, 1H), 1.95 (d, J = 6.3 Hz,
3H), 1.53 (t, J = 7.2 Hz, 3H); 13C NMR (75 MHz, CS2/DMSO-
d6 with Cr(acac)3 as relaxation reagent, all 1C unless indicated):
δ = 156.69, 154.58, 154.23, 153.42, 147.34 (2C), 146.98,
146.84, 146.70, 146.43, 146.39, 146.28, 146.26, 146.22, 146.17,
146.06 (2C), 145.92, 145.75, 145.65 (3C), 145.52, 145.43,
145.38 (2C), 145.33, 145.28, 144.90, 144.77, 144.52 (2C),
143.28, 143.16, 142.80, 142.77, 142.74, 142.72, 142.38 (3C),
142.34, 142.29, 142.28, 142.20, 142.18, 142.17, 142.06, 141.85,
141.81, 140.39, 140.32 (2C), 139.88, 137.72, 136.68, 136.13,
135.93, 75.89 (sp3-C of C60), 71.77 (NCHCH3), 69.20 (sp3-C of
C60), 66.49 (CH2N), 46.64 (NCH2CH3), 17.35 (CHCH3), 13.76
(NCH2CH3); FT-IR ν/cm−1 (KBr) 2962, 2923, 1506, 1451,
1426, 1377, 1332, 1186, 766, 699, 572, 525; UV-vis (CHCl3)
λmax/nm (lg ε) 257 (4.92), 323 (4.40), 431 (3.43), 704 (2.55);
HRMS (MALDI TF-ICR): m/z calcd for C65H11N: 805.0891;
found: 805.0895.
Thermal reaction of C60 with glycine
A 25 mL tube containing a mixture of C60 (36.1 mg,
0.05 mmol) and glycine (76.2 mg, 1.0 mmol) in ODCB (6 mL)
was wrapped with aluminum foil, sealed up, and then heated in
an oil bath preset at 200 °C for 24 h (monitored by TLC). After
removal of the solvent in vacuo, flash chromatography of the
residue on a silica gel column with carbon disulfide as the eluent
to afford unreacted C60 (22.7 mg, 63%), then with a mixture of
carbon disulfide and dichloromethane to give fulleropyrrolidine
1
2a12 (7.4 mg, 19%). H NMR (400 MHz, CS2/CDCl3): δ = 4.78
(s, 4H).
Thermal reaction of C60 with alanine
By the same procedure as above, the reaction of C60 (36.0 mg,
0.05 mmol) with alanine (46.0 mg, 0.52 mmol) at 200 °C for
48 h gave unreacted C60 (13.4 mg, 37%) and fulleropyrrolidine
2b13 (13.8 mg, 35%) as a mixture of cis and trans isomers in a
1
ratio of 3.6 : 1. cis-2b: H NMR (400 MHz, CS2/CDCl3): δ =
1
4.79 (q, J = 6.6 Hz, 2H), 2.02 (d, J = 6.6 Hz, 6H). trans-2b: H
NMR (400 MHz, CS2/CDCl3): δ = 5.06 (q, J = 6.8 Hz, 2H),
2.04 (d, J = 6.8 Hz, 6H).
Thermal reaction of C60 with phenylalanine
By the same procedure as above, the reaction of C60 (35.9 mg,
0.05 mmol) with phenylalanine (82.2 mg, 0.50 mmol) at 180 °C
for 12 h gave unreacted C60 (25.6 mg, 71%) and fulleropyrroli-
dine 2c13 (8.7 mg, 19%) as a mixture of cis and trans isomers in
Thermal reaction of C60 with N-benzylglycine
1
a ratio of 1 : 1.1. cis-2c: H NMR (400 MHz, CS2/CDCl3): δ =
By the same procedure as above, the reaction of C60 (36.1 mg,
0.05 mmol) with N-benzylglycine (82.5 mg, 0.50 mmol) at
180 °C for 28 h gave unreacted C60 (18.3 mg, 51%), fullero-
pyrrolidines 4c16 (9.4 mg, 22%) and 5c17 (10.7 mg, 23%). 4c:
1H NMR (300 MHz, CS2/CDCl3): δ = 7.69 (d, J = 7.1 Hz, 2H),
7.46 (t, J = 7.3 Hz, 2H), 7.36 (t, J = 7.3 Hz, 1H), 4.43 (s, 4H),
4.30 (s, 2H); 13C NMR (75 MHz, CS2/CDCl3 with Cr(acac)3 as
relaxation reagent): δ = 154.76 (4C), 147.13 (2C), 146.09 (4C),
145.90 (8C), 145.52 (2C), 145.30 (4C), 145.13 (4C), 144.41
(4C), 142.95 (2C), 142.49 (4C), 142.10 (4C), 141.92 (4C),
141.74 (4C), 140.04 (4C), 137.77 (1C, aryl C), 136.15 (4C),
128.68 (2C, aryl C), 128.65 (2C, aryl C), 127.55 (1C, aryl C),
7.40 (d, J = 7.2 Hz, 4H), 7.30 (t, J = 7.5 Hz, 4H), 7.21 (t, J =
7.3 Hz, 2H), 4.74 (dd, J = 10.6, 3.0 Hz, 2H), 3.87 (dd, J = 13.4,
3.0 Hz, 2H), 3.35 (dd, J = 13.4, 10.6 Hz, 2H), 2.63 (br s, 1H).
trans-2c: H NMR (400 MHz, CS2/CDCl3): δ = 7.27–7.24 (m,
10H), 5.11 (dd, J = 11.1, 3.5 Hz, 2H), 3.68 (dd, J = 13.3, 3.5
Hz, 2H), 3.50 (dd, J = 13.3, 11.1 Hz, 2H), 2.63 (br s, 1H).
1
Thermal reaction of C60 with phenylglycine
By the same procedure as above, the reaction of C60 (36.2 mg,
0.05 mmol) with phenylglycine (75.4 mg, 0.50 mmol) at 180 °C
8726 | Org. Biomol. Chem., 2012, 10, 8720–8729
This journal is © The Royal Society of Chemistry 2012

View Article Online
70.54 (2C, sp3-C of C60), 67.47 (2C, CH2), 58.82 (1C, NCH2).
5c: H NMR (400 MHz, CS2/CDCl3): δ = 7.93 (br s, 2H), 7.68
indicated): δ = 169.17 (CvO), 153.30, 149.98, 146.47, 145.69,
145.65, 145.35, 145.33, 145.18, 144.87, 144.83 (4C), 144.75,
144.66, 144.61, 143.79, 143.62, 142.46, 142.09, 142.05, 141.55
(6C), 141.49, 141.26, 141.18, 139.41, 138.94, 135.86, 134.92,
75.70 (sp3-C of C60), 73.09 (CHCO2), 61.24 (OCH2CH3), 13.91
(OCH2CH3); FT-IR ν/cm−1 (KBr) 2923, 1739, 1566, 1520,
1429, 1373, 1279, 1199, 1027, 942, 803, 751, 592, 527; UV-vis
(CHCl3) λmax nm (lg ε) 251 (4.97), 325 (4.50), 428 (3.47), 693
(2.62); HRMS (MALDI TF-ICR): m/z calcd for C68H13NO4:
907.0845; found: 907.0842.
1
(d, J = 7.4 Hz, 2H), 7.50–7.43 (m, 4H), 7.40–7.33 (m, 2H), 5.21
(s, 1H), 4.86 (d, J = 9.4 Hz, 1H), 4.59 (d, J = 13.4 Hz, 1H), 4.16
(d, J = 9.4 Hz, 1H), 3.67 (d, J = 13.4 Hz, 1H); 13C NMR
(75 MHz, CS2/CDCl3 with Cr(acac)3 as relaxation reagent, all
1C unless indicated): δ = 156.32, 154.04, 153.39, 153.33,
147.29 (2C), 146.79, 146.40, 146.29, 146.26, 146.20, 146.17,
146.14, 146.12, 146.09, 145.93 (2C), 145.74, 145.53 (2C),
145.51, 145.49, 145.35, 145.30, 145.25 (2C), 145.22, 145.13,
144.70, 144.62, 144.39, 144.36, 143.15, 142.98, 142.67, 142.58,
142.55 (2C), 142.34, 142.24, 142.14, 142.12, 142.10, 142.07,
142.01, 141.99, 141.92, 141.82, 141.64, 141.54, 140.18, 140.14,
139.89, 139.45, 137.85 (aryl C), 136.99 (aryl C), 136.84,
136.51, 135.92, 135.71, 129.46 (2C, aryl C), 128.86 (2C, aryl
C), 128.82 (2C, aryl C), 128.74 (2C, aryl C), 128.66 (aryl C),
127.56 (aryl C), 81.54 (CHPh), 77.37 (sp3-C of C60), 68.68
(sp3-C of C60), 66.69 (CH2), 56.84 (NCH2).
Thermal reaction of C60 with N-ethylglycine ethyl ester
By the same procedure as above, the reaction of C60 (36.1 mg,
0.05 mmol) with N-ethylglycine ethyl ester (67.8 mg,
0.52 mmol) at 180 °C for 2 h gave unreacted C60 (16.2 mg,
45%), 7b (14.5 mg, 31%), 8b (5.6 mg, 13%) and 9 as a mixture
of cis-isomer4a,17 (1.4 mg, 3.0%) and trans-isomer4a (0.4 mg,
1
0.9%). 7b: H NMR (400 MHz, CDCl3): δ = 5.92 (s, 2H), 4.41
(dq, J = 10.8, 7.1 Hz, 2H), 4.31 (dq, J = 10.8, 7.1 Hz, 2H), 3.52
(dq, J = 11.8, 7.5 Hz, 1H), 3.16 (dq, J = 11.8, 7.0 Hz, 1H), 1.57
(t, J = 7.3 Hz, 3H), 1.28 (t, J = 7.1 Hz, 6H); 13C NMR
(75 MHz, CDCl3 with Cr(acac)3 as relaxation reagent, all 2C
unless indicated): δ = 170.56 (CvO), 153.85, 151.14, 147.54,
146.51, 146.45, 146.19 (4C), 146.14, 145.95, 145.74 (4C),
145.61, 145.44, 145.39, 144.66, 144.62, 143.20, 142.84, 142.76,
142.33, 142.25, 142.11, 142.02, 141.95, 141.87, 140.12, 139.68,
137.28, 136.61, 74.21 (CHCO2), 71.64 (sp3-C of C60), 61.63
(OCH2CH3), 43.40 (1C, NCH2CH3), 14.45 (OCH2CH3), 13.71
(1C, NCH2CH3). FT-IR ν/cm−1 (KBr) 2970, 1727, 1509, 1427,
1366, 1337, 1263, 1166, 1093, 1012, 802, 573, 523; UV-vis
(CHCl3) λmax/nm (lg ε) 256 (5.00), 311 (4.52), 429 (3.53), 695
(2.84); HRMS (MALDI TF-ICR): m/z calcd for C70H17NO4:
935.1158; found: 935.1165. 8b: 1H NMR (400 MHz, CS2/
CDCl3): δ = 5.56 (s, 1H), 5.40 (q, J = 6.4 Hz, 1H), 4.35 (dq, J =
10.8, 7.1 Hz, 1H), 4.27 (dq, J = 10.8, 7.1 Hz, 1H), 3.43 (dq,
J = 12.0, 7.5 Hz, 1H), 3.10 (dq, J = 12.0, 7.0 Hz, 1H), 1.90 (d,
J = 6.4 Hz, 3H), 1.52 (t, J = 7.3 Hz, 3H), 1.29 (t, J = 7.1 Hz, 3H);
13C NMR (75 MHz, CS2/DMSO-d6 with Cr(acac)3 as relaxation
reagent, all 1C unless indicated): δ = 169.90 (CvO), 156.25,
154.46, 153.92, 151.48, 147.48, 147.35, 146.83, 146.81, 146.77,
146.45, 146.44, 146.41, 146.31, 146.20, 146.17, 146.09, 146.06,
145.82, 145.65, 145.63 (2C), 145.61, 145.49, 145.45, 145.42,
145.37 (2C), 145.24, 144.80 (2C), 144.63, 144.47, 143.26,
143.14, 142.84, 142.78 (2C), 142.70, 142.40, 142.33, 142.32
(2C), 142.23, 142.18 (2C), 142.07, 141.97, 141.91, 141.84 (2C),
140.42, 140.30, 139.89, 139.66, 137.67, 137.08, 136.51, 136.31,
75.05 (sp3-C of C60), 73.42 (CHCO2), 71.36 (sp3-C of C60),
67.20 (CHCH3), 61.02 (OCH2CH3), 42.25 (NCH2CH3), 18.15
(CHCH3), 14.99 (OCH2CH3), 14.73 (NCH2CH3). FT-IR ν/cm−1
(KBr) 2964, 2921, 1729, 1509, 1450, 1429, 1377, 1336, 1169,
574, 524; UV-vis (CHCl3) λmax/nm (lg ε) 256 (4.93), 320 (4.40),
430 (3.43), 701 (2.46); HRMS (MALDI TF-ICR): m/z calcd for
C68H15NO2: 877.1103; found: 877.1112.
Thermal reaction of C60 with N-phenylglycine
By the same procedure as above, the reaction of C60 (36.0 mg,
0.05 mmol) with N-phenylglycine (16.2 mg, 0.11 mmol) at
180 °C for 1.5 h gave unreacted C60 (17.6 mg, 49%) and fullero-
pyrrolidine 4d (15.1 mg, 36%). 1H NMR (300 MHz, CS2/
CDCl3): δ = 7.41 (t, J = 7.4 Hz, 2H), 7.25 (d, J = 7.7 Hz, 2H),
7.00 (t, J = 7.3 Hz, 1H), 5.14 (s, 4H); 13C NMR (75 MHz, CS2/
CDCl3 with Cr(acac)3 as relaxation reagent): δ = 153.77 (4C),
146.82 (2C), 145.74 (4C), 145.54 (4C), 145.30 (4C), 145.09
(2C), 145.02 (4C), 144.75 (4C), 144.01 (4C), 142.57 (2C),
142.12 (4C), 141.67 (4C), 141.54 (4C), 141.37 (4C), 139.72
(4C), 135.63 (4C), 129.14 (2C, aryl C), 120.18 (1C, aryl C),
116.10 (2C, aryl C), 69.23 (2C, sp3-C of C60), 62.63 (CH2N);
FT-IR ν/cm−1 (KBr) 2921, 1597, 1501, 1466, 1427, 1353, 1216,
1187, 751, 687, 558, 526; UV-vis (CHCl3) λmax/nm (lg ε) 256
(5.05), 308 (4.55), 430 (3.53), 703 (2.09); HRMS (MALDI
TF-ICR): m/z calcd for C68H9N: 839.0735; found: 839.0732.
Thermal reaction of C60 with sarcosine ethyl ester
By the same procedure as above, the reaction of C60 (36.0 mg,
0.05 mmol) with sarcosine ethyl ester (57.3 mg, 0.49 mmol) at
180 °C for 4 h gave unreacted C60 (16.2 mg, 45%), 7a17
(11.0 mg, 24%), 8a4a (3.9 mg, 9%) and 9 as a mixture of cis-
isomer4a,17 (1.1 mg, 2.4%) and trans-isomer4a (0.3 mg, 0.7%).
7a: 1H NMR (400 MHz, CDCl3): δ = 5.82 (s, 2H), 4.42 (dq, J =
10.8, 7.1 Hz, 2H), 4.33 (dq, J = 10.8, 7.1 Hz, 2H), 3.10 (s, 3H),
1.28 (t, J = 7.1 Hz, 6H). 8a: 1H NMR (300 MHz, CS2/CDCl3): δ
= 4.96 (d, J = 9.5 Hz, 1H), 4.79 (s, 1H), 4.41 (dq, J = 10.8, 7.2
Hz, 1H), 4.31 (dq, J = 10.8, 7.1 Hz, 1H), 4.25 (d, J = 9.5 Hz,
1H), 3.02 (s, 3H), 1.26 (t, J = 7.1 Hz, 3H). cis-9: 1H NMR
(400 MHz, CDCl3): δ = 5.51 (d, J = 13.6 Hz, 2H), 4.48 (t, J =
13.6 Hz, 1H, NH), 4.43 (dq, J = 10.8, 7.1 Hz, 2H), 4.32 (dq, J =
1
10.8, 7.1 Hz, 2H), 1.26 (t, J = 7.1 Hz, 6H). trans-9: H NMR
(400 MHz, CDCl3): δ = 6.01 (d, J = 9.0 Hz, 2H), 4.42 (dq, J =
10.8, 7.1 Hz, 2H), 4.33 (dq, J = 10.8, 7.1 Hz, 2H), 4.21 (t, J =
9.0 Hz, 1H, NH), 1.28 (t, J = 7.1 Hz, 6H); 13C NMR (75 MHz,
CS2/CDCl3 with Cr(acac)3 as relaxation reagent, all 2C unless
Thermal reaction of C60 with N-benzylglycine ethyl ester
By the same procedure as above, the reaction of C60 (35.9 mg,
0.05 mmol) with N-benzylglycine ethyl ester (96.7 mg,
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 8720–8729 | 8727

View Article Online
0.50 mmol) at 180 °C for 40 min gave unreacted C60 (16.9 mg,
47%), 7c (8.2 mg, 16%), 8c (5.0 mg, 10%) and 9 as a mixture
of cis-isomer4a,17 (4.5 mg, 10%) and trans-isomer4a (1.5 mg,
3%), and 10c18 (0.3 mg, 0.7%). 7c: 1H NMR (300 MHz,
CDCl3): δ = 7.76 (d, J = 7.2 Hz, 2H), 7.47 (t, J = 7.5 Hz, 2H),
7.39 (t, J = 7.2 Hz, 1H), 5.80 (s, 2H), 4.74 (d, J = 13.1 Hz, 1H),
4.37–4.30 (m, 4H), 4.14 (d, J = 13.1 Hz, 1H), 1.25 (t, J = 6.9
Hz, 6H); 13C NMR (75 MHz, CS2/CDCl3 with Cr(acac)3 as
relaxation reagent, all 2C unless indicated): δ = 169.82 (CvO),
153.49, 150.73, 147.17, 146.17, 146.11, 145.84 (4C), 145.78,
145.66, 145.41, 145.39, 145.27, 145.09, 145.03, 144.32, 144.28,
142.85, 142.51, 142.43, 141.98, 141.92, 141.82, 141.67, 141.62,
141.50, 139.79, 139.30, 137.35 (1C, aryl C), 136.82, 136.10,
128.78 (aryl C), 128.56 (aryl C), 127.68 (1C, aryl C), 73.05
(CHCO2), 71.26 (sp3-C of C60), 61.18 (OCH2CH3), 52.27 (1C,
NCH2), 14.27 (OCH2CH3); FT-IR ν/cm−1 (KBr) 2923, 1731,
1567, 1520, 1375, 1279, 1146, 1095, 1027, 941, 803, 751, 679,
592, 526; UV-vis (CHCl3) λmax nm (lg ε) 256 (5.01), 310 (4.52),
429 (3.36), 693 (1.81); HRMS (MALDI TF-ICR): m/z calcd for
C75H19NO4: 997.1314; found: 997.1308. 8c: 1H NMR
(300 MHz, CS2/CDCl3): δ = 7.89 (br s, 2H), 7.65 (d, J = 7.3 Hz,
2H), 7.46–7.28 (m, 6H), 6.64 (s, 1H), 5.43 (s, 1H), 4.50 (d, J =
13.8 Hz, 1H), 4.39 (dq, J = 10.7, 7.1 Hz, 1H), 4.28 (dq, J =
10.7, 7.1 Hz, 1H), 4.05 (d, J = 13.8 Hz, 1H), 1.26 (t, J = 7.1 Hz,
3H); 13C NMR (75 MHz, CS2/acetone-d6 with Cr(acac)3 as
relaxation reagent, all 1C unless indicated): δ = 170.34 (COO),
156.23, 154.25, 153.96, 151.31, 147.67, 147.54, 147.06, 146.87,
146.62 (2C), 146.51, 146.45, 146.37, 146.33, 146.29, 146.20
(2C), 145.92, 145.83 (3C), 145.78, 145.58, 145.53, 145.49,
145.45 (2C), 145.36, 144.96, 144.86, 144.74, 144.63, 143.40,
143.25, 142.98, 142.87 (2C), 142.82, 142.45 (3C), 142.41,
142.36, 142.29, 142.28 (2C), 142.03, 141.99, 141.90, 141.85,
140.41, 140.34, 139.80, 139.73, 138.25, 137.59, 137.34, 136.58,
136.42, 136.38, 129.35 (4C, aryl C), 129.22 (2C, aryl C),
128.87 (3C, aryl C), 128.22 (1C, aryl C), 77.32 (CHPh), 76.02
(sp3-C of C60), 73.04 (CHCO2), 70.89 (sp3-C of C60), 61.30
(OCH2CH3), 51.90 (NCH2), 14.93 (OCH2CH3); FT-IR ν/cm−1
(KBr) 1731, 1493, 1454, 1428, 1188, 1166, 1138, 1027, 906,
732, 700, 575, 527; UV-vis (CHCl3) λmax nm (lg ε) 257 (4.98),
311 (4.53), 431 (3.44), 703 (1.90); HRMS (MALDI TF-ICR):
m/z calcd for C78H19NO2: 1001.1416; found: 1001.1422. 10c:
1H NMR (400 MHz, CS2/CDCl3): δ = 7.76 (d, J = 7.2 Hz, 2H),
7.43 (t, J = 7.5 Hz, 2H), 7.35 (t, J = 7.4 Hz, 1H), 5.85 (s, 1H),
5.61 (s, 1H), 4.43 (dq, J = 10.8, 7.1 Hz, 1H), 4.34 (dq, J = 10.8,
7.1 Hz, 1H), 1.29 (t, J = 7.1 Hz, 3H).
144.26, 144.19, 142.82, 142.46, 142.38, 141.93, 141.86, 141.62
(4C), 141.50, 141.46, 139.89, 139.33, 136.57, 135.88, 129.21
(aryl C), 122.13 (1C, aryl C), 118.68 (aryl C), 74.12 (CHCO2),
70.75 (sp3-C of C60), 61.57 (OCH2CH3), 13.89 (OCH2CH3).
FT-IR ν/cm−1 (KBr) 2923, 1734, 1598, 1498, 1461, 1368, 1340,
1292, 1262, 1180, 1015, 757, 692, 576, 526; UV-vis (CHCl3)
λmax/nm (lg ε) 257 (5.03), 310 (4.52), 429 (3.41), 691 (2.14);
HRMS (MALDI TF-ICR): m/z calcd for C74H17NO4: 983.1158;
found: 983.1163.
Reaction of C60 with sarcosine, glycine and formaldehyde
A 25-mL tube containing a mixture of C60 (72.1 mg,
0.10 mmol), sarcosine (8.9 mg, 0.10 mmol), glycine (7.5 mg,
0.10 mmol) and formaldehyde solution (7.2 μL, 0.10 mmol) in
ODCB (10 mL) was wrapped with aluminum foil, sealed up,
and ultrasonicated until C60 was dissolved, and then heated in an
oil bath preset at 180 °C for 7 min. The usual workup gave
unreacted C60 (48.1 mg, 67%), 4a12 (14.3 mg, 18%) and a trace
amount of 2a.12
Reaction of C60 with N-benzylglycine, glycine and formaldehyde
By the same procedure as above, the reaction of C60 (72.1 mg,
0.10 mmol) with N-benzylglycine (16.5 mg, 0.10 mmol),
glycine (7.5 mg, 0.10 mmol) and formaldehyde solution
(7.2 μL, 0.10 mmol) at 180 °C for 7 min gave unreacted C60
(41.4 mg, 57%), 4c16 (13.8 mg, 16%), and a trace amount
of 2a.12
Reaction of C60 with N-benzylglycine, glycine and benzaldehyde
By the same procedure as above, the reaction of C60 (72.0 mg,
0.10 mmol), N-benzylglycine (16.5 mg, 0.10 mmol), glycine
(7.5 mg, 0.10 mmol) and benzaldehyde (10 μL, 0.10 mmol) at
180 °C for 7 min gave unreacted C60 (45.3 mg, 63%), 5c17
(25.4 mg, 27%) and a trace amount of the corresponding
N-unsubstituted fulleropyrrolidine.
Reaction of C60 with N-ethylglycine, formaldehyde and
acetaldehyde
By the same procedure as above, the reaction of C60 (71.9 mg,
0.10 mmol), N-ethylglycine (10.3 mg, 0.10 mmol), formal-
dehyde solution (7.2 μL, 0.10 mmol) and acetaldehyde solution
(12 μL, 0.10 mmol) at 180 °C for 6 min gave unreacted C60
(28.7 mg, 40%), 4b15 (18.5 mg, 23%) and 5b (6.4 mg, 8%).
Thermal reaction of C60 with ethyl N-phenylglycinate
By the same procedure as above, the reaction of C60 (36.3 mg,
0.05 mmol) with ethyl N-phenylglycinate (182.4 mg,
1.02 mmol) at 180 °C for 75 h gave unreacted C60 (18.1 mg,
50%) and fulleropyrrolidine 7d (5.3 mg, 11%). 1H NMR
(300 MHz, CDCl3): δ = 7.45 (t, J = 7.7 Hz, 2H), 7.33 (d, J = 7.8
Hz, 2H), 7.12 (t, J = 7.2 Hz, 1H), 6.54 (s, 2H), 4.26–4.16 (m,
4H), 1.15 (t, J = 7.1 Hz, 6H); 13C NMR (75 MHz, CDCl3 with
Cr(acac)3 as relaxation reagent, all 2C unless indicated): δ =
169.81 (CvO), 153.04, 150.17, 147.20, 146.13, 146.09, 145.82
(4C), 145.46, 145.40 (4C), 145.32 (4C), 145.06 (3C), 144.99,
Reaction of C60 with N-ethylglycine ethyl ester, ethyl glyoxalate
and acetaldehyde
By the same procedure as above, the reaction of C60 (72.0 mg,
0.10 mmol), N-ethylglycine ethyl ester (13.1 mg, 0.10 mmol),
ethyl glyoxalate solution (20.5 μL, 0.10 mmol) and acetaldehyde
solution (12 μL, 0.10 mmol) at 180 °C for 10 min gave
unreacted C60 (34.5 mg, 48%), 7b (35.5 mg, 38%) and 8b
(2.6 mg, 3%).
8728 | Org. Biomol. Chem., 2012, 10, 8720–8729
This journal is © The Royal Society of Chemistry 2012

View Article Online
63, 4240; (c) R. Skanji, M. B. Messaouda, Y. Zhang, M. Abderrabba,
H. Szwarc and F. Moussa, Tetrahedron, 2012, 68, 2713.
5 X. Zhang, L. Gan, S. Huang and Y. Shi, J. Org. Chem., 2004, 69, 5800.
6 G.-W. Wang, J.-X. Li, Y.-J. Li and Y.-C. Liu, J. Org. Chem., 2006, 71,
680.
7 G.-W. Wang, X.-P. Chen and X. Cheng, Chem.–Eur. J., 2006, 12, 7246.
8 (a) L. B. Gan, C. P. Luo, L. B. Xu, D. J. Zhou, C. H. Huang and
S. K. Zhao, Chin. Chem. Lett., 1994, 5, 275; (b) D. J. Zhou, L. B. Gan,
L. B. Xu, C. P. Luo and C. H. Huang, Fullerene Sci. Technol., 1995, 3,
127.
9 Preliminary results have been described in a few theses from our labora-
tory: (a) T.-H. Zhang, Ph.D. thesis, University of Science and Technology
of China, 2004; (b) X. Cheng, M.S. thesis, University of Science and
Technology of China, 2005; (c) C.-K. Mai, B.S. thesis, University of
Science and Technology of China, 2007.
Reaction of C60 with N-benzylglycine, formaldehyde and
benzaldehyde
By the same procedure as above, the reaction of C60 (36.1 mg,
0.05 mmol), N-benzylglycine (8.4 mg, 0.05 mmol), formal-
dehyde (3.6 μL, 0.05 mmol) and benzaldehyde (5.2 μL,
0.05 mmol) in ODCB (6 mL) at 180 °C for 4 min gave
unreacted C60 (27.1 mg, 75%), 4c16 (9.4 mg, 22%) and a trace
amount of 5c.17
Reaction of C60 with N-benzylglycine ethyl ester, ethyl glyoxalate
and benzaldehyde
10 Part of this work has been published in a communication: B. Jin,
R.-F. Peng, J. Shen, G.-W. Wang, B.-S. Tan and S.-J. Chu, Synth.
Commun., 2012, 42, 1532.
By the same procedure as above, the reaction of C60 (35.9 mg,
0.05 mmol), N-benzylglycine ethyl ester (9.7 mg, 0.05 mmol),
ethyl glyoxalate (10 μL, 0.05 mmol) and benzaldehyde (5.2 μL,
0.05 mmol) in ODCB (6 mL) at 180 °C for 3 min gave
unreacted C60 (21.0 mg, 58%), 7c (15.3 mg, 31%) and a trace
amount of 8c.
11 The inconsistency with the previous observation10 was probably caused
by the incomplete exclusion of oxygen in the system.
12 M. Maggini, G. Scorrano and M. Prato, J. Am. Chem. Soc., 1993, 115,
9798.
13 G.-W. Wang and Z.-G. Tian, Tetrahedron Lett., 2009, 50, 3860.
14 Adduct 2a was stable in dilute solution at room temperature, yet it par-
tially decomposed upon concentration after column separation,12 and thus
some extra peaks in the 1H NMR spectrum of 2a could be observed.
15 G.-W. Wang, T.-H. Zhang, E.-H. Hao, L.-J. Jiao, Y. Murata and
K. Komatsu, Tetrahedron, 2003, 59, 55.
Acknowledgements
We are grateful for financial support from the National Natural
Science Foundation of China (Nos. 21132007, 20972145),
National Basic Research Program of China (2011CB921402),
and Open Project of State Key Laboratory Cultivation Base for
Nonmetal Composites and Functional Materials (11zxfk15).
16 X. Zhang, M. Willems and C. S. Foote, Tetrahedron Lett., 1993, 34,
8187.
17 S.-H. Wu, D.-W. Zhang, G.-W. Wang, L.-H. Shu, H.-M. Wu, J.-F. Xu and
X.-F. Lao, Synth. Commun., 1997, 27, 2289.
18 R. Thayumanavan, B. C. Hawkins, P. A. Keller, S. G. Pyne and
G. E. Ball, Org. Lett., 2008, 10, 1315.
19 (a) A. Hirsch, Q. Li and F. Wudl, Angew. Chem., Int. Ed. Engl., 1991, 30,
1309; (b) H. Isobe, T. Tanaka, W. Nakanishi, L. Lemiègre and
E. Nakamura, J. Org. Chem., 2005, 70, 4826.
20 T. Konishi, M. Fujitsuka and O. Ito, Chem. Lett., 2000, 202.
21 (a) K. L. Waters, Chem. Rev., 1947, 41, 585; (b) J. Vicario, D. Aparicio
and F. Palacios, Tetrahedron Lett., 2011, 52, 4109.
22 B. R. Weiner and R. N. Rosenfeld, J. Org. Chem., 1983, 48, 5362.
23 (a) K. Okada, K. Okubo and M. Oda, J. Photochem. Photobiol., A, 1991,
57, 265; (b) W. Zhang, Y. Su, L. Gan, J. Jiang and C. Huang, Chem.
Lett., 1997, 1007.
24 J. Yin, M. Nigh, D. Vanderveer, Y. Jiang, X. Wang and F. Chen, Carbon,
2009, 47, 2883.
25 (a) M. Izquierdo, S. Osuna, S. Filippone, A. Martín-Domenech, M. Solà
and N. Martín, J. Org. Chem., 2009, 74, 1480; (b) M. Izquierdo,
S. Osuna, S. Filippone, A. Martín-Domenech, M. Solà and N. Martín,
J. Org. Chem., 2009, 74, 6253.
Notes and references
1 For selected reviews, see: (a) R. Taylor and D. R. M. Walton, Nature,
1993, 363, 685; (b) F. Wudl, J. Mater. Chem., 2002, 12, 1959;
(c) E. Nakamura and H. Isobe, Acc. Chem. Res., 2003, 36, 807;
(d) D. M. Guldi, F. Zerbetto, V. Georgakilas and M. Prato, Acc. Chem.
Res., 2005, 38, 38.
2 (a) M. Prato and M. Maggini, Acc. Chem. Res., 1998, 31, 519;
(b) N. Tagmatarchis and M. Prato, Synlett, 2003, 768.
3 K. Matsumoto, K. Hashimoto, M. Kamo, Y. Uetani, S. Hayase,
M. Kawatsura and T. Itoh, J. Mater. Chem., 2010, 20, 9226.
4 (a) L. Gan, D. Zhou, C. Luo, H. Tan, C. Huang, M. Lü, J. Pan and
Y. Wu , J. Org. Chem., 1996, 61, 1954; (b) L. Gan, J. Jiang, W. Zhang,
Y. Su, Y. Shi, C. Huang, J. Pan, M. Lü and Y. Wu, J. Org. Chem., 1998,
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 8720–8729 | 8729