Efficient synthetic access to cationic dendrons and their application for ZnO nanoparticles surface functionalization: New building blocks for dye-sensitized solar cells

Article abstract of DOI:10.1021/ja106076h

A new concept for the efficient synthesis of cationic dendrons, 4-tert-butyl-1-(3-(3,4-dihydroxybenzamido)benzyl)pyridinium bromide (17), 1,1′-(5-(3,4-dihydroxybenzamido)-1,3-phenylene)bis(methylene) bis(4-tert-butylpyridinium) bromide (18), N1,N7-bis(3-(

Full text of DOI:10.1021/ja106076h

Published on Web 12/01/2010
Efficient Synthetic Access to Cationic Dendrons and Their
Application for ZnO Nanoparticles Surface Functionalization:
New Building Blocks for Dye-Sensitized Solar Cells
Jan-Frederik Gnichwitz,† Renata Marczak,‡ Fabian Werner,§ Nina Lang,†
Norbert Jux,† Dirk M. Guldi,*,§ Wolfgang Peukert,*,‡ and Andreas Hirsch*,†
Department of Chemistry and Pharmacy & Interdisciplinary Center of Molecular Materials
(ICMM), Friedrich-Alexander-UniVersity Erlangen-Nuremberg, Henkestrasse 42,
91054 Erlangen, Germany, Institute of Particle Technology, Friedrich-Alexander-UniVersity
Erlangen-Nuremberg, Cauerstrasse 4, 91058 Erlangen, Germany, and Department of Chemistry
and Pharmacy & Interdisciplinary Center of Molecular Materials (ICMM),
Friedrich-Alexander-UniVersity Erlangen-Nuremberg, Egerlandstrasse 3,
91058 Erlangen, Germany
Received July 9, 2010; E-mail: w.peukert@lfg.uni-erlangen.de; guldi@chemie.uni-erlangen.de;
andreas.hirsch@chemie.uni-erlangen.de
Abstract: A new concept for the efficient synthesis of cationic dendrons, 4-tert-butyl-1-(3-(3,4-dihydroxy-
benzamido)benzyl)pyridinium bromide (17), 1,1′-(5-(3,4-dihydroxybenzamido)-1,3-phenylene)bis(methyl-
ene)bis(4-tert-butylpyridinium) bromide (18), N1,N7-bis(3-(4-tert-butyl-pyridium-methyl)phenyl)-4-(3-(3-(4-
tert-butyl-pyridinium-methyl)phenyl-amino)-3-oxopropyl)-4-(3,4-dihydroxybenzamido)heptanediamide tribromide
(19), and N1,N7-bis(3,5-bis(4-tert-butyl-pyridium-methyl)phenyl)-4-(3-(3,5-bis(4-tert-butyl-pyridinium-meth-
yl)phenylamino)-3-oxopropyl)-4-(3,4-dihydroxybenzamido)heptanediamide hexabromide (20), and their facile
binding to zinc oxide (ZnO) nanostructures is introduced. Dendrons containing highly reactive benzylic
bromides reacted readily with 4-tert-butyl-pyridine and resulted in cationic dendrons. Furthermore, these
permanently positively charged dendrons were equipped with a catechol anchor group. This enabled ZnO
surface functionalization by simple immersion. The adsorption of 17, 18, 19, and 20 on the colloidal
nanoparticles was monitored by Langmuir isotherms. The highest obtained experimental loadings correspond
to 99.5%, 98.6%, 99.1%, and 42.5% of the particle surface for 17, 18, 19, and 20, respectively. These
results indicate insufficient adsorption of the largest molecule 20 leading to reduced colloidal stability of
the nanoparticles, while an enhanced stability after grafting with 17, 18, and 19 was observed. Mesoporous
films suitable for the use as electrodes in dye-sensitized solar cells (DSSCs) were prepared. Subsequently,
the films were functionalized with 18, 19, or 20 and sensitized with zinc-5,15-bis-[2′,6′-bis-{2′′,2′′-bis-
(carboxy)-ethyl}-methyl-4′-tert-butyl-pheny]-10,20-bis-(4′-tert-butylphenyl)porphyrin-octasodium-salt. UV-vis
absorption spectra confirmed that 18, 19, and 20 are suitable for the stable electrostatic attachment of the
dye. Current-voltage characteristics of complete cells demonstrated that increasing positive functionalization
of the ZnO surface leads to decreased open circuit voltages (Voc). All Voc values were around 0.4 V with
a maximum for the 18 functionalized ZnO film of 0.45 V. The maximum cell efficiency obtained (0.31%) is
rather high, considering the narrow spectral absorption of the dye and the rather thin ZnO films used.
Finally, incident photon to current efficiency (IPCE) measurements confirmed photoinduced electron injection
from the dye. These features are important assets for applications in particle technology and even facilitated
advanced devices like a supramolecular DSSC complete with a subsequent layer of negatively charged
porphyrins.
Introduction
as enzymes and DNA. Recently, supramolecular organization
of charged artificial polymers using noncovalent Coulomb
interactions has emerged as a very powerful construction
principle for novel hierarchically ordered molecular materials
with unprecedented mechanical1,2 and electronic3-5 properties.
As an example of this approach, we have recently reported the
Coulomb interactions are rather fundamental noncovalent
binding motifs in nature, where they play an important role as
key interactions, for example, in biological macromolecules such
† Department of Chemistry and Pharmacy & Interdisciplinary Center of
Molecular Materials (ICMM), Friedrich-Alexander-University Erlangen-
Nuremberg, Henkestrasse 42.
(1) Decher, G. Science 1997, 277, 1232.
‡ Institute of Particle Technology, Friedrich-Alexander-University Er-
langen-Nuremberg.
(2) Kaschak, D. M.; Lean, J. T.; Waraksa, C. C.; Saupe, G. B.; Usami,
H.; Mallouk, T. E. J. Am. Chem. Soc. 1999, 121, 3435.
(3) Pitigala, P. K. D. D. P.; Seneviratne, M. K. I.; Perera, V. P. S.;
Tennakone, K. Langmuir 2004, 20, 5100.
§ Department of Chemistry and Pharmacy & Interdisciplinary Center of
Molecular Materials (ICMM), Friedrich-Alexander-University Erlangen-
Nuremberg, Egerlandstrasse 3.
(4) Carlisle, J. R.; Weck, M. Funct. Org. Mater. 2007, 261.
9
17910 J. AM. CHEM. SOC. 2010, 132, 17910–17920
10.1021/ja106076h  2010 American Chemical Society

Efficient Synthetic Access to Cationic Dendrons
A R T I C L E S
Figure 1. Dispersion of unfunctionalized nanoparticles without stabilization (on the top) resulting in sedimentation and stabilized nanoparticles functionalized
with charged dendrons (on the bottom) resulting in a very stable suspension.
strong interaction between oppositely charged fullerenes and
porphyrins/cytochrome c.6 In fact, the resulting hybrid structures
possess sufficiently strong electronic coupling between their
constituent components to permit internal charge separation
processes to occur. In another contribution, the interplay of
electrostatic attractions and repulsions was used as the modus
operandi between charged polyions to form self-assembled
multilayers, where the assembly and degradation behavior were
tested.7 Moreover, we have also reported on the strong binding
between metalloporphyrins equipped with catechol anchoring
groups and zinc oxide (ZnO) nanoparticles used for the
construction of dye-sensitized solar cells (DSSCs).8 Because
of its direct bandgap as well as excellent thermal, chemical,
and structural features, ZnO has emerged as a fascinating
material for a broad range of electronic, optical, and piezoelectric
applications. In particular, recent interest in ZnO has focused
on its application in light emitting diodes,9 transparent
electrodes,10,11 sensors,12 and photovoltaics.13-18
The successful utilization of nanoparticles is often limited
by their tendency to aggregate, a feature that stems from their
high surface area-to-volume ratio. The improvement of stability
can directly influence the ability of nanoparticles to form densely
packed layers (Figure 1), a morphology essential to the
successful operation of, for example, thin-film transistors,19,20
(9) Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370,
354.
(10) Dura´n, P.; Capel, F.; Tartaj, J.; Moure, C. AdV. Mater. 2002, 14, 137.
(11) Yang, M.; Wang, D.; Peng, L.; Zhao, Q.; Lin, Y.; Wei, X. Sens.
Actuators, B 2006, 117, 80.
(12) Rout, C. S.; Raju, A. R.; Govindaraj, A.; Rao, C. N. R. Solid State
Commun. 2006, 138, 136.
(13) Gra¨tzel, M. J. Photochem. Photobiol., C: Photochem. ReV. 2003, 4,
145.
(14) O’Regan, B.; Gratzel, M. Nature 1991, 353, 737.
(15) Quintana, M.; Edvinsson, T.; Hagfeldt, A.; Boschloo, G. J. Phys. Chem.
C 2006, 111, 1035.
(5) Ponader, S.; Rosenlehner, K.; Vairaktaris, E.; von Wilmowsky, C.;
Schlegel, K.; Neukam, F.; Schmidt, C.; Schunk, T.; Hirsch, A.;
Nkenke, E. J. Mater. Sci.: Mater. Med. 2009, 20, 2455.
(6) Guldi, D. M.; Zilbermann, I.; Anderson, G.; Li, A.; Balbinot, D.; Jux,
N.; Hatzimarinaki, M.; Hirsch, A.; Prato, M. Chem. Commun. 2004,
726.
(16) Redmond, G.; Fitzmaurice, D.; Graetzel, M. Chem. Mater. 1994, 6,
686.
(17) Zeng, L.; Dai, S.; Xu, W.; Wang, K. Plasma Sci. Technol. 2006, 8,
172.
(18) Werner, F.; Gnichwitz, J.-F.; Marczak, R.; Palomares, E.; Peukert,
W.; Hirsch, A.; Guldi, D. M. J. Phys. Chem. B 2010, 114, 14671.
(19) Walther, S.; Scha¨fer, S.; Jank, M. P. M.; Thiem, H.; Peukert, W.;
Frey, L.; Ryssel, H. Microelectron. Eng. 2010, 87, 2312.
(20) Sun, B.; Sirringhaus, H. Nano Lett. 2005, 5, 2408.
(7) Rosenlehner, K.; Schunk, T.; Jux, N.; Brettreich, M.; Hirsch, A. Org.
Biomol. Chem. 2008, 6, 2697.
(8) Marczak, R.; Werner, F.; Gnichwitz, J.-F.; Hirsch, A.; Guldi, D. M.;
Peukert, W. J. Phys. Chem. C 2009, 113, 4669.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17911

A R T I C L E S
Gnichwitz et al.
Figure 2. Illustration of electrostatic interaction between the negatively charged dye 21 [2′,6′-bis-{2′′,2′′-bis-(carboxy)-ethyl}-methyl-4′-tert-butyl-pheny]-
10,20-bis-(4′-tert-butylphenyl)porphyrin-octasodium-salt and the ZnO surface modified with positively charged dendron 18 in the supramolecular dye-
sensitized solar cell.
transparent conductive films,21,22 or solar cells.3,23 Various
strategies exist that impart colloidal stability on nanoparticles
and that rely mainly on the provision of electrostatic or steric
repulsions between the particles.24,25 A highly effective com-
bination of these effects, often referred to as electrosteric
stabilization, can be achieved by modifying the surface of
nanoparticles with ionic surfactants, charged oligomers, or
polyelectrolytes. This procedure may lead to an increase,
decrease, or even inversion of the apparent particle surface
charge, while the conformational entropy of the adsorbate
provides additional protection against aggregation. Besides
improving the colloidal stability, the modification of solid
surfaces by such charged moieties is of high technological
significance because the electrostatic interaction allows differ-
ently functionalized particles to behave as self-assembling
“building blocks”.
occur.33,34 The second general method for surface modification
circumvents this by anchoring the absorbate through the form-
ation of covalent bonds with the substrate (chemisorption).35-37
One issue that still remains, however, is that the degree of
charging depends on the pH, which may limit the effectiveness
of the functionalized particle to form self-assembled structures.
It is therefore a particular challenge to the synthetic community
to produce chemisorbing moieties that possess pH-independent
permanent charges.
In the present work, we report the synthesis of cationic
dendrons, which can be readily anchored to ZnO nanoparticles
via a catechol group and which possess dendrons with a pH-
independent charge. This approach leads to the formation of
more homogenously coated particles and surfaces than methods
where chemical post treatments are performed after the grafting
of precursors.35-37 Our cationic dendrons provide not only an
excellent colloidal stability to the particulate systems, but also
a modified ZnO surface, which is suitable for assembling with
oppositely charged entities in a layer-by-layer fashion.1 We
demonstrate this possibility by assembling an anionic dye
molecule onto films of the cationic functionalized ZnO nano-
particles and applying the structure as a photoanode in a dye-
sensitized solar cell (Figure 2). We believe that this is the first
example of a sensitizer being attached to the inorganic host in
such a fashion. We observe that the approach is successful due
to the built-in driving force for electrons to be drawn to the
metal oxide surface, given by the positive charges attached
between the acceptor (ZnO) and the dye.
There are two general approaches for attaching charged
molecules, oligomers, or macromolecules to particle surfaces.
First, polyelectrolytes can be bound physically (physisorption)
utilizing the inherent, but often pH dependent, surface charge
of the substrate26-31 or following its charge-inducing etching.32
One drawback of this approach is that the stability of the
physisorbed layer can be strongly influenced by the environment,
that is, solvents and pH, and desorption of the layer may even
(21) Reindl, A.; Mahajeri, M.; Hanft, J.; Peukert, W. Thin Solid Films 2009,
517, 1624.
(22) Mahajeri, M.; Voigt, M.; Klupp Taylor, R. N.; Reindl, A.; Peukert,
W. Thin Solid Films 2010, 518, 3373.
Results and Discussion
(23) Guldi, D. M.; Rahman, G. M. A.; Sgobba, V.; Kotov, N. A.; Bonifazi,
D.; Prato, M. J. Am. Chem. Soc. 2006, 128, 2315.
(24) Derjaguin, B. V.; Landau, L. D. Acta Physicochimica URSS 1941,
14, 633.
Syntheses of the Cationic Dendrons. Only few examples are
known in which permanent cationic charges are included in
functional molecules or dendrimers. These include the reaction of
tertiary amines with halogenides,38-42 the substitution of primary
(25) Verwey, E. J. W.; Overbeek, T. G. Theory of the Stability of Lyophobic
Colloids; Elsevier: Amsterdam, 1948.
(26) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210-
211, 831.
(27) Fendler, J. H.; Meldrum, F. C. AdV. Mater. 1995, 7, 607.
(28) Gao, M.; Richter, B.; Kirstein, S.; Mohwald, H. J. Phys. Chem. B
1998, 102, 4096.
(33) Ramachandran, R.; Somasundaran, P. J. Colloid Interface Sci. 1987,
120, 184.
¨
(34) Tanaka, H.; Odberg, L.; Wågberg, L.; Lindstro¨m, T. J. Colloid
(29) Pastoriza-Santos, I.; Koktysh, D. S.; Mamedov, A. A.; Giersig, M.;
Kotov, N. A.; Liz-Marza´n, L. M. Langmuir 2000, 16, 2731.
(30) Ariga, K.; Hill, J. P.; Ji, Q. Phys. Chem. Chem. Phys. 2007, 9, 2319.
(31) Vaccaro, A.; Hierrezuelo, J.; Skarba, M.; Galletto, P.; Kleimann, J. R.;
Borkovec, M. Langmuir 2009, 25, 4864.
Interface Sci. 1990, 134, 229.
(35) Mir, Y.; Auroy, P.; Auvray, L. Phys. ReV. Lett. 1995, 75, 2863.
(36) Biesalski, M.; Ruhe, J. Macromolecules 1999, 32, 2309.
(37) Zhang, H.; Ruhe, J. Macromolecules 2003, 36, 6593.
(38) Chen, C. Z.; Beck-Tan, N. C.; Dhurjati, P.; van Dyk, T. K.; LaRossa,
R. A.; Cooper, S. L. Biomacromolecules 2000, 1, 473.
(32) Fujita, S.; Shiratori, S. Nanotechnology 2005, 16, 1821.
9
17912 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Efficient Synthetic Access to Cationic Dendrons
A R T I C L E S
Scheme 1. (A) Stoddard’s Prototype of a Polycationic Dendrimer Obtained via a Substitution of Benzylic Bromides with Tertiary Amines and
Subsequent Counter Ion Exchange; (B) Cationic Charged Porphyrin of Jux Synthesized by a Condensation Reaction of a Benzaldehyde
with Benzylic Methylether Groups in the Ortho Position, Which Are Replaced by Benzylic Bromides with HBr/Acetic Acid and Subsequently
Substituted with tert-Butylpyridine, Leading to an Octa-cationic Porphyrin
bromide by pyridine,43,44 or the alkylation of polypyridines.45,46
Scheme 2. Synthesis of the Two Main Building Blocks,
3-(Methoxymethyl)aniline (on the Top, 1) and
Stoddart et al. published in 1997 a new method to synthesize
3,5-Bis(methoxymethyl)aniline (on the Bottom, 2)
polycationic dendrimers47 by using benzylic bromides as a
substitution center. These highly reactive centers react with
tertiary amines nearly quantitatively to yield architectures
decorated with numerous cationic charges. Jux modified this
method by employing a methoxyether-substituted benzaldehyde
for the synthesis of porphyrins, Scheme 1.48 The methoxyether
protecting groups preserve the reactivity of the benzylic bromide
during subsequent reaction steps. Acidic conditions using
hydrogen bromide in glacial acetic acid were selected to perform
the quantitative deprotection to yield the corresponding benzylic
bromide. The final replacement was performed with numerous
to the corresponding aniline derivative was performed by means
nucleophiles. Among the latter, pyridine is of great importance,
of iron in an aqueous solution of sodium chloride. The desired
because its use results in a quarternized nitrogen center and a
product was obtained in 55% yield without appreciable amounts
cationic charge balanced by bromide. The high reactivity of the
of any byproduct. Thus, the syntheses of 3-(methoxymethyl)a-
benzylic bromides due to their preference for substitution
niline and 3,5-bis(methoxymethyl)aniline (2) were successfully
reactions in the benzylic position as well as the nearly
quantitative conversion of the benzylic methoxyether to the
benzylic bromide inspired us to develop a system that displays
carried out. Here, the reaction starts from 1,3-dimethyl-5-
nitrobenzene, which was treated with N-bromosuccinimide
(NBS) and dibenzoyl peroxide (DBPO) in refluxing carbontet-
these features and can easily be used for the addition of extra
rachloride to obtain 1,3-bis(bromomethyl)-5-nitrobenzene.50
functionalities.
Light-induced bromination of the nitro-xylene leads mainly to
the monobrominated product, whereas the bromination with
Our synthesis started from 3-(methoxymethyl)aniline (1). As
described in the literature,49 1-(chloromethyl)-3-nitrobenzene
NBS afforded the target compound in 20% yield. To avoid
was treated with sodiummethanolate to obtain 1-(methoxym-
oligomeric products, the benzylic bromide was transformed into
ethyl)-3-nitrobenzene in very good yields (83%). The reduction
ether moieties prior to the reduction of the nitro-group. In this
regard, the ether synthesis was accomplished by using sodium-
(39) Schlapak, R.; Armitage, D.; Saucedo-Zeni, N.; Latini, G.; Gruber, H. J.;
methanolate, which yielded 1,3-bis(methoxymethyl)-5-nitroben-
zene. The final reduction to 3,5-bis(methoxymethyl)aniline (2)
was carried out in the same conditions as the procedure for
3-(methoxymethyl)aniline, Scheme 2.
The two main building blocks (1 and 2) are versatile
precursors to control the number of charges. The connection
between the aniline-derivatives and the compound to be
functionalized should preferably be an amide bond for stability
reasons. Thus, modified Steglich conditions51,52 with 1-(3-
dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC)
and 1-hydroxybenzotriazole (HOBT) as coupling agents are
used. Because of the decreasing yields in multiple coupling
reactions, we decided to develop a dendrimer containing
Mesquida, P.; Samotskaya, Y.; Hohage, M.; Cacialli, F.; Howorka,
S. Langmuir 2007, 23, 8916.
(40) Almarsson, O.; Adalsteinsson, H.; Bruice, T. C. J. Am. Chem. Soc.
1995, 117, 4524.
(41) Virboul, M. A. N.; Lutz, M.; Siegler, M. A.; Spek, A. L.; van Koten,
G.; Klein Gebbink, R. J. M. Chem.-Eur. J. 2009, 15, 9981.
(42) Plault, L.; Hauseler, A.; Nlate, S.; Astruc, D.; Ruiz, J.; Gatard, S.;
Neumann, R. Angew. Chem., Int. Ed. 2004, 43, 2924.
(43) Liu, G.-F.; Filipovi, M.; Ivanovi, I.; Beuerle, F.; Witte, P.; Hirsch, A.
Angew. Chem., Int. Ed. 2008, 47, 3991.
(44) Kuhnen-Clausen, D.; Hagedorn, I.; Bill, R. J. Med. Chem. 1979, 22,
177.
(45) Shifrina, Z. B.; Kuchkina, N. V.; Rutkevich, P. N.; Vlasik, T. N.;
Sushko, A. D.; Izumrudov, V. A. Macromolecules 2009, 42, 9548.
(46) Miskelly, G. M.; Webley, W. S.; Clark, C. R.; Buckingham, D. A.
Inorg. Chem. 1988, 27, 3773.
(47) Stoddart, J. F.; Ashton, P. R.; Shibata, K.; Shipway, A. N. Angew.
Chem., Int. Ed. Engl. 1997, 36, 2781.
(48) Jux, N. Org. Lett. 2000, 2, 2129.
(50) In, S.; Cho, S. J.; Lee, K. H.; Kang, J. Org. Lett. 2005, 7, 3993.
(51) Hoefle, G.; Steglich, W. Chem. Ber. 1972, 105, 1368.
(52) Neises, B.; Steglich, W. Angew. Chem. 1978, 90, 556.
(49) Wan, P.; Muralidharan, S.; McAuley, I.; Babbage, C. A. Can. J. Chem.
1987, 65, 1775.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17913

A R T I C L E S
Gnichwitz et al.
Scheme 3. Two Possible Ways To Synthesize the Dendrimers 7 and 8
the building blocks 1 and 2, which can multiply the amount of
charges with just one reaction. A useful dendrimer53 for
this purpose is the deprotected first generation Newkome-
Dendrimer54-56 because it possesses three carboxylic acids that
cannot only be utilized in basic media for negative charges but
also for further reactions. The focal nitro-group can be reduced
in a final reaction step and provides a new functional group for
further functionalization. We synthesized the target dendrimers
7 and 8 in two different ways (Scheme 3). In one case, the
reduction of the nitro group was done first, followed by its
protection with N-(9H-fluoren-2-ylmethoxycarbonyloxy)suc-
cinimide (FMOC-ONSu). The tert-butylester was cleaved with
formic acid followed by a 3-fold amide coupling with the
corresponding aniline 1 or 2. Finally, the fmoc protecting group
was removed by basic treatment with pyridine to obtain the
desired compounds 7 and 8. The other route leads first to the
3-fold amide coupling after acidic deprotection followed by
reduction of the focal nitro-group to the corresponding target
dendrimers. This route is two steps shorter than that aforemen-
tioned, and the overall yields (54% and 50% in comparison to
59% and 55%) are only a little lower. The amines 7 and 8 are
not stable over longer periods of time, so that they have to be
kept cold and stored in the dark.
3,4-Diphenylmethylenedioxyprotocatechuic acid57 was re-
acted with the amines 1, 2, 7, and 8 to obtain the intermediates
9, 10, 11, and 12, respectively (Scheme 4). Thereby, one
anchoring group can bear one, two, three, or six charges per
molecule, which allows control of the amount of charges on
the surface when these molecules are combined with dispersed
nanoparticles or surfaces of films. The dihydroxy functionality
of the catechol has to be protected because at least one
purification step via column chromatography on silica is needed.
Besides the high affinity of the catechol to ZnO- or TiO2-
surfaces, the attachment to SiO2 is also very strong, so that
purification of the reaction mixture via column chromatography
with silica leads to a high loss of product, which was avoided
conveniently via this protection step.
The diphenylmethane protecting group can be removed easily
with formic acid resulting in quantitative yields. Interestingly,
as the benzylic methyl ether is cleaved by treatment with
hydrobromic acid in glacial acetic acid, the deprotection of the
catechol unit and the generation of the benzylic bromide can
be achieved in one step. Both transformations are quantitative
including a simple workup. The benzylic bromides of the
resulting compounds 13, 14, 15, and 16 are very reactive. THF
was chosen for the quaternization of 13 and 14, but in the case
of 15 and 16 partly quarternized compounds precipitated
preventing complete conversion. In contrast, partially and fully
charged compounds remain dissolved in DMF, so that all target
cationic compounds 17, 18, 19, and 20 were accessible (Scheme
5 and Table 1).
(53) Astruc, D.; Boisselier, E.; Ornelas, C. T. Chem. ReV. 2010, 110, 1857.
(54) Newkome, G. R.; Behera, R. K.; Moorefield, C. N.; Baker, G. R. J.
Org. Chem. 1991, 56, 7162.
(55) Newkome, G. R.; Nayak, A.; Behera, R. K.; Moorefield, C. N.; Baker,
G. R. J. Org. Chem. 1992, 57, 358.
(57) Iacazio, G.; Pe´rissol, C.; Faure, B. J. Microbiol. Methods 2000, 42,
(56) Brettreich, M.; Hirsch, A. Synlett 1998, 1396.
209.
9
17914 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Efficient Synthetic Access to Cationic Dendrons
A R T I C L E S
Scheme 4. Synthesis of the Intermediates 9, 10, 11, and 12 via Modified Steglich Coupling Reaction
The progress of the quaternization can be followed easily
via H NMR spectroscopy. The signal of the protons of the
benzylic methylene group shifts upon substitution with the
pyridine unit to 5.8 ppm. Thus, if the signal at 4.5 ppm
corresponding to the protons of the benzylic bromide has
completely vanished, a full conversion is achieved.
estimation of the width of the particle size distribution (PSD).
In particular, the ZnO nanoparticles are characterized by a
narrow volume-weighted size distribution as represented in
Figure 4b. This PSD was obtained via an algorithm developed
by Segets et al.63 Thereby the algorithm converts the measured
absorption spectra into their single particle contributions using
previously published work of Bergstro¨m64 and Viswanatha et
al.65 From this PSD, a mean size of 4.5 nm was calculated
coinciding with the average particles size obtained by XRD.
Achieving uniform and stable dispersions of particles is a
key requisite for practical applications of colloidal ZnO nano-
particles. For example, mixing ZnO quantum dots with charged
molecules is expected to form stable colloids with a narrow
size distribution. To check the effect of the surface modification,
the dispersibilities of 17, 18, 19, and 20 grafted onto ZnO
nanoparticles were compared to that of as-synthesized ZnO. To
this end, the ZnO particles were simply mixed with 17, 18, 19,
and 20 in ethanol at 25 °C; see Experimental Section. Adsorption
isotherms of the oligocationic molecules on the surface of the
ZnO quantum dots are presented in Figure 5. The adsorption
behavior follows a Langmuir isotherm for all four components.
The highest experimental loadings of 0.97, 0.96, and 0.92 mg
m-2 of the adsorbed 17, 18, and 19 molecules correspond to
the maximum particle surface coverage of 0.98, 0.97, and 0.92
mg m-2, respectively, which was estimated theoretically (Table
2). In case of 20, the highest experimental coverage of only
0.56 mg m-2 is more than twice lower that the calculated
maximum amount of 1.28 mg m-2 equivalent to a monolayer.
These results indicate the insufficient adsorption of the largest
oligocationic molecule.
1
ZnO Quantum Dot Synthesis and Functionalization. The
adsorption of polyelectrolytes on oxide nanoparticles and its
application in preparing colloidal dispersions have been the
subjects of intensive theoretical and experimental studies. For
example, coating of polymers such as poly(methyl methacrylate)
(PMMA), poly(styrene) (PSt), poly(ethylene glycol) (PEG), or
poly(acrylic acid) (PAA) on the surface of ZnO nanoparticles
has been widely investigated.58-62 However, many chemical
steps are required during the adsorption procedure. Moreover,
this approach provides only limited control over the thickness
and homogeneity of the attached polymers. Therefore, the
present study is focused on grafting small cationic charged
molecules with defined sizes onto surfaces of ZnO quantum
dots that occurs without any further chemical procedures.
Accordingly, the ZnO quantum dots were synthesized in
ethanol by hydrolysis of zinc acetate with lithium hydroxide.
Figure 3 shows the XRD pattern of the dots that were aged for
4 h at 30 °C. All of the diffraction peaks are assigned to the
standard hexagonal phase of ZnO with a wurtzite crystal
structure, as reported in the JCPDS card (no. 36-1451, a ) 3.249
Å, c ) 5.206 Å). Importantly, no significant contributions from
impurities arise in the XRD analysis. The widths of the XRD
peaks confirm the nanocrystalline nature and the size of the ZnO
nanocrystals. Figure 4a displays the absorption spectrum of the
ZnO quantum dot ethanolic suspension with an absorption onset
at 357 nm. From the absorption behavior, we have access to an
Absorption spectra of ethanolic suspensions of ZnO recorded
in the absence and in the presence of the oligocationic molecules
are shown in Figure 6. The spectra were normalized with respect
to the 336 nm peak, which corresponds to the absorption of the
(58) Hong, R. Y.; Chen, L. L.; Li, J. H.; Li, H. Z.; Zheng, Y.; Ding, J.
Polym. AdV. Technol. 2007, 18, 901.
(59) Chibowski, S.; Paszkiewicz, M. Adsorpt. Sci. Technol. 2001, 19, 397.
(60) Liufu, S.; Xiao, H.; Li, Y. Powder Technol. 2004, 145, 20.
(61) Shim, J.-W.; Kim, J.-W.; Han, S.-H.; Chang, I.-S.; Kim, H.-K.; Kang,
H.-H.; Lee, O.-S.; Suh, K.-D. Colloids Surf., A 2002, 207, 105.
(62) Xiong, H.-M.; Wang, Z.-D.; Liu, D.-P.; Chen, J.-S.; Wang, Y.-G.;
Xia, Y.-Y. AdV. Funct. Mater. 2005, 15, 1751.
(63) Segets, D.; Gradl, J.; Taylor, R. K.; Vassilev, V.; Peukert, W. ACS
Nano 2009, 3, 1703.
(64) Bergstrom, L.; Meurk, A.; Arwin, H.; Rowcliffe, D. J. J. Am. Ceram.
Soc. 1996, 79, 339.
(65) Viswanatha, R.; Sapra, S.; Satpati, B.; Satyam, P. V.; Dev, B. N.;
Sarma, D. D. J. Mater. Chem. 2004, 14, 661.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17915

A R T I C L E S
Gnichwitz et al.
Scheme 5. Acidic Ether Cleavage and Final Quaternization Reaction To Obtain the Cationic Charged Target Molecules
ZnO quantum dots. In addition, the features of 17, 18, 19, and
20 (Figure 7) are also discernible. Two peaks at about 305 and
330 as well as a broad absorption with an onset at about 380
nm correspond to 17 and 18, whereas the peak around 290 nm
is a characteristic feature of 19 and 20. It is interesting, however,
that in the case of the ZnO quantum dots with 17, 18, and 19
(Figure 6) the absorption behaviors indicate that the particles
are stable in suspension without noticeable aggregation. Thus,
with a mean diameter below 10 nm clearly situated in the
Rayleigh regime, they are too small to scatter light. When the
diameter of particles increases due to their growth and/or
aggregation, scattering occurs and the extinction at λ > 364 nm
is nonzero.65 In line with our previous observations,8,66 the ZnO
particle growth for particles exceeding a size of 12 nm is
neglected. Thus, the only process that takes place in the
suspensions is particle aggregation, which is the origin for
scattering as observed in the case of 20 that is adsorbed onto
ZnO surface (Figure 6). Transparent suspensions of 17, 18, and
19 (Figure 8) grafted onto ZnO surfaces were rapidly obtained
after mixing the reactants. The corresponding suspensions
remained stable at room temperature for several weeks. How-
Figure 3. X-ray diffraction pattern of the ZnO nanoparticles aged at 30 °C
for 4 h. All peaks correspond to the ZnO wurtzite crystal structure. The broad-
ness of the XRD peaks reveals the nanocrystalline nature of the ZnO powders.
(66) Marczak, R.; Segets, D.; Voigt, M.; Peukert, W. AdV. Powder Technol.
2010, 21, 41.
9
17916 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Efficient Synthetic Access to Cationic Dendrons
A R T I C L E S
Figure 6. Normalized absorption spectra of 17 (green line), 18 (red line),
19 (black line), and 20 (blue line) adsorbed on the surface of ZnO quantum
dots measured in ethanol at 25 °C.
Figure 4. (a) Absorption spectra of the ZnO nanoparticles in ethanolic
solution recorded after 4 h of aging at 30 °C. (b) Volume density
distributions of the ZnO nanoparticles in ethanolic solution obtained by
inversion of the UV-vis absorption spectrum (Figure 4a).
Figure 7. Extinction spectra of 17 (green line), 18 (red line), 19 (black
line), and 20 (blue line) in ethanol at 25 °C.
particles have large negative or positive potentials, repulsion
dominates without appreciable aggregation. However, if particles
have low zeta potential values, the absence of repulsive forces
gives way to aggregation.67 Acetate groups, which originate
from the precursor materials, are adsorbed on the surface and
initially stabilize the ZnO nanoparticles. However, during the
three times repeated washing steps, when the still unreacted
reaction precursors (zinc acetate and lithium hydroxide) and the
reaction byproduct (lithium acetate) are removed, the number
of the protecting acetate groups is reduced. Therefore, in
case of three times washed ZnO nanoparticles, the remaining
amount of the acetate on the surface is unable to maintain
sufficient stabilization against agglomeration of the ZnO nano-
particles in ethanol for longer than 30 min. Thus, zeta potentials
of -8.9 ( 2.8 mV for the 0.05 M ethanolic suspensions of the
ZnO quantum dots relate to the low stability of the correspond-
ing suspension and a negative surface charge.
Figure 5. Adsorption isotherms of 17 (green 9 and line), 18 (red b and
line), 19 (2 and line), and 20 (blue [ and line) on the surface of ZnO
quantum dots at 25 °C.
ever, the suspension containing ZnO and 20 turned turbid,
indicating that the ZnO nanoparticles are subject to significant
aggregation (Figure 8).
To obtain quantitative information about the colloidal stability
of the ZnO quantum dots grafted with the 17, 18, 19, and 20,
zeta potential measurements were performed. In general, the
absolute value of the zeta potential correlates with the colloidal
stability due to electrostatic repulsions between particles. If
Figure 9 shows the zeta potential values of ZnO nanoparticles
as a function of the concentration of various oligocationic
molecules in ethanol. The absolute value of the zeta potential
is influenced by the presence of both the remaining acetate
(67) Lagaly, G.; Schulz, O.; Zimehl, R. Dispersionen und Emulsionen;
Steinkopff: Darmstadt, 1997.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17917

A R T I C L E S
Gnichwitz et al.
Figure 8. Visual observation of the stability of ZnO quantum dots grafted with 17 (a), 18 (b), 19 (c), and 20 (d) in ethanol at 25 °C.
Figure 10. UV/vis absorption spectra of ZnO mesoporous films function-
alized with 18 (red line), 19 (black line), and 20 (blue line) after sensitization
with 21.
Figure 9. Zeta potential of ZnO quantum dots for different additives
concentrations: 17 (green 9 and solid line), 18 (red b and line), 19 (2 and
line), and 20 (blue [ and line).
20) were observed during our experiments. The results show the
enhanced stability after grafting the ZnO quantum dots with 17,
18, and 19, while the stability is reduced after grafting with 20.
groups as well as the grafted oligocationic molecules on the
surface of the ZnO nanoparticles. In all cases, increasing zeta
potential values were observed with increasing dendrimer
concentrations, indicating the specific adsorption of the mol-
ecules onto the ZnO nanoparticles surface. The curves thus
reflect the adsorption behavior of the dendrimer molecules
(Figure 5). The initially low magnitude of the zeta potential
values indicates weak colloidal stability of the suspensions.
Further, the zeta potentials rise up to maximum values of +33.8
( 3.7 and +32.6 ( 3.1 mV for 17 and 18, respectively, at about
1 × 10-3 M. The increase of lower zeta potential values, which
is up to +29.5 ( 3.3 and +21.2 ( 2.5 mV for 19 and 20,
respectively, at about 3.5 × 10-4 M was noted. When the
concentrations of 17, 18, and 19 on the surface of the ZnO
quantum dots increase, higher absolute zeta potentials give rise
to stronger electrostatic repulsion between the particles. In stark
contrast, low zeta potentials as they are obtained for different
concentrations of 20 indicate that this molecule changes the sign
of the surface charge but is unable to maintain the stability
against agglomeration of the ZnO quantum dots in ethanol. The
overall changes in stability are rationalized on the basis of the
electrical double layer, which decreases, for example, toward
the higher charge concentrations on the particle surface.
In general, changes in the colloidal stability as a function of the
size and charge of the oligocationic molecules (17, 18, 19, and
Photophysical Characterization of DSSCs. Despite the afore-
mentioned advantages in the control over morphology down to
the nanoscale regime, there are only few examples utilizing
supramolecular principles in solar cell construction and design.3
Introducing the concept of noncovalent bonds (i.e., electrostatic
interaction) in solar cells opens several interesting possibilities,
for example, easy exchange of the absorbing species, introducing
extra driving force via a charge gradient and the facile
construction of multilayers. This is, to the best of our knowledge,
the first example for utilizing structurally similar anchors bearing
a different number of charges to influence the open circuit
voltage (Voc) systematically. In a proof of principle, the
photophysical characterization (i.e., UV-vis absorption spec-
troscopy, incident photon to current efficiency (IPCE), and
current voltage measurements) of a supramolecular DSSC is
shown. The absorption spectra of ZnO films functionalized with
18, 19, or 20 and sensitized with zinc-5,15-bis-[2′,6′-bis-{2′′,2′′-
bis-(carboxy)-ethyl}-methyl-4′-tert-butyl-pheny]-10,20-bis-(4′-
tert-butylphenyl)porphyrin-octasodium-salt6 21 are gathered in
Figure 10. There is no appreciable difference in the amount of
dye adsorbed, leading us to the assumption that 18, 19, and 20
are all suitable for the stable electrostatic attachment of the dye.
9
17918 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

Efficient Synthetic Access to Cationic Dendrons
A R T I C L E S
Figure 12. IPCE spectra of differently functionalized ZnO films sensitized
with porphyrin, for ZnO electrodes grafted with 18 (red line), 19 (black
line), and 20 (blue line) before dye sensitization.
Figure 11. I-V characteristics of DSSCs under AM1.5 illumination at
100 mW/cm2 and in the dark (dashed lines), for ZnO electrodes grafted
with 18 (red line), 19 (black line), and 20 (blue line) before dye sensitization.
Next, current voltage characteristics were recorded for the
DSSCs (Figure 11). In terms of Voc, a clear trend evolves, which
suggests that the positive functionalization of the ZnO surface leads
to decreased Voc. Less concise is the trend for Isc, although the
values seem to decrease as well. All open circuit voltages are, as
is typical for porphyrin sensitized solar cells, around 0.4 V with a
maximum for the 18 functionalized ZnO film of 0.45 V. The
maximum efficiency is at 0.31%, which is, taking the low overall
absorption of the dye on the rather thin ZnO films into account,
outstanding. It is difficult to compare this value to other systems
due to the unique character of the supramolecular binding motif.
It is, however, in the range of systems using porphyrins covalently
bound to ZnO.68 The results of the current voltage measurements
are summarized in Table 1.
were prepared according to the procedures reported earlier.8,66
Briefly, 1.09 g of zinc acetate dihydrate (ACS grade, 98.0%, VWR,
Germany) was dissolved in 50 mL of boiling ethanol (99.98%,
VWR, Germany) at atmospheric pressure and cooled to room
temperature. Next, 0.17 g of lithium hydroxide (98%, VWR,
Germany) was dissolved in 50 mL of boiling ethanol and also
cooled to the synthesis temperature. Further, the lithium hydroxide
solution was added dropwise to the cold zinc acetate solution under
vigorous stirring. Immediately after synthesis, aging processes
commence with an evolving particle diameter.8,66 Next, after 4 h
of aging at 30 °C, the ZnO nanoparticles were three times washed
to remove the reaction byproduct lithium acetate by repeated
flocculation affected by addition of n-heptane (Rotisolve HPLC,
Carl Roth GmbH + Co. KG, Germany). White ZnO flocculates
precipitated immediately after the addition of heptane. Afterward,
the supernatant containing the lithium acetate was separated from
the white ZnO precipitate by centrifugation at 3500 rpm for 10
min (Labofuge 400, Heraeus Instruments GmbH, Germany) and
decantation. For colloidal characterization, the ZnO flocculates were
redispersed in ethanol.
Table 1. Summary of Device Characteristics of DSSCs under
AM1.5 Illumination at 100 mW/cm2, for ZnO Electrodes Grafted
with 18, 19, or 20 before Dye Sensitization
oligocationic
molecule
Isc (mA)
Voc (V)
FF (%)
η (%)
ZnO Quantum Dots Surface Functionalization. To examine
the capability for the surface functionalization of the ZnO surface,
three times washed ZnO quantum dots were mixed with a different
amount of the oligocationic molecules in ethanol (concentrations
from 5 × 10-6 to about 1 × 10-3 M). The mixture was sonicated
for 15 min for the sake of adsorption. The mass of the adsorbed
molecules on the surface of the ZnO nanoparticles was obtained
from the difference of the initial amount of the molecules used for
the suspensions preparation and the amount of the not adsorbed
molecules. Thus, the suspensions of ZnO nanoparticles and different
amount of the oligocationic macromolecules were centrifuged until
the covered ZnO nanoparticles completely settled down. Next, the
absorption spectra of the supernatants were measured at 25 °C to
determine the amount of the not adsorbed molecules. The data were
considered in terms of the mass of adsorbed oligocationic molecules
per unit area of the ZnO quantum dots as a function of the total
amount of molecules added to the suspension.
The surface of as-synthesized ZnO quantum dots is not bare,
but acetate groups, which originate from the precursor materials,
are adsorbed on the surface of the crystals.8,66 The amount of the
residual acetate of 10 wt % was determined by measuring the weight
loss that occurred upon heating the sample in thermogravimetric
analysis (TGA) (not shown).
Considering the particle size of 4.5 nm as calculated from the
absorption spectrum (Figure 3a) by an algorithm developed by
Segets et al.,63 the total surface area of a ZnO nanoparticle was
calculated to be 63.6 nm2. Thus, the average number of the acetates
after the third washing step was estimated to 5 molecules per nm2
18
19
20
1.39
0.9
1.02
0.45
0.42
0.40
63
38
52
0.31
0.20
0.21
Finally, in IPCE measurements (Figure 12), we confirmed
that electron injection is indeed evolving from the dye, because
the IPCE spectral overlap resembles the absorption features of
the dye. The slight red-shift of the IPCE value can be attributed
to down-conversion of higher energy photons due to scattering.
The photon to current conversion efficiency is disproportionately
favored in the region of the Q-band absorptions (i.e., 550-610
nm) when contrasted to that of the Soret band (i.e., around 430
nm) absorptions.
The electrostatic interaction is an efficient and versatile motif
for the construction of DSSCs. With the permanently charged
molecules synthesized, a stable interaction even in the high ionic
strength environment of the electrolyte could be achieved
between the mesoporous ZnO surface and the dye.
Experimental Section
Synthesis of ZnO Quantum Dots. All chemicals were analytical
grade reagents used without further purification. ZnO quantum dots
(68) Galoppini, E.; Rochford, J.; Chen, H.; Saraf, G.; Lu, Y.; Hagfeldt,
A.; Boschloo, G. J. Phys. Chem. B 2006, 110, 16159.
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17919

A R T I C L E S
Gnichwitz et al.
and is equivalent to 43% of a monolayer.66 Therefore, just 57%
(36.3 nm2) of the particle surface could be grafted with the
oligocationic molecules.
the conducting glass electrode, which was allowed to dry in air
prior to firing at 380 °C for 20 min. The ZnO and platinum counter
electrode was then sealed together with a transparent film of Surlyn
1472 (Dupont) cut as a frame around the nanocrystalline ZnO film.
A solution of 0.6 M 1-methyl-3-propylimidazolium iodide (MPII),
0.03 M I2 in methoxyacetonitrile, was employed as electrolyte. The
electrolyte was introduced through a hole drilled in the counter
electrode and immediately sealed.
Additionally, the average surface areas have been optimized
theoretically in a vacuum and using an implicit solvent environment
in MeOH and H2O within the semiempirical MO-theory using the
AM1 Hamiltonian. From the optimized geometries, the surface areas,
which are shown in Table 2, have been computed according to the
formula for triangular surfaces. Thus, to form a monolayer, the maximal
number of the adsorbed molecules per nanoparticle,66 concentrations
of the oligocationic molecules required to completely cover the ZnO
surface in a 0.05 M ethanolic suspensions, and the maximal amount
of the molecules adsorbed on the total particles surface could be
estimated. All of the values are summarized in Table 2.
Conclusions
In short, advanced inorganic nanostructures via stable attachment
of cationically charged monolayers onto ZnO nanoparticles, and
solid-state mesoporous ZnO surfaces were realized. Remarkable
improvement in colloidal stability ensured the integration into
photovoltaic devices. The key to our approach was a novel synthetic
route generating highly reactive benzylic bromides, which react
readily with nucleophiles like pyridine resulting in quarternized
N-atoms and generating permanent cationic charges in very high
yields and very good purity. 3-Methoxymethylaniline and 3,5-
bis(methoxymethyl)aniline were successfully connected to car-
boxylic acid containing building blocks and transformed quanti-
tatively into the corresponding benzylic bromides by means of
hydrobromic acid. The cationic charges are obtained by the reaction
with 4-tert-butylpyridine leading to the quantitative formation of
cationic charges. In addition to the cationic functionality, these
dendritic systems are equipped with a catechol anchoring group,
which forms a very strong covalent bond to ZnO nanostructures.
A distinctive factor of this approach is that once the molecules are
attached to the surface, a chemical post-treatment is not necessary
anymore delivering high homogeneity of the mono molecularly
surface layer of 17, 18, 19, and 20. Importantly, changes in the
colloidal stability of the ZnO nanoparticles grafted with the cationic
dendric molecules were observed and directly linked to the size
and charge of the attached molecules. Greatly enhanced stability
after grafting the ZnO quantum dots with 17, 18, and 19 is shown,
while the stability is reduced after grafting with 20. Motivated by
these results, the successful integration of molecularly modified
mesoporous ZnO nanostructures into a DSSC device has been
demonstrated in a proof of principle experiment. The advanced
design of the molecules allows stable attachment of oppositely
charged porphyrins via an LbL approach even in the high ionic
strength environment of an electrolyte commonly utilized in DSSC
systems. The relatively high efficiencies in regard to ZnO and
porphyrin-based DSSCs were explained by the build-in driving
force provided by the positive charge of the polyions. These already
very promising results (i.e., efficiencies of up to 0.31%) are
expected to be greatly enhanced by the utilization of dye multi-
layers. Thus, by this approach, the complementary spectral proper-
ties of various dyes can be utilized, and even a built-in redox
gradient can further enhance the efficiency.
Table 2. Properties of the Oligocationic Molecules Adsorbed on
the ZnO Quantum Dots Surface
average
surface,
nm2
max. number
of adsorbed
molecules
concentration
required to form
a monolayer, mol L-1
max. adsorbed
amount,
mg m-2
oligocationic
molecule
17
18
19
20
0.77
1.18
2.32
2.56
47
30
15
14
11.8 × 10-4
7.8 × 10-4
3.9 × 10-4
3.6 × 10-4
0.98
0.97
0.93
1.28
Characterization of ZnO Nanoparticles. Optical properties of
the ZnO nanoparticles were determined from UV-vis absorption
spectra recorded at 25 °C using a Cary 100 Scan UV-visible
spectrophotometer (Varian Deutschland GmbH, Germany) with a
cuvette of 10 mm optical path length. The zeta potentials of 0.05
mol L-1 ZnO ethanolic suspensions (with respect to the Zn2+
concentration) were obtained at 25 °C via dynamic light scattering
(DLS) by using a Malvern Nano ZS Instrument (Malvern Instru-
ments GmbH, Germany) using laser Doppler electrophoresis.
Structural analysis of the ZnO nanoparticles was performed with a
D8 Advance (Bruker AXS, Germany) X-ray diffractometer (XRD)
in the Bragg-Brentano geometry using Cu KR radiation (λ ) 0.154
nm). The XRD patterns were recorded in the range of 2θ from 20°
to 70°.
Device Fabrication. The ZnO nanoparticles suitable to be used
in a dye-sensitized solar cell were prepared by dissolving 7.8 g of
zinc acetate dihydrate (ACS grade, 98.0%, VWR, Germany) in 43.6
mL of methanol (99.9%, Carl Roth GmbH & Co KG, Germany).
Next, ZnO was precipitated by adding tetramethylammonium
hydroxide (25% (w/w) in methanol, Sigma-Aldrich, Germany). The
reaction mixture was stirred for 3 h at 75 °C under refluxing. During
this time, flocculation of the nanoparticles was observed. To remove
the remaining salts, the ZnO flocculates were centrifuged. The
supernatant was removed by decantation, and the ZnO flocculates
were washed with ethanol. The particles were redispersed by
sonication for 15 min.
ZnO paste was fabricated by mixing the obtained nanoparticles
with 10% ethanolic solutions of ethyl cellulose (5-15 mPa and
30-50 mPa, Aldrich) in a 1:1 ratio. The resulting transparent paste
was doctor bladed onto TEC 15 F/SnO2 conducting glass substrates
utilizing two layers of Scotch magic tape (approximately 100 µm)
as spacer. After evaporation of the solvent, the films were calcinated
at 450 °C, yielding transparent mesoporous films of 4 µm thickness.
The films were then immersed in 10-4 M ethanolic solutions of
18, 19, and 20 overnight. Afterward, the films were rinsed with
ethanol and dried in a stream of nitrogen. The functionalized films
where then immersed in a 10-5 M aqueous solution of zinc-5,15-
bis-[2′,6′-bis-{2′′,2′′-bis-(carboxy)-ethyl}-methyl-4′-tert-butyl-pheny]-
10,20-bis-(4′-tert-butylphenyl)porphyrin-octasodium-salt2 for 4 days
to guarantee efficient attachment and ion exchange.
Acknowledgment. We gratefully acknowledge the funding of
the German Research Council (DFG), which, within the framework
of its “Excellence Initiative”, supports the Cluster of Excellence
“Engineering of Advanced Materials” at the University of Erlangen-
Nuremberg. We are grateful to Prof. R. N. Klupp Taylor for
comments and discussion. We dedicate this article to Eiichi
Nakamura on the occasion of his 60th birthday.
Supporting Information Available: Detailed synthetic pro-
cedures and spectroscopic data for all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
DSSCs have been fabricated from these slides. For counter
electrode fabrication, uncoated conducting glass slides were cut into
pieces, and holes of 1 mm2 were drilled at the edge of the active
area. A thin film of H2PtCl6 (0.01 M in ethanol) was spread over
JA106076H
9
17920 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010