Aromatic tripodal receptors for (C60-Ih)[5,6] fullerene

Article abstract of DOI:10.1039/b910921h

Monotopic and ditopic tripodal benzene platforms featuring aromatic and perfluoroaromatic side-arms have been synthesized, and their binding properties toward C₆₀-fullerene have been investigated by HPLC examining retention times on a fullerene

Full text of DOI:10.1039/b910921h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Aromatic tripodal receptors for (C60-Ih)[5,6]fullerene†
Gian Maria Dell’Anna,a Rita Annunziata,a Maurizio Benaglia,a Giuseppe Celentano,b Franco Cozzi,*a
Oscar Francesconic and Stefano Roelens*c
Received 4th June 2009, Accepted 2nd July 2009
First published as an Advance Article on the web 24th July 2009
DOI: 10.1039/b910921h
Monotopic and ditopic tripodal benzene platforms featuring aromatic and perfluoroaromatic side-arms
have been synthesized, and their binding properties toward C60-fullerene have been investigated by
HPLC examining retention times on a fullerene-modified silica stationary phase, using highly polar
eluants (acetonitrile and acetonitrile/water). By comparison of structurally homogeneous sets of
receptors, a clear trend could be found, pointing to an increased retention for ditopic derivatives, in
which binding can occur on both sides of the benzene platform, over their monotopic counterparts.
Among the latter, monotopic receptors containing H-substituted aromatic residues showed stronger
retention than their perfluorinated analogues. This effect was ascribed to the greater availability of the
p-electrons in a H-substituted aromatic ring with respect to the corresponding F-substituted
counterpart in participating in a p–p interaction with the electron-poor surface of fullerene. Several
NMR experiments aimed to investigate binding interactions in solution, using the much less polar
solvents required by the fullerene solubility (1,1,2,2-tetrachloroethane, chloroform, toluene, and CS2),
did not provide any evidence of binding interactions. We concluded that p–p interactions between
fullerene and the investigated flexible tripodal receptors cannot compete with solvation in poorly polar
solvents, and that the binding interactions observed by HPLC were essentially forced by the strongly
polar eluant employed for the HPLC analysis.
Combining our interest in aromatic–aromatic11a–c and aromatic–
perfluoroaromatic11d–f interactions with that in the development of
Introduction
tripodal receptors for neutral molecules,12 we decided to synthesize
a series of flexible, benzene-based tripodal receptors featuring
phenyl or pefluorophenyl rings, and to test them in the binding of
(C60-Ih)[5,6]-fullerene. Rather than the discovery of new effective
receptors for fullerene, one of the goals of this work was to
ascertain, through the comparison of aromatic receptors with their
perfluorinated counterparts, whether “pure” p–p convex–concave
interactions7a may be the driving force in the recognition of
fullerene. We wish to report here some results of this study, showing
that in media of good solvating properties toward fullerene and in
the absence of pre-organization of the receptor, p–p interactions
are not competing favourably for complexation. In contrast, in a
polar environment a clear evidence of binding has been obtained,
pointing to a stronger interaction of fullerene with H-substituted
aromatic receptors than with their corresponding perfluorinated
counterparts, in agreement with the higher electron availability of
the former with respect to the latter.
The identification of effective aromatic molecular receptors such as
tweezers,1 clips,2 and capsules3 for aromatic guests can be consid-
ered a probing test of our capacity of understanding arene–arene
interactions and successfully transferring this knowledge into
practical applications. In this context, the synthesis of aromatic
receptors for fullerenes extended aromatic–aromatic recognition
from an essentially two-dimensional phenomenon (that is, the
interaction between flat surfaces)4 to a three-dimensional event
(that is, the interaction between curved networks of conjugated
carbon–carbon bonds).5 Recent examples of curved aromatic
fullerene receptors include belt-shaped cyclic oligomers of ary-
lacetylenes (nanorings),5,6 a tweezer7a and a copper surface7b con-
taining corannulene subunits, several concave tetrathiafulvalene-
type donors,8 and a tris(trianthreno)-triquinacene.9 All of these
receptors exploit p–p interactions between the electron poor
convex surface of the fullerene guest and the electron rich concave
surface of the hosts.5,10
Results and discussion
aDipartimento di Chimica Organica
e Industriale, Universita` degli
Studi di Milano, via Venezian 21, I-20133, Milano, Italy. E-mail:
franco.cozzi@unimi.it; Fax: +39-02-50314159
In designing simple tripodal aromatic receptors for fullerene, we
took advantage of the so-called “facial segregation” experienced
by 1,3,5-CH2R-2,4,6-CH2R¢ hexasubstitued benzenes.13 Accord-
ing to seminal work by Mislow and Siegel on the static and dy-
namic behaviour of these compounds,14 hexasubstituted benzenes
adopt a preferred conformation with an alternated ababab (all-
trans) geometrical pattern, thermodynamically favoured by about
4 kcal mol-1 compared to the next most stable conformation. Thus,
when a benzene platform carries alternated CH2R and CH2R¢
bDipartimento di Scienze Molecolari Applicate ai Biosistemi, Universita` degli
Studi di Milano, via Venezian 21, I-20133, Milano, Italy
cCNR, Istituto di Metodologie Chimiche (IMC) and Dipartimento di
Chimica Organica, Universita` di Firenze, Polo Scientifico e Tecnologico,
I-50019, Sesto Fiorentino, Firenze, Italy. E-mail: stefano.roelens@unifi.it;
Fax: +39-055-4573570
† Electronic supplementary information (ESI) available: 1H and 13C NMR
spectra in toulene-d8 and ESI-MS spectra of (C60-Ih)[5,6]-fullerene with
3H. See DOI: 10.1039/b910921h
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3871–3877 | 3871
©

View Article Online
substituents, the R and R¢ groups point toward opposite sides
with respect to the benzene ring. The readily accessible 1,3,5-
tris(halomethyl)-2,4,6-triethylbenzenes (halo = Br and Cl)15 led
to a variety of interesting structures and opened the way to an
extensive exploitation of this scaffold for the development of
tripodal receptors for molecular recognition studies.13
Following a rational design of receptor’s geometry, aromatic
rings and their corresponding readily available fluorinated coun-
terparts were implemented as binding side-arms in the receptor’s
architecture, giving rise to a set of fullerene receptors of signifi-
cantly different electronic properties. Binding domains of different
size and shape were also obtained by implementing spacers of
different length and conformational mobility. For this purpose,
benzyloxymethyl side arms appeared to be particularly promising,
since they have been shown to be involved in binding interac-
tions with (C60-Ih)[5,6]-fullerene.16 In addition to 1,3,5-tris-CH2R-
2,4,6-triethylbenzenes, 1,2,3,4,5,6-hexa-CH2Ar-substituted ben-
zenes were also prepared to obtain ditopic receptors capable
of binding to both sides of the platform. The structures of the
receptors prepared for this study are reported in Fig. 1.
Hexabenzyl benzene 217 and hexabenzyloxy-methyl benzene 418
were known compounds and were prepared, according to liter-
ature procedures, by the Co2(CO)8-catalyzed cyclo-trimerization
of 1,4-diphenyl-2-butyne and 1,4-dibenzyloxy-2-butyne, respec-
tively. 1,3,5-Tribenzyl-2,4,6-triethylbenzene 1H was prepared as
described in Scheme 1. The tris-aldehyde 5,19 obtained in 69%
yield from 1,3,5-tribromomethyl-2,4,6-triethylbenzene12a by re-
action with DMSO and NaHCO3,20 was treated with excess
phenylmagnesium bromide in THF to afford the triol 6 as a
mixture of diastereoisomers in 62% yield. Deoxygenation with
Et3SiH/CF3COOH21 gave compound 1H in 59% yield. 1,3,5-Tris-
pentafluorophenylmethyl-2,4,6-triethylbenzene 1F was similarly
obtained (Scheme 1) by reaction of the tris-aldehyde 5 with
pentafluorophenylmagnesium bromide (53% yield) followed by
deoxygenation of the diastereoisomeric triols 7 with H3PO2/I2 in
acetic acid (60% yield).22
Scheme 1 Synthesis of hexasubstituted benzenes 1H, 1F, 3H, and 3F.
Finally, reaction of the tris-aldehyde 5 with the trimethylsilyl
ethers of benzyl and pentafluorobenzyl alcohol in the pres-
ence of trimethylsilyltriflate and triethylsilane afforded 1,3,5-
tribenzyloxymethyl-2,4,6-triethylbenzene 3H and its perfluori-
nated counterpart 3F, in 36 and 10% yield, respectively.23
With receptors 1–4 available, we reasoned that elution through
a HPLC column packed with fullerene-modified silica could
provide a fast way of testing their binding ability toward (C60-
Ih)[5,6]-fullerene simply by comparing retention times. Indeed,
use of the retention data obtained from HPLC analysis carried
out on modified silica gel has been demonstrated to reliably
correlate with the binding affinity of aromatic hosts toward
aromatic guests.24 Therefore silica 8 (Fig. 2, 0.135 mmol g-1 by
elemental analysis) was prepared following the procedure reported
Fig. 1 Structures of hexasubstituted benzenes 1–4.
3872 | Org. Biomol. Chem., 2009, 7, 3871–3877
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
increased retention for receptors in which binding can occur on
both sides of the central benzene ring, whereas poorer retention
was observed for perfluorinated derivatives. Thus, in the “benzyl”
series (1H, 1F, and 2), the hexabenzylsubstituted compound 2
was more strongly retained than the tribenzylsubstituted 1H,
which in turn was more retained than the tripentafluorobenzyl
analogue 1F. The benzyloxymethyl series showed analogous trend
of retention times (4 > 3H > 3F).
A
comparison across the two sets showed that the
benzyloxymethyl-substituted compounds were more strongly re-
tained than their benzyl-substituted counterparts. However, the
difference between each pair was larger for ditopic (4 vs. 2, 5.3 min)
than for monotopic (3H vs. 1H, DRt = 1.1 min) receptors, whereas
the difference between the fluorinated pair was negligibly small
(3F vs. 1F, DRt = 0.3 min). While indicating a more favourable
geometry/adaptability of the benzyloxy side-arms, likely due to
the ability of the latter of better achieving the most appropriate
geometry for binding,26 the significantly larger retention times
exhibited by the ditopic vs. monotopic receptors, exceeding the
values expected on the basis of mere additivity, strongly suggest
the occurrence of cooperative contributions from concomitant
binding, which may be ascribed to multivalent interactions on the
silica surface.
The observed trends were emphasized when water was added
to the mobile phase, eluting with an acetonitrile/water 90:10
mixture. The higher polarity of the eluant resulted in a general
and predictable increase of retention times for adducts 1–4 on
fullerene-modified silica 8. Indeed, 8 can be regarded a sort of
reversed phase silica, on which apolar substances generally show
larger affinities for a poorly polar stationary phase when the eluant
polarity is increased. Surprisingly, a similar effect was not observed
on the RPC18 column: the more polar eluant did not induce neither
an increase in retention times, nor a discrimination among the
solutes, clearly indicating binding interactions as the cause of the
observed discrimination occurring on the surface of the fullerene-
coated silica 8. In addition to polarity effects, it is also tempting
to suggest the occurrence of hydrophobic contributions, caused
by the added water, which may boost the discrimination observed
with respect to neat acetonitrile, resulting in enhanced interaction
between the immobilized fullerene on silica and the receptors in
the mobile phase.
Fig. 2 Fullerene-modified silica 8.
by Maggini et al.25 This fullerene-modified stationary phase has
been used to investigate the binding affinities of some cyclic
oligomeric hosts by HPLC, showing exceptional size selectivity
for calixarenes and cyclodextrins. It was also demonstrated that
silica 8 selectively binds to helical peptides, provided that the latter
would present aromatic residues properly located for interacting
with fullerene.25
Thus, compounds 1–4 were eluted through a HPLC column
packed with 8 (column size 125 ¥ 4.6 mm, internal diameter;
efficiency: N/m = 4000; for efficiency determination see Exper-
imental) using acetonitrile or acetonitrile/water 90:10 as eluant.
The observed retention times were reported in Table 1, together
with those observed using a commercial reversed phase RPC18
column having the same size (125 ¥ 4.6 mm) but much higher
efficiency (N/m = 30000).
The retention data obtained with acetonitrile as eluant showed
that, while the RPC18 column was not able to discriminate com-
pounds 1–4, elution of the latter through the column packed with
silica 8 resulted in significantly different retention times. It must
be emphasized that this latter column was able to selectively retain
1–4 even though its efficiency was much lower than that of the
RPC18 column (N/m 4000 vs. 30000). Comparing homogeneous
sets of compounds, a clear trend could be found pointing to an
Altogether, these data seem to suggest that the mechanism by
which silica 8 discriminates receptors 1–4 relies on aromatic–
aromatic interactions of different strength. In particular, the
interaction between the convex surface of the fullerene guest5,10
and the concave surface of the tripodal hosts is likely modulated
by the electron density available on the receptor surface, which is
certainly larger for the pair 1H/3H than for the pair 1F/3F because
of the presence on the latter of the strongly electronegative fluorine
atoms.27 Apart from electronic effects and geometric/adaptability
factors, the oxygen atom lone pairs of 3H and 4 may also provide
additional binding interactions with fullerene, as proposed for
the formation of the adduct between (C60-Ih)[5,6]-fullerene and
g-cyclodextrin.28 However, the very small difference in reten-
tion times observed between 1F and 3F seems to indicate that such
a contribution is negligible, being overwhelmed by the aromatic–
aromatic interaction.29
Table 1 HPLC Retention times Rt (min) on modified silica 8 and on
reversed phase RPC18 columns for compounds 1–4, 1,3,5-triethylbenzene
(TEB), naphthalene (N), 1,2,3,4-tetrafluoronaphthalene (TFN), and
octafluoronaphthalene (OFN)a
CH3CN
CH3CN/H2O 90:10
Compound
Rt (8)
Rt(RPC18)
Rt (8)
Rt(RPC18)
1H
1F
4.8
3.7
5.9
5.9
4.0
11.2
1.7
1.5
1.5
1.3
1.7
1.7
1.7
1.7
1.7
1.7
1.4
1.4
1.4
1.3
7.8
6.4
10.9
9.0
6.6
18.0
2.1
1.6
1.6
1.4
1.7
1.7
1.7
1.7
1.7
1.7
1.4
1.4
1.4
1.3
2
3H
3F
4
TEB
N
TFN
OFN
Further support to the hypothesis that the discrimination of the
tripodal receptors by the fullerene residues present on the surface
a Flow rate 1.0 mL/min; l = 254 nm.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3871–3877 | 3873
©

View Article Online
of silica 8 was mainly due to aromatic–aromatic interactions
involving the receptors’ side arms was obtained when 1,3,5-
triethylbenzene was eluted on the column containing silica 8 and
on the RPC18 column. Indeed, with these supports very short
retention times were observed (Table 1) both in acetonitrile (1.7
and 1.4 min on silica 8 and on RPC18, respectively) and in the
acetonitrile/water 90:10 mixture (2.1 and 1.4 min on silica 8 and
on RPC18, respectively). Similarly, when a series of flat aromatic
systems featuring different degrees of fluorine substituents such
as naphthalene, 1,2,3,4-tetrafluoronaphthalene, and octafluoro-
naphthalene were eluted through the two columns, again very
short and essentially identical retention times were observed with
both eluants (Table 1). These results concur to indicate that
the presence of properly arranged aromatic residues is essential
for the receptors to be discriminated by silica 8, and that the
discrimination mainly relies on aromatic–aromatic interactions
between the convex surface of fullerene and the concave surfaces
of adducts 1–4.
In order to support the results of the HPLC separations with
complementary experimental techniques, we attempted to investi-
gate the interaction of (C60-Ih)[5,6]-fullerene with the monotopic
receptor 3H by NMR spectroscopy in solution.30 Unfortunately,
in the highly polar solvents where HPLC data were acquired,
fullerene was completely insoluble, as was also the modified
fullerene employed for the synthesis of silica 8. Accordingly, the
NMR study had to be carried out in different solvents, and a
compromise between the solubility of fullerene, requiring non-
polar solvents, and the polarity of the medium, necessary to
emphasize aromatic–aromatic interactions, was thus mandatory.
NMR experiments carried out by adding up to a three-
fold molar excess of fullerene to 3H in deuterated 1,1,2,2-
tetrachloroethane (Cl2CDCDCl2, e = 8.2, m = 1.3 D) showed
no shift variation in the whole concentration range (see Fig. S1 in
the ESI†). The results were confirmed by following the 13C NMR
signal of fullerene through the addition of up to a 15-fold molar
excess of 3H in the same medium (Fig. S2). The lack of binding
evidence was ascribed to the good solvating properties and the
markedly lower polarity of Cl2CDCDCl2 compared to acetonitrile
(CD3CN, e = 37.5, m = 3.5 D), behaving as a strongly competitive
medium.
Fig. 3 1H-NMR spectra (400 MHz, CS2/CDCl3 80:20) of: (a) 3H
0.31 mM; (b) 3H 0.31 mM, fullerene 4.86 mM.
of the increased polarity of the medium. However, when the effect
of the solvent mixture was investigated by varying the solvent
ratio on a pre-formed 1:1 mixture of 3H and fullerene (Fig. 4), it
appeared evident that the observed spectral changes were due to
solvent-related shift variations rather than to binding interactions.
These results outlined that spectral variations caused by a mere
changing of the medium may bias experimental observations and
should therefore be taken with caution.
Independent evidence of binding was also sought through
MS spectrometry. The ESI-MS spectra of a 1:1 mixture of 3H
and fullerene in toluene/MeOH 80:20, scanned in positive and
negative ion mode, are reported in Fig. S5 and S6, respectively.†
While in the former only the receptor peak is detected (M + Na+),
in the latter only the fullerene peak can be observed (M-). Since
both species cannot be directly observed in neither one of the two
spectra, the absence of the peak of the complex is not conclusive,
indicating that mass spectrometry is not informative on binding
of fullerene to 3H.
Since binding of (C60-Ih)[5,6]-fullerene to a double concave
receptor was measured in toluene by Sygula and co-workers,7a
we also turned to this solvent to ascertain whether binding
interactions could be observed with receptor 3H. The results
shown in Fig. S3 and S4,† confirmed that NMR spectroscopy
did not provide evidence of binding in solution.
Continuing the search for a convenient medium to unravel
the occurrence of binding interactions, we finally turned to
CS2/CDCl3 mixtures, where fullerene binding by a concave
tetrathiafulvalene-type receptor could be observed by Martin and
co-workers.8a However, the addition of a 16-fold molar excess
of fullerene to a solution of 3H in CS2/CDCl3 80:20, followed
Conclusions
From the results obtained in the present investigation, we
concluded that the unambiguous interaction observed between
fullerene and the receptors of Fig. 1, determining the marked
discrimination of the receptors by the fullerene coating the surface
of the stationary phase of the HPLC column, was essentially
due to the “forced” interactions established between the aromatic
surfaces of the host and the guest in the hostile environment of a
strongly polar medium, such as acetonitrile or acetonitrile/water
90:10. The most reasonable interpretation of these data ascribes
the lack of detectable interactions in solution between fullerene
and receptor 3H to the solvation of fullerene, which cannot
1
through the H NMR signals of the receptor again showed no
evidence of interaction (Fig. 3). In the hope that the use of
a more polar solvent mixture could be beneficial for making
binding interactions to emerge, we change the solvent to a 50:50
CS2/CDCl3 mixture. Upon addition of 1 equiv of fullerene to 3H
we found that a small but unambiguous shift could be observed
on both the CH2O signals of the receptor indicating a clear effect
3874 | Org. Biomol. Chem., 2009, 7, 3871–3877
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 4 1H-NMR spectra (400 MHz) of: (a) 3H 1 mM in CS2/CDCl3 80:20; (b) 3H (1 mM) and fullerene (1 mM) in CS2/CDCl3 80:20; (c) 3H (1 mM)
and fullerene (1 mM) in CS2/CDCl3 62:38; (d) 3H (1 mM) and fullerene (1 mM) in CS2/CDCl3 50:50. Only the CH2O and the CH2(Et) signals of the
receptors are shown.
be overcome by the weak p–p interactions established with the
tripodal hosts in the absence of a compelling pre-organization of
the structure. This interpretation of the experimental data agrees
with recent results by Boyd et al.31 indicating that fullerene de-
solvation plays a major role in its binding by aromatic receptors
and that the strength of binding increases in a medium which is a
poor solvating agent for fullerene.
Synthesis of 1,3,5-tribenzyl-2,4,6-triethylbenzene 1H. To a
stirred solution of compound 5 (626 mg, 2.54 mmol) in dry
THF (25 mL) kept under nitrogen at 0 ◦C, a 3 M solution of
PhMgBr in diethyl ether (4.2 mL, 12.7 mmol) was slowly added.
The reaction mixture was allowed to slowly warm up to room
temperature and was then stirred overnight. The reaction was
quenched by a cautious addition of water (10 mL) and 1 N
aqueous HCl (15 mL) and the resulting mixture was extracted
with ethyl acetate (3 ¥ 20 mL). The combined organic phases
were dried and concentrated to give the crude product. This was
purified by flash chromatography with a 9:1 hexanes/ethyl acetate
mixture as eluant. Triol 6, a thick oil with Rf 0.2, was isolated as
a mixture of diastereoisomers (756 mg, 62% yield). These were
not further purified but directly subjected to de-oxygenation.21
To a stirred solution of compound 6 (596 mg, 1.24 mmol) in
dry dichloromethane (30 mL) kept under nitrogen, trifluoroacetic
acid (1.11 mL, 14.5 mmol) and triethylsilane (4.66 mL, 29.2 mmol)
were added in this order and the mixture was stirred for 70 h at
room temperature. The reaction was quenched by the addition
of a saturated aqueous solution of NaHCO3, and the organic
phase was separated, dried and concentrated under vacuum.
The residue was purified by flash chromatography with a 98:2
hexane/dichloromethane mixture as eluant to afford the prod◦uct
as a white solid (Rf 0.8, 319 mg, 59% yield). It had m.p. 158 C.
1H-NMR (CDCl3): d 7.26 (d, J = 7.6 Hz, 2H), 7.17 (d, J = 7.2 Hz,
1H), 7.03 (d, J = 7.2 Hz, 2H), 4.17 (s, 2H), 2.48 (q, J = 7.5 Hz,
2H), 1.08 (t, J = 7.5 Hz, 3H). 13C-NMR (CDCl3): d 141.5, 141.4,
134.0, 128.3, 127.8, 125.6, 34.6, 23.7, 15.1. Elemental analysis:
C33H36 requires: C, 91.61; H, 8.39; found: C, 91.79; H, 8.21%.
Experimental
Materials
All commercially available reagents including dry solvents were
used as received. Organic extracts were dried over sodium sul-
fate, filtered, and concentrated under vacuum using a rotatory
evaporator. Non-volatile materials were dried under high vacuum.
Reactions were monitored by thin-layer chromatography on pre-
coated Merck silica gel 60 F254 plates and visualized either by
UV or by staining with a solution of cerium sulfate (1 g) and
ammonium heptamolybdate tetrahydrate (27 g) in water (469 mL)
and concentrated sulfuric acid (31 mL). Flash chromatography
was performed on Fluka silica gel 60. The NMR spectra were
obtained at 500 or 300 MHz for 1H-, 125.6 or 75.3 MHz
for 13C-, and 470.5 or 282 MHz for 19F-NMR. The NMR
spectra for affinity measurements of (C60-Ih)[5,6]-fullerene with
3H were obtained at 400 and 200 MHz for 1H-, 100 and
50 MHz for 13C-NMR. The conditions for NMR analysis of
the receptors/fullerene mixtures were reported in the text. Mass
spectra were recorded on a LTQ-IT-Orbitrap equipped with
a nano-spray-injector, in positive and negative mode. Adducts
217 and 4,18 1,3,5-triformyl-2,4,6-triethylbenzene 5,19 N-[3-
(triethoxysilyl)-propyl]-2-carbomethoxy-3,4-fulleropyrrolidine25
and 1,2,3,4-tetrafluoronaphthalene11f were prepared according to
literature procedures and had m.p. and spectral data in agreement
with those reported in the literature.
Synthesis of 1,3,5-tris-pentafluorophenylmethyl-2,4,6-triethyl-
benzene 1F. To a stirred solution of compound 5 (713 mg,
2.89 mmol) in dry THF (30 mL) kept under nitrogen at 0 ◦C, a 3 M
solution of C6F5MgBr in diethyl ether (4.8 mL, 14.5 mmol) was
slowly added. The reaction mixture was allowed to slowly warm up
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3871–3877 | 3875
©

View Article Online
to room temperature and was then stirred overnight. The reaction
was quenched by a cautious addition of water (10 mL) and 1 N
aqueous HCl (15 mL) and the resulting mixture was extracted with
ethyl acetate (3 ¥ 20 mL). The combined organic phases were dried
and concentrated to give the crude product. This was purified by
flash chromatography with a 85:15 hexanes/ethyl acetate mixture
as eluant. Triol 7, a pale yellow solid with Rf 0.44 and m.p. 78–
80 ◦C, was isolated as a mixture of diastereoisomers (1.149 g, 53%
yield). These were not further purified but directly subjected to
de-oxygenation.22 To a stirred solution of compound 7 (500 mg,
0.66 mmol) in glacial acetic acid (17 mL) kept under nitrogen,
iodine (503 mg, 1.98 mmol) and a 50% w/w aqueous solution of
H3PO2 (0.42 mL, 3.96 mmol) were added in this order and the
mixture was stirred 20 h at 60 ◦C. Ethyl acetate (20 mL) and water
(20 mL) were added and the organic phase was separated. This was
washed with a saturated aqueous solution of NaHCO3, and then
with a saturated aqueous solution of Na2S2O3. The organic phase
was separated, dried and concentrated under vacuum. The residue
was purified by flash chromatography with a 95:5 hexanes/diethyl
ether mixture as eluant to afford the product as a white solid
(Rf 0.7, 278 mg, 60% yield). It had m.p. 91–92 ◦C. 1H-NMR
(CDCl3): d = 4.14 (s, 2H), 2.67 (q, J = 7.5 Hz, 2H), 1.00 (t,
J = 7.5 Hz, 3H). 13C-NMR (CDCl3): d = 145.4 (d, J = 248 Hz),
141.9, 139.6 (dt, J = 248, 15 Hz), 137.5 (dt, J = 248, 15 Hz),
131.9, 114.5, 23.8, 23.7, 14.5. 19F-NMR (CDCl3): d = -142.8 (d,
J = 19.4 Hz), -157.8 (t. J = 18.9 Hz), 163.1 (t, J = 21.0 Hz).
Elemental analysis: C33H21F15 requires: C, 56.42; H, 3.01; found:
C, 56.28; H, 2.94%.
with water (20 mL). The organic phase was dried and
concentrated to give the crude product. This was purified
by two flash chromatography columns, the first with a 70:30
hexanes/ethyl acetate mixture as eluant, and the second with a
95:5 pentane/diethyl ether mixture as eluant. Product 3F (63 mg),
a pale yellow solid wi◦th Rf 0.35, was isolated in 10% yield. It
1
had m.p. 135.8–136.6 C. H-NMR (CDCl3): d = 4.66 (s, 2H),
4.56 (s, 2H), 2.72 (q, J = 7.5 Hz, 2H), 1.11 (t, J = 7.5 Hz, 3H).
13C-NMR (CDCl3): d = 145.5 (d, J = 245 Hz), 145.3, 141.1 (dt,
J = 247, 18 Hz), 136.8 (dt, J = 250, 21 Hz), 131.2, 111.2, 67.2,
59.4, 22.8, 16.4. 19F-NMR (CDCl3: d = -143.4 (d, J = 22.8 Hz),
-154.1 (t, J = 19.2 Hz), -162.5 (t, J = 15.2 Hz). Elemental
analysis: C36H27F15O3 requires: C, 55.45; H, 3.43; found: C, 54.39;
H, 3.36%.
Synthesis of silica 8. This was prepared as described.25
A suspension of ApexPrepSil (particle size 8 mm, 3.0 g)
in dry toluene (80 mL) was heated with stirring under a
continuous stream of nitrogen until the volume of the sol-
vent was reduced to 60 mL. To this azeotropically-dried
mixture, solid N-[3-(triethoxysilyl)-propyl]-2-carbomethoxy-3,4-
fulleropyrrolidine (727 mg, 0.7 mmol)25 was added, and the
mixture was refluxed for 20 h. After cooling to room tem-
perature, the modified silica was collected by filtration and
washed sequentially with 200 mL portions of toluene, chloroform,
methanol, chloroform, and hexane, and subsequently dried under
high vacuum at 60 ◦C to afford 3.5 g of pale brown material.
Carbon elemental analysis (C = 11.13%) showed a loading of
0.135 mmol g-1. Evaporation of the solvents employed to wash the
silica allowed recovery of 58 mg of fulleropyrrolidine.
Synthesis of 1,3,5-trisphenylmethoxymethyl-2,4,6-triethyl-
benzene 3H. To a stirred solution of compound 5 (260 mg,
1.06 mmol) in dry dichloromethane (10 mL) kept under nitrogen at
room temperature, benzyloxytrimethylsilane (582 mg, 3.23 mmol)
in dichloromethane (5 mL) was added followed by triethylsilane
(0.52 mL, 3.23 mmol). The mixture was then cooled at 0 ◦C
and trimethylsilyl triflate (0.077 mL, 0.424 mmol) was added.
The reaction mixture was slowly allowed to warm up to room
temperature and was then stirred for 2 h. Dichloromethane
(50 mL) was then added and the organic phase was washed with
water (20 mL). The organic phase was dried and concentrated to
give the crude product. This was purified by flash chromatography
with a 8:2 hexanes/ethyl acetate mixture as eluant. Product 3H
(200 mg), a pale yellow solid with Rf 0.35, was isolated in 36%
yield. It had m.p. 100–102 ◦C. 1H-NMR (CDCl3): d = 7.30–7.43
(m, 5H), 4.64 (s, 2H), 4.55 (s, 2H), 2.81 (q, J = 7.5 Hz, 2H), 1.14
(t, J = 7.5 Hz, 3H). 13C-NMR (CDCl3): d = 145.2, 138.4, 131.9,
128.3, 128.1, 127.6, 73.0, 66.4, 22.8, 16.4. Elemental analysis:
C36H42O3 requires: C, 82.72; H, 8.10; found: C, 82.58; H, 8.19%.
HPLC experiments. Silica 8 was packed in a 125 ¥ 4.6 mm i.d.
glass-lined stainless steel column by the slurry procedure (MeOH,
140 atm); efficiency test gave N/m = 4000 for 1,3-dinitrobenzene
eluting with a 9:1 hexane/chloroform mixture at a flow rate of
0.2 mL/min at 25 ◦C. The retention times observed using a
commercial RPC18 column (C18 Resolve, Waters, 125 ¥ 4.6 mm;
N/m = 30000) were obtained with a flow rate of 1.0 mL/min
at 25 ◦C and the eluants indicated in Table 1. The retention times
observed with the column packed with silica 8 were obtained under
the same conditions. The results were reported in Table 1.
Acknowledgements
We are grateful to Prof. G. Moneti of the Centro Interdipartimen-
tale di Spettrometria di Massa (CISM) of the Universia` di Firenze
for the ESI-MS spectra.
References
Synthesis of 1,3,5-tris-(pentafluorophenyl)-methoxymethyl-
1 M. Harmata, Acc. Chem. Res., 2004, 37, 862–873.
2,4,6-triethylbenzene
3F. To
a
stirred
solution
of
2 F.-G. Kla¨rner and B. Kahlert, Acc. Chem. Res., 2003, 36, 919–932.
3 L. C. Palmer and J. Rebek, Jr., Org. Biomol. Chem., 2004, 2, 3051–3059.
4 For recent examples of aromatic receptors for fullerenes featuring “flat”
aromatic binding sites see: (a) Y. Matsuo, K. Morita and E. Nakamura,
Chem. Asian J., 2008, 3, 1350–1357; (b) L. H. Tong, J.-L. Wietor, W.
Clegg, P. R. Raithby, S. I. Pascu and J. K. M. Sanders, Chem. Eur. J.,
2008, 14, 3035–3044.
compound 5 (200 mg, 0.8 mmol) in dry dichloromethane
(10 mL) kept under nitrogen at room temperature,
pentafluorophenylmethoxytrimethylsilane (654 mg, 2.42 mmol)
in dichloromethane (5 mL) was added followed by triethylsilane
(0.4 mL, 2.5 mmol). The mixture was then cooled at 0 ◦C
and trimethylsilyl triflate (0.06 mL, 0.32 mmol) was added.
The reaction mixture was slowly allowed to warm up to room
temperature and was then stirred for 2 h. Dichloromethane
(50 mL) was then added and the organic phase was washed
5 For recent reviews, see: (a) T. Kawase and H. Kurata, Chem. Rev., 2006,
106, 5250–5273; (b) E. M. Perez and N. Martin, Chem. Soc. Rev., 2008,
37, 1512–1519.
6 T. Kawase, K. Tanaka, N. Shiono, Y. Seirai and M. Oda, Angew. Chem.,
Int. Ed., 2004, 43, 1722–1724; and references cited therein.
3876 | Org. Biomol. Chem., 2009, 7, 3871–3877
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
7 (a) A. Sygula, F. R. Fronczek, R. Sygula, P. W. Rabideau and M. M.
Olmstead, J. Am. Chem. Soc., 2007, 129, 3842–3843; (b) W. Xiao, D.
Passerone, P. Ruffieux, K. A¨ıt-Mansour, O. Gro¨ning, E. Tosatti, J. S.
Siegel and R. Fasel, J. Am. Chem. Soc., 2008, 130, 4767–4771.
8 (a) E. M. Perez, M. Sierra, L. Sanchez, M. R. Torres, R. Viruela, P. M.
Viruela, E. Ort`ı and N. Martin, Angew. Chem., Int. Ed., 2007, 46, 1847–
1851; (b) E. M. Perez, A. L. Capodilupo, G. Fernandez, L. Sanchez,
P. M. Viruela, R. Viruela, E. Ort`ı, M. Bietti and N. Martin, Chem.
Commun., 2008, 4567–4569; (c) J. Santos, B. Grimm, B. M. Illescas,
D. M. Guldi and N. Martin, Chem. Commun., 2008, 5993–5995; (d) S.
Shankara Gayatri, M. Wielopolski, E. M. Perez, G. Fernandez, L.
Sanchez, R. Viruela, E. Ort`ı, D. M. Guldi and N. Martin, Angew.
Chem., Int. Ed., 2009, 48, 815–819.
9 P. E. Georghiou, L. N. Dawe, H.-A. Tran, J. Stru¨be, B. Neumann, H.-G.
Stammler and D. Kuck, J. Org. Chem., 2008, 73, 9040–9047.
10 For discussions on the electronic properties of aromatic molecules
with convex-concave topology see: (a) M. Kamieth, F.-G. Kla¨rner
and F. Diederich, Angew. Chem., Int. Ed., 1998, 37, 3303–3306. For
applications to supramolecular fullerene chemistry see: (b) F. Diederich
and M. Gomez-Lopez, Chem. Soc. Rev., 1999, 28, 263–277. For
supramolecular (C60-Ih)[5,6]- fullerene chemistry on surfaces see: (c) D.
Bonifazi, O. Enger and F. Diederich, Chem. Soc. Rev., 2007, 36, 390–
414.
11 For a review describing early work, see: (a) F. Cozzi and J. S. Siegel, Pure
Appl. Chem., 1995, 67, 683–689. For recent work, see: (b) F. Cozzi, R.
Annunziata, M. Benaglia, M. Cinquini, L. Raimondi, K. K. Baldridge
and J. S. Siegel, Org. Biomol. Chem., 2003, 1, 157–162; (c) F. Cozzi,
R. Annunziata, M. Benaglia, K. K. Baldridge, G. Aguirre, J. Estrada,
Y. Sritana-Anant and J. S. Siegel, Phys. Chem. Chem. Phys., 2008,
10, 2686–2694; (d) F. Cozzi, F. Ponzini, R. Annunziata, M. Cinquini
and J. S. Siegel, Angew. Chem., Int. Ed. Engl., 1995, 34, 1019–1020;
(e) S. Bacchi, M. Benaglia, F. Cozzi, F. Demartin, G. Filippini and A.
Gavezzotti, Chem. Eur. J., 2006, 12, 3538–3546; (f) F. Cozzi, S. Bacchi,
G. Filippini, T. Pilati and A. Gavezzotti, Chem. Eur. J., 2007, 13, 7177–
7184.
12 (a) A. Vacca, C. Nativi, M. Cacciarini, R. Pergoli and S. Roelens, J. Am.
Chem. Soc., 2004, 126, 16456–16465; (b) O. Francesconi, A. Ienco, G.
Moneti, C. Nativi and S. Roelens, Angew. Chem., Int. Ed., 2006, 45,
6693–6696; (c) C. Nativi, M. Cacciarini, O. Francesconi, A. Vacca,
G. Moneti, A. Ienco and S. Roelens, J. Am. Chem. Soc., 2007, 129,
4377–4385; (d) M. Cacciarini, E. Cordiano, C. Nativi and S. Roelens,
J. Org. Chem., 2007, 72, 3933–3936; (e) C. Nativi, M. Cacciarini, O.
Francesconi, G. Moneti and S. Roelens, Org. Lett., 2007, 9, 4685–4688.
13 (a) G. Hennrich and E. V. Anslyn, Chem. Eur. J., 2002, 8, 2218–2224;
and references cited therein; (b) A. T. Wright, M. J. Griffin, Z. Zhong,
S. C. McCleskey, E. V. Anslyn and J. T. McDevitt, Angew. Chem., Int.
Ed., 2005, 44, 6375–6378; (c) G. Haberhauer, T. Oeser and F. Rominger,
Chem. Eur. J., 2005, 11, 6718–6726.
19 A. J. Lampkins, O. Abdul-Rahim and R. K. Castellano, J. Org. Chem.,
2006, 71, 5815–5818.
20 T. Mizutani, K. Wada and S. Kitagawa, J. Am. Chem. Soc., 1999, 121,
11425–11431.
21 S. Simaan, J. S. Siegel and S. E. Biali, J. Org. Chem., 2003, 68, 3699–
3701.
22 P. E. Gordon and A. J. Fry, Tetrahedron Lett., 2001, 42, 831–833.
23 P. Sarri, F. Venturi, F. Cuda and S. Roelens, J. Org. Chem., 2004, 69,
3654–3661.
24 (a) S. C. Zimmerman, K. W. Saionz and Z. Zeng, Proc. Natl. Acad. Sci.
U. S. A., 1993, 90, 1190–1193; (b) S. C. Zimmerman and K. W. Saionz,
J. Am. Chem. Soc., 1995, 117, 1175–1176.
25 (a) A. Bianco, F. Gasparrini, M. Maggini, D. Misiti, A. Polese, M.
Prato, G. Scorrano, C. Toniolo and C. Villani, J. Am. Chem. Soc.,
1997, 119, 7550–7554; (b) F. Gasparrini, D. Misiti, F. Della Negra,
M. Maggini, G. Scorrano and C. Villani, Tetrahedron, 2001, 57, 6997–
7002.
26 A referee suggested the possibility that the better binding provided
by receptors 2 and 4 could be due to a more rigid disposition
of the side-arms resulting from the presence of six relatively bulky
substituents. On the basis of previous14c and more recent work
(G. Hennrich, V. M. Lynch, E. V. Anslyn, Chem. Eur. J., 2002, 8, 2274–
2279) this hypothesis can be ruled out, because the rotation around the
bond between the substituent and the benzene platform in compounds
very similar to 1–4 can be slowed down only at very low temperature
(< -80 ◦C).
27 We believe that an interpretation based on the availability of the
aromatic p-electrons is more realistic than one based on the electron
“richness” or “poorness” of the aromatic residues of the receptors.
Indeed, it seems difficult to consider a perfluorophenyl residue pos-
sessing six p-electrons an electron poor moiety; rather, the availability
of p-electrons is being limited by the presence of the fluorine atoms.
We think that the p-electrons’ availability is a useful concept to
describe aromatic–aromatic interactions. For reviews on the subject,
see: (a) C. A. Hunter, K. R. Lawson, J. Perkins and C. J. Urch,
J. Chem. Soc., Perkin Trans. 2, 2001, 651–669; (b) E. A. Meyer, R. K.
Castellano and F. Diederich, Angew. Chem., Int. Ed., 2003, 42, 1210–
1250; (c) S. L. Cockroft, J. Perkins, C. Zonta, H. Adams, S. E. Spey,
C. M. R. Low, J. G. Vinter, K. R. Lawson, C. J. Urch and C. A. Hunter,
Org. Biomol. Chem., 2007, 5, 1062–1080. In this context, according to
the Oxford Dictionary of English, which defines generous as “free to
give” and avaricious as “eager to keep”, the expressions “electron rich”
and “electron poor” may be more appropriately replaced by “electron
generous” and “electron avaricious”. As a matter of fact, several recent
publications dealing with the interaction between ions and p-systems
emphasized the importance of electron availability of the latter. For a
review, see: B. L. Schottel, H. T. Chifotides and K. R. Dunbar, Chem.
Soc. Rev., 2008, 37, 68–83.
14 (a) D. J. Iverson, G. Hunter, J. F. Blount, J. R. Damewood and K.
Mislow, J. Am. Chem. Soc., 1981, 103, 6073–6083; (b) K. V. Kilway and
J. S. Siegel, J. Am. Chem. Soc., 1991, 113, 2332–2333; (c) K. V. Kilway
and J. S. Siegel, J. Am. Chem. Soc., 1992, 114, 255–261.
15 (a) C. Walsdorff, W. Saak and S. Pohl, J. Chem. Res. (M), 1996, 1601–
1609; (b) A. Metzger, V. M. Lynch and E. V. Anslyn, Angew. Chem.,
Int. Ed. Engl., 1997, 36, 862–864; (c) K. V. Kilway and J. S. Siegel,
Tetrahedron, 2001, 57, 3615–3627.
16 A. L. Balch, V. J. Catalano, J. W. Lee and M. M. Olmstead, J. Am.
Chem. Soc., 1992, 114, 5455–5457.
17 C. S. Frampton, J. H. Gall and D. D. MacNicol, Acta Crystallogr.,
2000, c56, e22.
18 S. Hecht and J. M. J. Frechet, J. Am. Chem. Soc., 1999, 121, 4084–4085.
28 Z. Yoshida, H. Takekuma, S. Takekuma and Y. Matsubara, Angew.
Chem., Int. Ed. Engl., 1994, 33, 1597–1599.
29 For a recent discussion on the interaction between oxygen lone pairs
and aromatic rings see: B. W. Gung, Y. Zou, Z. Xu, J. C. Amicangelo,
D. G. Irwin, S. Ma and H.-C. Zhou, J. Org. Chem., 2008, 73, 689–693
and references cited therein.
30 The choice of this compound instead of receptor 4, which should be
a better fullerene binder on the basis of the HPLC data, rested on
the fact that 3H could only give a 1:1 adduct with fullerene, whereas
ditopic receptor 4 could give both 1:1 and 1:2 adducts, thus potentially
complicating the NMR study.
31 A. Hosseini, S. Taylor, G. Accorsi, N. Armaroli, C. A. Reed and P. D. W.
Boyd, J. Am. Chem. Soc., 2006, 128, 15903–15913.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3871–3877 | 3877
©