Reversible photochromism of a ferrocenylazobenzene monolayer controllable by a single green light source

Article abstract of DOI:10.1039/b713107k

A 3-ferrocenylazobenzene monolayer on an ITO electrode exhibits reversible azobenzene isomerization using a single green light source, assisted by electrochemical control of the ferrocene redox state. The Royal Society of Chemistry.

Full text of DOI:10.1039/b713107k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Reversible photochromism of a ferrocenylazobenzene monolayer
controllable by a single green light source{
Kosuke Namiki, Aiko Sakamoto, Masaki Murata, Shoko Kume and Hiroshi Nishihara*
Received (in Cambridge, UK) 28th August 2007, Accepted 9th October 2007
First published as an Advance Article on the web 16th October 2007
DOI: 10.1039/b713107k
3-ferrocenyl-49-carboxylazobenzene (1), as a single-light-controlla-
ble chromophore to be immobilized on an ITO electrode.
A 3-ferrocenylazobenzene monolayer on an ITO electrode
exhibits reversible azobenzene isomerization using a single
green light source, assisted by electrochemical control of the
ferrocene redox state.
1 was newly synthesized via the acid-catalyzed coupling of
3-ferrocenylaniline with 4-nitrosobenzyl alcohol and subsequent
oxidation (Scheme S1). The photoreactivity of 1 in ethanol
solution was tested, and reversible photoisomerization behaviour
was observed (the cis molar ratio in its photostationary state (PSS)
was 59% with 365-nm UV light, 26% with 546-nm green light, and
13% with 436-nm blue light, see Fig. S1). Photoisomerization
behaviours on ITO nanoparticles were also examined by diffuse
reflectance UV–vis spectroscopy (Figs. S2 and S3) because the
aggregation of chromophores often prevents photoisomerization.8
Effective photoisomerization of 1 was observed on particles of ITO
as received, but was not observed on ITO particles hydrophilized
by H2O2–NH3 aq.–H2O = 1 : 1 : 5 solution to improve
immobilization.9 This result suggests that 1 was too crowded on
the hydrophilic surface and caused aggregation between azoben-
zene chromophores or ferrocene parts, and that photoisomeriza-
tion of 1 was prevented by the intermolecular stacking effects.
ITO–glass electrodes used in the following studies were therefore
treated without a hydrophilization process.
Optical data storage is a key issue for today’s information
technology, and photochromic molecules have been intensely
studied for their application to molecular-sized devices.1
Assemblies of photochromic azobenzene derivatives have attracted
much attention because they can achieve immobilization and high
density of the switching units.2–5 Azobenzene-based photoswitches
are usually manipulated with two different light sources, UV light
for a trans-to-cis isomerization and blue light for the reverse
isomerization. For the application to high-density data storage
manipulated with fine-focused light irradiation, focusing dual light
sources on the same small region can be a problem from the
perspective of both technical difficulty and cost performance.
Therefore, reversible isomerization of azobenzene in monolayers
with a single light source would be greatly advantageous.
We have previously reported the redox-conjugated reversible
photoisomerization of 3-ferrocenylazobenzene (3-FcAB).6,7 As
shown in Scheme 1a, 3-FcAB in the reduced Fe(II) state undergoes
trans-to-cis isomerization by the excitation of an MLCT band with
546-nm green light in addition to the normal UV and blue light
response (trans-to-cis isomerization by excitation of the azo p–p*
band with 365-nm UV light, and the reverse cis-to-trans
isomerization by excitation of the cis-azo n–p* band with
436-nm blue light). Once this compound is chemically or
electrochemically oxidized to 3-FcAB+ in the Fe(III) state, the
MLCT band disappears and cis-to-trans reverse isomerization
occurs with the same 546-nm light. This compound therefore
shows reversible isomerization with a single green light coupled
with the reversible redox reaction between Fe(II) and Fe(III). It is
thus interesting to assemble 3-FcAB derivatives on an electrode
surface with the aim of creating molecular-sized optical data
storage. We selected an indium-tin oxide (ITO) substrate as a
transparent electrode compatible with the characteristics of
The monolayer of 1 on an ITO–glass electrode (1/ITO) was
prepared by the following method. An ITO–glass substrate was
washed with acetone, detergent solution, water, and ethanol
3-FcAB, and designed
a
carboxylate-modified 3-FcAB,
Department of Chemistry, School of Science, The University of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan.
E-mail: nisihara@chem.s.u-tokyo.ac.jp; Fax: +81(03)5841-8063;
Tel: +81(03)5841-4346
{ Electronic supplementary information (ESI) available: Synthetic scheme
of 1 (Scheme S1); UV–vis spectra of 1 (Fig. S1) and 1 on ITO particles
(Figs. S2 and S3); immersion time dependence of cyclic voltammograms of
1/ITO (Fig. S4); a sketch of the spectroscopic cell (Fig. S5); UV–vis
spectral changes of 1/ITO in redox coupled isomerization measurement
(Fig. S6); experimental section. See DOI: 10.1039/b713107k
Scheme 1 Schematic images of the redox-coupled single-light isomeriza-
tion of 1/ITO. (a) Relation of trans–cis isomerization and a redox cycle. (b)
The procedures to confirm the photoisomerization with the green light in
its Fe(III) state.
4650 | Chem. Commun., 2007, 4650–4652
This journal is ß The Royal Society of Chemistry 2007

View Article Online
(30 min each with sonication) and sonicated in a 1 mM ethanol
solution of 1 at 20 uC, washed with ethanol and finally dried by
nitrogen blow.
signal, and the absorption of bare ITO substrates in the blank
measurement was subtracted.
Fig. 1b shows the UV–vis absorption spectral changes upon
photoirradiation. In the initial spectrum of trans-1/ITO, the p–p*
transition of the azobenzene moiety was observed at 352 nm. A
27-nm red shift of the band compared with that in its ethanol
solution and an increase of the band intensity (from emax = 2.1 6
Cyclic voltammograms of 1/ITO were measured to confirm the
immobilization of 1 and to evaluate the degree of the molecular
aggregate in the monolayer. In the cyclic voltammograms of the
monolayer in Bu4NClO4–ethanol shown in Fig. 1a, a reversible
peak appeared at 0.05 V vs. ferrocenium/ferrocene (Fc+/Fc),
almost equal to the value in ethanol solution (0.06 V vs. Fc+/Fc).
The peak current was proportional to the scan rate (Fig. 1a),
indicating that the peak can be ascribed to the Fe(III)/Fe(II) couple
of the ferrocene moieties of surface-confined 1.10
104 dm3 mol21 cm21 in ethanol to ca. 2.7 6 104 dm3 mol21 cm21
,
which was evaluated from the C value obtained from the cyclic
voltammogram) were derived from the J-aggregation of the
azobenzene moiety.12 The azo p–p* band decreased in intensity
with green (546 nm) light and further decreased with UV (365 nm)
light, and the band intensity was recovered with blue (436 nm)
light. Reproducibility was tested for 3 cycles. These behaviours are
characteristic of the trans-to-cis and cis-to-trans isomerization,
respectively, and are similar to the behaviours in solution. The cis
molar ratios of the azo moieties in the PSS were estimated to be
20% with the UV light, 10% with the green light, and almost 0%
with the blue light. These values are approximately one-third of
those in solution, probably because the isomerization of azo
moieties was restrained by the effect of J-aggregation.8
Surface coverage of redox active molecules (C) in the monolayer
was estimated from the anodic charge integration in the
voltammogram at a sweep rate of 0.1 V s21. The dependence of
C on the period of sonication indicates that the value was saturated
to ca. 1.7 6 10210 mol cm22 within 20 min (Fig. S4). The
relatively small C value against the saturated value of the
monolayer of usual ferrocene derivatives, ca.
6
6
10210 mol cm22, and a small peak shift on the formation of the
monolayer suggest that molecules of 1 in the prepared monolayer
were loosely packed and may not have had too much
intermolecular interaction for the photoisomerization reaction,11
even if local molecular aggregation was observed in the UV–vis
spectrum as shown below. The contact angle of water on the
substrate was also measured, and the angle was increased from
31(1)u to 63(2)u after sonication for 1 h with the solution of 1. This
increase suggests that the substrate was surely covered with
hydrophobic molecules. The 1/ITO samples used in the following
experiments were prepared by immersion for 24 h and sonication
for 1 h. All UV–vis spectra of 1/ITOs shown below were measured
in a cell shown in Fig. S5 for the convenience of the redox- and
photoirradiation-coupled measurement. These spectra are the sum
of two layers on the two substrates to increase the absorption
The single-light photoisomerization cycle of 1/ITO was achieved
by the combination of green light and electrochemical or chemical
redox reactions. Four states of 1/ITO during a single-light
photoisomerization cycle and the operations to change these states
are summarized in Scheme 1a. In the case of electrochemical
redox-coupled measurements, large UV–vis spectral changes of the
bare ITO substrate and the electrolyte solution were observed
when the potential was applied. In order to eliminate this
perturbation, the absorption spectra obtained from two different
sets of experiments at the same potential, route 1 (‘dark’) and route
2 (‘illuminated’) in Scheme 1b, were compared to prove the single-
light photoisomerization event.
A trans-1/ITO in the reduced Fe(II) state in Bu4NClO4–MeCN
(A state in Scheme 1a) was irradiated with the green light to reach
the PSS (B state), and the resulting mixture of trans- and cis-forms
was electrochemically oxidized to the Fe(III) state by holding the
potential at 0.32 V vs. Fc+/Fc for 5 min in the dark (C state),
followed by a reduction to the Fe(II) state by holding the potential
at 20.08 V vs. Fc+/Fc for 30 s (B9 state), and the spectrum was
measured. The next run was carried out under the same conditions
and with the same procedures, but the monolayer was irradiated
with the green light during 5 min oxidation, which should cause
isomerization to the D state. After the reduction to the Fe(II) state
(A9 state), the spectrum was measured. The differences in the UV–
vis spectral changes between the ‘dark’ and ‘illuminated’ conditions
are shown in Fig. 2a-A (see Fig. S6 for difference absorption
spectra in both conditions). Based on the amount of increase in the
absorption intensity in the p–p* region, approximately 10% of 1
was converted to the trans-form by the green light. This ratio is
almost equal to the ratio in the trans-to-cis photoisomerization
with green light in the Fe(II) state (see Fig. 1b). This spectrum
therefore indicates that the cis-form generated by the green light
irradiation to 1/ITO in the reduced Fe(II) state did not return to
the trans-form with oxidation to the Fe(III) state and the following
re-reduction, and returned to the trans-form by the photoreaction
of 1+/ITO with the irradiation of green light in the oxidized Fe(III)
state. It is noteworthy that the trans-1/ITO in the re-reduced Fe(II)
state could again be converted to the cis-form by irradiation with
Fig. 1 Electrochemical and photochemical properties of 1/ITO. (a)
Cyclic voltammograms of 1/ITO in 0.1 M Bu4NClO4–ethanol vs. Fc+/Fc
at a scan rate of 25, 50, 75, 150, 200, 300, 400, and 500 mV s21 (from inside
to outside). Inset: The peak current plot vs. scan rate. The monolayer was
prepared by the immersion method with 1 h sonication. (b) UV–vis
absorption spectrum of 1/ITO (solid line) and its 546-nm (dashed line),
365-nm (dotted line), and 436-nm (dash-dotted line) photoirradiated
spectra. These spectra are the sum of two substrates. Inset: Differences in
the UV–vis absorption spectra of 1/ITO between before irradiation and
after 546-nm (dashed line, A), 365-nm (dotted line, B), and 436-nm (dash-
dotted line, C) irradiation. Immersion time: 25 h (including 1 h sonication).
Irradiation time: 1 h.
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 4650–4652 | 4651

View Article Online
cis-to-trans isomerization in the Fe(III) state truly occurred with
546-nm light-irradiation, and not with a redox reaction.
In summary, we have succeeded in constructing a single-light
controllable azobenzene monolayer system by electrochemical
oxidation, which affords easy and clean manipulation of the
isomerization behaviour. This single green light photoisomeriza-
tion system is suitable to be combined with fine-focused light
irradiation and to be applied to a new high-density optical data
storage device.
Notes and references
1 (a) S. Kobatake, S. Takami, H. Muto, T. Ishikawa and M. Irie, Nature,
2007, 446, 778–781; (b) S. Kume, M. Murata, T. Ozeki and
H. Nishihara, J. Am. Chem. Soc., 2005, 127, 490–491; (c)
S. Nagashima, M. Murata and H. Nishihara, Angew. Chem., Int. Ed.,
2006, 45, 4298–4301; (d) T. Niazov, B. Shlyahovsky and I. Willner,
J. Am. Chem. Soc., 2007, 129, 6374–6375.
2 (a) J. Hessemann, J. Am. Chem. Soc., 1980, 102, 2167–2176; (b)
K. Ichimura, S. Oh and M. Nakagawa, Science, 2000, 288, 1624–1626;
(c) R. Micheletto, M. Yokokawa, M. Schroeder, D. Hobara, Y. Ding
and T. Kakiuchi, Appl. Surf. Sci., 2004, 228, 265–270.
3 (a) K. Ichimura, J. Photochem. Photobiol., A, 2003, 158, 205–214; (b)
M. ZahangirAlam, T. Yoshioka, T. Ogata, T. Nonaka and S. Kurihara,
Chem.–Eur. J., 2007, 13, 2641–2647.
4 (a) S. Kubo, Z. Gu, K. Takahashi, Y. Ohko, O. Sato and A. Fujishima,
J. Am. Chem. Soc., 2002, 124, 10950–10951; (b) M. Suda, M. Nakagawa,
T. Iyoda and Y. Einaga, J. Am. Chem. Soc., 2007, 129, 5538–5543.
5 Y. Yu, M. Nakano and T. Ikeda, Nature, 2003, 425, 145.
6 M. Kurihara, A. Hirooka, S. Kume, M. Sugimoto and H. Nishihara,
J. Am. Chem. Soc., 2002, 124, 8800–8801.
Fig. 2 Reversible photoisomerization of 1/ITO using a single 546-nm
light source and redox reaction of ferrocene moieties. (a) Differences in the
UV–vis absorption spectra of 1/ITO upon 546-nm photoirradiation in its
oxidized state (A) and in its reduced state (B). (b) Differences in the UV–
vis absorption spectra of 1+/ITO in ethanol between before and after in the
dark with iodine (A) and upon 546-nm photoirradiation with iodine (B),
and upon 546-nm photoirradiation in its re-reduced state by washing with
ethanol (C). Horizontal lines are shifted for clarity. (c) Absorbance change
of 1/ITO at 352 nm upon addition and removal of iodine coupled with
546-nm light irradiation. These spectra are the sum of two substrates.
the green light (see Fig. 2a-B). We can thus conclude that the
isomerization behaviour of 1/ITO can easily be controlled by an
electrochemical redox reaction of the ferrocene moiety, which does
not pollute the photoswitching system and is totally repeatable.
Direct observation of cis-to-trans photoisomerization of 1+/ITO
in the Fe(III) state by UV–vis spectral changes was also achieved
with chemical oxidation by iodine. The cis-1+/ITO in ethanol,
which was prepared by 546-nm photoirradiation and subsequent
addition of an excess amount of iodine (500 equiv.), was stable in
the dark (Fig. 2b-A) and showed cis-to-trans photoisomerization
with 546-nm light (Fig. 2b-B). These behaviours indicate that
7 A. Sakamoto, A. Hirooka, K. Namiki, M. Kurihara, M. Murata,
M. Sugimoto and H. Nishihara, Inorg. Chem., 2005, 44, 7547–7558.
8 H. Akiyama, K. Tamada, J. Nagasawa, K. Abe and T. Tamaki, J. Phys.
Chem. B, 2003, 107, 130–135.
9 D. H. P. Hedges, D. J. Richardson and D. A. Russell, Langmuir, 2004,
20, 1901–1908.
10 A. J. Bard and L. R. Faulkner, in Electrochemical Methods:
Fundamentals and Applications, John Wiley and Sons, New York, 2nd
edn, 2001, pp. 590–593.
11 B. Vercelli, G. Zotti, G. Schiavon, S. Zecchin and A. Berlin, Langmuir,
2003, 19, 9351–9356.
12 R. Wang, T. Iyoda, L. Jiang, D. A. Tryk, K. Hashimoto and
A. Fujishima, J. Electroanal. Chem., 1997, 438, 213–219.
4652 | Chem. Commun., 2007, 4650–4652
This journal is ß The Royal Society of Chemistry 2007