Radical reaction of [60]fullerene with phosphorus compounds mediated by manganese(III) acetate

Article abstract of DOI:10.1021/jo2007384

Radical reaction of [60]fullerene with phosphonates or phosphine oxide mediated by manganese(III) acetate dihydrate in chlorobenzene under three different conditions afforded three different types of phosphorylated fullerenes: singly bonded fullerene dime

Full text of DOI:10.1021/jo2007384

ARTICLE
pubs.acs.org/joc
Radical Reaction of [60]Fullerene with Phosphorus Compounds
Mediated by Manganese(III) Acetate
Guan-Wu Wang,*,†,‡ Cong-Zhou Wang,† and Jian-Ping Zou§
†Hefei National Laboratory for Physical Sciences at Microscale, CAS Key Laboratory of Soft Matter Chemistry, Joint Laboratory of
Green Synthetic Chemistry, and Department of Chemistry, University of Science and Technology of China, Hefei, Anhui 230026,
People’s Republic of China
‡State Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou, Gansu 730000, People’s Republic of China
§College of Chemistry and Chemical Engineering, Suzhou University, 199 Renai Street, Suzhou, Jiangsu 215123,
People’s Republic of China
S Supporting Information
b
ABSTRACT: Radical reaction of [60]fullerene with phosphonates or
phosphine oxide mediated by manganese(III) acetate dihydrate in chloro-
benzene under three different conditions afforded three different types of
phosphorylated fullerenes: singly bonded fullerene dimers 2, hydropho-
sphorylated fullerenes 3, and acetoxylated fullerene derivatives 4. In
addition, interconversions among the three types of phosphorylated full-
erene derivatives have also been investigated. A possible reaction mechan-
ism was proposed to explain experimental results.
’ INTRODUCTION
Studies on the reaction of phosphorus radicals with fullerenes
are rare. In continuation of our interest in manganese(III) acetate
Mn(OAc)3-mediated radical reactions of C60,6,7 we recently
reported the reaction of C60 with phosphorus radicals formed
from phosphonate esters in the presence of Mn(OAc)3 and
obtained singly bonded fullerene dimers.8 Later on, we found
that the reaction could be extended to phosphine oxide. More
interestingly, we uncovered that another two different types of
products could be generated by simply adjusting the reaction
conditions. Herein we disclose these intriguing results: the Mn-
(OAc)3-mediated reaction of C60 with phosphonate esters or
phosphine oxide affords selectively one of the three types of
phosphorylated fullerene derivatives, that is, singly bonded full-
erene dimers, hydrofullerenes, and acetoxylated fullerenes, sim-
ply under three different reaction conditions. In addition, inter-
conversions among the three types of phosphorylated fullerene
derivatives have also been explored.
Among the myriad of fullerene derivatives reported today,
only a limited number of fullerene compounds containing a phos-
phorus atom are known.1 Scarce examples of phosphorylated
fullerene derivatives with the phosphorus atom attached directly
to the fullerene skeleton have been reported and are limited to
hydrofullerenes only.2,3 Wu et al. explored the reaction of C60
with trialkylphosphine oxides and obtained hydrofullerenes R2-
(O)PC60H.2 Nakamura and co-workers3a described the reaction
of C60 with a lithiated phosphineꢀborane or a lithiated phos-
phinite borane followed by acid quenching and removal of the
BH3 group to afford a phosphine or a phosphinite bearing a ful-
lerene substituent. They also found that the addition of (MeO)2-
(O)PLitoC60 followedbyacidquenchinggave(MeO)2(O)PC60H.
A few years later, they improved the synthesis of RR0(O)PC60H
by performing the reaction under neutral conditions.3b When
Chuang and co-workers studied the reaction of C60 with dimethyl
acetylenedicarboxylate and P(NMe2)3 or P(NEt2)3, they ob-
tained hydrofullerene (R2N)2(O)PC60H (R = Me, Et) as a
byproduct, which could be formed directly from the reaction
of C60 with P(NMe2)3 or P(NEt2)3, albeit still in low yields (7%
and 15%, respectively).3c Some of the phosphorus-containing
fullerene derivatives were shown to have active biological activities2,4
and optical properties.5 Therefore, it is demanding to achieve the
synthesis of other types of phosphorylated fullerene derivatives
for the purpose of diversity and potential application.
’ RESULTS AND DISCUSSION
We previously reported that the Mn(OAc)3-mediated reac-
tion of C60 with phosphorus radicals generated from dimethyl
phosphonate (1a), diethyl phosphonate (1b), and 5,5-dimethyl-
1,3,2-dioxaphosphorinan-2-one (1c) afforded meso and racemic
Received: April 8, 2011
Published: June 14, 2011
r
2011 American Chemical Society
6088
dx.doi.org/10.1021/jo2007384 J. Org. Chem. 2011, 76, 6088–6094
|

The Journal of Organic Chemistry
ARTICLE
Table 1. Reaction Conditions, Product Yields, and Isomeric Ratios of Dimers 2aꢀd along with Recovered C60
substrate 1
molar ratio of C60/1/Mn(III)
product
temp (°C)a
reation time (min)
yield of 2 (%)
isometric ratiob
recovered C60 (%)
HP(O)(OMe)2
HP(O)(OEt)2
1:2:2
1:2:2
1:2:4
2a
2b
2c
135
135
100
50
50
90
29
28
34
1.4:1
1.4:1
1.5:1
67
68
65
HP(O)Ph2
1:2:6c
2d
60
20
29
1.2:1
68
a Oil bath temperature. b Major/minor ratio determined by 31P MR spectrum. c 1 equiv of DMAP was added.
Table 2. Reaction Conditions and Product Yields of Hydrofullerenes 3 along with Recovered C60
substrate 1
molar ratio of C60/1/Mn(III)
product
temp (°C)a
reaction time (h)
yield of 3 (%)
recovered C60 (%)
HP(O)(OMe)2
HP(O)(OEt)2
1:10:2
1:10:2
1:5:1
3a
3b
3c
135
135
100
12
26
11
32
41
31
62
37
42
HP(O)Ph2
1:5:1
3d
60
0.7
32
40
a Oil bath temperature.
isomers of singly bonded fullerene dimers 2aꢀc.8 When the
reaction was extended to diphenylphosphine oxide (1d), dimer
2d could be isolated in 29% yield when the molar ratio of C60:1d:
consistent with its dimeric structure. One doublet at 67.45 ppm
1
with JCꢀP = 61.9 Hz and another doublet at 67.51 ppm with
1JCꢀP = 61.8 Hz for the fullerenyl sp3-carbon connecting with the
1
Mn(OAc)3 2H2O = 1:2:6 was employed in the presence of 1 equiv
phosphorus atom were seen. It is noteworthy that the JCꢀP
3
of 4-(dimethylamino)pyridine (DMAP). Table 1 lists the reaction
conditions, product yield, and isomeric ratio for the Mn(OAc)3-
mediated reaction of C60 with 1d together with those with 1aꢀc.
The isolated yield of dimer 2d was comparable to those (28ꢀ
34%) of 2aꢀc. Diphenylphosphine oxide 1d was more reactive
toward C60 than phosphonate esters 1aꢀc, and its reaction with
C60 could proceed at 60 °C, much lower than the reaction tem-
peratures for 1a,b. Nevertheless, a larger amount of Mn(OAc)3
(6 equiv) and 1 equiv of additional DMAP were required to
suppress the formation of another product 3d (vide infra).
Singly bonded dimer 2d was characterized by its HRMS, 1H
NMR, UVꢀvis, and FT-IR spectra. The HRMS (ꢀESI) of 2d
displayed a peak at 921.0478, corresponding to its half molecular
ion. Similar phenomenon was described for 2aꢀc8 as well as
values (61.8ꢀ61.9 Hz) for the fullerenyl sp3-C of 2d is much
smaller than those (144.6ꢀ146.9 Hz)8 of 2aꢀc. Nakamura and co-
workers also recorded the dramatic effect of the substituents
1
attached to the phosphorus atom on the JCꢀP values.3a,b The
multiple peaks at 68.09ꢀ68.24 ppm were assigned to the pivot
carbons joining the two fullerene skeletons. The pivot carbons
were shown as multiple peaks because of the coupling with the
phosphorus atoms from two dimeric isomers. The 31P NMR
spectrum of 2d with 85% H3PO4 as the reference displayed two
signals at 29.07 and 29.70 ppm in a ratio of 1.2:1. The observed
broad absorption at 444 nm in the UVꢀvis spectrum of 2d
suggested that it adopted the 1,4-addition pattern.7a,8ꢀ10
It was interesting to find that when excess phosphorus com-
pound was used, another type of products, i.e., hydrophosphory-
lated fullerenes 3aꢀd, were selectively obtained from the Mn-
(OAc)3-mediated reaction of C60 with 1aꢀd. The reaction
conditions and isolated yields of hydrofullerenes 3aꢀd along
with recovered C60 from the Mn(OAc)3-mediated reaction of
C60 with 1aꢀd are summarized in Table 2.
1
other singly bonded dimers.7a The H NMR spectrum of 2d
showed two sets of signals for the aromatic protons, agreeing
with the existence of two isomers of 2d. The 13C NMR spectrum
of 2d exhibited 71 peaks in the 137ꢀ157 ppm range for the sp2-
carbons of the C60 skeleton, indicating that it had two C60 moieties,
6089
dx.doi.org/10.1021/jo2007384 |J. Org. Chem. 2011, 76, 6088–6094

The Journal of Organic Chemistry
ARTICLE
Table 3. Reaction Conditions and Product Yields of Acetoxylated Fullerenes 4 along with Recovered C60
substrate 1
molar ratio of C60/1/Mn(III)
product
temp (°C)a
reaction time (min)
yield of 4 (%)
recovered C60 (%)
HP(O)(OMe)2
HP(O)(OEt)2
1:2:10
1:2:10
1:2:10
4a
4b
4c
135
135
135
50
50
50
32
27
24
53
58
71
HP(O)Ph2
1:2:10b
4d
70
60
31
66
a Oil bath temperature. b 20 equiv of AcOH was added.
Hydrofullerenes 3aꢀd were formed exclusively from the
reaction with 1d, the reaction temperature could be lowered to
70 °C, yet 20 equiv of acetic acid was added to facilitate the
acetoxylation process and suppress the formation of dimer 1d.
The structures of acetoxylated fullerenes 4aꢀd were identified
by the HRMS, 1H NMR, 13C NMR, 31P NMR, UVꢀvis, and FT-
IR spectra. The 1H NMR spectra of products 4aꢀd exhibited a
singlet at 2.22ꢀ2.52 ppm for the acetoxy group besides the peaks
due to the phosphonate or phosphine oxide moieties. The 13C
NMR spectra of 4aꢀd displayed at least 52 peaks in the 137ꢀ
151 ppm range owing to the 58 sp2-carbons of the C60 skeleton
and two peaks at 60.10ꢀ66.64 ppm (1JCꢀP = 142.3ꢀ145.0 Hz
4.6ꢀ5.0 Hz for 4aꢀc and 3.7 Hz for 4d) for the two sp3-carbons
of the C60 cage, consistent with its C1 molecular symmetry. The
observed smaller coupling constants (JCꢀP) for both sp2- and
sp3-carbons of 4d bearing P(O)Ph2 had precedent in the lit-
erature.3a,b The chemical shift (ca. 78.5 ppm) of the fullerenyl
sp3-carbon connected to the acetoxy group in 4aꢀd was very
close to that of similar 1,4-adducts reported previously by us.7d,10
In addition, the 13C NMR and FT-IR spectra of 4aꢀd showed
peaks at 169.55ꢀ170.06 ppm and 1745ꢀ1752 cmꢀ1, further con-
firming the presence of an ester group.
reaction of C60 with 1aꢀd and Mn(OAc)3 2H2O in a molar
3
ratio of 1:10:2 or 1:5:1. Compounds 1c and 1d showed higher
reactivity toward C60, and the reaction could take place at lower
reaction temperature and less amount of phosphorus compound
and Mn(OAc)3, yet gave comparable product yields at less reaction
time. No formation of hydrofullerenes 3aꢀc and only ∼2% yield
of 3d was observed for the reaction of C60 with 1aꢀd in the
absence of Mn(OAc)3 under otherwise the same conditions. There-
fore, Mn(OAc)3 was essential for the successful formation of
hydrofullerenes 3aꢀd from the reaction of C60 with 1aꢀd in the
absence of DMSO, DMF, or HMPA, which may promote a
charge transfer or an electron transfer between C60 and the phos-
phorus reagent.3b Although Nakamura and co-workers developed
an elegant procedure for the synthesis of hydrophosphorylated
fullerenes,3b our protocol using less amount of phosphonate esters
offers an alternative access to these hydrofullerenes, which are
expected to excellent precursors for further modifications (vide
infra).7d
for 4aꢀc and 58.3 Hz for 4d) and 78.20ꢀ78.58 ppm (4JCꢀP
=
Products 3a,3a,b 3c,3b and 3d3a,b are known compounds, and
their identities were confirmed by comparison of their spectral
data with the reported ones. Product 3b is fully characterized by
the HRMS, 1H NMR, 13C NMR, 31P NMR, UVꢀvis, and FT-IR
spectra. In the 1H NMR spectrum of 3b, the proton attached to
the fullerene skeleton appeared at 7.22 ppm as a doublet with a
large PꢀH coupling (3JPꢀH = 29.6 Hz); the ethoxy group showed a
multiplet for the CH2 moiety due to the coupling with both the
methyl group and the phosphorus atom in the region of 4.66ꢀ
Further investigations revealed that both singly bonded di-
mers 2aꢀd and hydrofullerenes 3aꢀd could be converted
efficiently to acetoxylated fullerenes 4aꢀd, and hydrofullerenes
3aꢀd could be transformed to singly bonded dimers 2aꢀd. The
reaction conditions and product yields are summarized in
Tables 4ꢀ6.
4.76 ppm and a triplet for the CH3 group at 1.66 ppm. In its 31
P
It should be noted that the addition of acetic acid was required to
increase the yield of acetoxylated product 4d starting from both
dimer 2d and hydrofullerene 3d, yet some amount of C60 was
formed in these cases (Tables 4 and 5). In contrast, DMAP was
added in the conversion of hydrofullerene 3d to dimer 2d to achieve
higher product yield, while C60 (13%) was also isolated (Table 6).
Inert atmosphere was necessary to get the clean products for
the reaction of C60 with 1, especially 1d. We propose a possible
mechanism for the reaction of C60 with 1aꢀd giving three
different types of products (Scheme 1) based on the previous
literature.7,8,11
NMR spectrum, only one peak at 21.55 ppm was observed. In the
13C NMR spectrum, there were 27 peaks, of which 3 peaks were
overlapped and 5 peaks showed PꢀC couplings, in the 135ꢀ
152 ppm range for the 58 sp2-carbons of the C60 cage, and one
doublet at 65.48 ppm (1JPꢀC = 149.9 Hz) and a singlet at 56.30 ppm
for the two sp3-carbons of the fullerene moiety, fully consistent
with its Cs molecular symmetry.
Gratifyingly, when excess Mn(OAc)3 was employed, acetoxy-
lated fullerenes 4aꢀd were selectively obtained from the Mn-
(OAc)3-mediated reaction of C60 with 1aꢀd. The reaction
conditions and isolated yields of acetoxylated fullerenes 4aꢀd
along with recovered C60 from the Mn(OAc)3-mediated reaction
of C60 with 1aꢀd are listed in Table 3.
Phosphorus radical 5, generated from a phosphonate ester or
phosphine oxide by Mn(OAc)3, addstoC60 to give fullerenyl radical
6, which is in equilibrium with fullerenyl radical 7. Homocoupling of
radical 7 affords singly bonded dimer 2, while hydrogen abstraction
of radical 6 produces hydrophosphorylated fullerene 3. Reaction of
radical 7 with excess Mn(OAc)3 furnishes acetoxylated fullerene 4.
The synthesis of acetoxylated fullerenes 4aꢀd was achieved in
24ꢀ32% isolated yields by performing the reaction of C60 with
1aꢀd and Mn(OAc)3 2H2O in a molar ratio of 1:2:10. For the
3
6090
dx.doi.org/10.1021/jo2007384 |J. Org. Chem. 2011, 76, 6088–6094

The Journal of Organic Chemistry
ARTICLE
Table 4. Reaction Conditions and Product Yields for the Conversion of Dimers 2 to Acetoxylated Fullerenes 4
substrate 2
molar ratio of 2/Mn(III)
product
temp (°C)a
reaction time (min)
yield of 4 (%)
2a
2b
2c
2d
1:20
1:20
1:20
1:20b
4a
4b
4c
4d
135
135
135
70
20
30
20
30
67
68
73
56c
a Oil bath temperature. b 40 equiv of AcOH was added. c 22% of C60 was also obtained.
Table 5. Reaction Conditions and Product Yields for the Conversion of Hydrofullerenes 3 to Acetoxylated Fullerenes 4
substrate 3
molar ratio of 3/Mn(III)
product
temp (°C)a
reaction time (min)
yield of 4 (%)
3a
3b
3c
3d
1:10
1:10
1:10
1:10b
4a
4b
4c
4d
135
135
135
70
20
25
20
60
70
76
78
32c
a Oil bath temperature. b 20 equiv of AcOH was added. c 39% of C60 was also obtained, and 22% of 3d was recovered.
Table 6. Reaction Conditions and Product Yields for the Conversion of Hydrofullerenes 3 to Dimers 2
substrate 3
molar ratio 3/Mn(III)
product
temp (°C)a
reaction time (min)
yield of 2 (%)
3a
3b
3c
3d
1:1.5
1:1.5
1:1.5
1:1.5b
2a
2b
2c
2d
135
135
100
60
20
30
20
5
88
89
90
84c
a Oil bath temperature. b 1 equiv of DMAP was added. c 13% of C60 was also obtained.
The postulated mechanisms shown in Scheme 1 can also explain
all phenomena observed in the aforementioned experiments.
Thin-layer chromatography (TLC) monitoring indicated that
dimer 2 was formed and eventually disappeared in the reaction of
C60 with 1aꢀd, affording both hydrofullerenes 3a,b and acet-
oxylated fullerenes 4aꢀd. As shown in Scheme 1, both hydro-
fullerene 3 and acetoxylated fullerene 4 are formed via fullerenyl
radicals 6 and 7, which are in equilibrium with dimer 2. The
existence of the equilibrium was evident from the detection of
the ESR signal of the dissociated monomer radical C60P(O)R2
3
by heating dimer 2.8 As the reaction proceeds to completion,
consumption of fullerenyl radicals 6 and 7 will result in the
disappearance of dimer 2. The failure to observe dimer 2c and 2d
in the formation of hydrofullerenes 3c and 3d probably arose
6091
dx.doi.org/10.1021/jo2007384 |J. Org. Chem. 2011, 76, 6088–6094

The Journal of Organic Chemistry
ARTICLE
Scheme 1. Proposed Possible Reaction Mechanism
from the faster hydrogen abstraction of 1c and 1d by fullerenyl
radical 6 due to the weaker PꢀH bond in the substrates. The
transformations from 3 to 2, 3 to 4, and 2 to 4 can also be
elucidated by the proposed mechanism. Hydrofullerene 3 contains
a weak acidic C60ꢀH bond; thus, fullerenyl radical 6 could be
formed from hydrofullerene 3 in the presence of Mn(OAc)3. The
subsequent processes are the same as the pathways leading to
2 and 4 from radical 7. Another evidence for the proposed
mechanism came from the fact that dimer 2 was also formed and
then consumed during the conversion of 3aꢀd to 4aꢀd.
’ EXPERIMENTAL SECTION
Mn(OAc)3-Mediated Reaction of C60 with 1d Affording
Dimer 2d. A mixture of C60 (36.0 mg, 0.05 mmol), 1d (20.2 mg,
0.10 mmol), Mn(OAc)3 2H2O (80.4 mg, 0.30 mmol), and 4-
3
(dimethylamino)pyridine (6.1 mg, 0.05 mmol) was dissolved in
chlorobenzene (10 mL) and stirred at 60 °C (oil bath temperature)
under argon atmosphere for 20 min. The reaction mixture was
poured onto a silica gel column and eluted with CS2 to afford
recovered C60 (24.4 mg, 68%) and then with CS2/CH2Cl2 (10:1)
to give dimer 2d (13.2 mg, 29%).
The addition of DMAP increased the product yield of 2d
(Table 1) and also helped consume the most facilely generated
byproduct 3d from the Mn(OAc)3-mediated reaction of C60
with 1d because DMAP could deprotonate 3d to give anion 3dꢀ,
which was oxidized to radical 6d, and then coupled to afford
dimer 2d via radical 7d. The required addition of DMAP for the
Mn(OAc)3-mediated conversion of 3d to 2d (Table 6) can be
explained in the same way. Acetic acid was demanded for the
formation of acetoxylated fullerene 4d in the Mn(OAc)3-pro-
moted reaction of C60 with 1d (Table 3) probably due to the
reason that acetic acid could facilitate the acetoxylation of radical
7d and thus diminished the formation of dimer 2d. The necessary
addition of AcOH in the conversion from 2d to 4d (Table 4) and
from 3d to 4d (Table 5) could be interpreted similarly. In
addition, some amount of C60 was generated in these conversions
probably because radical 7d tended to decompose more rapidly
to give C60 than 7aꢀc.
Spectral Data of 2d. 1H NMR (300 MHz, CS2ꢀCDCl3) δ 7.16ꢀ
7.62 (m, 12H, meso and racemic), 8.27ꢀ8.37 (m, 8H ꢁ 0.49, minor),
8.50ꢀ8.61 (m, 8H ꢁ 0.51, major); 13C NMR (100.6 MHz, CS2ꢀCDCl3
with Cr(acac)3 as relaxation reagent) δ 67.45 (d, JPꢀC = 61.9 Hz, sp3-C
of C60, meso or racemic), 67.51 (d, JPꢀC = 61.8 Hz, sp3-C of C60, meso
or racemic), 68.09ꢀ68.24 (sp3-C of C60, meso and racemic), 128.03 (d,
JPꢀC = 11.5 Hz, o-CH, meso or racemic), 128.04 (d, JPꢀC = 11.4 Hz,
o-CH, meso or racemic), 128.09 (d, JPꢀC = 11.6 Hz, o-CH, meso and
racemic), 129.29 (d, JPꢀC = 99.3 Hz, PC, meso and racemic), 132.17 (d,
JPꢀC = 2.1 Hz, p-CH, meso or racemic), 132.25 (d, JPꢀC = 2.2 Hz, p-CH,
meso or racemic), 132.51 (p-CH, meso and racemic), 133.62 (d, JPꢀC
=
7.5 Hz, m-CH, meso or racemic), 133.82 (d, JPꢀC = 8.8 Hz, m-CH, meso
and racemic), 134.00 (d, JPꢀC = 8.4 Hz, m-CH, meso or racemic), 137.76
(d, JPꢀC = 4.3 Hz), 137.79 (d, JPꢀC = 4.6 Hz), 138.09, 138.18 (d, JPꢀC
=
3.1 Hz), 138.27, 138.52, 138.88, 139.14 (d, JPꢀC = 4.3 Hz), 139.29 (d,
JPꢀC = 4.1 Hz), 140.62, 140.69, 141.41, 141.58, 141.67, 141.69, 142.33,
142.37, 142.41, 142.51, 142.56, 142.66, 142.82, 142.90, 142.99, 143.01,
143.08, 143.12, 143.33, 143.36, 143.71, 143.87, 143.92, 143.94, 144.16,
144.21, 144.31, 144.35, 144.41, 144.58, 144.61, 145.06, 145.08, 145.11,
145.19, 145.42, 145.46, 145.52, 146.65, 146.69, 146.72, 146.93, 147.24,
147.36, 147.40, 147.44, 147.80 (d, JPꢀC = 4.9 Hz), 148.43, 148.50, 148.60,
148.64, 149.11 (d, JPꢀC = 5.6 Hz), 149.30, 149.51 (d, JPꢀC = 6.8 Hz),
150.09, 150.21, 150.40, 150.93, 152.42, 152.52, 156.51, 156.89; 31P NMR
(162 MHz, CDCl3) δ 29.07 (major), 29.70 (minor); FT-IR ν/cmꢀ1
(KBr) 2920, 1461, 1435, 1210, 1190, 1115, 1098, 876, 843, 818, 747,
723, 697, 613, 558, 526; UVꢀvis λmax/nm (log ε) (CHCl3) 258 (5.28),
324 (4.83), 444 (4.08); HRMS (ꢀESI): calcd for C72H10PO [M/2]ꢀ
921.0469, found 921.0478.
’ CONCLUSION
We have successfully realized the reaction of C60 with phos-
phorus-centered radicals, generated from phosphonate esters
or phosphine oxide by Mn(OAc)3. Depending on the reaction
conditions, three different types of phosphorylated fullerene
derivatives, that is, singly bonded fullerene dimers 2, hydro-
fullerenes 3, and acetoxylated fullerenes 4, could be selectively
obtained. Interconversions among the three types of phos-
phorylated fullerene derivatives were also achieved. It was
found that both singly bonded dimers 2 and hydrofullerenes 3
could be converted efficiently to acetoxylated fullerenes 4, and
hydrofullerenes 3 could be transformed to singly bonded
dimers 2. A possible reaction mechanism was proposed to
explain the formation and interconversion of the fullerene
products.
General Procedure for the Mn(OAc)3-Mediated Reaction
of C60 with 1aꢀd Affording Hydrofullerenes 3aꢀd. A mixture
of C60 (36.0 mg, 0.05 mmol), 1 (0.50 mmol for 1a and 1b; 0.25 mmol for
1c and 1d), and Mn(OAc)3 2H2O (0.10 mmol for 1a and 1b; 0.05
3
mmol for 1c and 1d) was dissolved in chlorobenzene (10 mL) and
stirred at a given temperature for the desired time under argon atmo-
sphere (monitored by TLC). The reaction mixture was poured onto a
6092
dx.doi.org/10.1021/jo2007384 |J. Org. Chem. 2011, 76, 6088–6094

The Journal of Organic Chemistry
ARTICLE
silica gel column and eluted with CS2 to afford recovered C60 and then
with CS2/AcOEt (10:1) to give hydrofullerenes 3aꢀd.
144.58 (d, JPꢀC = 5.7 Hz), 144.65, 144.81 (d, JPꢀC = 1.9 Hz), 145.03,
145.34, 145.37 (d, JPꢀC = 4.6 Hz), 145.46, 145.56, 145.87, 146.49,
146.56, 146.71, 146.88, 146.92, 146.95, 147.18, 147.28 (d, JPꢀC = 10.0
Hz), 147.53 (d, JPꢀC = 1.7 Hz), 147.82 (d, JPꢀC = 9.5 Hz), 148.12 (2C),
148.18 (d, JPꢀC = 2.4 Hz), 148.76, 149.57 (d, JPꢀC = 11.7 Hz), 170.06
(CdO); 31P NMR (162 MHz, CDCl3) δ 13.80; FT-IR ν/cmꢀ1 (KBr)
2976, 1748, 1429, 1389, 1364, 1258, 1223, 1160, 1126, 1042, 1015, 990,
763, 670, 639, 624, 561, 524; UVꢀvis λmax/nm (log ε) (CHCl3) 259
(5.01), 320 (4.51), 445 (3.78); HRMS (ꢀESI): calcd for C66H13PO5
[M]ꢀ 916.0501, found 916.0485.
Spectral Data of 3b. 1H NMR (300 MHz, CDCl3) δ 1.66 (t, J =
7.0 Hz, 6H), 4.66ꢀ4.76 (m, 4H), 7.22 (d, 3JPꢀC = 29.6 Hz, 1H); 13
C
NMR (75.5 MHz, CS2ꢀCDCl3 with Cr(acac)3 as relaxation reagent)
(all 2C unless indicated) δ 16.69 (d, 3JPꢀC = 5.5 Hz, CH3), 56.30 (1C,
sp3-C of C60), 64.72 (d, 2JPꢀC = 7.3 Hz, OCH2), 65.48 (1C, d, 1JPꢀC
=
149.9 Hz, sp3-C of C60), 135.20 (d, JPꢀC = 1.8 Hz), 135.96 (d, JPꢀC = 5.8
Hz), 140.00, 140.23, 141.24 (4C), 141.42, 141.67, 141.72 (4C), 142.28,
142.36, 142.95, 144.11, 144.37, 145.10, 145.18, 145.25 (4C), 145.56,
145.96, 146.03, 146.10, 146.17, 146.68, 146.94 (d, JPꢀC = 5.8 Hz),
Spectral Data of 4c. 1H NMR (300 MHz, CDCl3) δ 1.23 (s, 3H,
CCH3), 1.35 (s, 3H, CCH3), 2.50 (s, 3H, CH3), 4.39ꢀ4.56 (m, 4H,
CH2); 13C NMR (75.5 MHz, CS2ꢀCDCl3 with Cr(acac)3 as relaxation
reagent) (all 1C unless indicated) δ 21.42 (OOCCH3), 21.47 (CCH3),
21.59 (CCH3), 32.96 (d, JPꢀC = 7.8 Hz, CCH3), 60.37 (d, JPꢀC = 142.3
Hz, sp3-C of C60), 78.20 (d, JPꢀC = 7.6 Hz, CH2), 78.41 (d, JPꢀC = 7.6
Hz, CH2), 78.49 (d, JPꢀC = 5.0 Hz, sp3-C of C60), 137.63 (d, JPꢀC = 3.4
Hz), 138.86 (d, JPꢀC = 11.8 Hz), 139.21 (d, JPꢀC = 5.5 Hz), 140.20 (d,
JPꢀC = 3.2 Hz), 140.40, 141.27, 141.35 (d, JPꢀC = 1.4 Hz), 141.75,
142.01, 142.08, 142.13, 142.68, 142.72, 142.75, 142.76 (d, JPꢀC = 1.5 Hz),
142.81 (d, JPꢀC = 1.6 Hz), 142.86, 143.13, 143.20 (d, JPꢀC = 4.8 Hz),
143.21, 143.28, 143.30, 143.43, 143.62, 143.78 (2C), 143.80, 143.83 (d,
JPꢀC = 2.2 Hz), 143.94, 144.00 (d, JPꢀC = 1.5 Hz), 144.03, 144.20,
147.27 (d, JPꢀC = 1.5 Hz), 148.81 (1C), 148.96 (1C), 151.58 (d, JPꢀC
=
6.2 Hz); 31P NMR (121.5 MHz, CDCl3) δ 21.44; FT-IR ν/cmꢀ1 (KBr)
2975, 2901, 1512, 1462, 1428, 1389, 1365, 1254, 1183, 1160, 1095, 1046,
1019, 974, 960, 802, 763, 744, 702, 669, 632, 582, 572, 563, 548, 526;
UVꢀvis λmax/nm (log ε) (CHCl3) 256 (5.11), 308 (4.63), 324 (4.63),
434 (3.66), 637 (2.72), 705 (2.72); HRMS (ꢀESI): calcd for C64H10PO3
[M-H]ꢀ 857.0368, found 857.0382.
General Procedure for the Mn(OAc)3-Mediated Reaction
of C60 with 1aꢀd Affording Acetoxylated Fullerenes 4aꢀd. A
mixture of C60 (36.0 mg, 0.05 mmol), 1 (0.10 mmol), Mn(OAc)3 2H2O
3
(134.5 mg, 0.50 mmol), and AcOH (57 μL, 1.0 mmol in the case of 1d)
was dissolved in chlorobenzene (10 mL) and stirred at a given tem-
perature for the desired time under argon atmosphere (monitored by
TLC). The reaction mixture was poured onto a silica gel column and
eluted with CS2 to afford recovered C60 and then with CS2/AcOEt
(10:1) to give acetoxylated fullerenes 4.
144.58 (d, JPꢀC = 5.7 Hz), 144.63 (2C), 144.80, 145.02 (d, JPꢀC
=
2.1 Hz), 145.11, 145.43 (d, JPꢀC = 5.6 Hz), 145.47, 145.58, 145.70,
146.00, 146.36 (d, JPꢀC = 12.7 Hz), 146.62, 146.66, 146.82, 147.00,
147.03 (d, JPꢀC = 9.2 Hz), 147.04, 147.09, 147.20 (d, JPꢀC = 2.5 Hz),
Spectral Data of 4a. 1H NMR (300 MHz, CDCl3) δ 2.52 (s, 3H,
CH3), 4.22 (d, 3H, JPꢀH = 11.1 Hz, OCH3), 4.24 (d, 3H, JPꢀH = 11.1 Hz,
OCH3); 13C NMR (75.5 MHz, CS2ꢀCDCl3 with Cr(acac)3 as relaxa-
147.47 (d, JPꢀC = 1.9 Hz), 148.10 (d, JPꢀC = 1.6 Hz), 148.23 (d, JPꢀC
=
P
2.6 Hz), 148.27, 148.76 (d, JPꢀC = 11.9 Hz), 148.88, 170.03 (CdO); 31
NMR (162 MHz, CDCl3) δ 5.47; FT-IR ν/cmꢀ1 (KBr) 2920, 1752,
1512, 1463, 1429, 1366, 1273, 1221, 1061, 1003, 834, 767, 644, 622, 559,
524; UVꢀvis λmax (log ε) (CHCl3) 257 (4.97), 321 (4.48), 446 (3.75);
HRMS (ꢀESI): calcd for C67H13PO5 [M]ꢀ 928.0501, found 928.0522.
Spectral Data of 4d. 1H NMR (400 MHz, CDCl3) δ 2.22 (s, 3H,
CH3) 7.62ꢀ7.74 (m, 6H), 8.37ꢀ8.43 (m, 4H); 13C NMR (75.5 MHz,
CS2ꢀCDCl3 with Cr(acac)3 as relaxation reagent) (all 1C unless in-
dicated) δ 21.52 (CH3), 66.64 (d, JPꢀC = 58.3 Hz, sp3-C of C60), 78.55
(d, JPꢀC = 3.7 Hz, sp3-C of C60), 128.51 (2C, d, JPꢀC = 12.1 Hz, o-CH),
128.56 (2C, d, JPꢀC = 12.1 Hz, o-CH), 128.62 (d, JPꢀC = 99.4 Hz, PC),
128.99 (d, JPꢀC = 99.0 Hz, PC), 132.89 (2C, d, JPꢀC = 8.9 Hz, m-CH),
132.93 (d, JPꢀC = 2.2 Hz, p-CH), 132.96 (2C, d, JPꢀC = 8.7 Hz, m-CH),
132.97 (d, JPꢀC = 3.6 Hz, p-CH), 137.53 (d, JPꢀC = 2.5 Hz), 139.13 (d,
JPꢀC = 4.4 Hz), 139.54 (d, JPꢀC = 9.5 Hz), 140.20 (d, JPꢀC = 2.8 Hz),
tion reagent) (all 1C unless indicated) δ 21.19 (CH3), 54.86 (d, JPꢀC
7.4 Hz, OCH3), 55.11 (d, JPꢀC = 7.4 Hz, OCH3), 60.10 (d, JPꢀC
=
=
145.0 Hz, sp3-C of C60), 78.27 (d, JPꢀC = 4.6 Hz, sp3-C of C60), 137.35
(d, JPꢀC = 3.2 Hz), 138.38 (d, JPꢀC = 11.5 Hz), 138.73 (d, JPꢀC
5.4 Hz), 139.69 (d, JPꢀC = 3.0 Hz), 140.08, 140.93, 141.08, 141.49,
141.74, 141.79, 141.86, 142.40 (2C), 142.42, 142.45, 142.53 (d, JPꢀC
=
=
1.4 Hz), 142.57, 142.83, 142.91 (d, JPꢀC = 1.4 Hz), 142.97, 142.99,
143.03 (d, JPꢀC = 2.9 Hz), 143.12, 143.30, 143.47, 143.49, 143.51,
143.55 (d, JPꢀC = 2.2 Hz), 143.64, 143.71, 143.72 (d, JPꢀC = 1.6 Hz),
143.92, 144.31 (d, JPꢀC = 5.0 Hz), 144.32, 144.34, 144.49, 144.69 (d,
JPꢀC = 1.8 Hz), 144.82, 145.14 (d, JPꢀC = 4.7 Hz), 145.17, 145.28,
145.39, 145.69, 146.31, 146.37, 146.52, 146.61 (d, JPꢀC = 12.8 Hz),
146.69, 146.73, 146.78, 146.91 (d, JPꢀC = 2.4 Hz), 147.16 (d, JPꢀC
7.8 Hz), 147.24, 147.89 (d, JPꢀC = 1.4 Hz), 147.96, 147.98 (d, JPꢀC
=
=
140.39, 140.90, 140.94, 141.81, 142.05, 142.08, 142.11, 142.54 (d, JPꢀC
=
3.1 Hz), 148.57, 148.82 (d, JPꢀC = 11.7 Hz), 169.55 (CdO); 31P NMR
(121.5 MHz, CDCl3) δ 16.23; FT-IR ν/cmꢀ1 (KBr) 2947, 2847, 1749,
1429, 1363, 1263, 1221, 1187, 1046, 1027, 991, 841, 765, 670, 638, 624,
561, 525; UVꢀvis λmax/ nm (log ε) (CHCl3) 256 (4.94), 320 (4.46),
445 (3.73); HRMS (ꢀESI): calcd for C64H9PO5 [M]ꢀ 888.0188, found
888.0203.
1.4 Hz), 142.68, 142.69, 142.72, 142.73, 142.85, 143.05, 143.14, 143.27,
143.33, 143.38, 143.48, 143.61, 143.72, 143.76, 143.77 (d, JPꢀC = 1.9 Hz),
143.85, 143.89, 143.95 (d, JPꢀC = 2.4 Hz), 144.09 (d, JPꢀC = 1.2 Hz),
144.21, 144.39, 144.64, 144.71, 144.75 (d, JPꢀC = 1.8 Hz), 145.07 (d,
JPꢀC = 4.4 Hz), 145.19, 145.36 (d, JPꢀC = 4.1 Hz), 145.45, 145.57,
145.64, 145.84, 146.61, 146.63, 146.82, 146.94, 146.98, 147.03, 147.37
Spectral Data of 4b. 1H NMR (300 MHz, CDCl3) δ 1.56 (t, J =
6.6 Hz, 3H, OCH2CH3), 1.59 (t, J = 6.9 Hz, 3H, CH2CH3), 2.51 (s,
3H, CH3), 4.51ꢀ4.68 (m, 4H, OCH2CH3); 13C NMR (75.5 MHz,
CS2ꢀCDCl3 with Cr(acac)3 as relaxation reagent) (all 1C unless in-
dicated) δ 16.63 (d, JPꢀC = 5.8 Hz, OCH2CH3), 16.71 (d, JPꢀC = 5.4 Hz,
OCH2CH3), 21.42 (CH3), 60.85 (d, JPꢀC = 143.9 Hz, sp3-C of C60),
64.71 (d, JPꢀC = 7.0 Hz, OCH2CH3), 64.80 (d, JPꢀC = 6.9 Hz,
(d, JPꢀC = 5.8 Hz), 147.43, 148.04 (d, JPꢀC = 1.5 Hz), 148.21 (d, JPꢀC
=
5.7 Hz), 148.22, 148.44 (d, JPꢀC = 1.9 Hz), 148.47, 148.82, 150.05 (d,
JPꢀC = 7.2 Hz), 170.40 (CdO); 31P NMR (162 MHz, CDCl3) δ 27.08;
FT-IR ν/cmꢀ1 (KBr) 1745, 1435, 1362, 1222, 1205, 1184, 1114, 1097,
1012, 989, 725, 696, 614, 577, 560, 549, 526; UVꢀvis λmax/nm (log ε)
(CHCl3) 259 (5.10), 326 (4.60), 447 (3.82); HRMS (ꢀESI): calcd for
C74H13PO3 [M]ꢀ 980.0602, found 980.0591.
OCH2CH3), 78.58 (d, JPꢀC = 4.7 Hz, sp3-C of C60), 137.47 (d, JPꢀC
=
General Procedure for the Converson of Dimer 2aꢀd to
3.2 Hz), 138.50 (d, JPꢀC = 11.6 Hz), 138.73 (d, JPꢀC = 5.3 Hz), 139.92
(d, JPꢀC = 3.0 Hz), 140.26, 141.03, 141.22, 141.71, 141.95, 141.98,
142.07, 142.59 (3C), 142.61 (d, JPꢀC = 1.6 Hz), 142.70 (d, JPꢀC = 1.4
Hz), 142.75, 143.04 (d, JPꢀC = 1.6 Hz), 143.06, 143.16 (2C), 143.31,
Acetoxylated Fullerenes 4aꢀd. A mixture of 2 (6.5ꢀ7.5 μmol),
Mn(OAc)3 2H2O (20 equiv), and AcOH (40 equiv in the case of 1d)
3
was dissolved in chlorobenzene (3 mL) and stirred at a given tempera-
ture for the desired time under argon atmosphere (monitored by TLC).
The reaction mixture was poured onto a silica gel column and eluted
with CS2/AcOEt (10:1) to give acetoxylated fullerenes 4.
143.35 (d, JPꢀC = 3.1 Hz), 143.48, 143.66, 143.70 (2C), 143.81 (d, JPꢀC
=
1.8 Hz), 143.87 (2C), 143.95 (d, JPꢀC = 1.5 Hz), 144.11, 144.50, 144.54,
6093
dx.doi.org/10.1021/jo2007384 |J. Org. Chem. 2011, 76, 6088–6094

The Journal of Organic Chemistry
ARTICLE
General Procedure for the Converson of Hydrofullerenes
3aꢀd to Acetoxylated Fullerenes 4aꢀd. A mixture of 3 (12ꢀ16
(9) (a) Yoshida, M.; Sultana, F.; Uchiyama, N.; Yamada, T.; Iyoda,
M. Tetrahedron Lett. 1999, 40, 735. (b) Cheng, F.; Murata, Y.; Komatsu,
K. Org. Lett. 2002, 4, 2541.
(10) Wang, G.-W.; Lu, Y.-M.; Chen, Z.-X. Org. Lett. 2009, 11, 1507.
(11) (a) Kagayama, T.; Nakano, A.; Sakaguchi, S.; Ishii, Y. Org. Lett.
2006, 8, 407. (b) Mu, X.-J.; Zou, J.-P.; Qian, Q.-F.; Zhang, W. Org. Lett.
2006, 8, 5291. (c) Pan, X.-Q.; Zou, J.-P.; Zhang, G.-L.; Zhang, W. Chem.
Commun. 2010, 46, 1721.
μmol), Mn(OAc)3 2H2O (10 equiv), and AcOH (20 equiv in the case
3
of 1d) was dissolved in chlorobenzene (3 mL) and stirred at a given
temperature for the desired time under argon atmosphere (monitored
by TLC). The reaction mixture was poured onto a silica gel column and
eluted with CS2/AcOEt (10:1) to give acetoxylated fullerenes 4.
General Procedure for the Converson of Hydrofullerenes
3aꢀd to Dimers 2aꢀd. A mixture of 3 (12ꢀ15 μmol), Mn-
(OAc)3 2H2O (1.5 equiv), and 4-(dimethylamino)pyridine (1 equiv
3
in the case of 1d) was dissolved in chlorobenzene (3 mL) and stirred at a
given temperature for the desired time under argon atmosphere
(monitored by TLC). The reaction mixture was poured onto a silica
gel column and eluted with CS2/AcOEt (10:1) (CS2/CH2Cl2 = 10:1 in
the case of 1d) to get dimer 2.
’ ASSOCIATED CONTENT
Supporting Information. 1H NMR, 13C NMR, and 31
P
S
b
NMR spectra of 2d, 3b, and 4aꢀd. This material is available free
of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: gwang@ustc.edu.cn.
’ ACKNOWLEDGMENT
We are grateful for the financial support from National Natural
Science Foundation of China (No. 20972145, 91021004) and
National Basic Research Program of China (2011CB921402).
’ REFERENCES
(1) For selected reviews, see: (a) Taylor, R.; Walton, D. R. M. Nature
1993, 363, 685. (b) Hirsch, A. Synthesis 1995, 895. (c) Thilgen, C.;
Diederich, F. Chem. Rev. 2006, 106, 5049. (d) Matsuo, Y.; Nakamura, E.
Chem. Rev. 2008, 108, 3016.
(2) Wu, S.-H.; Sun, W.-Q.; Zhang, D.-W.; Shu, L.-H.; Wu, H.-M.;
Xu, J.-F.; Lao, X.-F. Tetrahedron Lett. 1998, 39, 9233.
(3) (a) Yamago, S.; Yanagawa, M.; Mukai, H.; Nakamura, E. Tetra-
hedron 1996, 52, 5091. (b) Isobe, H.; Chen, A.-J.; Solin, N.; Nakamura,
E. Org. Lett. 2005, 7, 5633. (c) Chen, S.-Y.; Cheng, R.-L.; Tseng, C.-K.;
Lin, Y.-S.; Lai, L.-H.; Venkatachalam, R. K.; Chen, Y.-C.; Cheng, C.-H.;
Chuang, S.-C. J. Org. Chem. 2009, 74, 4866.
(4) Cheng, F.; Yang, X.; Fan, C.; Zhu, H. Tetrahedron 2001,
57, 7331.
(5) (a) Cheng, F.; Yang, X.; Zhu, H.; Song, Y. Tetrahedron Lett. 2000,
41, 3947. (b) Liu, Z. B.; Tian, J. G.; Zang, W. P.; Zhou, W. Y.; Zhang,
C. P.; Zheng, J. Y.; Zhou, Y. C.; Zu, H. Appl. Opt. 2003, 42, 7072.
(6) For a review, see: Wang, G.-W.; Li, F.-B. J. Nanosci. Nanotechnol.
2007, 7, 1162.
(7) (a) Zhang, T.-H.; Lu, P.; Wang, F.; Wang, G.-W. Org. Biomol.
Chem. 2003, 1, 4403. (b) Wang, G.-W.; Zhang, T.-H.; Cheng, X.; Wang,
F. Org. Biomol. Chem. 2004, 2, 1160. (c) Wang, G.-W.; Li, F.-B. Org.
Biomol. Chem. 2005, 3, 794. (d) Chen, Z.-X.; Wang, G.-W. J. Org. Chem.
2005, 70, 2380. (e) Cheng, X.; Wang, G.-W.; Murata, Y.; Komatsu, K.
Chin. Chem. Lett. 2005, 16, 1327. (f) Wang, G.-W.; Yang, H.-T.; Miao,
C.-B.; Xu, Y.; Liu, F. Org. Biomol. Chem. 2006, 4, 2595. (g) Wang, G.-W.;
Li, F.-B.; Zhang, T.-H. Org. Lett. 2006, 8, 1355. (h) Li, F.-B.; Liu, T.-X.;
Huang, Y.-S.; Wang, G.-W. J. Org. Chem. 2009, 74, 7743. (i) Li, F.-B.;
Zhu, S.-E.; Wang, G.-W. Chin. Sci. Bull. 2010, 55, 2909.
(8) Wang, G.-W.; Wang, C.-Z.; Zhu, S.-E.; Murata, Y. Chem. Com-
mun. 2011, 47, 6111.
6094
dx.doi.org/10.1021/jo2007384 |J. Org. Chem. 2011, 76, 6088–6094