Supramolecular multicomponent self-assembly of shape-adaptive nanoprisms: Wrapping up C60 with three porphyrin units

Article abstract of DOI:10.1021/ol800796h

(Figure Presented) Self-assembly of a C_3v symmetric trisphenanthroline and linear bisterpyridines in the presence of Cu⁺ did not furnish the expected supramolecular nanoprisms in quantitative yield. With an accurately sized tripyridine as a stabilizing template, the nanoprism formed exclusively. Furthermore, an adaptive constriction of the nanoprism was seen with C₆₀ as template: as a result of the smaller size of C ₆₀ the nanoframework wrapped up around the guest like an accordion-type host system.

Full text of DOI:10.1021/ol800796h

ORGANIC
LETTERS
2008
Vol. 10, No. 12
2513-2516
Supramolecular Multicomponent
Self-Assembly of Shape-Adaptive
Nanoprisms: Wrapping up C60 with
Three Porphyrin Units
Michael Schmittel,* Bice He, and Prasenjit Mal
Center for Micro- and Nanochemistry and Engineering, Organische Chemie I,
UniVersita¨t Siegen, D-57068 Siegen, Germany
schmittel@chemie.uni-siegen.de
Received April 7, 2008
ABSTRACT
Self-assembly of a C3v symmetric trisphenanthroline and linear bisterpyridines in the presence of Cu+ did not furnish the expected supramolecular
nanoprisms in quantitative yield. With an accurately sized tripyridine as a stabilizing template, the nanoprism formed exclusively. Furthermore,
an adaptive constriction of the nanoprism was seen with C60 as template: as a result of the smaller size of C60 the nanoframework wrapped
up around the guest like an accordion-type host system.
Self-assembly, a burgeoning section of supramolecular
chemistry, has been a prolific tool for the construction of
diverse fascinating hollow 3D structures, such as Fujita’s
cages1 and hexahedra;2 Atwood’s snub cube;3 Stang’s prism,4
cuboctahedra,5 and dodecahedra;6 Raymond’s tetrahedra;7
etc. These structures are spectacular not only because of their
size and shape but also because of their ability to encapsulate
neutral7 or charged guest8 molecules via noncovalent interac-
tions. Yet, almost all known host-guest complexes formed
from nanosized, self-assembled aggregates are rather static,
indicating that our ability to set up systems for adaptive shape
transformations is still limited.9 Herein, we disclose how an
added guest decisively modulates the shape of a supramo-
lecular three-component assembly affording either a nano-
prism through templation or a constricted nanoframework
wrapped up around a C60 unit.
(1) (a) Fujita, M.; Oguro, D.; Miyazawa, M.; Oka, H.; Yamaguchi, K.;
Ogura, K. Nature 1995, 378, 469–471. (b) Kumazawa, K.; Biradha, K.;
Kusukawa, T.; Okano, T.; Fujita, M. Angew. Chem., Int. Ed. 2003, 42, 3909–
3913.
In contrast to Su¨ss-Fink and Therrien,10 Hupp,11 and Su,12
we focused on a three-component self-assembly protocol
toward nanoprisms utilizing the HETTAP (Heteroleptic
Terpyridine And Phenanthroline Metal Complexes) concept.
(2) Takeda, N.; Umemoto, K.; Yamaguchi, K.; Fujita, M. Nature 1999,
398, 794–796.
(3) MacGillivray, L. R.; Atwood, J. L. Nature 1997, 389, 469–472.
(4) Kuehl, C. J.; Yamamoto, T.; Seidel, S. R.; Stang, P. J. Org. Lett.
2002, 4, 913–915.
(9) Lehn, J.-M. Chem. Soc. ReV. 2007, 36, 151–160.
(10) (a) Govindaswamy, P.; Linder, D.; Lacour, J.; Su¨ss-Fink, G.;
Therrien, B. Chem. Commun. 2006, 4691–4693. (b) Govindaswamy, P.;
Su¨ss-Fink, G.; Therrien, B. Organometallics 2007, 26, 915–924.
(11) Lee, S. J.; Mulfort, K. L.; O’Donnell, J. L.; Zuo, X.; Goshe, A. J.;
Wesson, P. J.; Nguyen, S. T.; Hupp, J. T.; Tiede, D. M. Chem. Commun.
2006, 4581–4583.
(5) Olenyuk, B.; Whiteford, J. A.; Fechtenko¨tter, A.; Stang, P. J. Nature
1999, 398, 796–799.
(6) Olenyuk, B.; Levin, M. D.; Whiteford, J. A.; Shield, J. E.; Stang,
P. J. J. Am. Chem. Soc. 1999, 121, 10434–10435.
(7) Biros, S. M.; Bergman, R. G.; Raymond, K. N. J. Am. Chem. Soc.
2007, 129, 12094–12095.
(8) Fiedler, D.; Leung, D. H.; Bergman, R. G.; Raymond, K. N. J. Am.
Chem. Soc. 2004, 126, 3674–3675.
(12) Liu, Z.-M.; Liu, Y.; Zheng, S.-R.; Yu, Z.-Q.; Pan, M.; Su, C.-Y.
Inorg. Chem. 2007, 46, 5814–5816.
10.1021/ol800796h CCC: $40.75
Published on Web 05/27/2008
2008 American Chemical Society

This strategy (Figure 1) exploits steric and electronic effects
originating from bulky aryl substituents at the phenanthroline
each other in dichloromethane for 5 min, and then solutions
of BT1, BT2 (in dichloromethane), and BT3 (in chloroform)
were added to form the heteroleptic nanoprisms P1-P3.
From solutions of P1 ) [Cu6(TP)2(BT1)3]6+ and P2 )
[Cu6(TP)2(BT2)3]6+ in dichloromethane, ESI-MS signals of
notable intensity were observed arising from the 4+, 5+,
1
and 6+ charged prisms, while their H NMR spectra (see
Figure 2) were characterized by atypically broad absorptions.
1
Figure 2
.
Partial H NMR spectra of complexes R1, P1-P3, and
P3·template with P1
[Cu6(TP)2(BT2)3]6+, P3 ) [Cu6(TP)2(BT3)3]6+, and R1 )
[Cu3(TP)(Terpy)3]3+
)
[Cu6(TP)2(BT1)3]6+
,
P2
)
Figure 1. Ligands, such as trisphenanthroline, tripyridines, and
terpyridines, used in the self-assembly of nanoprisms P1-P3.
.
to control the coordination equilibrium at the metal ion.13
Use of Cu+ should be optimal for generating mixed [M(phen)-
(terpy)]n+ complexes, as it avoids having too strong a driving
force for bishomoleptic [M(terpy)2]n+ complexes, quite in
contrast to Zn2+.14
Solutions of P1 and P2 in acetonitrile, after being heated to
various temperatures between room temperature and 79 °C,
furnished identical spectra at room temperature, precluding
kinetic control in the assembly process. For P1 and P2 signals
sharpened notably at -30 °C but still represented more than
one species, thus suggesting the occurrence of other ther-
modynamically competitive species in the mixture. When
we used the longest bisterypridine, BT3, in the self-assembly,
the alleged P3 ) [Cu6(TP)2(BT3)3]6+ was formed in even
smaller amounts, as noted from both ESI-MS and 1H NMR
data. However, the signal at ca. 6.2 ppm in all spectra of
P1-P3 clearly pointed to the compelling formation of the
desired heteroleptic copper(I) complexes.
Trisphenanthroline TP (see Figure 1) was synthesized by
a Sonogashira cross-coupling of 1,3,5-triiodobenzene15 and
the terminal alkynylphenanthroline, the latter prepared by a
procedure described earlier.16 Bisterpyridine BT1 was pur-
chased, while BT2 was prepared as reported.13c Bisterpyri-
dine BT3 was synthesized by a Suzuki cross-coupling of
[2,2′:6′,2′′]-terpyridinyl-4′-boronic acid and zinc(II)-5,15-
bis(4-bromophenyl)-10,20-dimesitylporphyrin (see Support-
ing Information).
To slow down formation of the undesired [Cu(terpy)2]+
complexes, [Cu(MeCN)4]PF6 and TP were first reacted with
In contrast, the simple complex [Cu3(TP)(Terpy)3]3+
)
R1 was formed in a clean manner after mixing [Cu(Me-
CN)4]PF6, TP, and the parent terpyridine (Terpy) as
evidenced by the clean ESI-MS signal set and sharp
resonances in the 1H NMR spectrum. After comparing these
results with those of the prisms, we reasoned that complete
formation of P1-P3 was impeded due to the thermodynami-
cally competitive formation of some small oligomeric
aggregates.
(13) Steric stoppers at the phenanthroline’s coordination site in 2,9-
position prevent any competitive bishomoleptic metal phenanthroline
complexes, therefore leading preferably to hetero combinations: (a) Schmit-
tel, M.; Kalsani, V.; Fenske, D.; Wiegrefe, A. Chem. Commun. 2004, 490–
491. (b) Schmittel, M.; Kalsani, V.; Kishore, R. S. K.; Co¨lfen, H.; Bats,
J. W. J. Am. Chem. Soc. 2005, 127, 11544–11545. (c) Schmittel, M.;
Kalsani, V.; Mal, P.; Bats, J. W. Inorg. Chem. 2006, 45, 6370–6377. (d)
Schmittel, M.; He, B.; Kalsani, V.; Bats, J. W. Org. Biomol. Chem. 2007,
5, 2395–2403.
(14) Schmittel, M.; Mal, P. Chem. Commun. 2008, 960–962.
(15) Vatsadze, S. Z.; Titanyuk, I. D.; Chernikov, A. V.; Zyk, N. V. Russ.
Chem. Bull., Int. Ed. 2004, 53, 471–473.
(17) (a) Anderson, H. L.; Walter, C. J.; Vidal-Ferran, A.; Hay, R. A.;
Lowden, P. A.; Sanders, J. K. M. J. Chem. Soc., Perkin Trans. 1 1995,
2275–2279. (b) Anderson, S.; Anderson, H. L.; Bashall, A.; McPartlin, M.;
Sanders, J. K. M. Angew. Chem., Int. Ed. 1995, 34, 1096–1099.
(16) (a) Schmittel, M.; Michel, C.; Wiegrefe, A.; Kalsani, V. Synthesis
2001, 1561–1567. (b) Schmittel, M.; Michel, C.; Wiegrefe, A. Synthesis
2005, 367–373.
2514
Org. Lett., Vol. 10, No. 12, 2008

Since zinc porphyrin can form strong complexes with
pyridines11,17 and fullerenes,18 we chose the tripyridines
tpy119 and tpy2 as well as C60 as templates to stabilize the
prism P3. Not unexpectedly, all attempts to fit the large tpy1
(requesting dZn-Zn ) 1.78 nm in a trigonal arrangement of
three zinc porphyrins, see Figure 3) into prism P3 (providing
Figure 3. Zn-Zn distances in an idealized, prism-like trigonal
setting of three zinc porphyrins complexing to tpy1, tpy2, and
fullerene C60 (from MM+ minimized molecular models; a typical
ZnPor-NPy coordination distance of 0.22 nm was assumed17b).
Figure 4. Side view of space-filling presentation of prism P5 )
P3•tpy2 as generated from force field modeling (Hyperchem). The
atoms/ligand are color coded for clarity: carbon, cyan; nitrogen,
blue; bromine, yellow; copper, red; zinc, white; tpy2 ligand, green.
Hydrogens are removed for clarity.
dZn-Zn ) 1.45 nm) failed to afford P4 ) P3·tpy1 in good
yield as judged by ESI-MS and NMR.
In contrast, tpy2, requiring a dZn-Zn of 1.35 nm, is a highly
suitable template for P3 since its size closely matched that
of P3 as shown in Figure 4. After addition of tpy2 to P3,
the ESI-MS spectrum (see Supporting Information) exhibited
two main signals at 1284.3 and 1570.5 Da representing the
6+ and 5+ charged prism P5 ) P3·tpy2, i.e., [Cu6(TP)2-
(BT3)3(tpy2)]6+ and [Cu6(TP)2(BT3)3(tpy2)(PF6)]5+, respec-
tively. Absence of peaks at 1233.4, 1509.1, and 1922.7 Da
for the 6+, 5+, and 4+ charged prism P3 suggested
Notably, the smaller C60 requiring a Zn-Zn distance of
only 1.09 nm for complexation (see Figure 3) also proved
to be a good template for P3. Due to the low solubility of
C60 in acetonitrile, C60 (1 equiv per P3) was added in carbon
disulfide to an acetonitrile solution of P3. After removal of
all solvents, the solid residue, now P6 ) P3·C60, was
1
dissolved in CD3CN and analyzed by ESI-MS, H NMR,
13C NMR, UV-vis, CV, and DPV without further purifica-
tion. The 1H NMR spectrum of P6 was significantly
simplified compared to that of P3 (Figure 2). Its 13C NMR
spectrum showed a strong signal at δ ) 139.4 ppm,
unmistakably indicating the presence of bound C60, as free
C60, studied in a control experiment, exhibited a signal shifted
downfield by ∆δ ∼4 ppm.20 The ESI-MS (Supporting
Information) displayed two main signals at 1652.0 and
1352.8 Da from the 5+ and 6+ charged P6, i.e., [Cu6-
(TP)2(BT3)3(C60) (PF6)]5+ and [Cu6(TP)2 (BT3)3(C60)]6+,
respectively. Also their isotopic splitting was in good
agreement with those from simulations. If harsher ionization
conditions were applied (offset increased, higher tempera-
ture), two extra signals at 1233.5 Da for [Cu6(TP)2(BT3)3]6+
and at 1511.5 Da for [Cu6(TP)2(BT3)3(PF6)]5+ were ob-
served, suggesting that P6 can release C60. It is noteworthy
that the porphyrin-fullerene interaction has rarely been
documented by ESI-MS analysis,18a,f hence suggesting an
unusually strong binding between P3 and C60 in P6.
1
exclusive formation of P5. The H NMR analysis of P5 in
CD2Cl2/CD3CN (4:1) revealed only one set of signals with
the chemical shift of the 8-H mesityl proton of TP being
diagnostically shifted from 6.96 to 6.33 ppm (see Figure 2).
Additionally, the chemical shifts of the tpy2 pyridine protons,
initially located at 8.59 and 7.59 ppm in CDCl3, were
dramatically shifted to 5.71 and 5.13 ppm. Existence of P5
as a single species was conclusively established by DOSY
1H NMR (see Supporting Information). Hence, one has to
conclude that the strong binding and good spatial match
between the three pyridine nitrogens of tpy2 and the three
zinc porphyrin units in P3 transfers a dynamic mixture into
a single species P5.
(18) (a) Tashiro, K.; Aida, T.; Zheng, J.-Y.; Kinbara, K.; Saigo, K.;
Sakamoto, S.; Yamaguchi, K. J. Am. Chem. Soc. 1999, 121, 9477–9478.
(b) Zheng, J.-Y.; Tashiro, K.; Hirabayashi, Y.; Kinbara, K.; Saigo, K.; Aida,
T.; Sakamoto, S.; Yamaguchi, K. Angew. Chem., Int. Ed. 2001, 40, 1858–
1861. (c) Wang, Y.-B.; Lin, Z. J. Am. Chem. Soc. 2003, 125, 6072–6073.
(d) Boyd, P. D. W.; Reed, C. A. Acc. Chem. Res. 2005, 38, 235–242. (e)
Satake, A.; Kobuke, Y. Tetrahedron 2005, 61, 13–41. (f) Hosseini, A.;
Taylor, S.; Accorsi, G.; Armaroli, N.; Reed, C. A.; Boyd, P. D. W. J. Am.
Chem. Soc. 2006, 128, 15903–15913. (g) Ouchi, A.; Tashiro, K.; Yamagu-
chi, K.; Tsuchiya, T.; Akasaka, T.; Aida, T. Angew. Chem., Int. Ed. 2006,
45, 3542–3546. (h) Olmstead, M. M.; Nurco, D. J. Cryst. Growth Des.
2006, 6, 109–113. (i) Hosseini, A.; Hodgson, M. C.; Tham, F. S.; Reed,
C. A.; Boyd, P. D. W. Cryst. Growth Des. 2006, 6, 397–403.
Cyclic (CV) and differential pulse voltammograms (DPV)
further confirmed that C60 was entrapped in P6. Due to
(20) As C60 has a very low solubility of 8 µg mL-1 in acetonitrile (
Marcus, Y.; Smith, A. L.; Korobov, M. V.; Mirakyan, A. L.; Avramenko,
N. V.; Stukalin, E. B. J. Phys. Chem. B 2001, 105, 2499–2506. ) we
measured its 13C signal in a 1:1 mixture of toluene/acetonitrile. See
Supporting Information, Figures S31 and S32.
(19) Asselberghs, I.; Hennrich, G.; Clays, K. J. Phys. Chem. A 2006,
110, 6271–6275.
Org. Lett., Vol. 10, No. 12, 2008
2515

solubility problems in presence of the electrolyte, CV and
DPV measurements of P3 and P6 were conducted in THF/
acetonitrile (1:1) using tetra-n-butyl-ammonium hexafluo-
there are several atypical proton shifts and splittings in P6
that are not there in P3, P5, R1, TP, and BT3: (i) Protons
1-H of TP are shifted highfield in P6 by about 0.2-0.4 ppm
as compared to those in R1 and P5. (ii) Protons 8-H of the
phenanthroline unit seen at identical δ 6.3 ppm in P3, P5,
and R1 are shifted 0.3 ppm downfield and split into two
singlets with an integration ratio of 1:1 for P6. Whereas the
high field shift in P3, P5, and R1 advocates a positioning of
8-H in the shielding region of the terpyridine ligand due to
the complex geometry, the lowfield shift in P6 in contrast
suggests a partial egression from this spatial region. The
signal splitting moreover indicates that the phenanthroline
is not any more representing a plane of symmetry in P6,
obviously due to its twisted structure. The sum of data thus
strongly supports the adaptive complexation of C60 in a
constricted inner cavity of P6.
rophosphate as electrolyte. The CV and DPV spectra showed
0/-
two waves corresponding to the reduction C60
and
C60-/2- at -0.94 and -1.38 V vs ferrocene (Fc) confirming
the presence of C60 in P6 (Figure 5, Figure S3). Besides,
In summary, we have demonstrated that nanoprisms can
be prepared from multicomponent self-assembly of the
trisphenanthroline TP and various bisterpyridines BT1-BT3
following the HETTAP strategy. Different from the situation
with the analogous but quantitatively formed nanoladder
systems investigated earlier,13 the nanoprisms P1-P3 are
formed in equilibrium with smaller oligomers. After adding
templates of suitable size, such as tpy2 or C60, interaction
between the latter and the porphyrin unit of BT3 provided
an extra driving force that shifted the equilibrium completely
toward the desired structures. Thus, with tpy2 the prism P5
) P3·tpy2 was formed. In contrast, with C60 as template the
nanoprism P3 acted as an adaptive host leading to the
constricted prism P6 ) P3·C60. As such, P3 represents an
interesting type of host system changing its longitudinal
dimensions in order to adapt laterally to the size of the guest,
somewhat similar to an accordion. Clearly, its modus
operandi is different from that of longitudinally adaptive
hosts.18a
Figure 5. DPV of P3 (black) and P6 ) P3·C60 (red) measured in
acetonitrile/THF (1:1) using ferrocene as internal standard.
the oxidation potential of Cu+/2+ was shifted from 0.43 to
0.36 V, the ZnP0/+ wave was shifted from 0.79 to 0.74 V,
and that of ZnP+/2+ was shifted from 1.02 to 0.96 V. These
small cathodic shifts of about 60 mV exhibited by the zinc
porphyrin units in P6 are obviously due to their interaction
with C60.
The sum of evidence indicates that C60 is strongly bound
inside the prism P6 ) P3·C60: (i) a templating role of C60
can only become effective inside P3, and (ii) ESI-MS, CV,
and DPV advocate strong interactions between P3 and C60.
However, in order to make use of three porphyrin units,18g–i
P6 needs to have a constricted geometry properly adapting
to the size of the encapsulated fullerene. A model study
(MM+ in Hyperchem) suggests that this is easily possible
by twisting the C3V top versus the C3V bottom deck (see
Graphical Abstract and Supporting Information). This adap-
tive distortion of the host is well evinced in the NMR! A
Acknowledgment. We are grateful to the Deutsche
Forschungsgemeinschaft, the Humboldt Foundation and the
Fonds der Chemischen Industrie for financial support.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for all new ligands and
nanoprisms. This material is available free of charge via the
Internet at http://pubs.acs.org.
1
detailed comparison of the H NMRs (Figure 2) indicates
OL800796H
2516
Org. Lett., Vol. 10, No. 12, 2008