Physicochemical study of spiropyran-terthiophene derivatives: Photochemistry and thermodynamics

Article abstract of DOI:10.1039/c2cp41137g

The photochemistry and thermodynamics of two terthiophene (TTh) derivatives bearing benzospiropyran (BSP) moieties, 1-(3,3″-dimethylindoline- 6′-nitrobenzospiropyranyl)-2-ethyl 4,4″-didecyloxy-2,2′: 5′,2″-terthiophene-3′-acetate (BSP-2) and 1-(3,3″- dimethylindoline-6′-nitrobenzospiropyranyl)-2-ethyl 4,4″-didecyloxy- 2,2′:5′,2″-terthiophene-3′-carboxylate (BSP-3), differing only by a single methylene spacer unit, have been studied. The kinetics of photogeneration of the equivalent merocyanine (MC) isomers (MC-2 and MC-3, respectively), the isomerisation properties of MC-2 and MC-3, and the thermodynamic parameters have been studied in acetonitrile, and compared to the parent, non-TTh-functionalised, benzospiropyran derivative, BSP-1. Despite the close structural similarity of BSP-2 and BSP-3, their physicochemical properties were found to differ significantly; examples include activation energies (E_a(MC-2) = 75.05 kJ mol^-1, E_a(MC-3) = 100.39 kJ mol^-1) and entropies of activation (ΔS?MC-2 = 43.38 J K^-1 mol^-1, ΔS?MC-3 = 37.78 J K^-1 mol^-1) for the thermal relaxation from MC to BSP, with the MC-3 value much closer to the unmodified MC-1 value (46.48 J K^-1 mol ^-1) for this latter quantity. The thermal relaxation kinetics and solvatochromic behaviour of the derivatives in a range of solvents of differing polarity (ethanol, dichloromethane, acetone, toluene and diethyl ether) are also presented. Differences in the estimated values of these thermodynamic and kinetic parameters are discussed with reference to the molecular structure of the derivatives.

Full text of DOI:10.1039/c2cp41137g

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
PCCP
Cite this: Phys. Chem. Chem. Phys., 2012, 14, 9112–9120
www.rsc.org/pccp
PAPER
Physicochemical study of spiropyran–terthiophene derivatives:
photochemistry and thermodynamicsw
Michele Zanoni,a Simon Coleman,a Kevin J. Fraser,a Robert Byrne,a
Klaudia Wagner,b Sanjeev Gambhir,b David L. Officer,b Gordon G. Wallaceb and
Dermot Diamond*a
Received 9th April 2012, Accepted 11th May 2012
DOI: 10.1039/c2cp41137g
The photochemistry and thermodynamics of two terthiophene (TTh) derivatives bearing
benzospiropyran (BSP) moieties, 1-(3,300-dimethylindoline-60-nitrobenzospiropyranyl)-2-ethyl
4,400-didecyloxy-2,20:50,200-terthiophene-30-acetate (BSP-2) and 1-(3,300-dimethylindoline-60-
nitrobenzospiropyranyl)-2-ethyl 4,400-didecyloxy-2,20:50,200-terthiophene-30-carboxylate (BSP-3),
differing only by a single methylene spacer unit, have been studied. The kinetics of
photogeneration of the equivalent merocyanine (MC) isomers (MC-2 and MC-3, respectively), the
isomerisation properties of MC-2 and MC-3, and the thermodynamic parameters have been
studied in acetonitrile, and compared to the parent, non-TTh-functionalised, benzospiropyran
derivative, BSP-1. Despite the close structural similarity of BSP-2 and BSP-3, their
physicochemical properties were found to differ significantly; examples include activation
energies (Ea(MC-2) = 75.05 kJ molꢀ1, Ea(MC-3) = 100.39 kJ molꢀ1) and entropies of activation
(DSMz C-2 = 43.38 J Kꢀ1 molꢀ1, DSMz C-3 = 37.78 J Kꢀ1 molꢀ1) for the thermal relaxation from
MC to BSP, with the MC-3 value much closer to the unmodified MC-1 value (46.48 J Kꢀ1 molꢀ1
for this latter quantity. The thermal relaxation kinetics and solvatochromic behaviour of the
derivatives in a range of solvents of differing polarity (ethanol, dichloromethane, acetone, toluene
and diethyl ether) are also presented. Differences in the estimated values of these thermodynamic
and kinetic parameters are discussed with reference to the molecular structure of the derivatives.
)
Photochromism is
a phenomenon that occurs when
Introduction
electromagnetic radiation is absorbed by a molecule that
undergoes a reversible rearrangement between two forms, such as
1-(2-hydroxyethyl)-3,3-dimethylindoline-60-nitrobenzospiropyran
(BSP-1, Fig. 1)1–13 and its equivalent merocyanine form (MC-1),
with dramatic accompanying changes in the absorption spectra
in the visible region.13
Nitrobenzospiropyran (BSP) is member of a family of photo/
thermochromic molecules that have been widely studied due to
their ability to be switched reversibly under various external
stimuli between two molecular states that exhibit dramatically
different properties.1–22 This behaviour has attracted great
interest in terms of possible applications such as optical memory,12
biomaterials for targeted therapeutic agents,8 components for
ophthalmic lenses,9 molecular functionalization of micro channels
and as optical probes for the analysis of nano-structures in
ionic liquids (ILs).14
Benzospiropyran is a molecule that exists as two hetero-
cyclic functional units connected by a tetrahedral sp3 hybri-
dised carbon atom. This atom co-joins two orthogonal planes
comprising an indoline subunit and a benzopyran subunit, the
common component for all the spiropyran moieties.
Photo-induced isomerisation processes are well known in
the literature.6,7,18–21 In spiropyrans, heterocyclic ring cleavage
occurs at the CspiroꢀO bond when exposed to UV light,6,13
leading to the formation of the merocyanine (MC) species that
can be detected via a striking colour change due to the strong
absorbance of the MC isomer in the visible region (Fig. 1).
Upon irradiation with white light, the system reverts back to
the original BSP form. The MC isomer absorbance is sensitive
to its immediate molecular environment, leading to solvent
dependant colour changes, and the solvent can also mediate
the equilibrium between the isomers.
a CLARITY: Centre for Sensor Web Technologies, National Centre
for Sensor Research, Dublin City University, Dublin 9, Ireland.
E-mail: dermot.diamond@dcu.ie; Fax: +353 1 700 7995;
Tel: +353 1 700 5404
b ARC Centre of Excellence for Electromaterials Science and
Intelligent Polymer Research Institute, University of Wollongong,
Northfields Avenue, Wollongong, NSW 2522, Australia
w Electronic supplementary information (ESI) available: Characteri-
zation of BSP-3. Baseline corrected conversion of BSP-1 (10ꢀ4 M in
ACN) to the isomer MC-1. Thermal relaxation profiles of synthesized
MC-2 and MC-3 in various solvents. Thermal relaxation profile of
MC-1 and MC-3 in ACN at various temperatures. Solvatochromic
effect of a range of solvents on BSP-2. See DOI: 10.1039/c2cp41137g
c
9112 Phys. Chem. Chem. Phys., 2012, 14, 9112–9120
This journal is the Owner Societies 2012

View Online
Fig. 1 Molecular structure and isomerisation of 1-(2-hydroxyethyl)-
3,3-dimethylindoline-60-nitrobenzospiropyran (BSP-1) spiropyran.
The solvatochromism of various benzospiropyran derivatives
has been reported previously including nitro- and methoxy-
substituted spiro-moieties.16,17,22 Recently, the well-known
photo-, thermo- and solvato-chromic properties of benzo-
spiropyran (BSP) and spirooxazine derivatives have been used
by our group to provide new insights into the nanostructure of
ILs through in situ mediation of their switching behaviour.16,17
Furthermore the presence of the phenolate anion in the MC
isomer provides a weak binding capacity for certain metal ions
(d- and f-elements), and the resulting complexes generate new
absorbance bands in the visible region, which are directly
influenced by the type of metal ion present.18–21
Polythiophenes (PThs) are a class of organic polymeric
semiconductors with extremely interesting conductivity, electro-
chromism and electroluminescence23 properties, which have
been explored for applications such as electrochemical sensors
and energy harvesting devices.24–27 From a materials chemistry
perspective, they are particularly attractive due to their high
chemical stability and relative ease of isolation.25 Other
interesting PTh properties previously reported in the literature
include thermochromism, ionochromism, photochromism and
bio-chromism.24–37 Chromic variations are generated by con-
formational changes or folding of the polymer chains, with
consequent dramatic variation in the conjugation length of the
whole system, which can be monitored using spectroscopy.
Substituents connected to the three-aromatic rings backbone
can have a considerable effect on the overall conjugation of the
system, influencing the properties and the distribution of the
p-electrons in the structure.
As part of our ongoing interest in functionalised PThs, we
recently reported the first synthesis of a BSP-functionalised
poly(terthiophene) and undertook a detailed study of its
complex electrochemical and photochemical properties.38 The
BSP was attached to the terthiophene monomer (acTTh, Fig. 2)
by an acetate linkage (BSP-2, Fig. 2). Given the multiple effects
of the spiropyran on the electrochemical behaviour of the
poly(terthiophene), we were interested in how these might vary
depending on the distance of the spiropyran from the polymer
backbone. Therefore, we synthesised the BSP–terthiophene
monomer in which the ester linkage was directly attached to
the terthiophene moiety (BSP-3, Fig. 2), differing from BSP-2
only by one carbon atom. Prior to studying its electrochemical
polymerisation and resulting polymer properties, we deemed it
important to determine the photochemical properties of BSP-3
and compare them to BSP-1 and BSP-2.
Fig. 2 Structures of the TTh and TTh-BSP derivatives studied:
methyl 4,400-didecyloxy-2,20:50,200-terthiophene-30-acetate (AcTTh);
4,400-didecyloxy-2,20:50,200-terthiophene-30-carboxylic acid (cbTTh); 1-(3,300-
dimethylindoline-60-nitrobenzospiropyranyl)-2-ethyl 4,400-didecyloxy-
2,20:50,200-terthiophene-30-acetate (BSP-2) and 1-(3,300-dimethylindoline-
60-nitrobenzospiropyranyl)-2-ethyl 4,400-didecyloxy-2,20:50,200-terthio-
phene-30-carboxylate (BSP-3).
1-(3,300-dimethylindoline-60-nitrobenzospiropyranyl)-2-ethyl
4,400-didecyloxy-2,20:50,200-terthiophene-30-carboxylate (BSP-3)
is reported, along with a detailed study of the thermodynamics
and kinetics of their photo-isomerization. Their photoswitchable
behaviour is reported in comparison to the original parent
BSP-1, as this is important for some innovative applications,
such as photo-actuated valves incorporated into micro-fluidic
manifolds11 or light-modulated surfaces for selective cell
adhesion/detachment.39
Experimental
Materials
1-(2-Hydroxyethyl)-3,3-dimethylindolino-60-nitrobenzopyrylospiran
(BSP-1) was purchased from TCI chemicals, Ireland, and was
used without further purification. Ethyl (3,300-dimethylindoline-
60-nitrobenzospiropyranyl)-4,400-didecyloxy-2,20:50,200-terthio-
phene-30-acetate (BSP-2) was synthesised as described
previously.38
Therefore, in this work, the photochemical behaviour
of the two BSP-functionalised terthiophene monomers,
1-(3,300-dimethylindoline-60-nitrobenzospiropyranyl)-2-ethyl
4,400-didecyloxy-2,20:50,200-terthiophene-30-acetate (BSP-2) and
Ethyl (3,300-dimethylindoline-60-nitrobenzospiropyranyl)-
4,400-didecyloxy-2,20:50,200-terthiophene-30-carboxylate (BSP-3)
c
This journal is the Owner Societies 2012
Phys. Chem. Chem. Phys., 2012, 14, 9112–9120 9113

View Online
was synthesized as follows: 4,400-didecyloxy-2,20:50,200-terthio-
phene-30-carboxylic acid (cbTTh)40 (0.083 g, 0.138 mmol),
dicyclohexylcarbodiimide (DCC) (0.03 g, 0.143 mmol) and
4-dimethylaminopyridine (0.0013 g, 0.011 mmol) were dissolved
in DCM (12 ml) at 0 1C. 1-(2-Hydroxyethyl)-3,3-dimethyl-
indoline-60-nitrobenzospiropyran BSP-1 (0.04 g, 0.11 mmol)
in DCM (10 ml) was added dropwise over 30 min to the
stirring solution. After BSP-1 was completely added, the
reaction mixture was allowed to reach 20 1C. After a further
90 min, 50% of the DCM was removed under reduced
pressure, cold ether (10 ml) was added to precipitate the urea
salt of DCC. This precipitate was removed by filtration and
the filtrate was reduced to give a brown oil, which was purified
on a silica column, eluting with DCM. The resulting product
was a yellowish-brown oil, which solidified on standing
(0.081 g, 77%). After purification via column chromato-
graphy, the BSP-3 was stored in the dark in a sealed glass
vial under nitrogen. Characterisation data are available in the
ESIw section.
before the absorbance measurements were acquired. Thermal
relaxation analysis of the thermodynamically favoured rever-
sion of MC to BSP was measured at 5 1C increments from 283
to 308 K over a period of 1 hour. Samples were irradiated in
the dark with the UV-LED system in a sealed quartz cuvette at
room temperature for 3 minutes prior to acquisition of the
spectra. The study of the BSP " MC equilibrium was
performed using the LED system in the spectrometer chamber
in order to minimise any influence of external light. The
MC - BSP thermal relaxation typically follows first order
kinetics in molecular solvents and the rate is not sensitive to
oxygen, but is strongly medium dependent.18–21
First-order rate constants for the thermal relaxation were
determined in this work using Microsoft Excel Solver, by
fitting the absorbance data to eqn (2), where y is the absor-
bance value at the lmax, a is the absorbance at t = 0, k is the
rate constant, t is time and b is a baseline offset, as for eqn (1)41
y = aeꢀkt + b
(2)
The solvents used for the photo- and thermo-chromic
analyses (acetonitrile (ACN), ethanol (EtOH), toluene (Tol),
acetone (Acet), dichloromethane (DCM) and diethyl ether
(DetEth)) were Sigma Aldrich HPLC grade, and used without
further purification.
The Arrhenius equation (3) was used to plot the linear
temperature dependence of the rate constant for MC - BSP
thermal relaxation and find the activation energy (Ea) and the
pre-exponential factor A.
ln k = Ea/RT + ln A
(3)
Photochemical methods
where k = rate constant (sꢀ1), Ea = activation energy
(kJ molꢀ1), R = gas constant (J molꢀ1 Kꢀ1), T = temperature
(K) and A = pre-exponential factor (sꢀ1).
To obtain the rate constants of the ring opening process for
BSP-1, BSP-2 and BSP-3 to their MC-1, MC-2 and MC-3
forms, the derivatives were continuously monitored during
UV irradiation at 375 nm. The samples were kept in the dark
for 24 hours to ensure equilibrium in the solvent was reached.
Diluted solutions (10ꢀ4 M) of BSP-1, BSP-2 and BSP-3 were
exposed to UV light and the spectrum of each was taken at
intervals from t = 1 minute to t = 45 minutes.
Following the generalised BSP " MC system (Fig. 1), the
Eyring equation (4) was used to determine the thermodynamic
properties at the transition state equilibrium of activation (Kz),
entropy of activation (DSz), and the enthalpy of activation
(DHz). A derivative of this equation, eqn (5), can be used to
derive the equilibrium of activation (Kz):16,17,22
A single exponential model (eqn (1)) was used to determine
the ring opening rate constants for BSP-1, BSP-2 and BSP-3.
ln(k/T) = ꢀDHz/RT + ln(k/h) + DSz/R
k = (kBT/h)Kz
(4)
(5)
y = a ꢁ (1 ꢀ eꢀ(kt)) + b
(1)
where k = rate constant of the forward reaction, T =
temperature (K), DHz = enthalpy of activation, DSz
entropy of activation, R = gas constant, kB = Boltzmann
where y is the absorbance at lmax (assumed to be proportional
to the concentration of the merocyanime isomer), a is a scaling
factor, k is the first order rate constant (sꢀ1), b is the baseline
offset and t is time (s).
=
constant and h = Planck’s constant.
Thermodynamic activation properties such as DSz, DHz and
Kz for each MC derivative were calculated using the Eyring
equation. Eqn (6) was used to calculate the Gibbs free energy
of activation (DGz).
Absorption spectra were recorded using a Perkin Elmer
Lambda 900 UV-Vis/NIR spectrometer equipped with a
thermostated temperature control system (PTP-1 Peltier System).
Photo-conversion of BSP to MC was achieved using an in-house
fabricated UV light source consisting of three UV LEDs
(Roithner LaserTechnik Gmbh, emission lmax = 375 nm).
The LEDs were powered from a 3.75 V current source, which
provided an optical power output of 6 mW cmꢀ2 from each
LED. The white light radiation was provided via a DDL
150 W source obtained from Polytec GmbH Waldbronn.
To determine the thermodynamic parameters of the samples
and obtain an estimation of the transition state equilibrium of
activation (Kz) for the MC - BSP equilibrium, standard
solutions of BSP-1, BSP-2 and BSP-3 were prepared in
concentrations up to 10ꢀ4 M in acetonitrile. The solutions
were stored in volumetric flasks in the dark under nitrogen
DGz = DHz ꢀ TDSz
(6)
where DGz = Gibbs free energy of activation (kJ molꢀ1), DHz =
enthalpy of transition state (kJ molꢀ1), T = temperature (K) and
DSz = entropy of the transition state (J Kꢀ1 molꢀ1).
The solvatochromic study was performed at room temperature
in 6 different solvent systems (ethanol, toluene, acetone,
acetonitrile, diethyl ether and dichloromethane). The solutions
(all 10ꢀ4 M) were sealed and stored under nitrogen overnight
before the analysis was performed. Spectra were taken under
controlled temperature conditions using the LED system
described above to illuminate the samples.
c
9114 Phys. Chem. Chem. Phys., 2012, 14, 9112–9120
This journal is the Owner Societies 2012

View Online
Fluorescence (emission) spectra were recorded with a Perkin
Elmer Luminescence Spectrometer LS 50B. Solutions diluted
up to 10ꢀ5 M in acetonitrile of the two freestanding TTh
molecules, BSP-1, BSP-2 and BSP-3 were made. Excitation
wavelengths were set as follows: 358 nm (acTTh), 362 nm
(cbTTh), 360 nm (BSP-2 and BSP-3). The solutions were
stored in sealed volumetric flasks for 24 hours under nitrogen
in the dark. Photo-isomerisation to the relatives MC-1, MC-2
and MC-3 was carried out with in situ LED illumination
(lmax 375 nm) for 180 seconds.
Results and discussion
Material synthesis
The synthesis of BSP-3 in 77% yield by a base-catalyzed
substitution of BSP-1 onto cbTTh was straightforward and
its spectral data (see ESIw) was consistent with this structure.
As expected, its 1H NMR spectrum was near identical to that of
BSP-2 except for the absence of the methylene signal at d = 7.3.
Photochromic properties
It is well known that MC derivatives typically exhibit negative
solvatochromism, meaning that the absorption maximum
undergoes a hypsochromic (blue) shift as the solvent polarity
increases, due to interactions between the solute and the solvent
molecules that increase the energy gap between the MC ground
and the excited states.18–21,42 The focus of this study is whether
the spectroscopic properties and switching behaviour of the BSP
group is affected by attachment of the differing TTh sub-units.
UV-vis spectra of the two spiropyran-functionalised terthio-
phenes show two major absorbance bands with maxima at
352 nm and 362 nm for BSP-2 and BSP-3, respectively (TTh
absorbance, BSP contribution not shown38) and 568 nm and
572 nm for MC-2 and MC-3, respectively (MC absorbance), as
shown in Fig. 3(a) and (b). All solutions were kept in the dark
for 24 hours at 293 K to establish thermodynamic equilibrium
prior to UV illumination. In Fig. 3(a), the absorbance maximum
at ca. 360 nm is greater for MC-2 compared to BSP-2, whereas
they are the same for BSP-3 and MC-3 (Fig. 3(b)). This suggests
that there is a change in the electronic structure of MC-2 relative
to BSP-2, which does not occur with BSP-3/MC-3.
Fig. 3 Baseline corrected UV-vis spectra before and after illumina-
tion of 10ꢀ4 M solutions of (a) BSP-2 with its MC-2 absorption band
at 568 nm and (b) BSP-3 with its MC-3 absorption band at 572 nm,
using the 3 UV LED light source with lmax emission = 375 nm;
exposure time = 3 minutes.
constant at ca. 0.50 a.u. after about 200 min. This would be
expected due to BSP-2 having a faster ring-opening rate than
BSP-3 (Fig. 4(c)). Clearly, there is a significant difference in the
steady state absorbance of BSP-1 relative to BSP-2 and BSP-3:
the detected maximum absorbance for BSP-1 after 45 minutes of
exposure to the LED device was 1.84 a.u. at 563 nm versus
0.31 a.u. at 568 nm and 0.39 a.u. at 572 nm for BSP-2 and BSP-3,
respectively. This is most likely due to competing absorbance by
the TTh moiety present in BSP-2 and BSP-3, which reduces the
availability of UV-photons for driving the ring-opening process.
Further indications supporting this statement might be
found by comparing the ring opening of BSP-2 and BSP-3.
Extrapolation of the curves (Fig. 4(c)) suggests that the
process reaches steady state after ca. 150 minutes for MC-2
and ca. 200 minutes for MC-3. This model predicts final
maximum absorbances for 10ꢀ4 M acetonitrile solutions of
ca. 0.39 a.u. for MC-2, and ca. 0.5 a.u. for MC-3, considerably
smaller if compared to derivatives already reported in litera-
ture for BSP-1 (1.8 a.u.).42 The calculated rate of ring opening
in the described process, BSP - MC, is fastest for BSP-2
compared to BSP-3, but both are slower than BSP-1 by an
order of magnitude.
Kinetics of photo-induced ring opening
Fig. 4(a) and (b) show the spectra used to obtain the lmax
values for substitution into the model (eqn (1)), while Fig. 4(c)
gives the growth in absorbance over time (293 K) at the lmax
for MC-1, MC-2 and MC-3 (spectra for MC-1 are given in
ESIw, Fig. S1). The ring opening process analysed using
eqn (1) occurs B1.5 times faster for BSP-2 compared to BSP-3
(kBSP-2 = 4.45 ꢁ 10ꢀ4 sꢀ1 versus kBSP-3 = 2.85 ꢁ 10ꢀ4
s
ꢀ1).
For BSP-1 the equivalent rate constant was higher again
(kBSP-1 = 1.09 ꢁ 10ꢀ3
s
ꢀ1), giving an order in terms of speed
of ring opening of BSP-1 > BSP-2 > BSP-3.
Extrapolation of the fitted curves using eqn (1) allowed
estimation of the steady state absorbance values for the derivatives
in ACN. For the formation of MC-2, the model suggests the
steady state (ca. 0.4 a.u.) is reached in about 150 minutes while
for the formation of MC-3 steady state absorbance is relatively
Kinetics and thermodynamics of ring closure
Table 1 shows the rate constants for ring closure (MC - BSP) over
the temperature range 283–308 K for BSP-1, BSP-2 and BSP-3,
c
This journal is the Owner Societies 2012
Phys. Chem. Chem. Phys., 2012, 14, 9112–9120 9115

View Online
Fig. 5 Thermal relaxation profiles for MC-3 (10ꢀ4 M in ACN) over
the temperature range 283–308 K. Prior to acquisition of the
spectra, the solutions were irradiated with the 3 UV-LED device
(lmax = 375 nm) for 180 s.
and the ratios of the rate constants for each form. The
MC-3–MC-2 ratio steadily increases from 0.5 at 283 K to
1.4 at 308 K, meaning that at 283 K, MC-2 relaxation is faster,
whereas at 308 K, MC-3 is faster. This pattern of behaviour is
repeated for MC-1 compared to MC-2, with the ratio = 0.42 at
283 K, increasing to 1.15 at 308 K. In contrast, the MC-1/MC-3
ratio remains relatively constant over the temperature range
studied, at about 0.8. These results suggest that the temperature
affects ring closure kinetics in a similar manner for BSP-3
compared to the parent BSP-1 (increases around 3-fold), but a
different type of dependency is observed with BSP-2 (remains
relatively constant). Furthermore, at 283 K MC-2 reversion to
BSP-2 is the fastest for the three derivatives studied, whereas
at 308 K, it is the slowest.
Fig. 5 shows the absorbance decrease at 568 nm for MC-3 in
ACN (10ꢀ4 M) at different temperatures (similar spectra for
MC-2 and MC-1 are given in the ESIw, Fig. S3 and S4,
respectively), from which the rate constants for ring closing
(thermal relaxation) at each temperature were obtained by
fitting a first order rate equation.
The resulting Arrhenius and Eyring plots are shown in
Fig. 6(a) and (b). From the Arrhenius plots, the activation
energy (Ea) and the pre-exponential factor were derived
(Table 2) for the thermal relaxation process. Using Eyring’s
transition state theory (eqn (4)), the activation thermodynamic
parameters, DSz (entropy of activation), DHz (enthalpy of
activation) and DGz (Gibbs free energy of activation), for the
thermal relaxation process can be estimated (Table 2, Fig. 6(b)).
The Arrhenius and Eyring plots both show that MC-1 and MC-3
exhibit strikingly similar behaviour, with almost identical slopes,
Fig. 4 Exposure effects to 375 nm UV LEDs: (a) baseline corrected
conversion of BSP-2 (10ꢀ4 M in ACN) to the isomer MC-2.
(b) Baseline corrected conversion of spiropyran BSP-3 (10ꢀ4 M in
ACN) to the isomer MC-3. (c) Exponential curves extrapolated
to steady state for generation of MC isomers versus time of UV
irradiation with experimental results superimposed for MC-2 and
MC-3. Values were taken at lmax MC-1 = 563 nm, MC-2 = 568 nm,
MC-3 = 572 nm.
Table 1 Rate constants k and their ratios for MC - BSP thermal relaxation for MC-1, MC-2 and MC-3 (10ꢀ4 M in ACN) over 283–308 K
averaged from 3 replicates. Standard deviations reported in brackets
k (10ꢀ3 ꢀ1) in ACN
s
lmax (nm)
283 K
288 K
293 K
298 K
303 K
308 K
MC-1
MC-2
MC-3
Ratios MC 1/2
Ratios MC 1/3
Ratios MC 3/2
563
568
572
0.5 (ꢂ0.002)
1.2 (ꢂ0.002)
0.6 (ꢂ0.004)
0.42
0.83
0.50
1.06 (ꢂ0.001)
2.6 (ꢂ0.002)
1.5 (ꢂ0.002)
0.40
0.70
0.57
2.2 (ꢂ0.001)
4.3 (ꢂ0.0009)
3.6 (ꢂ0.001)
0.50
0.62
0.69
4.5 (ꢂ0.0009)
8.3 (ꢂ0.002)
5.9 (ꢂ0.001)
0.52
0.76
0.71
9.04 (ꢂ0.0009)
11.7 (ꢂ0.008)
10.9 (ꢂ0.001)
0.77
0.82
0.93
18.3 (ꢂ0.0014)
15.8 (ꢂ0.007)
21.6 (ꢂ0.002)
1.15
0.84
1.36
c
9116 Phys. Chem. Chem. Phys., 2012, 14, 9112–9120
This journal is the Owner Societies 2012

View Online
is related to the degree of order of the system as the MC
molecules reorganise within the solvent system due to confor-
mational changes as they pass into the transition state, MCz.
The negative value for MC2 - MC2z suggests that the
formation of the transition state results in a more ordered system
compared to the equivalent processes for MC-1 and MC-3.
Finally, the enthalpy of activation for MC-2 is 73.05 kJ molꢀ1
,
about 25% lower than the values for MC-3 and MC-1. These
data suggest that the difference in kinetics of ring closing
observed for MC-2 may arise primarily for entropic reasons,
possibly due to differences in the conformation adopted by
MC-2/BSP-2 compared to the other derivatives.
One explanation for these results is the additional methylene
group in the BSP-2 linker chain provides a degree of increased
flexibility, and the two functional units (BSP and TTh) may be
better able to align, for example, through intramolecular p–p
interactions. This would be particularly so when the planar
merocyanine isomer is present, and this would favour the
adoption of a more ordered configuration, which is consistent
with the negative value of DSz for BSP-2.
This explanation is also supported by the difference in DHz
between the three spiropyrans; DHz for BSP-2 is B20 kJ molꢀ1
lower than BSP-1 and BSP-3. This suggests that the MC-2
adopts a ground state conformation somewhat closer to the
transition state leading to the spiropyran (BSP-2) than MC-3
does on isomerising to BSP-3. Given that the formation of
spiropyran from merocyanine typically requires rotation
about the three bonds linking the indoline and phenolate
components of the merocyanine, as discussed extensively by
Wojtyk et al.,43 a p–p interaction between the terthiophene
and the phenolate component of MC-2 that brought the
phenolate oxygen (in polar solvents) or quinone oxygen
(in non-polar solvents) closer to the indoline nitrogen should
lower DHz. Such an MC conformation could be determined by
p–p interactions during the ring opening of the spiropyran to
form the merocyanine initially.
Fig. 6 Comparative (a) Arrhenius plots and (b) Eyring plots for the
thermal relaxation process MC - BSP. All solution concentrations
were 10ꢀ4 M in ACN. The values for MC-2, MC-1 and MC-3 were
estimated from the thermal relaxation profiles in Fig. S3, S4, ESIw and
Fig. 5, respectively.
and a slight offset, whereas MC-2 has a lower slope in both
cases, with a higher rate constant at lower temperatures that
gradually converges and eventually crosses the other plots.
This pattern of behaviour is reflected in the values of the rate
constants (Table 1) discussed above, and in the thermo-
dynamic and kinetic parameters extracted from the Arrhenius
and Eyring plots (Table 2). Taking each in turn, analysis by
Arrhenius theory returns an activation energy for MC-2 that
is ca. 25% lower (75.05 kJ molꢀ1 for BSP-2 compared to
100.39 kJ molꢀ1 for BSP-3, and 101.49 kJ molꢀ1 for BSP-1),
and a pre-exponential factor that is smaller by 4 orders of
magnitude (ca. 1011 vs. 1015); Eyring theory returns similar
Gibb’s free energy of activation (85–87 kJ molꢀ1) and equilibrium
constants of activation (ca. 4–7 ꢁ 10ꢀ16) in all three cases, but the
Solvatochromic effects
Samples of BSP-2 and BSP-3 were dissolved in selected
solvents and kept in the dark for 24 hours before solvato-
chromic studies were carried out. For this study, a polar protic
solvent (ethanol) and polar aprotic solvents (acetone, acetonitrile)
were used and their effect compared to non-polar solvents
(toluene, dichloromethane and diethyl ether). Fig. 7 shows
spectra obtained for MC-3 in a number of solvents after
irradiation for 180 seconds with UV light (375 nm), along
with inset photographs of the resulting solutions. In addition,
the thermal relaxation rate constants were estimated at 293 K
(Fig. 8(a) and (b), extrapolated from Fig. S2(a) and (b), ESIw)
entropy of activation (DSz) is negative for MC-2 (ꢀ43.38 kJ molꢀ1
)
and positive for MC-3 (37.78 kJ molꢀ1
) and MC-1
(46.48 kJ molꢀ1). This quantity (for the process MC - MCz)
Table 2 Thermodynamic parameters expressed as a mean of 3 replicates and standard deviation (standard deviations values are reported in
brackets) for MC-2 " BSP-2 and MC-3 " BSP-3 thermal relaxation compared to the results obtained from a reference solution of BSP-1 10ꢀ4 M
in ACN
Arrhenius
Eyring
MC
lmax (nm) k293 (10ꢀ3 sꢀ1
)
Ea (kJ molꢀ1
)
A (sꢀ1
)
DSz (J Kꢀ1 molꢀ1
)
DGz293 (kJ molꢀ1
)
DHz (kJ molꢀ1 Kz293 (10ꢀ16
) )
MC-1 563
MC-2 568
MC-3 572
2.2
4.3
3.6
101.49 (ꢂ2.06) 3.51 ꢁ 1015 (ꢂ2.83)
100.39 (ꢂ3.06) 8.62 ꢁ 1015 (ꢂ0.24)
46.48 (ꢂ2.92)
86.59 (ꢂ0.082)
85.17 (ꢂ0.18)
85.94 (ꢂ0.23)
99.04 (ꢂ2.06)
73.05 (ꢂ1.5)
96.74 (ꢂ1.42)
3.56 (ꢂ0.108)
6.57 (ꢂ0.73)
5.93 (ꢂ0.74)
75.05 (ꢂ1.5) 1.35 ꢁ 1011 (ꢂ0.9) ꢀ43.38 (ꢂ2.07)
37.78 (ꢂ2.81)
c
This journal is the Owner Societies 2012
Phys. Chem. Chem. Phys., 2012, 14, 9112–9120 9117

View Online
Table 3 Photo and thermal properties of MC-2 and MC-3 derivatives
(10ꢀ4 M) in a selection of solvents. lmax measured after 15 hours in the
dark. The thermal 15 relaxation rate constants were calculated by
plotting ln k vs. time, where k = slope
MC-2
MC-3
lmax
Solvent
(nm) Abs k (10ꢀ3 sꢀ1) lmax (nm) Abs k (10ꢀ3 sꢀ1
)
Ethanol
Acetonitrile
Acetone
Dichloromethane 587 0.13 9.2
Diethyl ether
Toluene
552 0.42 2.4
568 0.12 4.3
570 0.44 6.3
557
572
574
584
599
605
0.3 0.7
0.1 3.6
0.4 6.3
0.4 17.3
0.2 19.3
0.5 25.7
597 0.29 21.9
605 0.42 24.3
moment on electronic excitation and an increase in the dipolar
nature of the ground state.43 It has also been shown that this
tends to be independent of the nature of the N-substituent,
and this is indeed the case for MC-2 and MC-3, which show
near identical solvatochromic behaviour.
Fig. 7 Solvatochromic effects of MC-3 in the following solvents:
EtOH, Acet, DCM, DetEth and Tol, listed and inset in terms of
decreasing polarity. Prior to acquisition of the spectra, the solutions
were irradiated with UV-LED device (lmax = 375 nm) for 180 s at
293 K. [BSP-3] = 10ꢀ4 M in all cases.
In contrast, the rate constant for ring closure (k) decreases by
an order of magnitude from 24.3 ꢁ 10ꢀ3 sꢀ1 to 2.4 ꢁ 10ꢀ3 sꢀ1 for
MC-2 for increasing solvent polarity, while for MC-3 an even
greater rate decrease occurs (25.7 ꢁ 10ꢀ3 sꢀ1 to 0.7 ꢁ 10ꢀ3 sꢀ1).
The values for the rate constants for ring closure are similar in
acetonitrile, acetone, diethyl ether and toluene, implying that
solvation of the MC isomers must be quite similar. However,
in ethanol this is not the case, as ring closure is four times
faster for MC-2.
This implies that, even though the dipolar nature of the
MC-2 ground state is similar to that of MC-3 (similar lmax
values), the ethanol is acting like a less polar solvent with
regard to ring closure rate. If intramolecular p–p interactions
do indeed control the ground state conformation of MC-2,
then MC-2 is in effect partly solvated by the p bonds of
terthiophene, a less polar ‘‘solvent’’, diminishing the full effect
of the polar ethanol solvent. Somewhat more surprising is the
contrary effect of dichloromethane for which ring closure is
almost twice as fast for MC-3 relative to MC-2, even though
they appear to be electronically similar in this solvent. A more
detailed study needs to be undertaken to probe the reasons for
this behaviour.
Fluorescence spectroscopy
The presence of intramolecular p–p interactions in MC-2
would be expected to have a significant impact on the emissive
properties of the molecule. The emissions of the terthiophene
components of both BSP-2 and BSP-3 and their corre-
sponding merocyanines were investigated since excitation of
the molecules at 360 nm would lead to typical terthiophene
emissions26 around 450 nm that overlapped with the edge of
the merocyanine absorptions (Fig. 3). The corresponding
emission spectra, as well as those of BSP-1/MC-1 and the
terthiophene acids are presented in Fig. 9(a) and (b), and the
relevant data given in Table 4. A comparison of the emission
spectra of the two terthiophenes shows, as expected, that
acTTh is blue shifted by 28 nm (Fig. 9(a), lmax,em = 439 nm)
compared to cbTTh (Fig. 9(b), lmax,em = 467 nm) as a result of
the effect of the loss of conjugation of the carboxylic acid.
However, both BSP-2 (lmax,em = 464 nm) and MC-2 have
Fig. 8 The ln k vs. time plots used to calculate the rate of thermal
relaxation of 10ꢀ4 M solutions of BSP-2 (a) and BSP-3 (b) in common
organic solvents at 293 K.
in each solvent. The formation of J-aggregates (Fig. 7) in non-
polar solvents typical of many BSP derivatives is evident from
the appearance of new absorbance bands at 605 nm and 599 nm
for MC-3, and at 605 nm and 597 nm for MC-2 (Fig. S5, ESIw),
in toluene and diethyl ether, respectively.
Table 3 shows clearly that for both MC-2 and MC-3, the
lmax is blue-shifted as the solvent polarity increases (negative
solvatochromism) from toluene through to ethanol (605 nm to
552 nm for MC-2, 552 nm to 557 nm for MC-3). As is well
established in the literature, merocyanines exhibit negative
solvatochromism as a result of both a decrease in the dipole
c
9118 Phys. Chem. Chem. Phys., 2012, 14, 9112–9120
This journal is the Owner Societies 2012

View Online
Fig. 10 (a) Side view of MC-2 in a ‘U’ arrangement showing the
distance ‘d’ (3.5–3.7 A) between the aligned MC and TTh moieties and
(b) alternative view of the same arrangement of the MC and TTh
moieties. Images generated using Chem-3d Ultra V10.0. Atom types
carbon (grey), oxygen (red), nitrogen (blue), sulphur (yellow); hydrogen
atoms not displayed for clarity.
Fig. 9 (a) Emission spectra of acTTh, BSP-1, BSP-2 and their related
MC isomers. (b) Emission spectra of cbTTh, BSP-1, BSP-3 and their
related MC isomers. All the solutions were 10ꢀ5 M in ACN and all
excitation wavelengths were 360 nm.
which does not manifest when the same experiment is per-
formed with BSP-3, suggesting that a new interaction occurs
when BSP-2 is switched to MC-2.
Table 4 Parameters derived from the analysis of the emission spectra
of acTTh, cbTTh, BSP-2 and BSP-3. Stokes shifts are also indicated
Therefore, these results suggest that the extra methylene
group present in the BSP-2 linker chain provides a degree of
additional flexibility that leads to enhanced intramolecular
interaction between the MC and TTh units, most likely
through alignment of p-electrons, as the molecule folds around
the linker chain. A 3-d representation of this intramolecular
alignment of the MC and TTh p-electrons is shown in
Fig. 10(a) and (b). This was produced by generating the energy
minimised MC and TTh sub units in Chem-3D Ultra using
standard molecular mechanics and arranging these with the
appropriate linker chain, while applying typical constraints for
the various bond lengths and angles. This tentatively suggests
that is possible for the sub-units to align, with a separation
as close as 3.5–3.7 A, which is well within the range for
p-stacking. As indicated by the shift in the emission spectra,
such an interaction is also likely between the BSP and TTh
moieties in BSP-2, although perhaps not as strongly given the
non-planar nature of the BSP unit. In the case of BSP-3/MC-3,
the shorter linker group, coupled with the extended conjugation
of the TTh-p electrons into the carbonyl group prevents this
interaction from occurring and the fluorescence and absor-
bance spectra are not affected in the same manner. A more
rigorous ab initio study using molecular orbital theory calcu-
lations is currently underway and will be the subject of a
separate publication.
lmax UV-vis l Excitation lmax Emission Stokes shift
(nm)
Molecules cmꢀ1 (nm)
cmꢀ1 (nm)
(cmꢀ1
)
acTTh
cbTTh
BSP-2
BSP-3
28 090 (356) 358
27 472 (364) 362
28 090 (356) 360
27 472 (364) 360
22 727 (440)
21 413 (467)
21 598 (463)
21 276 (470)
5363
5874
6492
6196
notably red-shifted (B25 nm) terthiophene emissions (Fig. 9(a)
and Table 4) whilst those of BSP-3 and MC-3 are largely
unchanged (Fig. 9(b) and Table 4).
Clearly, there is a much larger structural effect on the
emission properties of BSP-2 and MC-2 than can be accounted
for by a single carbon atom, and this is fully consistent with
the influence of intramolecular p–p interactions. However, it
should be noted that since the emission shift is similar for both
BSP-2 and MC-2, similar interactions must be present in both
isomers.
Equally significant in this regard is the change in emission
intensity for the BSP-2/MC-2 isomers, which is absent for
BSP-3/MC-3. While the quantum yields of the molecules were
not measured and therefore, it is difficult to draw any conclu-
sions from the relative intensities of the two sets of molecules,
the apparent quenching of the MC-2 terthiophene emission is
again consistent with a intramolecular p–p interaction between
the planar merocyanine moiety and the terthiophene. This
may also explain the increase in the merocyanine absorbance
at 355 nm (Fig. 3(a)) when BSP-2 is exposed to UV light,
This explanation is also consistent with the observed differences
in the kinetic and thermodynamic behaviour of the BSP-2 and
BSP-3. For example, BSP-3 is similar to BSP-1 in terms of
thermal relaxation behaviour. Table 1 shows that the ratio of
c
This journal is the Owner Societies 2012
Phys. Chem. Chem. Phys., 2012, 14, 9112–9120 9119

View Online
the rate constants remains virtually constant over the
temperature range studied (283–308 K), whereas for BSP-2,
the ratio increases markedly towards the higher end of the
range. This could be explained in terms of disruption of the
MC and TTh p-electron alignment as the system energy
increases. Similar explanations have been advanced for observed
patterns in the kinetics the thermal relaxation of the merocyanine
form of spironaphtooxazine derivatives.44
8 I. Willner, S. Rubin and Y. Cohen, J. Am. Chem. Soc., 1993, 115,
4937–4938.
9 J. C. Crano, T. Flood, D. Knowles, A. Kumar and B. VanGemert,
Pure Appl. Chem., 1996, 68, 1395–1398.
10 S. Stitzel, R. Byrne and D. Diamond, J. Mater. Sci., 2006, 41,
5841–5844.
11 F. Benito-Lopez, R. Byrne, A. M. Radu-ta, N. E. Vrana,
G. McGuinness and D. Diamond, Lab Chip, 2010, 10, 195.
12 A. S. Dvornikov, J. Malkin and P. M. Rentzepis, J. Phys. Chem.,
1994, 98, 6746–6752.
13 H. Bouas-Laurent and H. Dvorr, in Photochromism, ed. D. R.
Heinz and B.-L. Henri, Elsevier Science, Amsterdam, 2003,
pp. XXVII–LIII.
Conclusions
14 R. Byrne, K. J. Fraser, E. Izgorodina, D. R. MacFarlane,
M. Forsyth and D. Diamond, Phys. Chem. Chem. Phys., 2008,
10, 5919–5924.
15 S. Coleman, R. Byrne, S. Minkovska and D. Diamond, J. Phys.
Chem. B, 2009, 113, 15589–15596.
16 R. Byrne, S. Coleman, K. J. Fraser, A. Raduta, D. R. MacFarlane
and D. Diamond, Phys. Chem. Chem. Phys., 2009, 11, 7286–7291.
17 R. Byrne, S. Coleman, S. Gallagher and D. Diamond, Phys. Chem.
Chem. Phys., 2010, 12, 1895–1904.
18 A. K. Chibisov and H. Gorner, Chem. Phys., 1998, 237,
425–442.
19 H. Gorner, Chem. Phys. Lett., 1998, 288, 589.
20 H. Gorner, Phys. Chem. Chem. Phys., 2001, 3, 416–423.
21 H. Gorner and A. K. Chibisov, J. Chem. Soc., Faraday Trans.,
1998, 94, 2557–2564.
22 K. J. M. Laidler and J. H. Meiser, Physical Chemistry, Houghton
Mifflin, Boston, MA, 3rd edn, 1999.
23 J. Roncali, Chem. Rev., 1992, 92, 711–738.
24 J. Roncali, Chem. Rev., 1997, 97, 173–205.
25 S. Gambhir, K. Wagner and D. L. Officer, Synth. Met., 2005, 154,
117–120.
26 P. Wagner and D. L. Officer, Synth. Met., 2005, 154, 325–328.
27 M. C. Gallazzi, L. Castellani, R. A. Marin and G. Zerbi, J. Polym.
Sci., Part A: Polym. Chem., 1993, 31, 3339–3349.
28 C. H. E. Zhou, Z. Tan, C. Yang and Y. Li, J. Polym. Sci., Part B:
Polym. Phys., 2006, 44, 4916–4922.
The photochemical and thermodynamic properties of two
derivatives, BSP-2 and BSP-3, incorporating terthiophene
units attached to spiropyrans via the indoline nitrogen atom
and ester linkers that differ by one methlyene unit have been
investigated. In BSP-2, the pyran is separated from the ester
group by a methylene group, whereas in BSP-3, the ester
group is directly linked to the terthiophene group. This
apparently rather small structural difference results in quite
striking differences in the characteristics of the derivatives,
such as the fluorescence emission spectra (dominated by the
terthiophene unit) and the thermal relaxation kinetics of
the spiropyran unit. We believe these differences arise from
the ability of the p-electron clouds of the thiophene and
spiropyran units to align and form intramolecular p–p inter-
actions in BSP-2 due to the flexibility of the longer linker
chain. In BSP-3, the terthiophene conjugation extends into the
ester group of the linker, and this additional rigidity, coupled
with the shorter linker length, prevents a similar type of
interaction from occurring. We are now developing a new
generation of polymers based on these derivatives that could
exploit such effects.
29 S. H. Hosseini and A. A. Entezami, Polym. Adv. Technol., 2001, 12,
524–534.
30 L. Kumpumbu-Kalemba and M. Leclerc, Chem. Commun., 2000,
1847–1848.
31 I. Levesque, P. Bazinet and J. Roovers, Macromolecules, 2000, 33,
2952–2957.
Acknowledgements
MZ, RB and DD acknowledge funding from Science Founda-
tion Ireland (SFI) under the CLARITY CSET award (Grant
07/CE/I1147). KJF acknowledges the European Commission
for financial support through a Marie Curie Actions Inter-
national Re-integration Grant (IRG) (PIRG07-GA-2010-
268365) and Irish Research Council for Science, Engineering
and Technology. We also acknowledge support from the
European Commission for funding under grant PIRSES-
GA-2010-269302. Financial support from the Australian
Research Council and access to the Australian National
Fabrication Facility are also gratefully acknowledged.
32 I. Levesque and M. Leclerc, Chem. Mater., 1996, 8, 2843–2849.
33 I. Levesque and M. Leclerc, Macromolecules, 1997, 30, 4347–4352.
34 F. Raymond, N. Di Cesare, M. Belletete, G. Durocher and
M. Leclerc, Adv. Mater., 1998, 10, 599–602.
35 A. Yassar, C. Moustrou, H. K. Youssoufi, A. Samat,
R. Guglielmetti and F. Garnier, Macromolecules, 1995, 28,
4548–4553.
36 K. Yoshino, S. Nakajima, M. Onoda and R. Sugimoto, Synth.
Met., 1989, 28, 349–357.
37 G. Zotti, R. A. Marin and M. C. Gallazzi, Chem. Mater., 1997, 9,
2945–2950.
38 K. Wagner, R. Byrne, M. Zanoni, S. Gambhir, L. Dennany,
R. Breukers, M. Higgins, P. Wagner, D. Diamond,
G. G. Wallace and D. L. Officer, J. Am. Chem. Soc., 2011, 133,
5453–5462.
39 A. Higuchi, A. Hamamura, Y. Shindo, H. Kitamura, B. O. Yoon,
T. Mori, T. Uyama and A. Umezawa, Biomacromolecules, 2004, 5,
1770–1774.
40 J. M. Locke, S. Gambhir, R. D. Breukers, K. Wagner,
K. W. Jolley, G. G. Wallace and D. L. Officer, 2012, manuscript
in preparation.
41 D. Diamond and V. C. A. Hanratty, Spreadsheet Applications
for Chemistry Using Microsoft Excel, John Wiley and Sons,
New York, 1997.
References
1 Y. Hirshberg, J. Am. Chem. Soc., 1956, 78, 2304–2312.
2 E. Berman, R. E. Fox and F. D. Thomson, J. Am. Chem. Soc.,
1959, 81, 5605–5608.
3 C. A. Heller, D. A. Fine and R. A. Henry, J. Phys. Chem., 1961,
65, 1908–1909.
4 J. G. Calvert and J. N. J. Pitts, Photochemistry, John Wiley and
Sons, Inc., New York, N.Y., 1966.
5 R. Guglielmetti, Stud. Org. Chem. (Amsterdam), 1990, 40,
855–879.
6 V. I. Minkin, Chem. Rev., 2004, 104, 2751–2776.
7 T. Yoshida and A. Morinaka, J. Photochem. Photobiol., A, 1994,
78, 179–183.
42 J. B. Flannery, J. Am. Chem. Soc., 1968, 90, 5660–5671.
43 J. T. C. Wojtyk, A. Wasey, P. M. Kazmaier, S. Hoz and E. Buncel,
J. Phys. Chem. A, 2000, 104, 9046–9055.
44 T. W. Shin, Y. S. Cho, Y. D. Huh, K. D. Lee, W. K. Yang, J. Park
and I. J. Lee, J. Photochem. Photobiol., A, 2000, 137, 163–168.
c
9120 Phys. Chem. Chem. Phys., 2012, 14, 9112–9120
This journal is the Owner Societies 2012