Incarceration of Higher-Order Fullerenes within Cyclotriveratrylene-Based Hemicarcerands Allows Selective Isolation of C76, C78, and C84from a Commercial Fullerene Mixture

Article abstract of DOI:10.1002/chem.201603451

Size-complementary cyclotriveratrylene (CTV)-based hosts can incarcerate C₇₆, C₇₈, and C₈₄, thus allowing the selective isolation of these higher-order fullerenes from a commercially available mixture of fullerenes. The hemicarceplexes, formed after the encapsulation of the size-complementary fullerenes within the hosts, are isolated by column chromatography and released at elevated temperature, thereby leading to the isolation of C₇₆/C₇₈and C₈₄in good purities (up to 95 and 88 %, respectively).

Full text of DOI:10.1002/chem.201603451

DOI: 10.1002/chem.201603451
Full Paper
&
Host–Guest Systems |Hot Paper|
Incarceration of Higher-Order Fullerenes within
Cyclotriveratrylene-Based Hemicarcerands Allows Selective
Isolation of C76, C78, and C84 from a Commercial Fullerene Mixture
Kuang-Shun Liu,[a] Ming-Jhe Li,[a] Chien-Chen Lai,[b] and Sheng-Hsien Chiu*[a]
Abstract: Size-complementary cyclotriveratrylene (CTV)-
based hosts can incarcerate C76, C78, and C84, thus allowing
the selective isolation of these higher-order fullerenes from
a commercially available mixture of fullerenes. The hemicar-
ceplexes, formed after the encapsulation of the size-comple-
mentary fullerenes within the hosts, are isolated by column
chromatography and released at elevated temperature,
thereby leading to the isolation of C76/C78 and C84 in good
purities (up to 95 and 88%, respectively).
Introduction
higher-order fullerenes in commercial mixtures are similar in
terms of abundances and sizes (compare C70, C76, C78, C82, C84,
and C86; Figure 1a),[6] thus making their separation through
host–guest approaches even more challenging and, therefore
requiring subtle changes in the host structures to afford rea-
sonable selectivities. Herein, we report kinetic and thermody-
namic studies of the incarceration and release of hemicarce-
plexes formed from size-complementary higher-order ful-
lerenes[7] and cyclotriveratrylene (CTV)-based hosts[8] and the
application of this approach to the selective isolation of C76/C78
and C84 from a commercially available mixture of higher-order
fullerenes. The higher-order fullerenes C76/C78 and C84 isolated
by using this approach had reasonably good purities (up to 95
and 88%, respectively), which simplified their subsequent
HPLC separations and improved their purities further (up to
ꢀ99.5%).
Since the discovery of buckminsterfullerene (C60) in 1985, vari-
ous fullerenes have found applications in photovoltaics, semi-
conductors, and biopharmacology.[1] Although low solubility in
common organic solvents makes it difficult to purify fullerenes,
at present we have access to relatively large supplies of high-
purity C60 and C70, the most abundant fullerenes in extracts of
carbon soot obtained from the combustion of graphite
through the arc discharge method as a result of the develop-
ment of several ingenious separation methods.[2] Nevertheless,
the higher-order fullerenes C76, C78, and C84, which have unique
stereochemical and electrochemical properties relative to C60
and C70, remain difficult to access, even though these species
are the most abundant in commercially available mixtures of
higher-order fullerenes obtained after the extraction of C60 and
C70. The purification of these higher-order fullerenes relies
mainly on multiple cycles of HPLC, thus limiting their supply in
high-purity (ꢀ98%) forms and significantly elevating the
prices (up to ca. US$360 mgÀ1).[3] Host–guest chemistry has
become a reliable means of selectively separating C60 and
C70,[2c,4] but has rarely been applied to the purification of
higher-order fullerenes.[5] The pioneering host–guest systems
for the separation of C60 and C70 from their mixtures have
relied mainly on size differences; unfortunately, many of the
Results and Discussion
Previously, we used molecular cage 1 (Figure 1d) to isolate C70
in high purity (ꢀ99.0%) directly from a fullerene extract,[9] thus
taking advantage of room-temperature-isolable hemicarceplex-
es (formed from the sequestration of complementary guests
within hemicarcerands) and releasing the guests at elevated
temperatures. Because the higher-order fullerenes are not
much bigger than C70, their selective complexation of these
compounds would require host molecules with only slightly
larger cavities and openings. Thus, we anticipated that hosts 2
and 3, with a greater numbers of bridging methylene units,
might be applicable for the selective isolation of C76, C78, and
C84 from a commercial mixture of higher-order fullerenes.
We synthesized (Scheme 1) molecular cages 2 and 3 in
seven steps from 3,4-dihydroxybenzaldehyde (4). Selective
monoalkylation of 4 at its most acidic phenolic group (i.e., the
para position) with 1,12-dibromododecane gave aldehyde 5,
which we subjected to NaBH4-mediated reduction of the
formyl group to give bromide 6. Another selective alkylation of
[a] K.-S. Liu, Dr. M.-J. Li, Prof. S.-H. Chiu
Department of Chemistry and
Center for Emerging Material and Advanced Devices
National Taiwan University
No. 1, Sec. 4, Roosevelt Road, Taipei, Taiwan, 10617 (R.O.C.)
Fax: (+886)2-33661677
E-mail: shchiu@ntu.edu.tw
[b] Prof. C.-C. Lai
Institute of Molecular Biology, National Chung Hsing University and
Department of Medical Genetics, China Medical University Hospital
Taichung, Taiwan (R.O.C.)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201603451.
Chem. Eur. J. 2016, 22, 1 – 10
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
chains, 2 and 3 are quite soluble in less-polar organic solvents
(e.g., CH2Cl2, CHCl3, and CHCl2CHCl2), but not very soluble in
more-polar organic solvents (e.g., CH3CN and THF). Thus, we
expected that the complexation of higher fullerenes within
these two hosts would significantly increase the solubility of
the fullerenes in less-polar solvents.
Figure 1. a–c) HPLC profiles of a) the commercially available mixture of
higher-order fullerenes and b,c) the residue mixtures after the first (b) and
second (c) extraction cycles of C76/C78 and C84. d) Chemical structures of mo-
lecular cages 1–3.
Figure 2. 1H NMR spectra (400 MHz, CDCl2CDCl2, 298 K) of a) the free host 2;
b,d) mixtures of host 2 (2.25 mm) and C76 (1.5 mm; b) and C78 (1.5 mm; d)
after sitting at 298 K for 48 h; c,e) the purified hemicarceplexes C76@2 (c)
and C78@2 (e).
1
The H NMR spectrum of a solution of host 2 (2.25 mm) and
C76 (1.5 mm) in CDCl2CDCl2 displayed a single new set of sig-
nals that correspond to the formation of the hemicarceplex
C76@2 after the mixture had been left at 298 K for 48 hours
(Figure 2). The stability of C76@2 was confirmed by isolation
with column chromatography (SiO2); there were no noticeable
1
signals that correspond to the free host 2 in the H NMR spec-
1
trum (Figure 2 c). In contrast, the H NMR spectrum of a solu-
tion of 2 (2.25 mm) and C78 (1.5 mm) displayed three sets of
new signals after the mixture had been left at 298 K for
48 hours (Figure 2d). It has been demonstrated previously that
a CTV-based molecular cage can distinguish between the two
structural isomers of the metallofullerene Sc3N@C80, thus re-
porting the differences in their incarcerated forms by using
1H NMR spectroscopic analysis.[11] Because three major structur-
al isomers (i.e., D3(3), C2v(4), C2v(5)) are generally present in sam-
ples of C78,[12] we suspected that the three new sets of signals
arose from the incarceration of these three major structural
isomers of C78 within 2. The observation of an intense signal at
m/z 2707.3 that corresponds to the [C78@2]+ ion in the mass
spectrum of the chromatographically isolated C78@2 and an
Scheme 1. Synthesis of molecular cages 2 and 3.
4 with bromide 6 afforded triol 7, which we treated with dibro-
mide 8 (or 9) to produce the macrocycle 10 (or 11). We ob-
tained a trialdehyde after the Sc(OTf)3-catalyzed condensation
of three units of macrocycle 10 (or 11) into the first CTV
unit.[10] NaBH4-mediated reduction afforded triol 12 (or 13),
which condensed under acidic conditions to form the second
CTV unit, thereby affording the molecular cage 2 (or 3). The
overall yields for the syntheses of hosts 2 and 3 were 0.2 and
0.3%, respectively. Because these cages present six long alkyl
1
isomeric distribution ratio in its H NMR spectrum (Figure 2e)
similar to that of the original mixture (Figure 2d) supported
the notion that C78@2 was indeed a mixture of three different
complexes. In contrast, a solution of 2 (2.25 mm) and C84
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
(1.5 mm) in CDCl2CDCl2 gave no noticeable signals for the cor-
responding hemicarceplex C84@2 in its H NMR spectrum after
rium had been reached, afforded association constants of Ka =
6800 and 9400mÀ1, respectively, for these hemicarceplexes.[14]
By monitoring the changes of the integration ratios of the free
host 2 and the hemicarceplexes, we determined the dissocia-
tion half-lives and rate constants of the hemicarceplexes at
298 K to be t1/2 =16.9 h and kd =1.1ꢁ10À5 sÀ1, respectively, for
C76@2 and t1/2 =33.6 h and kd =5.7ꢁ10À6 sÀ1, respectively, for
C78@2 (Figure 3). Therefore, the calculated association rate con-
stants of host 2 to C76 and C78 were ka =7.5ꢁ10À2 and 5.4ꢁ
1
sitting at 298 K for 48 hours, thus suggesting that squeezing
the relatively large C84 through the openings of host 2 requires
more energy than the passage of the relatively small C76 and
C78. Indeed, this process did not occur at a measurable rate
under these conditions.
Quan and Cram defined the activation energy required for
a guest to enter the cavity of a hemicarcerand and the free
energy for the formation of a hemicarceplex as the intrinsic
and constrictive binding energies, respectively.[13a] The use of
1H NMR spectroscopic analysis to monitor the dissociation of
purified C76@2 and C78@2 in CDCl2CDCl2 at 298 K, until equilib-
10À2 mÀ1 À1, respectively. Based on these numbers, we estimat-
s
ed the intrinsic and constrictive binding energies of the hemi-
carceplexes at 298 K to be E=5.2 and 19.0 kcalmolÀ1, respec-
tively, for C76@2 and E=5.4 and 19.2 kcalmolÀ1, respectively,
for C78@2. As expected, the smaller C76 molecule can squeeze
through the openings of host 2 more readily than C78, but the
larger C78 molecule can fill the cavity of the host better, thus
making C78@2 more stable than C76@2, as revealed by the
slower dissociation kinetics and larger association constant Ka
of the former hemicarceplex.
After confirming that molecular cage 2 can form hemicarce-
plexes selectively with C76 and C78, but not C84, at 298 K, we
used 2 (450 mg) to directly extract C76 and C78 from a mixture
of higher-order fullerenes (600 mg) in CHCl2CHCl2 (30 mL).
Figure 4 presents the flow chart of the isolation process. After
heating the mixture at 308 K for 40 hours to incarcerate the
fullerenes, we evaporated the solvent under reduced pressure,
suspended the solid residue in CH2Cl2, and filtered the suspen-
sion to recycle the insoluble higher-order fullerenes (“solid A”).
We concentrated the filtrate and purified it chromatographical-
ly to afford the free host 2, free higher-order fullerenes (“solid
B”), and a mixture of hemicarceplexes. We dissolved the free
molecular cage 2 from the chromatography step with solids A
Figure 3. Constrictive and intrinsic binding energies for the complexation of
C76 and C78 with host 2 in CDCl2CDCl2 at 298 K.
Figure 4. Flow chart for the isolation of C76/C78 and C84 from a commercial mixture of higher-order fullerenes. This sequence was performed twice, with the
“final solid C+D” used as the starting “higher-order fullerenes” (top left corner) the second time around.
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
and B in CHCl2CHCl2, and subjected this mixture to another
cycle of the extraction process. After three repeated extraction
cycles, we dissolved the combined hemicarceplexes (316 mg)
in CHCl3 and left the mixture at 298 K for 4 hours to release
the incarcerated fullerenes. After flash column chromatography
(SiO2), we isolated hemicarceplexes (177 mg), higher-order ful-
lerenes (40 mg), and 2 (65 mg). The HPLC trace of the ful-
lerenes indicated that 2 was quite efficient at incarcerating the
fullerenes C76 and C78 from the mixture because these species
were predominant among all the released fullerenes, with
a combined purity of approximately 94% (Figure 5a). The elec-
trospray ionization (ESI) mass spectrum of the hemicarceplex
mixture after extracting the higher-order fullerenes mixture
with 2 (solid B) revealed signals for hemicarceplexes C82@2 and
C84@2 in addition to signals for C76@2 and C78@2; thus, the in-
carceration of C84 by 2 proceeded at a reasonable rate at
308 K. Thus, the isolation of C76/C78 in high purity from the
higher-order fullerene mixture was presumably the result of
relatively slow association and dissociation kinetics for the in-
carceration and release of the relatively large C82 and C84 mole-
cules from their corresponding hemicarceplexes.
cule remained as the predominant species (ca. 64%) in the
second (Figure 5b). The difference in the dissociation kinetics
of the two hemicarceplexes appears to be too small for separa-
tion of these two fullerenes through selective release of the
less-sizable C76 from 2. The combined purity of C76 and C78 did
not, however, vary much between the first and second rounds
of the guest being released (ca. 94 and 95%, respectively),
thus suggesting reasonably good selectivity for host 2 toward
C76 and C78 during guest incarceration and release among all
of the higher-order fullerenes in the extract. The presence of
significant amounts of C76 and C78 in the released higher-order
fullerenes mixture suggests that these two fullerenes are so
similar in size that any differences in their complexation kinet-
ics to 2 are too insignificant to allow selective encapsulation or
release of one of them from the higher-order fullerene or hem-
icarceplex mixture, respectively.
Figure 6. 1H NMR spectra (400 MHz, CDCl2CDCl2, 298 K) of a) the free host 3;
b) a mixture of 3 (2.25 mm) and C84 (1.5 mm) after sitting at 298 K for 48 h;
c) purified C84@3.
1
In contrast to host 2, the H NMR spectrum of a solution of
Figure 5. HPLC profiles of the solids obtained after the first (a) and second
(b) guest-release processes from the hemicarceplex fullerene@2.
host 3 (2.25 mm) and C84 (1.5 mm) in CDCl2CDCl2 displayed
new sets of signals for hemicarceplex C84@3 after sitting at
298 K for 48 hour (Figure 6). Similar to hemicarceplex C78@2,
the encapsulation of many structural isomers of C84 by 3 was
reflected in the presence of several overlapping aromatic sig-
nals from the host. The appearance of two major, yet unequal,
signals at d=6.71 and 6.72 ppm and several minor signals for
the aromatic protons of the CTV units is consistent with C84
containing two major isomers (i.e., D2(IV) and D2d(II)) and sever-
al minor ones.[16] By using 1H NMR spectroscopic analysis to
monitor the dissociation kinetics of the chromatographically
purified hemicarceplex C84@3 in CDCl2CDCl2 at 298 K, until
equilibrium had been reached, we determined the association
constant and dissociation rate constant of the hemicarceplex
Next, we dissolved the mixture of hemicarceplexes (177 mg)
in THF and left the solution at 298 K for 16 hours to release
the incarcerated higher-order fullerene guests again. We then
centrifuged the dark suspension to precipitate both the free
fullerenes and the molecular cage 2. After dissolving the black
residue in CS2, we subjected the solution to flash chromatogra-
phy through a short column to separate the free fullerenes
(55 mg) from the free host 2 (83 mg).[15] A significant amount
(ca. 20%) of the larger C78 molecule was present in the fuller-
ene mixture released in the first run, yet the smaller C76 mole-
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
ing its components were similar to those procedures described
above for the extraction of C76/C78. The higher-order fullerene
C84 (80 mg) obtained from dissociation of hemicarceplex C84@3
(282 mg), collected from two repeated extraction cycles (via
“solid C” and “solid D”) and precipitation from THF, had
a purity of 88%, based on HPLC analysis (Figure 8).[17] The pres-
ence of considerable amounts of C82 and C86 in this isolated
sample of C84 suggests that host 3 could not differentiate
these three fullerenes efficiently.
HPLC analysis (Figure 1b) of the solid mixture of higher-
order fullerenes (“final solid C+D”) obtained after the sequen-
tial extractions of C76/C78 (with 2) and C84 (with 3) revealed sig-
nificant decreases in the contents of these three fullerenes by
using the signal of C70 as a reference. Nevertheless, these three
fullerenes remained the predominant species in the recycled
fullerene mixture; accordingly, we applied recycled hosts 2 and
3 to a second extraction cycle of this fullerene mixture. As ex-
pected, the purities of the C76/C78 and C84 precipitates obtained
from this second extraction cycle decreased (to 80 and 81%,
respectively) due to the lower molar fractions of these ful-
lerenes in the mixture. HPLC analysis (Figure 1c) of the solid
(“final solid C+D”) obtained in the second extraction cycle in-
dicated that C70 was the predominant species in the mixture
after the two extraction cycles. The presence of large signals
for C70 and C82O in this HPLC trace confirmed that large
amounts of C76, C78, and C84 had been removed from the com-
mercial mixture of higher-order fullerenes.
Figure 7. Constrictive and intrinsic binding energies for the complexation of
C84 with host 3 in CDCl2CDCl2 at 298 K.
C84@3 to be Ka =6500mÀ1[14] and kd =3.5ꢁ10À6 sÀ1, respectively.
These values correspond to a dissociation half-life and free
energy of
(Figure 7). Accordingly, the association rate constant and free
energy for the host incarcerating C84 were k=2.3ꢁ
t , respectively
1/2 =55.5 h and E=24.9 kcalmolÀ1
3
The use of the CTV-based hosts 2 (450 mg) and 3 (494 mg)
to form hemicarceplexes with suitably sized higher-order ful-
lerenes allowed us to isolate C76/C78 (total=126 mg) and C84
(total=112 mg) directly in purities of 80–95 and 81–88%, re-
spectively, from a commercially available mixture of higher-
order fullerenes (600 mg). Based on the HPLC profile of the
commercial mixture of higher-order fullerenes, the total con-
tents of C76/C78 and C84 were approximately 251 and 205 mg,
respectively. Therefore, the two extraction cycles allowed us to
isolate approximately 46 and 47% of the C76/C78 and C84, re-
spectively, from the commercial mixture of higher-order ful-
lerenes.
10À2 mÀ1 sÀ1 and E=19.7 kcalmolÀ1, respectively. Because the
complexes formed from relatively small C76 and C78 molecules
with host 3 in CDCl2CDCl2 were insufficiently stable to allow
isolation of their hemicarceplexes at room temperature, we ex-
pected significant dissociation of these complexes to occur
during chromatography when isolating hemicarceplex C84@3. If
so, we could use 3 to selectively isolate C84 from the higher-
order fullerene mixture.
Because the higher-order fullerenes obtained from our
simple extraction processes and column chromatography were
in good purities, there is no need to perform recycling HPLC
for further purification. Thus, by performing a single standard
HPLC separation (Cosmosil 5 m PBB, semipreparative, 10.0ꢁ
250 mm; sample injection volume=2.5 mL) using toluene (a
relatively poor solvent for higher-order fullerenes) as the
eluent, we obtained 16 and 8 mg of C76 and C78 (84 and 89%
separation yields), respectively, in high purity (ꢀ99.5%; Fig-
ure 9a,b) from 30 mg of the fullerene solid obtained from the
second guest-release process of hemicarceplex fullerene@2
(C76: 64.1, C78: 31.5%; Figure 5b). Similarly, 7 mg (78% separa-
tion yield) of high-purity C84 ꢀ99.5%; Figure 9c) was obtained
from the same HPLC separation from 10 mg of the fullerene
solid collected from the guest-release process of the hemicar-
ceplex fullerenes@3 (C84: 88%; Figure 8). The use of a prepara-
tive-scale HPLC column, a larger injection loop, and/or an
eluent more suitable for fullerenes would presumably increase
Figure 8. HPLC profile of the solid obtained after applying the guest-release
process to the fullerene@3 hemicarceplexes.
Because the selective extraction of C76 and C78 increased the
molar fraction of C84 in the recycled mixture of higher-order
fullerenes (“final solid A+B”; 494 mg), we attempted to extract
C84 from this mixture by treatment with the relatively large
molecular cage 3 (494 mg) in CHCl2CHCl2 at 303 K for 16 hours.
The processes for isolating hemicarceplex C84@3 and dissociat-
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Bromide 5:
A
mixture of 3,4-dihydroxybenzaldehyde (16.6 g,
120 mmol), KHCO3 (13.4 g, 134 mmol), and 1,12-dibromododecane
(39.4 g, 120 mmol) in acetone (1200 mL) was heated under reflux
for 40 h. After cooling to room temperature, the solvent was
evaporated under reduced pressure and the residue partitioned
between CH2Cl2 (3ꢁ200 mL) and H2O (600 mL). The combined or-
ganic phases were dried (MgSO4) and concentrated. The residue
was purified chromatographically on SiO2 (EtOAc/hexanes (1:10)) to
afford bromide 5 as a white solid (16.9 g, 37%). M.p. 78–798C;
1H NMR (400 MHz, CDCl3, 298 K): d=1.19–1.46 (m, 16H), 1.74–1.84
(m, 4H), 3.35 (t, J=6.8 Hz, 2H), 4.07 (t, J=6.8 Hz, 2H), 6.04 (s, 1H),
6.89 (d, J=8.4 Hz, 1H), 7.35 (dd, J=8.4, 2.0 Hz, 1H), 7.38 (d, J=
2.0 Hz, 1H), 9.77 ppm (s, 1H); 13C NMR (100 MHz, CDCl3, 298 K): d=
25.8, 28.0, 28.6, 28.8, 29.2, 29.2, 29.3, 32.7, 33.9, 69.2, 110.8, 113.9,
124.5, 130.2, 146.1, 151.3, 190.9 ppm (two signals were missing,
possibly because of signal overlap); HRMS (ESI): calcd for
C19H30BrO3+: m/z 385.1378; found: 385.1380 [M+H]+.
Bromide 6: NaBH4 (827 mg, 21.9 mmol) was added to a solution of
bromide 5 (16.9 g, 43.9 mmol) in MeOH (220 mL) and CH2Cl2
(220 mL) at 08C and then the mixture was stirred at room temper-
ature for 2 h. The organic solvents were evaporated under reduced
pressure and the residue partitioned between CH2Cl2 (3ꢁ200 mL)
and H2O (600 mL). The combined organic phases were washed
with brine (200 mL), dried (MgSO4), and concentrated to give bro-
mide 6 as a white solid (16.9 g, 99%), which was used directly in
the next step without further purification. M.p. 74–758C; 1H NMR
(400 MHz, CDCl3, 298 K): d=1.23–1.47 (m, 16H), 1.74–1.88 (m, 4H),
3.38 (t, J=6.8 Hz, 2H), 4.01 (t, J=6.8 Hz, 2H), 4.55 (d, J=5.2 Hz,
2H), 5.67 (s, 1H), 6.77–6.83 (m, 2H), 6.92 ppm (s, 1H); 13C NMR
(100 MHz, CDCl3, 298 K): d=26.0, 28.1, 28.7, 29.2, 29.3, 29.4, 29.5,
29.5, 32.8, 34.0, 65.1, 69.1, 111.6, 113.5, 118.8, 134.2, 145.5,
145.9 ppm (one signal was missing, possibly because of signal
overlap); HRMS (ESI): calcd for C19H30BrO2+: m/z 369.1429; found:
369.1436 [MÀOH]+.
Figure 9. HPLC profiles of high-purity samples of a) C76, b) C78, and c) C84, ob-
tained after subjecting the samples collected from the guest-release pro-
cesses of the hemicarceplex fullerene@2 (a,b) and fullerenes@3 (c) to a stan-
dard semipreparatory HPLC process.
the scale of this separation process; therefore, we believe that
our extraction method might have potential applicability in
the large-scale production of higher-order fullerenes.
Conclusion
Triol 7: A mixture of bromide 6 (16.9 g, 43.6 mmol), 3,4-dihydroxy-
benzaldehyde (6.64 g, 48.1 mmol), and KHCO3 (4.82 g, 48.1 mmol)
in DMF (440 mL) was stirred at 508C for 40 h and then the solvent
was evaporated under reduced pressure. The residue was washed
with H2O (400 mL) and Et2O (150 mL) to afford triol 7 as a pale-
brown solid (17.8 g, 92%). M.p. 130–1318C; 1H NMR (400 MHz,
CDCl3, 298 K): d=1.23–1.37 (m, 12H), 1.38–1.52 (m, 4H), 1.74–1.88
(m, 4H), 4.01 (t, J=6.4 Hz, 2H), 4.11 (t, J=6.4 Hz, 2H), 4.56 (s, 2H),
5.67 (br, 2H), 6.77–6.83 (m, 2H), 6.90–6.94 (m, 2H), 7.36–7.43 (m,
2H), 9.81ppm (s, 1H); 13C NMR (100 MHz, CDCl3, 298 K): d=25.9,
26.0, 29.0, 29.2, 29.3, 29.3, 29.5, 29.5, 65.2, 69.1, 69.3, 110.9, 111.6,
113.5, 114.1, 118.8, 124.4, 130.5, 134.2, 145.5, 145.9, 146.2, 151.3,
191.0 ppm (two signals were missing, possibly because of signal
overlap); HRMS (ESI): calcd for C26H35O6À: m/z 443.2434; found:
443.2445 [MÀH]À.
Macrocycle 10: A mixture of triol 7 (6.67 g, 15.0 mmol), 1,13-dibro-
motridecane (5.13 g, 15.0 mmol), and K2CO3 (16.6 g, 120 mmol) in
DMF (1500 mL) was stirred at 508C for 5 days, and then the solvent
was evaporated under reduced pressure. The residue was parti-
tioned between CHCl3 (4ꢁ250 mL) and H2O (400 mL). The com-
bined organic phases were dried (MgSO4) and concentrated. The
residue was purified chromatographically on SiO2 (CHCl3) to afford
macrocycle 10 as a white solid (1.86 g, 20%). M.p. 156–1578C;
1H NMR (400 MHz, CDCl3, 323 K): d=1.24–1.41 (m, 26H), 1.42–1.56
(m, 8H), 1.73–1.86 (m, 8H), 3.94–4.09 (m, 8H), 4.58 (d, J=5.6 Hz,
2H), 6.82–6.85 (m, 2H), 6.90–6.95 (m, 2H), 7.37–7.41 (m, 2H),
9.81 ppm (s, 1H); 13C NMR (100 MHz, CDCl3, 323 K): d=26.3, 26.3,
26.4, 26.4, 29.4, 29.4, 29.6, 29.6, 29.6, 29.7, 29.8, 29.8, 29.8, 29.9,
65.4, 69.3, 69.3, 69.5, 69.8, 111.7, 112.3, 113.5, 114.6, 119.7, 126.4,
The CTV-based host molecules 2 and 3 can form hemicarce-
plexes with higher-order fullerenes and, thus, can be used to
selectively isolate C76/C78 and C80, respectively, in high purities
(80–95 and 81–88%, respectively) from a commercially avail-
able mixture of higher-order fullerenes. By following simple ex-
traction/column chromatography processes, subsequent
straightforward HPLC separations could improve the purities
to ꢀ99.5%. Thus, host–guest chemistry remains a powerful
strategy for solving tough separation problems, as long as the
hosts are designed to suitably balance the complexation kinet-
ics and thermodynamics. The use of this approach should in-
crease the supplies of high-purity C76, C78, and C84, thus poten-
tially allowing the discovery of further interesting applications
for these higher-order fullerenes.
Experimental Section
General: All glassware, syringes, needles, and stirring bars were
either oven- or flame-dried prior to use. All reagents, unless other-
wise indicated, were obtained from commercial sources. Reactions
were conducted under N2 or Ar. Thin layer chromatography (TLC)
was performed on Merck 0.25 mm silica gel plates (Merck Art.
5715). Column chromatography was performed on Kieselgel 60
(Merck; 70–230 mesh). Melting points were determined by using
a Fargo MP-2D melting-point apparatus.
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
130.2, 134.1, 149.2, 149.9, 155.1, 190.8 ppm (eight signals were
missing, possibly because of signal overlap); HRMS (ESI): calcd for
C39H60O6+: m/z 624.4390; found: 624.4425 [M]+.
was washed with brine (20 mL) and concentrated. The residue was
dissolved in isopropyl alcohol (14 mL) and CH2Cl2 (42 mL) and then
NaBH4 (62 mg, 1.64 mmol) was added to the reaction mixture,
which was stirred at room temperature for 2 h. The organic sol-
vents were then evaporated under reduced pressure. The residue
was partitioned between CHCl3 (3ꢁ50 mL) and H2O (50 mL). The
combined organic phases were washed with brine (50 mL), dried
(MgSO4), and concentrated. The residue was purified chromato-
graphically on SiO2 (CH2Cl2/MeOH (99:1)) to afford a white solid
(475 mg, 21%). M.p. 133–1348C; 1H NMR (400 MHz, CDCl3, 298 K):
d=1.20–1.36 (m, 84H), 1.38–1.51 (m, 24H), 1.59 (t, J=5.6 Hz, 3H),
1.68–1.81 (m, 24H), 3.46 (d, J=13.6 Hz, 3H), 3.82–4.03 (m, 24H),
4.57 (d, J=5.6 Hz, 6H), 4.68 (d, J=13.6 Hz, 3H), 6.76–6.87 (m, 12H),
6.89 ppm (s, 3H); 13C NMR (100 MHz, CDCl3, 298 K): d=26.3, 26.4,
26.4, 29.6, 29.6, 29.6, 29.8, 29.8, 29.9, 36.4, 65.3, 69.2, 69.4, 69.6,
113.1, 114.0, 116.1, 116.1, 119.6, 132.3, 133.8, 147.9, 148.0, 148.9,
149.5 ppm (15 signals were missing, possibly because of signal
overlap); HRMS (ESI): calcd for C120H186O15Na: m/z 1890.3689;
found: 1890.3622 [M+Na]+.
Triol 12: A solution of macrocycle 10 (1.86 g, 2.98 mmol) and
Sc(OTf)3 (146 mg, 297 mmol) in CHCl3 (54 mL) and CH3NO2 (6 mL)
was heated under reflux for 16 h. After cooling to room tempera-
ture, water (30 mL) was added to the mixture. The organic phase
was washed with brine (30 mL) and concentrated. The residue was
dissolved in isopropyl alcohol (8 mL) and CH2Cl2 (24 mL) and then
NaBH4 (42 mg, 1.11 mmol) was added to the reaction mixture,
which was stirred at room temperature for 2 h. The organic sol-
vents were then evaporated under reduced pressure. The residue
was partitioned between CHCl3 (3ꢁ50 mL) and H2O (50 mL). The
combined organic phases were washed with brine (50 mL), dried
(MgSO4), and concentrated. The residue was purified chromato-
graphically on SiO2 (CH2Cl2/MeOH (99:1)) to afford triol 12 as
a white solid (350 mg, 19%). M.p. 114–1158C; 1H NMR (400 MHz,
CDCl3, 298 K): d=1.22–1.35 (m, 78H), 1.37–1.51 (m, 24H), 1.68–1.80
(m, 24H), 3.46 (d, J=13.6 Hz, 3H), 3.82–4.00 (m, 24H), 4.56 (s, 6H),
4.68 (d, J=13.6 Hz, 3H), 6.78–6.85 (m, 12H), 6.89 ppm (s, 3H);
13C NMR (100 MHz, CDCl3, 298 K): d=26.3, 26.4, 26.4, 29.6, 29.7,
29.8, 29.9, 30.0, 36.5, 65.4, 69.2, 69.6, 69.8, 113.0, 114.2, 115.9, 116.3,
119.6, 132.2, 132.3, 133.9, 147.9, 148.0, 148.9, 149.6 ppm (14 signals
were missing, possibly because of signal overlap) HRMS (ESI): calcd
for C117H180O15+: m/z 1825.3322; found: 1825.3373 [M]+.
Molecular cage 3: A solution of the triol 13 (475 mg, 254 mmol) in
CHCl3 (20 mL) was added to a solution of TFA (12.7 mL) in CHCl3
(200 mL) and CH3NO2 (50 mL) over a period of 12 h and then the
mixture was stirred at room temperature for 60 h. After the addi-
tion of saturated Na2CO3(aq) (100 mL), the organic phase was
washed with brine (100 mL), dried (MgSO4), and concentrated. The
residue was purified chromatographically on SiO2 (from CH2Cl2 to
CH2Cl2/MeOH (200:1)) to afford a white solid, which was recrystal-
lized (CH2Cl2/hexanes) to give 3 (73.7 mg, 16%). M.p. 248–2498C;
1H NMR (400 MHz, CDCl3, 298 K): d=1.14–1.47 (m, 108H), 1.60–1.93
(m, 24H), 3.46 (d, J=13.6 Hz, 6H), 3.76–3.90 (m, 12H), 3.96–4.08
(m, 12H), 4.68 (d, J=13.6 Hz, 6H), 6.77 (s, 6H), 6.82 ppm (s, 6H);
13C NMR (100 MHz, CDCl3, 298 K): d=25.9, 26.4, 29.0, 29.4, 29.8,
30.2, 30.5, 36.4, 68.2, 70.3, 114.4, 117.1, 131.7, 132.7, 147.3,
148.1 ppm (four signals were missing, possibly because of signal
overlap); HRMS (ESI): calcd for C120H180O12: m/z 1813.3475; found:
1813.3495 [M]+.
Molecular cage 2: A solution of triol 12 (296 mg, 162 mmol) in
CHCl3 (45 mL) was added to a solution of trifluoroacetic acid (TFA;
8.1 mL) in CHCl3 (85 mL) and CH3NO2 (32 mL) over a period of 12 h
and then the mixture was stirred at room temperature for 60 h.
After the addition of saturated Na2CO3(aq) (100 mL), the organic
phase was washed with brine (100 mL), dried (MgSO4), and concen-
trated. The residue was purified chromatographically on SiO2 (from
CH2Cl2 to CH2Cl2/MeOH (200:1)) to afford a white solid, which was
recrystallized (CH2Cl2/hexanes) to give 2 (29.5 mg, 16%). M.p. 267–
2688C; 1H NMR (400 MHz, CDCl3, 298 K): d=1.14–1.45 (m, 102H),
1.60–1.84 (m, 24H), 3.45 (d, J=13.6 Hz, 6H), 3.76–3.90 (m, 12H),
3.98–4.10 (m, 12H), 4.68 (d, J=13.6 Hz, 6H), 6.76 (s, 6H), 6.82 ppm
(s, 6H); 13C NMR (100 MHz, CDCl3, 298 K): d=26.4, 26.5, 29.1, 29.2,
29.5, 29.7, 29.9, 30.1, 30.5, 30.5, 36.4, 68.2, 70.0, 113.8, 116.5, 131.4,
Hemicarceplex C76@2: A solution of the molecular cage 2 (21 mg,
18 mmol) and C76 (11 mg, 12 mmol) in CHCl2CHCl2 (6 mL) was stirred
at 358C for 30 h and then the solvent was evaporated under re-
duced pressure. The solid residue was purified chromatographically
on SiO2 (CS2 then CH2Cl2/hexanes (4:1)) to afford hemicarceplex
C76@2 as a black solid (15 mg, 48%). 1H NMR (400 MHz, CDCl2CDCl2,
298 K): d=1.17–1.55 (m, 102H), 1.68–1.94 (m, 24H), 3.50 (d, J=
132.5, 147.1, 148.1 ppm (one signals was missing, possibly because
+
of signal overlap); HRMS (ESI): calcd for C117H174O12
z 1771.3005; found: 1771.3946 [M]+.
: m/
Macrocycle 11: A mixture of triol 7 (6.67 g, 15.0 mmol), 1,14-dibro-
motetradecane (5.34 g, 15.0 mmol), and K2CO3 (16.6 g, 120 mmol)
in DMF (1500 mL) was stirred at 508C for 5 days and then the sol-
vent was evaporated under reduced pressure. The residue was par-
titioned between CHCl3 (4ꢁ250 mL) and H2O (400 mL). The com-
bined organic phases were dried (MgSO4) and concentrated. The
residue was purified chromatographically on SiO2 (CH2Cl2/MeOH
(100:0.5)) to afford a white solid (2.27 g, 24%). M.p. 149–1508C;
1H NMR (400 MHz, CDCl3, 323 K): d=1.21–1.39 (m, 28H), 1.45–1.54
(m, 8H), 1.73–1.87 (m, 8H), 3.94–4.09 (m, 8H), 4.58 (d, J=6.0 Hz,
2H), 6.81–6.87 (m, 2H), 6.89–6.95 (m, 2H), 7.35–7.43 (m, 2H),
9.80 ppm (s, 1H); 13C NMR (100 MHz, CDCl3, 323 K): d=26.2, 26.3,
26.4, 29.2, 29.3, 29.4, 29.5, 29.6, 29.6, 29.7, 29.7, 29.8, 29.8, 29.8,
29.9, 65.4, 69.0, 69.2, 69.3, 69.4, 111.2, 111.8, 113.1, 114.0, 119.6,
126.6, 129.9, 133.8, 149.5, 149.6, 154.9, 191.0 ppm (eight signals
were missing, possibly because of signal overlap); HRMS (ESI):
calcd for C40H62O6Na: m/z 661.4444; found: 661.4420 [M+Na]+.
13.6 Hz, 6H), 3.66–3.82 (m, 12H), 3.95–4.10 (m, 12H), 4.74 (d, J=
+
13.6 Hz, 6H), 6.76 ppm (s, 12H); HRMS (ESI): calcd for C193H174O12
m/z 2683.3000; found: 2683.2757 [M]+.
:
Hemicarceplex C78@2: A solution of the molecular cage 2 (15 mg,
13 mmol) and C78 (8 mg, 9 mmol) in CHCl2CHCl2 (4 mL) was stirred
at 358C for 30 h and then the solvent was evaporated under re-
duced pressure. The solid residue was purified chromatographically
on SiO2 (CS2 then CH2Cl2/hexanes (4:1)) to afford hemicarceplex
C78@2 as a black solid (16 mg, 67%). 1H NMR (400 MHz, CDCl2CDCl2,
298 K): d=1.14–1.57 (m, 102H), 1.69–1.93 (m, 24H), 3.33 (d, J=
14.0 Hz, 1.3H), 3.42 (d, J=13.6 Hz, 3.1H), 3.56–3.84 (m, 13.6H),
3.91–4.10 (m, 12H), 4.57 (d, J=14.0 Hz, 1.3H), 4.66 (d, J=13.6 Hz,
3.1H), 4.84 (d, J=13.6 Hz, 1.6H), 6.56 (s, 2.6H), 6.66 (s, 6.2H),
+
6.87 ppm (s, 3.2H); HRMS (ESI): calcd for C195H174O12
z 2707.3000; found: 2707.3005 [M]+.
: m/
Hemicarceplex C84@3: A solution of the molecular cage 3 (16 mg,
9 mmol) and C84 (6 mg, 6 mmol) in CHCl2CHCl2 (3 mL) was stirred at
358C for 30 h and then the solvent was evaporated under reduced
pressure. The solid residue was purified chromatographically on
SiO2 (CS2 then CH2Cl2/hexanes (4:1)) to afford hemicarceplex C84@3
Triol 13: A solution of macrocycle 11 (2.27 g, 3.55 mmol) and
Sc(OTf)3 (87 mg, 0.177 mmol) in CHCl3 (35 mL) and CH3NO2 (3.5 mL)
was heated under reflux for 16 h. After cooling to room tempera-
ture, water (20 mL) was added to the mixture. The organic phase
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
as a black solid (4.0 mg, 24%). 1H NMR (400 MHz, CDCl2CDCl2,
298 K): d=1.21–1.61 (m, 108H), 1.68–1.92 (m, 24H), 3.47 (d, J=
13.6 Hz, 6H), 3.63–3.75 (m, 6H), 3.76–3.88 (m, 6H), 3.90–4.10 (m,
12H), 4.71 (d, J=13.6 Hz, 6H), 6.65–6.77 ppm (m, 12H); HRMS
(ESI): calcd for C204H180O12+: m/z 2821.3475; found: 2821.3499 [M]+.
MeOH (98:2)). The recovered higher-order fullerene mixture and
molecular cage 2 were mixed in CHCl2CHCl2 (10 mL) and then the
extraction process was repeated two more times. Combined, these
three repeated extractions provided a hemicarceplex mixture
(116 mg), free higher-order fullerenes (227 mg), and the free molec-
ular cage 2 (90 mg).
The hemicarceplex mixture (116 mg) was dissolved in CHCl3
(12 mL) and left at room temperature for 4 h to release the incar-
cerated higher-order fullerenes. The mixture was concentrated and
purified chromatographically on SiO2 (CS2, then CH2Cl2/hexanes
(4:1), followed by CH2Cl2/MeOH (98:2)) to afford a hemicarceplex
mixture (64 mg), free higher-order fullerenes (15 mg), and the free
molecular cage 2 (20 mg).
Typical extraction procedure
First extraction cycle: A black suspension of the molecular cage 2
(450 mg) and the higher-order fullerene mixture (600 mg) in
CHCl2CHCl2 (30 mL) was heated at 308 K for 40 h and then the sol-
vent was evaporated under reduced pressure. CH2Cl2 (40 mL) was
added to the residue and then the suspension was filtered. The fil-
trate was concentrated and the residue purified chromatographi-
cally on SiO2 (CS2, then CH2Cl2/hexanes (4:1), followed by CH2Cl2/
MeOH (98:2)). The recovered higher-order fullerene mixture and
the molecular cage 2 were mixed in CHCl2CHCl2 (10 mL); the ex-
traction process was repeated two more times. Combined, these
three repeated extractions provided a hemicarceplex mixture
(316 mg), free higher-order fullerenes (494 mg), and the free mo-
lecular cage 2 (198 mg).
The recovered hemicarceplex mixture (64 mg) was dissolved in
THF (6 mL) and then the solution was left at room temperature for
16 h to release the incarcerated higher-order fullerenes. The black
solid obtained after centrifugation of the suspension was purified
chromatographically on SiO2 (CS2, then CH2Cl2/hexanes (4:1), fol-
lowed by CH2Cl2/MeOH (98:2)) to afford free higher-order ful-
lerenes (16 mg) and the free molecular cage 2 (27 mg).
The higher-order fullerene mixture recovered from the first three
extractions (227 mg) was mixed with molecular cage 3 (151 mg) in
CHCl2CHCl2 (28 mL) and then the black suspension was heated at
303 K for 16 h. The organic solvent was evaporated under reduced
pressure and the residue suspended in CH2Cl2 (20 mL). After filtra-
tion, the filtrate was concentrated and the residue purified chroma-
tographically on SiO2 (CS2, then CH2Cl2/hexane (4:1), followed by
CH2Cl2/MeOH (98:2)). The recovered higher-order fullerene mixture
and molecular cage 3 were mixed in CHCl2CHCl2 (23 mL) and the
extraction process was repeated. Combined, these two repeated
The hemicarceplex mixture (316 mg) was dissolved in CHCl3
(32 mL) and left at room temperature for 4 h to release the incar-
cerated higher-order fullerenes. The mixture was concentrated and
the residue purified chromatographically on SiO2 (CS2, then CH2Cl2/
hexanes (4:1), followed by CH2Cl2/MeOH (98:2)) to afford a hemicar-
ceplex mixture (177 mg), free higher-order fullerenes (40 mg), and
the free molecular cage 2 (65 mg).
The recycled hemicarceplex mixture (177 mg) was dissolved in THF
(16 mL) and left at room temperature for 16 h to release the incar-
cerated higher-order fullerenes. The black solid obtained after cen-
trifugation of the suspension was purified chromatographically on
SiO2 (CS2, then CH2Cl2/hexanes (4:1), followed by CH2Cl2/MeOH
(98:2)) to afford free higher-order fullerenes (55 mg) and the free
molecular cage 2 (83 mg).
extractions provided
a hemicarceplex mixture (103 mg), free
higher-order fullerenes (167 mg), and the free molecular cage 3
(101 mg).
The hemicarceplex mixture (103 mg) was dissolved in THF (12 mL)
and then the solution was left at room temperature for 16 h. The
black suspension was centrifuged and the collected solid purified
chromatographically on SiO2 (CS2, then CH2Cl2/hexanes (4:1), fol-
lowed by CH2Cl2/MeOH (98:2)) to afford free higher-order ful-
lerenes (32 mg) and the free molecular cage 3 (40 mg). The recy-
cling recoveries of the free hosts 2 and 3 in this extraction cycle
were 75 and 94%, respectively.
The higher-order fullerene mixture recovered from the first three
extractions (494 mg) was mixed with the molecular cage
3
(494 mg) in CHCl2CHCl2 (91 mL) and then the black suspension was
heated at 303 K for 16 h. The solvent was evaporated under re-
duced pressure and the residue suspended in CH2Cl2 (40 mL). After
filtration, the filtrate was concentrated and the residue purified
chromatographically on SiO2 (CS2, then CH2Cl2/hexanes (4:1), fol-
lowed by CH2Cl2/MeOH (98:2)). The recovered higher-order fuller-
ene mixture and molecular cage 3 were mixed in CHCl2CHCl2
(78 mL) and then the extraction process was repeated. Combined,
these two repeated extractions provided a hemicarceplex mixture
(282 mg), free higher-order fullerenes (275 mg), and the free mo-
lecular cage 3 (268 mg).
HPLC analysis: Cosmosil-packed 5PBB analytical column, 4.6ꢁ
250 mm; mobile phase=toluene; UV detection=285 nm; elution
rate=1 mLminÀ1. Separation: Cosmosil-packed 5PBB semiprepara-
tive column, 10ꢁ250 mm; mobile phase=toluene; UV detection=
285 nm; elution rate=5 mLminÀ1
.
The hemicarceplex mixture (282 mg) was dissolved in THF (34 mL)
and then the solution was left at room temperature for 16 h. The
black suspension was centrifuged and the collected solid was puri-
fied chromatographically on SiO2 (CS2, then CH2Cl2/hexanes (4:1),
followed by CH2Cl2/MeOH (98:2)) to afford free higher-order ful-
lerenes (80 mg) and the free molecular cage 3 (121 mg). The recy-
cling recoveries of the free hosts 2 and 3 in this extraction cycle
were 77 and 79%, respectively.
Acknowledgements
This study was supported by the Ministry of Science and Tech-
nology (Taiwan) (MOST-104-2628M-002-012) and National
Taiwan University (NTU-105-R890913).
Keywords: cyclotriveratrylenes · cage compounds · fullerenes ·
hemicarceplexes · host–guest systems
Second extraction cycle: The recycled molecular cage 2 (183 mg)
and higher-order fullerene mixture (275 mg) were mixed in
CHCl2CHCl2 (12 mL) and then the black suspension was heated at
308 K for 40 h. After evaporating the solvent under reduced pres-
sure, the residue was suspended in CH2Cl2 (20 mL) and filtered. The
filtrate was concentrated and the residue purified chromatographi-
cally on SiO2 (CS2, then CH2Cl2/hexanes (4:1), followed by CH2Cl2/
[1] a) A. Hirsch, M. Brettreich, Fullerenes: Chemistry and Reactions, Wiley-
VCH, Weinheim, 2005; b) F. Langa, J.-F. Nierengarten, Fullerenes: Princi-
ples and Applications, Royal Society of Chemistry, Cambridge, 2007.
[2] a) W. A. Scrivens, P. V. Bedworth, J. M. Tour, J. Am. Chem. Soc. 1992, 114,
7917–7919; b) K. C. Khemani, M. Prato, F. A. Wudl, J. Org. Chem. 1992,
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
57, 3254–3256; c) J. L. Atwood, G. A. Koutsantonis, C. L. Raston, Nature
1994, 368, 229–231.
[3] Based on the prices listed on the Sigma–Aldrich website on July 20,
2016.
For examples of complexes generated from fullerenes and CTV-based
hosts, see: c) K. T. Holman, M. M. Halihan, S. S. Jurisson, J. L. Atwood,
R. S. Burkhalter, A. R. Mitchell, J. W. Steed, J. Am. Chem. Soc. 1996, 118,
9567–9576; d) H. Matsubara, A. Hasegawa, K. Shiwaku, K. Asano, M.
Uno, S. Takahashi, K. Yamamoto, Chem. Lett. 1998, 27, 923–924; e) H.
Matsubara, T. Shimura, A. Hasegawa, M. Semba, K. Asano, K. Yamamoto,
Chem. Lett. 1998, 27, 1099–1100; f) E. Huerta, E. Cequier, J. de Mendoza,
Chem. Commun. 2007, 5016–5018; g) H. Matsubara, S. Oguri, K. Asano,
K. Yamamoto, Chem. Lett. 1999, 28, 431–432; h) J. J. Henkelis, C. J. Car-
ruthers, S. E. Chambers, R. Clowes, A. I. Cooper, J. Fisher, M. J. Hardie, J.
Am. Chem. Soc. 2014, 136, 14393–14396; i) B. Chatelet, L. Joucla, D.
Padula, L. Di Bari, G. Pilet, V. Robert, V. Dufaud, J.-P. Dutasta, A. Martinez,
Org. Lett. 2015, 17, 500–503; j) B. A. Riggle, Y. Wang, I. J. Dmochowski,
J. Am. Chem. Soc. 2015, 137, 5542–5548.
[4] a) T. Andersson, K. Nilsson, M. Sundahl, G. Westman, O. Wennerstrom, J.
Chem. Soc. Chem. Commun. 1992, 604–605; b) J. L. Atwood, L. J. Bar-
bour, C. L. Raston, I. B. N. Sudria, Angew. Chem. Int. Ed. 1998, 37, 981–
983; Angew. Chem. 1998, 110, 1029–1031; c) D. Felder, B. Heinrich, D.
Guillon, J.-F. Nicoud, J.-F. Nierengarten, Chem. Eur. J. 2000, 6, 3501–
3507; d) J.-Y. Zheng, K. Tashiro, Y. Hirabayashi, K. Kinbara, K. Saigo, T.
Aida, S. Sakamoto, K. Yamaguchi, Angew. Chem. Int. Ed. 2001, 40, 1857–
1861; Angew. Chem. 2001, 113, 1909–1913; e) Y. Kubo, A. Sugasaki, M.
Ikeda, K. Sugiyasu, K. Sonoda, A. Ikeda, M. Takeuchi, S. Shinkai, Org. Lett.
2002, 4, 925–928; f) T. Kawase, K. Tanaka, N. Fujiwara, H. R. Darabi, M.
Oda, Angew. Chem. Int. Ed. 2003, 42, 1624–1628; Angew. Chem. 2003,
115, 1662–1666; g) H. Sato, K. Tashiro, H. Shinmori, A. Osuka, Y. Murata,
K. Komatsu, T. Aida, J. Am. Chem. Soc. 2005, 127, 13086–13087; h) A.
Hosseini, S. Taylor, G. Accorsi, N. Armaroli, C. A. Reed, P. D. W. Boyd, J.
Am. Chem. Soc. 2006, 128, 15903–15913; i) E. Huerta, G. A. Metselaar, A.
Fragoso, E. Santos, C. Bo, J. de Mendoza, Angew. Chem. Int. Ed. 2007, 46,
202–205; Angew. Chem. 2007, 119, 206–209; j) E. Huerta, H. Isla, E. M.
Perez, C. Bo, N. Martin, J. de Mendoza, J. Am. Chem. Soc. 2010, 132,
5351–5353; k) G. Gil-Ramꢂrez, S. D. Karlen, A. Shundo, K. Porfyrakis, Y.
Ito, G. A. D. Briggs, J. J. L. Morton, H. L. Anderson, Org. Lett. 2010, 12,
3544–3547; l) D. Canevet, E. M. Perez, N. Martin, Angew. Chem. Int. Ed.
2011, 50, 9248–9259; Angew. Chem. 2011, 123, 9416–9427; m) N.
Martin, J.-F. Nierengarten, Supramolecular Chemistry of Fullerenes and
Carbon Nanotubes, Wiley-VCH, Weinheim, 2012; n) N. Kishi, M. Akita, M.
Kamiya, S. Hayashi, H.-F. Hsu, M. Yoshizawa, J. Am. Chem. Soc. 2013, 135,
12976–12979; o) C. Garcꢂa-Simꢃn, M. Costas, X. Ribas, Chem. Soc. Rev.
2016, 45, 40–62.
[9] M.-J. Li, C.-H. Huang, C.-C. Lai, S.-H. Chiu, Org. Lett. 2012, 14, 6146–
6149.
[10] T. Brotin, V. Roy, J.-P. Dutasta, J. Org. Chem. 2005, 70, 6187–6195.
[11] M.-Y. Ku, S.-J. Huang, S.-L. Huang, Y.-H. Liu, C.-C. Lai, S.-M. Peng, S.-H.
Chiu, Chem. Commun. 2014, 50, 11709–11712.
[12] K. Kikuchi, N. Nakahara, T. Wakabayashi, S. Suzuki, H. Shiromaru, Y.
Miyake, K. Saito, I. Ikemoto, M. Kainosho, Y. Achiba, Nature 1992, 357,
142–144; b) R. Taylor, G. J. Langley, A. G. Avent, T. J. S. Dennis, H. W.
Kroto, D. R. M. Walton, J. Chem. Soc. Perkin Trans. 2 1993, 1029–1036.
[13] a) M. L. C. Quan, D. J. Cram, J. Am. Chem. Soc. 1991, 113, 2754–2755;
b) K. N. Houk, K. Nakamura, C. Sheu, A. E. Keating, Science 1996, 273,
627–629.
[14] Because of the low solubility and possible slow decomposition of the
higher fullerenes in organic solvents, we determined the equilibrium
constants from the dissociation pathways; nevertheless, slow decompo-
sition of the higher-order fullerenes released in solution remained pos-
sible and, therefore, the equilibrium constants and their derived kinetic
and thermodynamic parameters should be considered only as rough
estimates.
[5] a) Y. Shoji, K. Tashiro, T. Aida, J. Am. Chem. Soc. 2004, 126, 6570–6571;
b) T. Haino, C. Fukunaga, Y. Fukazawa, Org. Lett. 2006, 8, 3545–3548.
[6] Purchased from Nanjing XFNANO Materials Tech., Nanjing, China.
[7] a) A. Jasat, J. C. Sherman, Chem. Rev. 1999, 99, 931–967; b) S. Ro, S. J.
Rowan, A. R. Pease, D. J. Cram, J. F. Stoddart, Org. Lett. 2000, 2, 2411–
2414; c) R. Warmuth, J. Yoon, Acc. Chem. Res. 2001, 34, 95–105; d) J. C.
Sherman, Chem. Commun. 2003, 1617–1623; e) R. Warmuth, S. Mako-
wiec, J. Am. Chem. Soc. 2005, 127, 1084–1085; f) Y. Liu, R. Warmuth,
Org. Lett. 2007, 9, 2883–2886; g) N. M. Rue, J. Sun, R. Warmuth, Isr. J.
Chem. 2011, 51, 743–768.
[15] The total amount of free host 2 recycled from the full isolation cycle
was 346 mg, thus giving a recycling recovery of 77%.
[16] a) T. J. S. Dennis, H. Shinohara, Chem. Commun. 1998, 619–620; b) J. A.
Azamar-Barrios, T. J. S. Dennis, S. Sadhukan, H. Shinohara, G. E. Scuseria,
A. Penicaud, J. Phys. Chem. A 2001, 105, 4627–4632.
[17] The total amount of free host 3 recycled from the full isolation cycle
was 389 mg, thus giving a recycling recovery of 79%.
[8] a) T.-H. Wong, J.-C. Chang, C.-C. Lai, Y.-H. Liu, S.-M. Peng, S.-H. Chiu, J.
Org. Chem. 2014, 79, 3581–3586. For a recent review of CTVs, see:
b) J. J. Henkelis, M. J. Hardie, Chem. Commun. 2015, 51, 11929–11943.
Received: July 21, 2016
Published online on && &&, 0000
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
Host–Guest Systems
K.-S. Liu, M.-J. Li, C.-C. Lai, S.-H. Chiu*
&& – &&
Incarceration of Higher-Order
Fullerenes within Cyclotriveratrylene-
Based Hemicarcerands Allows
Selective Isolation of C76, C78, and C84
from a Commercial Fullerene Mixture
A successful release: Two cyclotrivera-
trylene (CTV)-based hosts can form
hemicarceplexes with higher-order ful-
lerenes, which allows the selective isola-
tion of C76/C78 and C80 in high purities
(80–95 and 81–88%, respectively) from
a commercially available mixture of
higher-order fullerenes (see picture).
&
&
Chem. Eur. J. 2016, 22, 1 – 10
www.chemeurj.org
10
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!