Supramolecular assembly of zinc salphen complexes: Access to metal-containing gels and nanofibers

Article abstract of DOI:10.1002/anie.200702680

(Chemical Equation Presented) Salphen the puzzle: Zn(salphen) complexes such as that shown organize into nanofibrils in the solid state. This assembly involves an intermolecular Zn...O interaction that stabilizes a polymeric structure with a (ZnO)_n

Full text of DOI:10.1002/anie.200702680

Communications
DOI: 10.1002/anie.200702680
Nanofibers and Gels
Supramolecular Assembly of Zinc Salphen Complexes: Access to
Metal-Containing Gels and Nanofibers**
Joseph K.-H. Hui, Zhen Yu, and Mark J. MacLachlan*
The supramolecular self-assembly of molecules into 1D
nanostructures is an important goal for developing future
nanoscale technologies, such as molecular circuitry and
machines.[1–5] Nanowires and nanotubes can be assembled
using noncovalent interactions, including hydrogen-bonding
and intermolecular p–p stacking.[6–10] The formation of nano-
fibrils is also relevant to biological systems and diseases (for
example, Alzheimerꢀs).[11] Moreover, there has been much
recent interest in the gelation of fiber-forming assemblies.[12]
Herein we report the first examples of nanofiber forma-
tion promoted by Zn···O interactions between Schiff-base
complexes. Such interactions between zinc and a phenolic
oxygen have been reported as the basis of the reversible
dimerization of [Zn(salen)]-type complexes (salen = N,N’-
bis(salicylidene)ethylenediamine dianion).[13,14] In these com-
plexes, the metal is unable to adopt a tetrahedral geometry
and therefore dimerizes to give a five-coordinate metal
center. We have ascribed the aggregation of large, zinc-
containing macrocycles in solution to Zn···O interactions.[15]
Batley and Graddon proposed in 1967 that [Zn(salen)]
complexes could form polymeric structures,[16] but definitive
evidence for aggregation has not been reported. We present a
newfamily of Schiff-base complexes that organize into gels
and 1D nanofibrils. Furthermore, we demonstrate the ability
to control the nanofibril widths by modification of the
substituents.
We prepared a series of salphen complexes 1–5 (salphen =
N,N’-bis(salicylidene)-o-phenylenediamine dianion) with
peripheral linear and branched alkoxy groups. We were
surprised to discover that zinc compound 4a forms lumines-
cent gels in methanol (ca. 3 mgmLÀ1), Figure 1. Gel forma-
tion was also observed in toluene, but not in DMF or
chloroform. In gel-forming solvents, formation was thermor-
eversible, and could be readily disrupted with addition of
pyridine (< 2%), which coordinates to the zinc center.
Table 1 shows the gel-forming abilities of compounds 1a–5a
in various solvents. Notably, the complexes form gels best in
aromatic solvents, such as benzene, toluene, and o-xylene.
[*] J. K.-H. Hui, Z. Yu, Prof. M. J. MacLachlan
Department of Chemistry
University of British Columbia
Figure 1. a) Photograph of gel 4a in methanol under visible light (left)
2036 Main Mall, Vancouver, BC V6T1Z1 (Canada)
and when irradiated with UV light (right). b) Fiber morphology of 4a
Fax: (+1)604-822-2847
cast from methanol, as observed by TEM. c) UV/Vis and fluorescence
E-mail: mmaclach@chem.ubc.ca
spectroscopy of 4a in methylene chloride and in the solid state (gel).
[**] The authors thank the Natural Sciences and Engineering Research
Council (NSERC) of Canada for a Special Research Opportunities
grant. Salphen=N,N’-bis(salicylidene)-o-phenylenediamine dia-
Transmission electron microscopy (TEM) revealed that
complexes 1a–5a all form fibers when cast from methanol
(Figure 2). In each case, the fibers are only tens of nanometers
nion.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
7980
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 7980 –7983

Angewandte
Chemie
Table 1: Gel-forming abilities of 1a–5a in various solvents.[a]
initially thought that the fiber morphology observed in the
[Zn(salphen)] complexes may be a consequence of the
hydrophobic alkoxy substituents on the salphen complexes,
which lead the complexes to aggregate in methanol. We
hypothesized that replacing the alkoxy substituents of 1–5
with hydrophilic substituents, such as monosaccharides, may
disrupt the organization or may favor the formation of single-
strand nanofibrils from methanol. To test this hypothesis,
compounds 6 and 7, functionalized with glucose and galac-
tose, respectively, were prepared. Drop-casting of Zn2+
complexes 6a and 7a from methanol, in which they are very
soluble, mostly led to the formation of nanofibrils on the
TEM grid (Figure 3). For both complexes, very narrow
structures with diameters of about 5–7 nm, but with lengths
that extend over micrometers, are formed. The fibers are all
of similar width, and are much narrower than those from 1a–
5a.
1a
2a
3a
4a
5a
MeOH
EtOH
Toluene
MeCN
EtOAc
Benzene
Pyridine
o-Xylene
DMF
–
–
–
–
–
–
–
–
–
–
–
Y
–
–
Y
–
Y
–
–
–
Y
–
–
Y
–
Y
–
Y
–
Y
–
–
Y
–
Y
–
–
–
Y
–
–
Y
–
Y
–
[a] Concentration 3 mgmLÀ1. Ygel observed; – no gel observed. OAc=
acetate.
Figure 3. TEM images of a),b) complex 6a and c),d) complex 7a
deposited from methanol.
We believe the 1D assembly observed for 1a–7a involves
Zn···O interactions rather than hydrogen-bonding or p–p
interactions between the salphen complexes. Although we
cannot directly observe Zn···O interactions, we have made
several model compounds (1b, 4b, 6b, 7b, 8, 9) and have
carried out control experiments that support this assign-
ment.[17] We rule out p–p interactions, which are frequently
Figure 2. TEM images of a) complex 1a, b) complex 2a, c) complex
4a, d) complex 5a, and e),f) complex 3a deposited from methanol.
in diameter but extend several microns in length. The
relatively large diameters and rope morphology of the fibers
suggest that they are not one molecule thick, but instead are
formed from the assembly of 1D fibers into bundles. Fiber
formation was observed for chain lengths ranging from
methoxy to hexadecyloxy.
The fibers on the TEM grids do not exhibit electron
diffraction and were too small to study by X-ray diffraction
(XRD). Powder XRD studies of dried films of the zinc
complexes cast from methanol do showsome crystallinity. In
the case of complex 2a, the peaks fit to a tetragonal unit cell
with a = b = 14.90 Š, c = 20.49 Š and a = b = g = 908. For the
other complexes, an insufficient number of peaks were
observed to determine the unit cell with confidence. We
Angew. Chem. Int. Ed. 2007, 46, 7980 –7983
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7981

Communications
seen in polycyclic aromatics, as the exclusive interaction, as
square-planar Ni2+ complexes 1b, 4b, 6b, and 7b do not show
similar organization to the Zn2+ complexes. TEM images of
the nickel complexes showonly ill-defined morphologies
when cast from methanol or solvent mixtures. Complexes 6
and 7 with copper or vanadyl in the place of zinc also show no
fiber assembly.
Complex 8 has bulky tert-butyl groups that should inhibit
aggregation of the zinc complexes. Indeed complex 8 does not
form a gel in methanol or in toluene, and TEM images of this
complex cast from methanol showed no fibrillar texture.
Complex 9, which is the thiol analogue to complex 4a, does
not self-assemble into fibers in methanol. In this case, the
aggregation is most likely inhibited by the bulky sulfur atoms
at the zinc center that prevent expansion of the coordination
sphere of the metal center. Additionally, the softer sulfur
atom would be expected to interact less strongly with zinc
than oxygen. These control compounds further support that
the aggregation is mediated at the metal center.
The UV/Vis spectra for films of Zn2+ complexes 1a–7a
exhibit small red-shifts and broadening relative to the
solution-phase spectra. Moreover, they are all luminescent,
emitting at 525–540 nm, red-shifted by 20 nm from the
solution phase. The similarity of the absorption and emission
spectra for solutions and films of 1a–7a indicates that a
similar aggregation mechanism is involved in both the
assembly of alkoxy- and carbohydrate-substituted complexes.
Further evidence for aggregation was obtained from
electrospray ionization mass spectra (ESI-MS) of the com-
plexes. All the spectra of the zinc complexes showed the
molecular ion, but also showed peaks corresponding to
aggregates. In fact, when the instrument was optimized for
measurements on 2a, we were able to observe singly and
doubly charged species up to nine monomers (Figure 4).
Although quantifying the individual species by ESI-MS is not
feasible, the technique is known to give a snapshot of species
present in solution. Thus, large aggregates of 2a are formed in
solution, supporting the strong tendency of these complexes
to aggregate.
Figure 4. ESI-MS spectrum of complex 2a exhibits singly (A–E) and
doubly (F–I) charged species. A: [2a+Na]+, B: [(2a)2 +Na]+, C:
[(2a)3 +Na]+, D: [(2a)4 +Na]+, E: [(2a)5 +Na]+, F: [(2a)3 +Na2]2+, G:
[(2a)5 +Na2]2+, H: [(2a)7 +Na2]2+, I: [(2a)9 +Na2]2+
.
Figure 5. Energy-minimized (PM3) calculated structure for a heptamer
of [Zn(salphen)]. a) The entire oligomer containing seven [Zn-
(salphen)] molecules connected into a 1D structure. b) An expanded
view of three [Zn(salphen)] units from the middle of the heptamer.
(Zn gray, O red, N blue, C green).
Semi-empirical (PM3) calculations were performed to
better understand the supramolecular organization in these
structures. Assuming the Zn2+ center is five-coordinate in the
fibers as observed in dimers of [Zn(salphen)]-type complexes,
an oligomer containing seven unsubstituted [Zn(salphen)]
was constructed and its energy minimized, starting from
several different conformations. Figure 5 shows an energy-
minimized structure for the 1D heptamer with five-coordinate
organization that may arise from the stacking of the salphen
moieties.
It is noteworthy that the aggregation behavior we have
observed for metal salphen complexes is distinct from that of
porphyrins.[10,18] Whereas porphyrins (and phthalocyanines)
are known to aggregate, their assembly is dominated by p–p
stacking or interactions of substituents, and the metal
dependence is generally less pronounced. As an example,
Shinkai and co-workers have observed improved gelation
properties of a copper-containing porphyrin over the analo-
gous zinc-containing porphyrin and have attributed this
improvement to the stabilization of H-aggregates.[18] The
coordination environment of salphen complexes is much
more flexible than that of porphyrin complexes, offering new
possibilities for self-assembly.
À
À
zinc ions. Calculated Zn O and Zn N distances are in
agreement with crystallographic studies of dimers. The
polymeric backbone contains a (ZnO)n polymer as illustrated
in Figure 5, with an insulating organic sheath.
An interesting revelation from the modeling is that the
complexes assembled in the 1D structures will likely assume a
helical conformation. We have observed regions of fibers of
the zinc complexes 1a–5a that appear to have helical
structure, and it is possible this motif derives from the helical
conformation of the polymer strand. Figures 2e,f show
regions of fibers formed from compound 3a. In these
images, there are segments that appear to have helical
7982 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 7980 –7983

Angewandte
Chemie
[8] S. Kawano, S. Tamaru, N. Fujita, S. Shinkai, Chem. Eur. J. 2004,
In conclusion, we have demonstrated the first supra-
molecular assembly of [Zn(salphen)] complexes into lumi-
nescent metal–organic gels and 1D nanofibers. This work
demonstrates the potential of reversible Zn···O interactions
for self-assembly of nanostructures. Structural changes to the
side groups enable control over the morphology and dimen-
sions of the nanofibers. Supramolecular assembly of salphen
complexes may be used for the controlled organization of
nanoscale patterns and wires, and the carbohydrate shell may
ultimately prove beneficial for biological integration. We are
nowstudying the scope of this fiber assembly.
10, 343 – 351.
[9] K. Balakrishnan, A. Datar, W. Zhang, X. Yang, T. Naddo, J.
Huang, J. Zuo, M. Yen, J. S. Moore, L. Zang, J. Am. Chem. Soc.
2006, 128, 6576 – 6577.
[10] For examples of aggregation of zinc porphyrins, see: a) R. J.
Abraham, S. C. Fell, H. Pearson, K. M. Smith, Tetrahedron 1979,
35, 1759 – 1766; b) T. Kishida, N. Fujita, O. Hirata, S. Shinkai,
Org. Biomol. Chem. 2006, 4, 1902 – 1909.
[11] W. H. Binder, O. W. Smrzka, Angew. Chem. 2006, 118, 7482 –
7487; Angew. Chem. Int. Ed. 2006, 45, 7324 – 7328.
[12] For recent examples of gelation, see: a) K. Sugiyasu, N. Fujita, S.
Shinkai, J. Mater. Chem. 2005, 15, 2747 – 2754; b) O. Lebel, M.-E.
Perron, T. Maris, S. F. Zalzal, A. Nanci, J. D. Wuest, Chem.
Mater. 2006, 18, 3616 – 3626; c) S. Kume, K. Kuroiwa, N.
Kimizuka, Chem. Commun. 2006, 2442 – 2444; d) W. Weng,
J. B. Beck, A. M. Jamieson, S. J. Rowan, J. Am. Chem. Soc. 2006,
128, 11663 – 11672; e) F. Camerel, L. Bonardi, M. Schmutz, R.
Ziessel, J. Am. Chem. Soc. 2006, 128, 4548 – 4549; f) G. John, G.
Zhu, J. Li, J. S. Dordick, Angew. Chem. 2006, 118, 4890 – 4893;
Angew. Chem. Int. Ed. 2006, 45, 4772 – 4775.
[13] a) J. Reglinski, S. Morris, D. E. Stevenson, Polyhedron 2002, 21,
2175 – 2182; b) M. Odoko, N. Tsuchida, N. Okabe, Acta Crys-
tallogr. Sect. E 2006, 62, m708 – m709; c) J. Sanmartꢂn Matalobos,
A. M. Garcia-Deibe, M. Fondo, D. Navarro, M. R. Bermejo,
Inorg. Chem. Commun. 2004, 7, 311 – 314; d) A. L. Singer, D. A.
Atwood, Inorg. Chim. Acta 1998, 277, 157 – 162.
[14] A. W. Kleij, M. Kuil, M. Lutz, D. M. Tooke, A. L. Spek, P. C. J.
Kamer, P. W. N. M. van Leeuwen, J. N. H. Reek, Inorg. Chim.
Acta 2006, 359, 1807 – 1814.
[15] C. T. L. Ma, M. J. MacLachlan, Angew. Chem. 2005, 117, 4250 –
4254; Angew. Chem. Int. Ed. 2005, 44, 4178 – 4182.
[16] G. E. Batley, D. P. Graddon, Aust. J. Chem. 1967, 20, 885 – 891.
[17] We cannot rule out the possibility that the zinc salphen
complexes dimerize then further assemble through p–p inter-
actions. However, MALDI-TOF mass spectrometry studies of
the fibers showthe monomeric zinc complex rather than the
dimer as the prevalent species.
[18] a) T. Kishida, N. Fujita, K. Sada, S. Shinkai, J. Am. Chem. Soc.
2005, 127, 7298 – 7299; b) T. Kishida, N. Fujita, K. Sada, S.
Shinkai, Langmuir 2005, 21, 9432 – 9439.
Received: June 19, 2007
Published online: September 12, 2007
Keywords: gels · nanostructures · Schiff bases ·
.
supramolecular chemistry · zinc
[1] C. Grave, A. D. Schlüter, Eur. J. Org. Chem. 2002, 3075 – 3098.
[2] Y. Yamaguchi, Z.-i. Yoshida, Chem. Eur. J. 2003, 9, 5430 – 5440.
[3] J. S. Moore, Acc. Chem. Res. 1997, 30, 402 – 413.
[4] S. Iijima, Nature 1991, 354, 56 – 58.
[5] A. M. Morales, C. M. Lieber, Science 1998, 279, 208 – 211.
[6] a) S.-i. Tamaru, M. Nakamura, M. Takeuchi, S. Shinkai, Org.
Lett. 2001, 3, 3631 – 3634; b) S.-i. Tamaru, M. Takeuchi, M. Sano,
S. Shinkai, Angew. Chem. 2002, 114, 881 – 884; Angew. Chem.
Int. Ed. 2002, 41, 853 – 856; c) M. Takeuchi, S. Tanaka, S. Shinkai,
Chem. Commun. 2005, 5539 – 5541; d) M.-a. Morikawa, M.
Yoshihara, T. Endo, N. Kimizuka, J. Am. Chem. Soc. 2005, 127,
1358 – 1359; e) C. OꢀDwyer, D. Navas, V. Lavayen, E. Bena-
vente, M. A. Sanata Ana, G. Gonzµlez, S. B. Newcomb, C. M.
Sotomayor Torres, Chem. Mater. 2006, 18, 3016 – 3022.
[7] For recent examples, see: a) S. Rosselli, A.-D. Ramminger, T.
Wagner, B. Silier, S. Wiegand, W. Häußler, G. Lieser, V.
Scheumann, S. Hꢁger, Angew. Chem. 2001, 113, 3233 – 3237;
Angew. Chem. Int. Ed. 2001, 40, 3137 – 3141; b) S. Hꢁger, K.
Bonard, S. Rosselli, A.-D. Ramminger, Macromol. Symp. 2002,
177, 185 – 191; c) D. Venkataraman, S. Lee, J. Zhang, J. S. Moore,
Nature 1994, 371, 591 – 593; d) D. Zhao, J. S. Moore, Chem.
Commun. 2003, 807 – 818.
Angew. Chem. Int. Ed. 2007, 46, 7980 –7983
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7983