Photochemical behavior in azobenzene having acidic groups. Preparation of magnetic photoresponsive gels

Article abstract of DOI:10.1016/j.jphotochem.2010.10.003

The photochemistry of three azobenzenes representing contrasting photochemical behaviors is described in the present work. Thus, Methyl Orange (MO, 4-[[(4-dimethylamino)phenyl]-azo]benzenesulfonic acid sodium salt, hereinafter (1) and 4-hydroxyazobenzene-

Full text of DOI:10.1016/j.jphotochem.2010.10.003

Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Photochemical behavior in azobenzene having acidic groups. Preparation of
magnetic photoresponsive gels
Gonzalo Abellána, Hermenegildo Garcíab,∗, Carlos J. Gómez-Garcíaa, Antonio Riberaa,c,∗∗
a Instituto de Ciencia Molecular, Universidad de Valencia, Catedrático José Beltrán 2, 46980, Paterna, Valencia, Spain
b Instituto de Tecnología Química CSIC-UPV, Universidad Politécnica de Valencia, Avenida de los Naranjos s/n, 46022 Valencia, Spain
c Fundación General de la Universidad de Valencia (FGUV), Amadeu de Savoia n◦ 4, 46010 Valencia, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The photochemistry of three azobenzenes representing contrasting photochemical behaviors is described
in the present work. Thus, Methyl Orange (MO, 4-[[(4-dimethylamino)phenyl]-azo]benzenesulfonic
acid sodium salt, hereinafter (1) and 4-hydroxyazobenzene-4ꢀ-sulfonic acid (2) undergo in water fast
photochemical proton shift, with decays in the microsecond timescale. In contrast to the previous
cases, azobenzene-4,4ꢀ-dicarboxylic acid (3) undergoes photoisomerization in water. This photochemical
behavior allows the preparation of aqueous gels with Aerosil as gelating agent (5% weight) exhibiting
high cyclability and photoreversible isomerization of the trans to cis (300 nm irradiation) and cis to trans
(visible white light irradiation). Gels containing azobenzene 3 as photoresponsive molecule and well-
dispersed magnetite (Fe3O4) nanoparticles exhibit simultaneously high photoresponse and magnetic
properties. Photoisomerization of compound 3 in these gels leads to small but reliable changes (25 G
coercive field) in the magnetic hysteresis loop of magnetite nanoparticles. This novel concept of optical
modulation of magnetism in iron oxide nanoparticles paves the way for the study of new systems, with
a stronger coupling between trans–cis photoisomerization and magnetic properties.
Received 22 July 2010
Received in revised form
30 September 2010
Accepted 2 October 2010
Available online 28 October 2010
Keywords:
Photoisomerization
Azobenzene
Laser flash photolysis
Photoresponsive gel
Magnetite nanoparticles
Multifunctional materials
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
One of the most notable consequences of the trans to cis iso-
merization is the dramatic change of molecule’s length, that for
˚
Photochemical isomerization of trans azobenzenes to the cis
isomers and the corresponding dark (thermal), or photochemical
relaxation to the most stable trans isomer of the azocompound
has attracted a considerable attention from the fundamental [1–4]
and applied points of view [5–14]. In general this photoisomer-
ization can be conveniently followed by monitoring the optical
spectrum since trans isomers absorb at around 350 nm, while the
cis isomers typically exhibit a less intense absorption band at longer
(around 450 nm) wavelengths [1,3]. The significant variation of the
absorption band from trans to cis is useful, not only to follow the
photoisomerization, but also to induce specifically the transforma-
tion of the trans into the cis (with UV light) and the conversion of
the cis into the trans (with visible light), that can be employed to
control at will the configuration of the molecule [1].
the parent azobenzene changes from 9.0 A for the trans isomer to
˚
5.5 A of the cis azobenzene isomer, this change in the molecular
dimensions is among the largest possible for a reversible reaction
[2].
Since the establishment of methods to control the narrow size
distribution of nanoparticles [15], materials including metal and
metal-oxide nanoparticles in combination with organic molecules
and/or polymers have been widely reported [16], searching for
new and enhanced properties relative to their individual compo-
nents, based on the interaction between the metal or metal-oxide
nanoparticles and the organic molecules and/or polymers [17].
Unfortunately, an important part of these materials are based on
organic matrix that fall in a low chemical durability and mechanical
resistance.
Recently, sol–gel derived inorganic matrices have been used
as hosts due to their good chemical and mechanical stability and
moreover, the synthesis is performed at low temperature, which is
appropriated for the thermal stability of organic molecules [18].
The aim of this study is to present data of photoisomerization
of three 4,4ꢀ-azobenzenes having acid groups as substituent (see
Scheme 1), showing the behavior in water and in a silica gel matrix.
Our study is complemented with laser flash photolysis data and
analysis of the photoreaction process to rationalize the results pre-
∗
Corresponding author. Tel.: +34 62095 2690; fax: +34 96387 7809.
Corresponding author at: Instituto de Ciencia Molecular, Universidad de Valen-
∗∗
cia, Catedrático José Beltrán 2, 46980, Paterna, Valencia, Spain.
Tel.: +34 96354 4419; fax: +34 96354 3273.
E-mail addresses: hgarcia@qim.upv.es (H. García), antonio.ribera@uv.es
(A. Ribera).
URLs: http://www.upv.es/itq (H. García), http://www.icmol.es (A. Ribera).
1010-6030/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2010.10.003

158
G. Abellán et al. / Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
Scheme 1. Variation in the molecular dimensions upon cis–trans isomerization and
structure of the three azocompounds under study.
Fig. 1. Absorption spectra of 1 under UV irradiation in water. The arrows indicate
the growth or decrease of the absorption band upon UV irradiation.
sented. The three azobenzenes under study have in common the
possession of acids groups, whose presence would make possi-
ble to complete the trans/cis isomerization with the use of these
molecules as ligand of magnetite nanoparticles. The novel concept
of our work is to couple the well-studied reversible cis/trans iso-
merization of azobenzenes with the magnetic properties of iron
oxide nanoparticles. The presence of acids groups as substituents
of azobenzenes would allow coordination of these photorespon-
sive molecules with the nanoparticles. In the long-term we are
interested in developing systems in which azobenzene photoi-
somerization can switch in a controllable manner the magnetic
properties of metal oxides. Towards this goal, herein we have pre-
pared gels that have been magnetically characterized before and
after UV irradiation showing, indeed, that the photoisomerization
modifies the hysteresis loop when azobenzenes with acid groups
are coordinated to the nanoparticles.
2. Results and discussion
Preliminary studies with the three azobenzenes in aqueous
solutions (about 10−5 M) were carried out under steady state
irradiation. Although some minor changes were observed in the
210–240 nm region for 1, this azocompound basically does not
undergo any change in the optical spectrum under these conditions
(Fig. 1). Considering the short wavelength in which the changes are
observed (<250 nm) these small variations could be due to some
impurities formed in small amounts during the photochemical
treatment and unrelated to the cis/trans isomerization.
Fig. 2. Absorption spectra of 2 under UV irradiation in water, showing some minor
variations in the 200–240 nm region upon UV irradiation.
In the case of 4-hydroxy, 4ꢀ-sulfonic substituted azocompound
2, irradiation at 355 nm with quasi monochromatic mercury
lamps in water (pH 6.3) leads to relatively minor changes in the
300–520 nm region that are compatible with the occurrence of
some photoisomerization to reach a photostationary trans/cis mix-
ture different from the initial state. In addition also some minor
variations in the 200–240 nm region were also observed (Fig. 2). As
commented in the case of compound 1, variations of the absorbance
in such short wavelength region, are most likely due to the forma-
tion of impurities, even in trace amounts, during the photochemical
irradiation.
More interesting from the point of view of the photochemical
isomerization under amenable conditions was the behavior of sym-
metric 4,4ꢀ-dicarboxylate azobenzene 3. In this case it has to be
noted that the pH of the aqueous solution was set to 11 to facil-
itate the solubility of the azobenzene 3 [19]. As it can be seen in
Fig. 3, 355 nm steady state irradiation of compound 3 leads to a
Fig. 3. Absorption spectra of 3 under UV irradiation in water. The peaks at about
330 nm are characteristic of the trans isomer, while the peaks at about 430 nm cor-
respond to the cis isomer. The series of spectra presented correspond to the visible
light photoisomerization of the cis isomer into the trans. The inset shows a first order
plot for cis to trans thermal isomerization.

G. Abellán et al. / Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
159
Fig. 4. Transient UV–visible absorption spectra given as variation of absorbance,
ꢀA, divided by the initial absorption of an O2-purged 1 solution recorded 1 ␮s after
355 nm laser excitation. The inset shows the temporal signal decay measured at
460 nm.
Fig. 5. Transient UV–visible absorption spectra given as variation of absorbance,
ꢀA, divided by the initial absorption of an O2-purged 2 solution recorded 250 ␮s
after 355 nm laser excitation. The inset shows the temporal signal decay measured
at 360 nm.
gradual decrease of the 330 nm band and 228 nm corresponding
to the trans isomer accompanied by a concomitant increase of the
430 nm and 250 nm bands of the cis isomer. Observation of four
isosbestic points at 226, 238, 284 and 393 nm indicates the inter-
conversion of the initial trans isomer to the cis. It should be noted
that compounds 1 and 2 even at pH 11 exhibit exactly the same
behavior (i.e. lack of evidence for trans-cis photoisomerization) that
was observed when these compounds were irradiated at the pH
that results from adding the compounds to solution. After photo-
chemical isomerization using quasi-monochromatic UV-lamps, the
course of thermal isomerization of the photogenerated cis isomer
to the more stable trans isomer was followed in the dark (∼25 ◦C) by
monitoring the band corresponding to the trans isomer (330 nm).
Fitting of the experimental data to a first order kinetics (see inset
of Fig. 3) has allowed an estimation of the value of the rate of the
larities with that previously commented from azobenzene 1, i.e.,
narrow absorption peaking at 260 nm, bleaching of the trans iso-
mer at 360 nm and a continuous absorption from 470 to 800 nm,
and it was again attributed to photoinduced proton transfer from
the phenolic to the azo group mediated by the solvent. One interest-
ing difference between azobenzenes 1 and 2 is the lifetime of the
initial trans isomer. This increased lifetime of the species arising
from prototropism of compound 2 has to be related with the sta-
bility of the corresponding phenolate, with respect to the acid/base
equilibrium corresponding to azobenzene 1. In any case for azoben-
zene 1 and 2 under the conditions studied, laser flash photolysis
clearly rules out the occurrence of trans to cis photoisomeriza-
tion.
Compounds 1, 2 and 3 were used to prepare gels that could
exhibit photoresponse as consequence of the trans/cis isomeriza-
tion. As observed in aqueous solution only compound 3 exhibits
reversible photoisomerization upon UV irradiation (see SI-1, SI-2
and SI-3, respectively). In contrast to solutions, the much higher
viscosity of gels allows the formation of thick films and other bulky
systems containing indefinitely persistent suspended micro-/nano
particles. The reduced diffusion and mobility of suspended parti-
cles occurring in gels is not possible in solution. In our case the gels
were obtained by suspending Aerosil (5% weight) in water contain-
ing 3, then the suspension was vigorous stirred and mild heated at
70 ◦C until thick gel was obtained. Irradiation of this gel contain-
ing 3 leads to changes in the optical spectrum, compatible with
the occurrence of trans–cis isomerization observed for this com-
pound in aqueous solution [22]. The two main differences of the
photoisomerization in gel and in liquid phase are the extent of the
decrease/increase of the bands corresponding to the trans–cis iso-
mers (which is smaller in gel than in solution), and the kinetics of
the reverse cis to trans isomerization that is much slower in the gel
than in solution [23,24]. Consecutive cycles of trans to cis isomer-
ization irradiating at 300 nm and cis to trans by irradiation in the
visible were performed up to 10 times, without observing signifi-
cant fatigue in the system. Considering that the increased viscosity
of gels could favor the trans to cis isomerization of azobenzenes
1 and 2, despite that these process do not occur in aqueous solu-
tion, we also studied the photochemical behavior of two analogous
gels prepared by incorporating azobenzene 1 or 2. However, like in
solution, also for the gels we have been unable to obtain any spec-
troscopic evidence that could support the occurrence of trans to cis
cis to trans thermal isomerization of 3.8 × 10−4 min−1
.
Since in the case of methyl orange (1) we have not been able
to obtain any evidence of the photoisomerization using conven-
tional spectroscopy, we have submitted an aqueous solution of 1
after 15 min purging with oxygen to laser flash photolysis using a
nanosecond system and operating at 355 nm as excitation wave-
length. The presence of oxygen during the experiment should
suppress the formation of any triplet excited state that could mask
the spectral changes of the isomerization. In principle trans–cis
isomerization should occur independently of the presence of oxy-
gen [20]. Under these conditions we have been able to record
a transient spectrum showing an absorption band from 340 to
400 nm accompanied by an intense bleaching of the ground state
absorption peaking at 455 nm and a continuous absorption from
500 to 800 nm (see Fig. 4). The temporal profile of the signals in
the 500–800 broad band is identical showing that this absorption
corresponds to a single species that do not correspond to the cis
isomer. For the cis isomer a narrower band from 500 to 600 nm
should have been recorded. Based on precedents in the literature
we attribute the continuous 500–800 nm absorption to a photo-
chemical proton shift mediated by the solvent in which, at the
pH studied, a proton from the dimethyl amino substituent move
to the azo group [21]. Similarly laser flash photolysis of azoben-
zene 2 in water was performed in the presence of oxygen to avoid
the interference of the triplet excited state. Fig. 5 shows the cor-
responding transient spectrum recorded upon 355 nm excitation.
As it can be seen there, this transient spectrum has many simi-

160
G. Abellán et al. / Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
Fig. 7. Absorption UV–vis spectra of the photoresponsive gel containing iron oxide
nanoparticles. The peak at about 330 nm is characteristic of the trans isomer, while
the peak at about 430 nm corresponds to the cis isomer. The inset shows the
reversibility data monitored at 324 nm to 327 nm after successive cycles of irra-
diation with UV and visible light.
Fig. 6. HRTEM micrograph of the photoresponsive magnetic gel containing azo-
compound 3. Scale bar 20 nm. The inset shows a HRTEM image of magnetite
nanoparticles obtained with a higher magnification. Scale bar 5 nm.
TB) changes [15d]. So, magnetite nanoparticles smaller than 20 nm
are usually superparamagnetic at room temperature, it means that
the magnetic moment of the nanoparticles fluctuates in response to
thermal energy, overcoming the anisotropy energy (EA) and there
is no coercive field (HC). However, upon cooling, this superpara-
magnetic behavior turns into typical field dependence, showing
HC.
On the other hand, one of the greatest influences on the mag-
netic behavior of nanoparticles may be exercised by the magnetic
interactions, the higher interactions the bigger coercive field. Addi-
tionally, the Stoner–Wolhfarth theory for single-domain particles
predicts that the HC of the nanosized materials depends also on
the anisotropy constant (K). Organic ligands used to stabilize the
magnetic nanoparticles have an influence on their magnetic prop-
erties, that is, ligands can modify the anisotropy of the metal atoms
located at the surface of the particles [28]. In consequence, when
the surfaces of the nanoparticles are modified or adsorb different
molecules, the anisotropy constant changes, leading to changes in
the HC.
The magnetic properties of the gel were measured before and
after irradiation of the sample (at room temperature in a photoreac-
tor at 350 nm) at 2 K, below the TB. Therefore non-zero coercive field
was expected, because as mentioned above, the normal superpara-
magnetic behavior of magnetite nanoparticle have been canceled,
due to the spin blockage, as a result of the low temperature.
Figs. 8 and 9 show the hysteresis loops obtained in both cases. Note
that the measurements were not performed under irradiation but
after irradiating the sample. As can be seen in Fig. 9 the hysteresis
loop obtained after irradiation shows a slightly higher coercive field
(ca. 2.5 mT = 25 G) but no change in the saturation magnetization.
Since between both measurements the sample is heated to room
temperature and kept at room temperature during the irradiation
process, in order to discard any interference due to the temperature
variation, we have also performed blank measurements repeating
the heating of the sample but without any irradiation. These mea-
surements (Fig. 8) show no change at all between both hysteresis
cycles, within experimental error, demonstrating that the changes
observed in the hysteresis loops cannot be attributed to any ther-
mal effect but to the irradiation itself. Since the differences in the
hysteresis loop are only observed after irradiation of the sample,
we can conclude that these reproducible changes in the hystere-
photoisomerization, and therefore, no further studies with these
two azobenzenes were pursued.
One advantage of gels with respect to liquid solutions is the
possibility to add, together with the azocompound 3, nanoparti-
cles homogeneously dispersed in the medium, providing in this
way, additional functionalities and properties coming from the
nanoparticles [16f,25], to the already existing photoresponse pro-
vided by the trans-cis isomerization of the azocompound. In the
present case, we have prepared a gel in which in addition to azo-
compound 3, magnetite nanoparticles (average particle size 10 nm)
were also incorporated. The purpose is to disclose if the photo-
chemical trans to cis isomerization observed for azocompound 3
can lead to changes in the magnetic properties of Fe3O4 nanoparti-
cles. The preparation of the material started by the synthesis of the
Fe3O4 nanoparticles, using the high temperature decomposition of
the Fe(acac)3 precursor in organic phase, following the method
described by Sun and Zeng [26]. This method guarantees the
small size and the monodispersity of the nanoparticles. However it
presents the disadvantage, from the point of view of the synthesis of
the gel, of requiring organic solvent. This inconvenience was solved
including an aqueous phase transfer post-synthesis process, using
hexadecyl trimethyl ammonium bromide (CTAB, cetyl trimethyl
ammonium bromide) as transfer agent. This larger synthetic proce-
dure assures the advantage of the organic phase method (small and
monodisperse nanoparticles) and the aqueous medium required
for the preparation of the gel material. Fig. 6 shows representative
HRTEM of the as-prepared gel. This gel containing 3 and magnetite
nanoparticles exhibits exactly the same photochemical cyclability
as that previously commented for gels containing azocompound 3
(Fig. 7 shows a set of optical spectra recorded for this gel together
with cyclability data), while simultaneously presenting magnetic
properties.
As the particles size decreases, a large percentage of all the
atoms in a nanoparticle are surface atoms, which implies that sur-
face and interface effects become more important. The two most
studied finite-size effects in nanoparticles are the single-domain
limit and the superparamagnetic limit [27]. It is well known that
spherical nanoparticles of magnetite form single-domain even for
a diameter of 128 nm and depending of this size the temperature
limit for the superparamagnetic behavior (blocking temperature,

G. Abellán et al. / Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
161
the dimensions of the trans and cis isomers makes the interactions
between the magnetic nanoparticles vary as consequence of the
distance between them.
3. Experimental
All reagents were of commercial grade and were used
as received without further purification. Methyl Orange
(4-[[(4-dimethylamino)phenyl]-azo]benzenesulfonic
acid
sodium salt), iron tris(acetylacetonate) (Fe[acac]3), 1,2-
hexadecanediol, oleic acid, oleylamine, were obtained from
Aldrich. 4-Hydroxyazobenzene-4ꢀ-sulfonic acid and azobenzene-
4,4ꢀ-dicarboxylic acid were obtained from TCI. Phenyl ether,
ammonium hydroxide, ethanol, hexane and dichloromethane
were obtained from Scharlau and silicon dioxide Aerosil (a fumed
silica constituted by nanoparticles of uniform size below 20 nm)
was obtained from Degussa. The synthesis of the materials and
the visible-light photoisomerization experiments were carried out
with Milli-Q water.
Fig. 8. Hysteresis loops at 2 K of the gel prepared with 3 before and after heating
the sample to room temperature and without irradiation.
3.1. Preparation of azocompound solutions
sis loop are due to the trans–cis isomerization of the azocompound
3. As mentioned above, the hysteresis loops of magnetic materi-
als are very sensitive to the distance between the nanoparticles
due to their interaction and the changes in the coordination shell
around themagneticcentersincethespindynamicsofthe magnetic
nanoparticles is strongly dependent on their size and environment
[29–31] and, therefore, it is proposed that those molecules of 3
that are coordinated to the iron oxide nanoparticles and undergo-
ing the trans–cis isomerization are responsible for inducing slight
changes in the distance between the nanoparticles and/or nanopar-
ticle environment and, consequently in the coercive field of the
hysteresis cycle.
These changes indicate that the original oleic acid/CTAB capping
around the nanoparticle has been partially exchanged by azocom-
pound 3 through the carboxylate groups analogous to those of oleic
acid and, thus, some azocompound 3 molecules are acting as a
new capping agent. Then, upon irradiation of the azocompound
3, occurrence of the trans–cis isomerization of the azocompound
and variation of the hysteresis loop of the iron oxide nanoparti-
cles, takes place. Apparently magnetite nanoparticles containing
trans azocompound 3 have somewhat different magnetic prop-
erties than those magnetite nanoparticles coordinated to the cis
isomer. In addition, other possibility would be that the changes in
Samples were prepared by dissolving the compounds 1–3 in
Milli-Q water, reaching a concentration 10−5 M. In the particular
case of compound 3, it was necessary to add a base to dissolve
the sample correctly. For this purpose we employed 2 M sodium
hydroxide.
3.2. Preparation of photoresponsive gels
The azobenzene-containing gels were synthesized by mixing
an aqueous solution of 1–3 (10−4 M), previously alkalinized with
sodium hydroxide (2 M) until the sample was completely dissolved
(final pH = 11, yellowish solution), in the case of azocompound 3.
Then 5 g of Aerosil were added to the initial solution (100 mL) and
the mixture was vigorously stirred during 5 min. Finally the sample
was placed on a preheated oven at 70 ◦C for 3 h to generate a yel-
lowish dense and viscous gel that does not fall down upon turning
the vial upside down.
3.3. Synthesis of Fe3O4 nanoparticles
The synthesis of Fe3O4 was carried out following the high tem-
perature organic-phase method described by Sun and Zeng [26].
According to this method, a mixture of iron tris(acetylacetonate)
(Fe[acac]3, 2 mmol), 1,2-hexadecanediol (10 mmol), oleic acid
(6 mmol) and oleylamine (6 mmol) was dissolved in phenyl ether
(20 mL), under argon and refluxed for 30 min. The mixture was
cooled to room temperature and the nanoparticles precipitated
after the addition of 80 mL of ethanol. The dark-brown solid was
recovered by centrifugation and dissolved in hexane (20 mL) in the
presence of 5 ␮L of oleic acid and 5 ␮L of oleylamine. The washing
process was repeated twice without the addition of oleic acid and
oleylamine and finally the solid nanoparticles were suspended in
chloroform (60 mL).
The transfer of the Fe3O4 nanoparticles to the aqueous phase
was performed using a modification of the method developed by
Fan et al. [32]. An aqueous solution of CTAB (0.1 M, 30 mL) was
added to the chloroform solution of the nanoparticles (15 mL),
forming a two-phase system. The chloroform was removed in a
rotavapor, resulting in a homogeneous brown aqueous solution of
Fe3O4 nanoparticles.
3.4. Preparation of the photoresponsive magnetic gels
The incorporation of magnetite into photoresponsive gels was
performed by mixing 5 mL of the aqueous iron oxide nanoparti-
Fig. 9. Hysteresis loops at 2 K of the gel prepared with 3 before and after UV irradi-
ation.

162
G. Abellán et al. / Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
cles suspension with 100 mL of a solution of azobenzene 3 (10−4 M,
pH = 11), resulting in a color change of the solution from yellowish
to brown. The so-obtained mixture clearly exhibited the Tyndall
effect when irradiated with a laser beam. Then, 5 g of Aerosil
were added to the azocompound/nanoparticle suspension and the
resulting mixture was vigorously stirred during 5 min, finally the
sample was placed on a preheated oven at 70 ◦C for 3 h, and a
brownish thick gel was obtained (∼5% of iron respect to Si deter-
mined by electron probe microanalysis, EPMA).
into liquid nitrogen, inserted in the SQUID magnetometer at 100 K
and cooled to 2 K in less than 30 min and the hysteresis cycle was
repeated. Note that in order to remove any possible influence of the
thermal history on the magnetic behavior of the samples, we have
performed exactly the same thermal treatment to all the samples.
Furthermore, the reproducibility of the measurements was checked
for up to three different samples that gave very similar results.
4. Conclusions
In the present work we have shown that azobenzene-4,4ꢀ-
dicarboxylic acid (3) is a suitable derivative that exhibits reversible
trans-cis isomerization triggered by irradiation of the most sta-
ble trans isomer in aqueous solutions. Other azobenzenes having
sulfonic acid groups do not undergo photoinduced trans/cis iso-
merization in water at the natural pH arising from the solution of
the compounds (pH values: 8.7 and 6.3 for azobenzenes 1 and 2,
respectively) as evidenced by fast spectroscopic techniques with
nanosecond resolution. Compound 3 shows photoresponse with
a high photochemical reversibility. We have taken advantage of
the reversible trans to cis isomerization of azobenzene 3 even in
gel, and its ability to bind a Fe3O4 nanoparticles to prepare for the
first time a soft matter in which the hysteresis loop is influenced
by light through the photoisomerization event. We propose that
the photoisomerization alters the distance between the nanopar-
ticles or the coordination sphere of the same and these changes
are reflected in the coercitive field of the iron oxide nanoparticles.
Our studies will be expanded in the future towards the preparation
of photo-responsive materials having stronger coupling between
their magnetic properties and photo-isomerization.
3.5. Photoisomerization experiments
A Luzchem photoreactor equipped with 10 tunable quasi-
monochromatic mercury lamps emitting between 300 nm and
355 nm were used for most of the trans–cis reactions, including
those performed on the gel for the magnetic measurements. For
cis–trans isomerization a visible light projector equipped with a
200 W mercury lamp was used. The temperature of the sample was
always lower than 35 ◦C by means a digital temperature controller.
3.6. Instrumental techniques
All UV–vis spectra were conducted using a Varian Cary-5G
UV–vis spectrophotometer. The samples were measured in 10 mm
light path Hellma quartz precision cells in the case of azoben-
zene solutions and in a 1 mm light path Hellma quartz precision
cell in the case of gel. Spectra of all the samples were mea-
sured in transmission mode against air in the reference path.
Laser flash photolysis experiments were carried out in a Luzchem
ns laser flash system using the third (355 nm) harmonic of a Q-
switched Nd:YAG laser for excitation (pulse ≤ 10 ns) and a 175 W
ceramic Xenon Fiberoptic Lightsource, Cermax, perpendicular to
the laser beam, as a probing light. The signal from the monochroma-
tor/photomultiplier detection system was captured by a Tektronix
TDS 3032B digitizer. Laser system and digitizer are connected
to a PC computer via GPIB and serial interfaces that controlled
all the experimental parameters and provided suitable process-
ing and data storage capabilities. The software package has been
developed in the LabVIEW environment from National Instru-
ments and compiled as a stand-alone application. The samples
contained on Suprasil quartz 0.7 × 0.7 cm2 cuvettes capped with
septa were purged with a N2 or O2 flow for at least 15 min before
laser experiments. Metallic atomic composition of bulk samples
was determined by means of EPMA analysis performed in a Philips
SEM-XL30 equipped with an EDAX microprobe. HRTEM studies
of photoresponsive magnetic gels were carried out on a JEM-
2010 microscope (JEOL, Japan) operating at 200 kV. Samples were
deposited on a carbon-coated copper grid and were tested when
they were dried. The digital analysis of the HRTEM micrographs
was done using DigitalMicrographTM 1.80.70 for GMS 1.8.0 by
Gatan. Magnetic measurements were carried out with a Quantum
Design (SQUID) Magnetometer MPMS-XL-5. The susceptibility data
were corrected for the diamagnetic contributions calculated using
the Pascal constants. The magnetization studies were performed
between −5 and +5 T at 2 K. The sample was placed on a transparent
sealed plastic bag that had been previously measured in the same
conditions. In a first run the hysteresis cycle of the magnetization of
the sample was measured at 2 K and then the sample was heated
at room temperature and allowed to stand at room temperature
during several minutes. After this heating, the sample was again
cooled to 2 K and the hysteresis cycle measurement was repeated
without noticing any difference between both cycles. In a second
run we performed a first hysteresis cycle at 2 K. Then the sample
was heated and irradiated at room temperature during ca. 15 min in
a UV reactor. After irradiation, the sample was immediately placed
Acknowledgements
This work has been supported by the EU (MolSpinQIP and the
ERC SPINMOL Advanced Grant), the Spanish Ministerio de Ciencia
e Innovación with FEDER cofinancing (Project Consolider-Ingenio
in Molecular Nanoscience CSD2007-00010 and projects MAT2007-
61584, CTQ2009-11583 and CTQ-2008-06720) and the Generalitat
Valenciana (PROMETEO program). The authors thank Maykel de
Miguel his kindly assistance with the laser flash photolysis experi-
ments.
References
[1] H. Rau, in: Z. Sekkat, W. Knoll (Eds.), Photoreactive Organic Thin Films, Aca-
demic Press, Amsterdam, 2002, pp. 3–47.
[2] A. Gilbert, J. Baggott, Essentials of Organic Photochemisry, Blackwell, Oxford,
1990.
[3] G.S. Kumar, D.C. Neckers, Chem. Rev. 89 (1989) 1915–1925.
[4] J. Griffiths, Chem. Soc. Rev. 1 (1972) 481–493.
[5] K. Ichimura, Chem. Rev. 100 (2000) 1847–1873.
[6] M. Alvaro, M. Benitez, D. Das, H. Garcia, E. Peris, Chem. Mater. 17 (2005)
4958–4964.
[7] A.S. Kumar, T. Ye, T. Takami, B.-C. Yu, A.K. Flatt, J.M. Tour, P.S. Weiss, Nano Lett.
8 (2008) 1644–1648.
[8] V. Ferri, M. Elbing, G. Pace, M.D. Dickey, M. Zharnikov, P. Samori, M. Mayor, M.A.
Rampi, Angew. Chem., Int. Ed. 47 (2008) 3455–3457.
[9] G. Pace, V. Ferri, G. Grave, M. Elbing, C. Hänisch, M. Zharnikov, M. Mayor, M.A.
Rampi, P. Samorì, Proc. Natl. Acad. Sci. U.S.A. 104 (2007) 9937–9942.
[10] J. Zeitouny, C. Aurisicchio, D. Bonifazi, R. De Zorzi, S. Geremia, M. Bonini, C.-A.
Palma, P. Samorì, A. Listorti, A. Belbakrab, N. Armaroli, J. Mater. Chem. 19 (2009)
4715–4724.
[11] N. Henningsen, K.J. Franke, G. Schulze, I. Fernández-Torrente, B. Priewisch, K.
Rück-Braun, J.I. Pascual, ChemPhysChem 9 (2008) 71–73.
[12] M. Han, D. Ishikawa, T. Honda, E. Itob, M. Hara, Chem. Commun. 46 (2010)
3598–3600.
[13] M.J. Comstock, N. Levy, A. Kirakosian, J. Cho, F. Lauterwasser, J.H. Harvey, D.A.
Strubbe, J.M.J. Frechet, D. Trauner, S.G. Louie, M.F. Crommie, Phys. Rev. Lett. 99
(2007) 038301.
[14] K.G. Yager, C.J. Barrett, J. Photochem. Photobiol. A 182 (2006) 250–261.
[15] (a) B.L. Cushing, V.L. Kolesnichenko, J. O’Connor, Chem. Rev. 104 (2004)
3893–3946;

G. Abellán et al. / Journal of Photochemistry and Photobiology A: Chemistry 217 (2011) 157–163
163
(b) Y. Jun, J. Choi, J. Cheon, Angew. Chem. Int. Ed. 45 (2006) 3414–3439;
(c) J. Park, J. Joo, S.G. Kwon, Y. Jang, T. Hyeon, Angew. Chem. Int. Ed. 46 (2007)
4630–4660;
(d) A. Lu, E.L. Salabas, F. Schüth, Angew. Chem. Int. Ed. 46 (2007) 1222–1244;
(e) S.G. Kwon, T. Hyeon, Acc. Chem. Res. 41 (12) (2008) 1696–1709;
(f) S. Laurent, D. Forge, M. Port, A. Roch, C. Robic, L.V. Elst, R.N. Muller, Chem.
Rev. 108 (2008) 2064–2110.
[19] N.J. Dunn, W.H. Humphries, A.R. Offenbacher, T.L. King, J.A. Gray, J. Phys. Chem.
A 113 (47) (2009) 13144–13151.
[20] A. Dirksen, E. Zuidema, R.M. Williams, L. De Cola, C. Kauffmann, F. Vögtle, A.
Roque, F. Pina, Macromolecules 25 (2002) 2743–2747.
[21] A.M. Sanchez, M. Barra, R.H. de Rossi, J. Org. Chem. 64 (1999) 1604–1609.
[22] W. Freyer, D. Brete, R. Carley, R. Schmidt, C. Gahl, M. Weinelt, J. Photochem.
Photobiol. A 204 (2009) 102–109.
[16] (a) R. Gangopadhyay, A. De, Chem. Mater. 12 (2000) 608–622;
(b) K. Hervé, L. Douziech-Eyrolles, E. Munnier, S. Cohen-Jonathan, M. Soucé,
H. Marchais, P. Limelette, F. Warmont, M.L. Saboungi, P. Dubois, I. Chourpa,
Nanotechnology 19 (2008) 465–608;
[23] T. Ikeda, M. Nakano, Y. Yu, O. Tsutsumi, A. Kanazawa, Adv. Mater. 15 (2003)
201–205.
[24] Y.-L. Zhao, J.F. Stoddart, Langmuir 25 (15) (2009) 8442–8446.
[25] A. Pich, S. Bhattacharya, Y. Lu, V. Boyko, H.-J.P. Adler, Langmuir 20 (24) (2004)
10706–10711.
[26] S. Sun, H. Zeng, J. Am. Chem. Soc. 124 (2002) 8204–8205.
[27] (a) D. Peddis, C. Cannas, A. Musinu, G. Piccaluga, Chem. Eur. J. 15 (2009)
7822–7829;
(c) L. Wang, J. Sun, J. Mat. Chem. 18 (2008) 4042–4049;
(d) C. Tu, Y. Yang, M. Gao, Nanotechnology 19 (2008) 105601;
(e) S. Li, j. Qin, A. Fornara, M. Toprak, M. Muhammed, D.K. Kim, Nanotechnology
20 (2009) 185607;
(f) J. Li, L. Guo, L. Zhang, C. Yu, L. Yu, P. Jiang, C. Wei, F. Qin, J. Shi, Dalton Trans.
(2009) 823–831.
[17] (a) S. Nie, S.R. Emory, Science 275 (1997) 1102–1106;
(b) R.G. Ispasoiu, L. Balogh, O.P. Varnavski, D.A. Tomalia, T. Goodson III, J. Am.
Chem. Soc. 122 (2000) 11005–11006;
(b) Y.W. Jun, J.W. Seo, J. Cheon, Acc. Chem. Res. 41 (2) (2008) 179–189.
[28] P.M. Paulus, H. Bönnemann, A.M. van der Kraan, F. Luis, J. Sinzig, L.J. de Jongh,
Eur. Phys. J. D 9 (1999) 501–504.
[29] K. Takeda, K. Agawa, Phys. Rev. B 54 (1997) 14560.
[30] M. Clemente-León, H. Soyer, E. Coronado, C. Mingotaud, C.J. Gómez-García, P.
Delhaès, Angew. Chem. 37 (1998) 2842–2845.
[31] M. Wei, K. Okabe, H. Arakawa, Y. Teraoka, New J. Chem. 26 (2002) 20–23.
[32] H. Fan, K. Yang, D.M. Boyle, T. Sigmon, K.J. Malloy, H. Xu, G.P. López, C.J. Brinker,
Science 304 (2004) 567–571.
(c) W.U. Huynh, J.J. Dittmer, A.P. Alivisatos, Science 295 (2002) 2425–2427;
(d) A.H. Yuwono, J. Xue, J. Wang, H.I. Elim, W. Ji, Y. Li, T.J. White, J. Mat. Chem.
13 (2003) 1475–1479.
[18] (a) J.D. MacKenzie, E.P. Bescher, Acc. Chem. Res. 40 (9) (2007) 810–818;
(b) M. Niederberger, Acc. Chem. Res. 40 (9) (2007) 793–800.