Hydroxyalkylation of [60]fullerene: Free radical addition of alcohols to C60

Article abstract of DOI:10.1039/c0cc02836c

A new general reaction for the preparation of isomerically pure hydroxyalkylated C₆₀ monoadducts, via the free radical photochemical addition of alcohols to [60]fullerene, is described.

Full text of DOI:10.1039/c0cc02836c

View Online / Journal Homepage / Table of Contents for this issue
This article is part of the
Carbon Nanostructures web
themed issue
This issue showcases high quality research in the field of nanotechnology,
specifically research that deals with carbon nanostructures
Guest Editors: Professors Nazario Martín, Dirk Guldi and Luis Echegoyen
Please visit the website to access the other articles in this issue:-
http://www.rsc.org/chemcomm/carbonnanostructures

COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Hydroxyalkylation of [60]fullerene: free radical addition
of alcohols to C60wz
Manolis D. Tzirakis, Mariza N. Alberti and Michael Orfanopoulos*
Received 27th July 2010, Accepted 9th September 2010
DOI: 10.1039/c0cc02836c
A new general reaction for the preparation of isomerically pure
hydroxyalkylated C60 monoadducts, via the free radical photo-
chemical addition of alcohols to [60]fullerene, is described.
monoadduct 1a was obtained in moderate yield after 30 min
of irradiation (Table 1, entry 1). Encouraged by this initial
result, we further explored the scope of this novel transformation
with a series of structurally diverse alcohols, including primary
alcohols (2–3), a secondary alcohol (4), propargyl alcohol (5),
and benzyl alcohol (6), under the same reaction conditions.
Indeed, these reactions proceeded rapidly to afford the corres-
ponding HAF monoadducts 2a–6a in good yields. At longer
reaction times an increase of the multi-addition byproducts
was observed. A summary of the reaction times and yields is
given in Table 1. The low yield observed for alcohol 6a is
essentially due to further side-reactions that produce C60H2
(at ca. 20% yield).8 The formation of this side-product (C60H2)
may be rationalized by the in situ conversion of 6a to C60H2,
via a base catalyzed elimination of benzaldehyde from 6a.9,10
In this case, C60ꢀ anion acts as a good leaving group by virtue
of its increased electrophilicity. At this point, it is worth
mentioning that C60H2 has been also detected, albeit in a
low yield (up to 5% based on HPLC analysis), during the
reaction of C60 with alcohols 1–5.
Free-radical reactions were one of the first investigated
reactions in fullerene chemistry,1 although the high affinity
of free-radicals with [60]fullerene (a ‘‘radical sponge’’) has
limited their utility as a practical tool for the selective
functionalization of C60. However, under appropriately controlled
conditions a single-addition of free-radicals to C60 can be
achieved. For example, we have recently shown that tetra-
butylammonium decatungstate [TBADT, (n-Bu4N)4W10O32],
a well-established free-radical initiator with several applications
in organic synthesis,2 can be utilized for the selective, radical
mono-functionalization of C60.3 Building upon this work, we
report herein a powerful free-radical method for the preparation
of structurally diverse hydroxyalkylated [60]fullerene mono-
adducts, thus broadening the scope of TBADT complex as an
efficient catalyst in fullerene chemistry.
The synthesis of hydroxyl-containing [60]fullerene
derivatives is an area of intense research interest, due to the
great application potential that this class of compounds
possess in materials science and medicinal chemistry.4 These
compounds can be roughly divided into two types based on
their motifs: (i) polyhydroxylated fullerenes (fullerenols) where
the –OH group is located on the C60 core,5 and (ii) hydroxy-
alkyl(aryl)fullerenes (HAFs) where the hydroxyl group is
located on the organic addend bound to the fullerene core.6,7
Although synthetic methods for the preparation of fullerenols
have been extensively investigated,5 hydroxyalkylation of C60
is still in its infancy; to the best of our knowledge this is the
first example concerning a general method for the selective,
and yet direct mono-hydroxyalkylation of C60 from commercially
available alcohols.
The structure of compounds 1a–6a was unambiguously
established by 1H NMR, 13C NMR and UV-vis spectroscopy,
as well as by MALDI mass spectrometry. The 1H NMR
spectra showed the expected proton coupling pattern for these
C1 or Cs symmetric fullerene derivatives. As a diagnostic
signal, compounds 1a–6a exhibit the characteristic fullerenyl
C60-H proton resonance at 6.60–6.98 ppm. The Ca-H protons
Table 1 TBADT-photocatalyzed reaction of alcohols 1–6 with C60
Initially, the potential of the aforementioned mono-
hydroxyalkylation of [60]fullerene was assessed by studying
the TBADT-photocatalyzed reaction of C60 with one of the
simplest commercially available alcohols, namely MeOH (1).
This reaction involved irradiation of a degassed solution of
Entry Substrate R1R2CHOH
Irradiation time/min Yield (%)a
1
2
1
2
CH3OH
CH3CH2OH
30
25
30 (70)
30 (70)
C
60 in a mixture of chlorobenzene/CH3CN (85 : 15), 0.5 equiv.
3
4
3
4
20
10
30 (65)
25 (60)
TBADT, and 200 equiv. of MeOH (1) under an argon
atmosphere. Gratifyingly, hydroxyalkylfullerene (HAF)
Department of Chemistry, University of Crete, 71003 Voutes,
Heraklion (Crete), Greece. E-mail: orfanop@chemistry.uoc.gr;
Fax: 30 2810 545001; Tel: 30 2810 545030
w This article is part of the ‘Carbon Nanostructures’ web-theme issue
for ChemComm.
z Electronic supplementary information (ESI) available: Experimental
procedures, spectral data, 1H/13C NMR and UV/vis spectra of all new
compounds. See DOI: 10.1039/c0cc02836c
5
6
a
5
6
40
25
35 (65)
15 (35)
Isolated yield. Yield in parentheses is based on consumed C60
.
c
8228 Chem. Commun., 2010, 46, 8228–8230
This journal is The Royal Society of Chemistry 2010

appear at 5.13–5.57 ppm for adducts 1–4. This resonance is
low-field shifted at 6.11 and 6.52 ppm, for the propargyl and
benzyl protons of 5a and 6a, respectively. The 13C NMR
spectra of C1 symmetric compounds 2a, 3a, 5a, and 6a exhibit
almost all 60 resonances of the fullerene cage carbons. On the
other hand, Cs symmetric compound 4a exhibits 30 13C NMR
signals for the sp2 carbons of the C60 core and two signals for
the sp3-hybridized fullerenyl carbons. More importantly, the
UV-vis absorption spectra of compounds 1a–6a showed the
diagnostic weak absorption of 1,2-adducts of C60 at around
432 nm.
experiment mentioned above, that the intermediate C60 anion
is mainly trapped by a proton originating from residual
moisture in the reaction mixture.
At this point it is worth noting that the reaction of alcohols
1
and 2 with C60 has been previously studied under
co-sensitization conditions.7 In particular, Mattay and coworkers
found that irradiation of C60 in the presence of a sensitizer
(e.g., DCA) and biphenyl (BP) as co-sensꢂitizer leads to the
formation of the radical cation of C60 (C60 +), which in turn
reacts with alcohols 1 or 2 by abstracting a hydrogen atom
from the a-C–H and/or O–H bond of 1 or 2.7 In this case, the
corresponding adducts 1a and 2a are identical with those
observed in the present study. Hence, it was reasonable to
assume that a similar mechanism may be operative in the
present study, with TBADT acting as a strong photochemical
The reaction mechanism was next investigated by performing
a series of deuterium labeling experiments. Initially, the origin
of the fullerenyl H atom in RC60H was determined by
performing the photocatalyzed reaction of 3 with C60 in a
mixture of C6H5Cl/CH3CN containing 0.5% D2O. A high
percentage of deuterium incorporation (>80%) at the C60 sp3
+
ꢂ
oxidant to produce C60
. This assumption is further
supported by considering the relative reduction potential order of
the involved species. Thus, photoexcited decatungstate (TBADT*)
has a higher reduction potential (Ered = 2.26–2.61 V vs. SCE)12
compared to that of ground or excited state C60 [Ered(3C60*) =
1
carbon of the reaction product 3a was observed by H NMR
spectroscopy.11 Apparently, this result suggests that the
reaction mechanism involves the intermediacy of fullerene
anion, which can be effectively trapped by D2O.
1.14
V
vs. SCE or
E
red(C60)
=
ꢀ0.42
V vs. SCE,
To probe this mechanism further and identify the rate-
limiting transition state, we measured the intramolecular
primary kinetic isotope effect (KIE) of this reaction. To this
end, we studied the reaction of C60 with butan-1-ol-d1 (3-d1)
(Scheme 1). This reaction was performed similarly to those
described previously in this work. The ratio of products 3a-d1
and 3b-d1,which is the result of an intramolecular isotopic
competition between the Ca–H and Ca–D bonds of 3-d1, is
proportional to the primary isotope effect of kH/kD. 1H NMR
integration of the methylene –CH2– signals of both 3a-d1 and
3b-d1 as well as the Ca-H proton of 3b-d1 determined the
primary isotope effect kH/kD = 2.23 ꢁ 0.10 (Scheme 1). This
substantial KIE indicates an extensive C–H(D) bond breaking
in the transition state of the first slow radical-forming step.
Another important finding from this experiment concerns the
source of the fullerenyl hydrogen. Thus, the adduct 3b-d1,
from which the kD is derived, is partially deuterated (RC60-H/D).
This result indicates, in accordance with the D2O-quenching
respectively],13 which makes an electron transfer oxidation
of C60 by TBADT* energetically feasible. On the other hand,
another possible reaction pathway that could also be operative
in the present study would involve H-atom abstraction from
+
ꢂ
a-C–H alcohol bond by TBADT* instead of by C60
.
In order to clarify this issue, we studied the TBADT-
catalyzed reaction of C60 with ethylene glycol (7, eqn (1)).
This substrate was chosen as a mechanistic probe by virtue of
its characteristic product distribution upon reaction with
C60 +. In particular, as first shown by Mattay and coworkers,
ꢂ
the DCA/BP-catalyzed reaction of C60 with
7 affords
exclusively adduct 1a.7b The formation of this adduct has been
rationalized by considering a H-atom abstraction from the
–OH group of 7 by C60 +, followed by extrusion of
ꢂ
formaldehyde (CH2O) and coupling of the resultant radical
ꢂ
species (ꢂCH2OH and HC60
) (eqn (1)). Accordingly,
the formation of this product in the corresponding
+
ꢂ
TBADT-catalyzed reaction would reveal the formation of C60
.
ð1Þ
In the present study, however, the TBADT-catalyzed reaction
of C60 with 7 afforded exclusively product 7a instead of 1a
(Scheme 2). This important result indicates that alcohol 7a is
derived through a H-atom abstraction from the a-C–H
alcohol bond of 7 by TBADT* (vide infra). The absence of
compound 1a in this case strongly argues against the reaction
+
ꢂ
pathway that would involve the intermediacy of C60
.
Similarly, the otherwise possible TBADT-mediated formation
+
ꢂ
of C60
can now be safely excluded from the previously
reported TBADT-mediated reactions of C60.3 This important
conclusion is further supported by the fact that no TBADT-
catalyzed reaction of C60 with either propionic acid or methyl
formate was observed, although these substrates are known to
be reactive with C60 upon DCA/BP catalysis.7b Also, TBADT
photocatalyzed oxidation of aryl alkanols to form aryl ketones
Scheme 1 Determination of the intramolecular primary KIE in the
1
reaction of 3-d1 with C60 by H NMR spectroscopy.
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 8228–8230 8229

materials, and the unique structure of the final adducts, this
new reaction paves the way for the preparation of a variety of
new hydroxyalkylated fullerenes of interest in materials
science and medicinal chemistry.
The Foundation for Education and European Culture is
acknowledged for providing a one year fellowship to M.N.A.
We are also grateful to Prof. T. Drewello at the University of
Erlangen for performing the MALDI MS analyses.
Scheme 2 TBADT-photocatalyzed reaction of C60 with 7.
is known to proceed through a hydrogen abstraction mechanism
via the formation of the coresponding radical intermediate
similar to the present case.14
Notes and references
1 A. Hirsch and M. Brettreich, Fullerenes, Chemistry and Reactions,
Wiley-VCH, Weinheim, Germany, 2005.
2 M. D. Tzirakis, I. N. Lykakis and M. Orfanopoulos, Chem. Soc.
Rev., 2009, 38, 2609–2621.
According to the above-mentioned results, the reaction
mechanism of the photocatalyzed hydroxyalkylation of C60
is summarized in Scheme 3. Initially, the excited state of
decatungstate (W10O324ꢀ*), generated upon light excitation,
3 (a) M. D. Tzirakis and M. Orfanopoulos, Org. Lett., 2008, 10,
873–876; (b) M. D. Tzirakis and M. Orfanopoulos, J. Am. Chem.
Soc., 2009, 131, 4063–4069; (c) M. D. Tzirakis and
M. Orfanopoulos, Angew. Chem., Int. Ed., 2010, 49, 5891–5893.
4 For selected examples, see: (a) A. R. Wielgus, B. Zhao,
C. F. Chignell, D.-N. Hu and J. E. Roberts, Toxicol. Appl.
Pharmacol., 2010, 242, 79–90; (b) R. Injac, M. Perse, M. Cerne,
N. Potocnik, N. Radic, B. Govedarica, A. Djordjevic, A. Cerar
and B. Strukelj, Biomaterials, 2009, 30, 1184–1196;
(c) A. R. Badireddy, E. M. Hotze, S. Chellam, P. Alvarez and
M. R. Wiesner, Environ. Sci. Technol., 2007, 41, 6627–6632;
(d) B. Vileno, P. R. Marcoux, M. Lekka, A. Sienkiewicz,
T. Feher and L. Forro, Adv. Funct. Mater., 2006, 16, 120–128.
5 For selected examples, see: (a) G. Zhang, Y. Liu, D. Liang, L. Gan
and Y. Li, Angew. Chem., Int. Ed., 2010, 49, 5293–5295;
(b) K. Kokubo, K. Matsubayashi, H. Tategaki, H. Takada and
T. Oshima, ACS Nano, 2008, 2, 327–333; (c) G. C. Alves,
L. O. Ladeira, A. Righi, K. Krambrock, H. D. Calado, R. P. F. Gil
and M. V. B. Pinheiro, J. Braz. Chem. Soc., 2006, 17, 1186–1190.
6 For selected examples, see: (a) S. Chopin, J. Delaunay and
J. Cousseau, Tetrahedron Lett., 2005, 46, 373–376;
(b) Y. Nakamura, K. O-kawa, S. Minami, T. Ogawa, S. Tobita
and J. Nishimura, J. Org. Chem., 2002, 67, 1247–1252;
(c) X. Zhang and C. S. Foote, J. Org. Chem., 1994, 59,
5235–5238; (d) S. Shi, K. C. Khemani, Q. C. Li and F. Wudl,
J. Am. Chem. Soc., 1992, 114, 10656–10657.
7 (a) C. Siedschlag, H. Luftmann, C. Wolff and J. Mattay, Tetra-
hedron, 1997, 53, 3587–3592; (b) C. Siedschlag, H. Luftmann,
C. Wolff and J. Mattay, Tetrahedron, 1999, 55, 7805–7818.
8 M. D. Tzirakis, M. N. Alberti, L. C. Nye, T. Drewello and
M. Orfanopoulos, J. Org. Chem., 2009, 74, 5746–5749.
9 A similar substitution of a –CH2OH group on the C60 core by
H atom has been reported previously. See: P. Timmerman,
H. L. Anderson, R. Faust, J.-F. Nierengarten, T. Habicher,
P. Seiler and F. Diederich, Tetrahedron, 1996, 52, 4925–4947.
10 This transformation was also observed during the purification
procedure with column chromatography on silica gel of adducts
4a–6a.
abstracts a H-atom from the Ca–H bond of alcohol affording
5ꢀ
the one-electron reduced form of decatungstate (W10O32
)
and the corresponding a-hydroxy C-centered radical. Isotope
effect studies showed that this step determines the rate of this
reaction. The triplet excited state of C60 (3C60*), generated
upon light excitation followed by an effective intersystem
4ꢀ
crossing (FISC E 1), regenerates W10O32 through an electron
transfer process, thus closing the catalytic cycle. Since the free
5ꢀ
energy change of electron transfer (DG1et) from W10O32
3
[E1(W10O324ꢀ/W10O325ꢀ) = –1.215 V vs. SCE] to C60*
(E1red = 1.14 V vs. SCE)13 is negative (DG1et = ꢀ226 kJ molꢀ1),
3
the electron transfer reduction of C60* is thermodynamically
feasible to give C60 ꢀ. Similarly, however, the electron
ꢂ
transfer reduction of ground state C60 (E1red = ꢀ0.42 V vs.
5ꢀ
SCEꢀ)13 by W10O32 is also energetically feasible to form
C60 (DG1et = ꢀ76.7 kJ molꢀ1). Finally, radical coupling
ꢂ
ꢀ
ꢂ
followed by protonation of the resulting radical anion (C60
affords the observed fullerene adducts.
)
In summary, the photochemical addition of a-hydroxy
C-centered radicals to C60 represents a powerful method for
the hitherto unexplored mono-hydroxyalkylation of C60.
Considering the inexpensive catalytic system, the simplicity
of this method, the convenient availability of the starting
11 Compound 3a does not exchange its C60-H proton in a measurable
amount, since no deuterium incorporation could be detected after
a sample of 3a in an 85 : 15 mixture of C6H5Cl/CH3CN had been
agitated with D2O for several hours or even upon irradiation in the
presence or absence of TBADT. Similarly, the use of dry C6H5Cl/
CD3CN as the reaction solvent in the presence of TBADT did not
alter this result.
12 E1(X/W10O325ꢀ) = E1(W10O324ꢀ/W10O325ꢀ) + E*(W10O324ꢀ - X),
where X = transient excited state W10O324ꢀ*. See: I. Texier,
J. A. Delaire and C. Giannotti, Phys. Chem. Chem. Phys., 2000, 2,
1205–1212.
13 J. W. Arbogast, C. S. Foote and M. Kao, J. Am. Chem. Soc., 1992,
114, 2277–2279.
14 I. N. Lykakis, C. Tanielian and M. Orfanopoulos, Org. Lett., 2003,
5, 2875–2878.
Scheme 3 Proposed mechanism for the TBADT-catalyzed reaction
of alcohols with fullerene C60
.
c
8230 Chem. Commun., 2010, 46, 8228–8230
This journal is The Royal Society of Chemistry 2010