Void and filled supramolecular nanoprisms-notable differences between seemingly identical construction principles

Article abstract of DOI:10.1039/b808988d

Void and filled supramolecular nanoprisms (void: 4900 A³) were furnished in quantitative yield utilising the (terpyridine)-Zn ²⁺-(phenanthroline) complex as a dynamic and heteroleptic building motif (HETTAP approach), but only if units serving as panels and pillars in the self-assembly were optimised with regard to their kinetic behaviour. The Royal Society of Chemistry.

Full text of DOI:10.1039/b808988d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Void and filled supramolecular nanoprisms—notable differences between
seemingly identical construction principlesw
Michael Schmittel* and Bice He
Received (in Cambridge, UK) 28th May 2008, Accepted 4th August 2008
First published as an Advance Article on the web 28th August 2008
DOI: 10.1039/b808988d
Void and filled supramolecular nanoprisms (void: 4900 A3) were
furnished in quantitative yield utilising the (terpyridine)-Zn2+
2,9-positions of each phenanthroline unit. These prevent associa-
tion to homoleptic [Zn(phenanthroline)2]2+ complexes. While BP1
and BP2 have been reported earlier,5c ligands TP and TT were
synthesised via sequential Sonogashira coupling protocols (ESIw).
It is important to note that method 1 did not reproducibly lead
to quantitative formation of the projected prism P1. For example,
in an attempt to prepare P1 3 equivalents of Zn(CF3SO3)2 were
added to TP in dichloromethane–acetonitrile (8 : 2) followed by
heating at 40 1C for 5 minutes. Thereafter, bisterpyridine BT in
chloroform was added.7 Although ESI-MS spectra of this solu-
tion displayed two weak signals indicative of prism P1, i.e. at
1173.2 and 1501.9 Da for the 5+ and 4+ charged P1, the main
set of signals arose from [Zn4(TP)2(BT)2]n+ and [Zn2(BT)2]n+
attesting incomplete formation of the prism (Fig. S13, ESIw).
Moreover, the 1H-NMR spectrum of P1 provided a set of
strongly broadened signals (Fig. S14, ESIw), witnessing that P1
was formed in equilibrium with other structures. Equally, clean
preparation of the analogous copper(I) prism P2 failed, in spite of
the fact that a (terpyridine)-Cu+-(phenanthroline) complex is
more stable and kinetically labile than the alike Zn2+ complex.5a,d
Besides, it should be noted that self-assembly along method 1 fails
for prism structures independent of size.6
-
(phenanthroline) complex as a dynamic and heteroleptic building
motif (HETTAP approach), but only if units serving as panels
and pillars in the self-assembly were optimised with regard to
their kinetic behaviour.
Metallosupramolecular prisms obtained through self-assembly1
have been studied not only for their interesting architecture2
but also for their ability to act as host–guest systems.3 Using
the 2,4,6-trispyridyl-1,3,5-triazine as panels, an assortment of
bipyridines as pillars, and palladium(II) ions as coordination
corners Fujita et al. prepared triangular prisms, which allowed
accommodation of exciting p–p stacked pyrenes3d or por-
phyrin dimers3g in their cavity. Recently, even nonameric
aromatic stacks were realised from interpenetrated coordina-
tion cages.4 As a rule, host–guest interactions in those prisms
were designed via non-covalent bonding.
Herein, we want to report void and filled nanoprisms con-
ceived on the basis of the HETTAP concept (HETeroleptic
Terpyridine And Phenanthroline aggregation).5 Our concept
relies on the formation of heteroleptic pentacoordinated metal
centres, such as zinc(II) or copper(I) ions, surrounded by
[1,10]-phenanthroline and terpyridine ligands. Due to the
heteroleptic arrangement, there exist two seemingly identical
strategies for constructing the projected prisms: the first pro-
tocol uses a trisphenanthroline as the panel, e.g. TP, and a
bisterpyridine, such as BT, as the pillar (Fig. 1a). By contrast,
the second method utilises a tristerpyridine as the panel, e.g.
TT, and a bisphenanthroline such as BP1 as the pillar (Fig. 1b).
While both approaches to nanoprisms are thermochemically
alike with regard to enthalpic and entropic contributions, as six
identical heteroleptic (terpyridine)-Zn2+-(phenanthroline)
complex units are formed, we see remarkable differences in
their reliability. The present study identifies the disparities of
both protocols and analyses reasons why method 2 is much
more successful than approach 1.6
In contrast, method 2 led to quantitative formation of
nanoprisms P3 and P4 without any problems. By following
the same self-assembly protocol as for P1 but now using TT as
the panel and BP1 and BP2 as pillars, P3 and P4 were
afforded. In both cases, ESI-MS spectra witnessed the clean
formation of the prisms. Accordingly, in the ESI-MS spectrum
of P3 the full series of signals, being quite characteristic
Ligands used in the present report are depicted in Scheme 1. To
assist the heteroleptic self-assembly process conceived on the basis
of the HETTAP concept,5 building blocks TP, BP1 and BP2 were
encoded with the required bulky methylaryl groups in the
Center of Micro and Nanochemistry and Engineering, Organische
Chemie I, Universitat Siegen, Adolf-Reichwein-Str. 2, D-57068
¨
Siegen, Germany. E-mail: schmittel@chemie.uni-siegen.de;
Fax: (+49) 271 740 3270
w Electronic supplementary information (ESI) available: Synthesis
and characterization of ligands TP, TT and nanoprisms P3, P4. See
DOI: 10.1039/b808988d
Fig. 1 Pictorial representation of two concepts for the assembly of
nanoprisms by the HETTAP approach. P1 = [Zn6(TP)2(BT)3]12+
,
P2
[Zn6(TT)2(BP2)3]12+
= , P3 = , P4 =
[Cu6(TP)2(BT)3]6+ [Zn6(TT)2(BP1)3]12+
.
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4723–4725 | 4723

View Article Online
Fig. 3 1H-NMR spectra of ligand BP1 and prism P3 and P4.
NMR signal of the central benzene ring in BP1, BP2 was shifted
from 7.20 ppm to 6.54 ppm in P3 (Dd = 0.66 ppm) but much less
in P4 to 6.88 ppm. The remarkable shift difference is attributed to
the different environment of the central BP1, BP2 benzene ring in
P3 and P4 caused by the ferrocene units (see Fig. 4).
The electronic situation of the ferrocene units in prism P4 was
evaluated by cyclic voltammetry (CV) and differential pulse
voltammetry (DPV). An earlier report5c about ferrocene-filled
nanoladders had indicated that the redox potential reflected the
average distances of the ferrocene iron atoms held together in the
supramolecular framework. The smaller the average distance, i.e.
29.5 A, 18.5 A and 15.2 A, the more anodically shifted was the
redox potential of the ferrocene units, i.e. 0.462, 0.480 and 0.491 V
vs. DMFc (decamethyl ferrocene). CV and DPV results of P4
illustrated that the ferrocene units had an oxidation potential of
0.477 V vs. DMFc (Fig. S5, ESIw), suggesting an average distance
of the six ferrocene iron atoms in prism P4 of around 22 A.
Evaluation of the average distances of the six ferrocene iron atoms
in prism P4 from a HyperChems energy minimised structure
provided a value of 20.8 A (Fig. 4), being very close to 22 A. As
the HyperChems structure of prism P4 revealed a cavity with a
large void of about 4900 A3 (calculated from the distances of the
Zn2+ coordination centres), in principle the inner space should be
sufficiently large to accommodate all six ferrocene units inside
the prism. However, due to the large pores and flexible hexa-
methylene linkers the attached redox centres assume the biggest
possible distances between the positively charged ferroceniums.
What is the reason for the failure of method 1 and success of
method 2 in forming nanoprisms? Why do nanoprisms form
quantitatively along pathway 2 but only as a component in an
equilibrium in route 1? Based on thermochemical reasoning, there
is no obvious difference between the two pathways: both are
equivalent with regard to enthalpic (six identical terpyridine-
Zn2+-phenanthroline complex units are formed) and overall
Scheme 1 Ligands used to generate nanoprisms.
because of its charge pattern covering all species from 4+ to
11+, was unequivocally assigned to the nanoprism (Fig. 2).
Besides, the isotopic splitting of all charged species matched
exactly the theoretical ones (see inset of Fig. 2). The ESI-MS
spectrum of P4 (Fig. S20, S21, ESIw) showed the same
characteristic pattern as that of P3, again indicative of the
nanoprism structure.
The 1H-NMR spectra of P3 and P4 (see Fig. 3) displayed only
one set of signals, witnessing equally the successful and clean
preparation of the prisms. Proton signals from the bisphenanthro-
line units in P3 or P4 were shifted significantly downfield as
compared to those of free BP1, with shifts being comparable for
both prisms. Protons of the phenanthroline 2,9-aryl groups in
BP1, BP2 were shifted from 6.60/6.98 ppm in the free ligands to
6.08/6.28 ppm in P3 and 5.96/6.31 ppm in P4. Interestingly, the 1H
Fig. 2 ESI-MS spectrum of P3. The inset shows the experimental
(black) and theoretical (red) isotopic distributions, charged from 5+
to 8+.
Fig. 4 HyperChems structure of prism P4. The atoms/groups are
colour coded for clarity: carbon: cyan; nitrogen: blue; oxygen: red;
zinc: hidden; ferrocene group: green. Left: top view; right: side view.
ꢀc
This journal is The Royal Society of Chemistry 2008
4724 | Chem. Commun., 2008, 4723–4725

View Article Online
of the prism structure generates a kinetic barrier of thermo-
dynamic origin. Formation of [Znn(TT)m]2n+ complexes by
dissociation of the prisms is highly unlikely, as the concentra-
tions of [Zn(TT)]2+ and TT remain too small.
In conclusion, the novel nanoprisms P3 and P4, i.e.
[Zn6(TT)2(BP13)]12+ and ([Zn6(TT)2(BP23)]12+) were gene-
rated quantitatively using a HETTAP guided approach along
route 2. Formation of the ferrocene-containing prism P4
further illustrated the utility of our approach for generating
internally functionalised supramolecular structures.
Fig. 5 Illustration of the final step of the self-assembly of P1 along
route 1 (left and middle) and P3 along route 2 (right). The atoms and
the last ligands are colour coded for clarity: carbon: cyan; nitrogen:
blue; oxygen: red; zinc: white; last BT and BP1: green.
We are grateful to the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie. We thank Dr Jan W.
Bats (Universitat Frankfurt) for the verification of the struc-
¨
entropic contributions. As a consequence, one is apt to assume
that route 1, which does not work well, has a kinetic barrier. As
there is no kinetic barrier in the generation of simple [2 + 2]-
ladders from bisphenanthrolines and bisterpyridines,5 the most
likely reason for a kinetic impediment lies in the last step of the
self-assembly process along method 1: herein one needs the last
bisterpyridine to clip in sideways against some steric bulk (Fig. 5,
left and middle).
ture of TP by X-ray analysis.
Notes and references
1 M. Schmittel and V. Kalsani, Top. Curr. Chem., 2005, 245, 1–53.
2 (a) C. Piguet, J.-C. G. Bunzli, G. Bernardinelli, G. Hopfgartner and
¨
A. F. Williams, J. Am. Chem. Soc., 1993, 115, 8197–8206; (b) C. J.
Kuehl, T. Yamamoto, S. R. Seidel and P. J. Stang, Org. Lett., 2002,
4, 913–915; (c) B. Manimaran, P. Thanasekaran, T. Rajendran,
R.-T. Liao, Y.-H. Liu, G.-H. Lee, S.-M. Peng, S. Rajagopal and
K.-L. Lu, Inorg. Chem., 2003, 42, 4795–4597; (d) C.-Y. Su, Y.-P.
Cai, C.-L. Chen, M. D. Smith, W. Kaim and H.-C. zur Loye,
J. Am. Chem. Soc., 2003, 125, 8595–8613; (e) Y. K. Kryschenko, S.
R. Seidel, D. C. Muddiman, A. I. Nepomuceno and P. J. Stang,
J. Am. Chem. Soc., 2003, 125, 9647–9652; (f) D. Kim, J. H. Paek,
M.-J. Jun, J. Y. Lee, S. O. Kang and J. Ko, Inorg. Chem., 2005, 44,
7886–7894; (g) Z.-M. Liu, Y. Liu, S.-R. Zheng, Z.-Q. Yu, M. Pan
and C.-Y. Su, Inorg. Chem., 2007, 46, 5814–5816.
3 (a) N. Fujita, K. Biradha, M. Fujita, S. Sakamoto and K.
Yamaguchi, Angew. Chem., Int. Ed., 2001, 40, 1718–1721;
(b) J. D. Crowley, A. J. Goshe and B. Bosnich, Chem. Commun.,
2003, 2824–2825; (c) K. Kumazawa, K. Biradha, T. Kusukawa, T.
Okano and M. Fujita, Angew. Chem., Int. Ed., 2003, 42,
3909–3913; (d) M. Yoshizawa, J. Nakagawa, K. Kumazawa, M.
Nagao, M. Kawano, T. Ozeki and M. Fujita, Angew. Chem., Int.
Ed., 2005, 44, 1810–1813; (e) P. Govindaswamy, D. Linder, J.
Assuming a kinetic barrier in the last step of route 1 and no
kinetic barrier for the last step of pathway 2 seems to make
sense at first: in approach 2 the binding direction of panel TT
is pointing outside allowing the third bisphenanthroline to
slide on without any steric hindrance (Fig. 5, right). In
contrast, self-assembly along method 1 requires an approach
of BT sideways from the binding direction of the free phenan-
throline unit in TP as any other approach is sterically impeded
by the 2,9-aryl groups. Although the angle between the bind-
ing sites of TP and BT can arrange at anything in between 45
and 1351,8 this flexibility does not help much for a sideways
slip-in motion (Fig. 5).
While the above rationalisation seems to be convincing at
first, it can not be maintained after an in-depth evaluation.
Control experiments for approach 1 showed that raising the
temperature up to 79 1C did not increase the yield. Neither 1H
NMR at higher temperature, nor at room temperature (after
various times at elevated temperatures) showed any significant
changes in the spectra. Such finding clearly argues against a
kinetic barrier. Equally, the finding of a templating effect in
nanoprisms formed along pathway 1 argues against a kinetic
barrier; such effect should not lower a kinetic barrier.6
A hypothesis consistent with all findings suggests that
oligomeric terpyridine complexes, such as [Znn(BT)m]2n+ or
[Znn(TT)m]2n+, are thermodynamically competitive with the
nanoprisms. Along route 1, side products [Znn(BT)m]2n+ are
able to form because the last step to P1 is slow. This allows the
components to explore the global energy hypersurface for
thermochemically competitive structures, such as mononuc-
lear and oligomeric zinc(II) bisterpyridine complexes. In that
way, a reasonably high concentration of [Zn(BT)]2+ builds up
initiating formation of oligomeric complexes. Along pathway
2, oligomeric complexes of TT do not arise as the global
minimum structure P3, P4 is reached rapidly. Fast formation
Lacour, G. Suss-Fink and B. Therrien, Chem. Commun., 2006,
¨
4691–4693; (f) S. Tashiro, M. Kobayashi and M. Fujita, J. Am.
Chem. Soc., 2006, 128, 9280–9281; (g) K. Ono, M. Yoshizawa, T.
Kato, K. Watanabe and M. Fujita, Angew. Chem., Int. Ed., 2007,
46, 1803–1806; (h) K. Ono, M. Yoshizawa, T. Kato and M. Fujita,
Chem. Commun., 2008, 2328–2330.
4 Y. Yamauchi, M. Yoshizawa and M. Fujita, J. Am. Chem. Soc.,
2008, 130, 5832–5833.
5 HETTAP: this strategy makes use of steric and electronic effects
originating from bulky aryl substituents at the bisimine coordina-
tion sites to control the coordination equilibrium both kinetically
and thermodynamically. The steric stoppers at coordination sites
(2, 9 positions) of phenanthroline prevent any competitive homo-
leptic combination with itself, therefore leading to just hetero
combinations; (a) M. Schmittel, V. Kalsani, R. S. K. Kishore,
J. W. Bats and H. Colfen, J. Am. Chem. Soc., 2005, 127,
¨
11544–11545; (b) M. Schmittel, V. Kalsani, P. Mal and J. W. Bats,
Inorg. Chem., 2006, 45, 6370–6377; (c) M. Schmittel, B. He, V.
Kalsani and J. W. Bats, Org. Biomol. Chem., 2007, 5, 2395–2403;
(d) M. Schmittel and P. Mal, Chem. Commun., 2008, 960–962.
6 M. Schmittel, B. He and P. Mal, Org. Lett., 2008, 10, 2513–2516.
7 This protocol helps to suppress the formation of terpy-Zn(II)-terpy
complexes after addition of tristerpyridine ligand BT.
8 S. Bonnet, J.-P. Collin and J.-P. Sauvage, Inorg. Chem., 2007, 46,
10520–10533.
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4723–4725 | 4725