Cyanide sensing with organic dyes: Studies in solution and on nanostructured Al2O3 surfaces

Article abstract of DOI:10.1002/chem.200700412

The synthesis of two new azo phenyl thiourea compounds and their optical response to different anions is reported herein. Solution studies in methanol indicate that cyanide induces a colour change in these dyes (whereas no changes are observed in the presence of other anions, such as F^-, Cl ^-, Br^-, CH₃COO^-, H ₂PO₄^-, HSO₄^-). Interestingly, in DMSO these dyes are responsive not only to cyanide, but also to fluoride, acetate and dihydrogen phosphate. Each of these anions induces a different colour change. In the second part of the paper, we report the attachment of one of these dyes onto nanostructured TiO₂ and Al ₂O₃ films. The stability of these sensitised films to pH was studied and we concluded that the sensitised Al₂O₃ films are more robust, and hence, better than the TiO₂ for anion sensing. The dye-sensitised Al₂O₃ films were immersed in solutions of different anions and their response studied. The films can detect cyanide down to 3 ppm in aqueous solution with relatively good selectivity over other anions.

Full text of DOI:10.1002/chem.200700412

DOI: 10.1002/chem.200700412
Cyanide Sensing with Organic Dyes: Studies in Solution and on
Nanostructured Al2O3 Surfaces
NØlida Gimeno, Xiaoe Li, James R. Durrant, and Ramón Vilar*[a]
Abstract: The synthesis of two new azo
phenyl thiourea compounds and their
optical response to different anions is
reported herein. Solution studies in
methanol indicate that cyanide induces
a colour change in these dyes (whereas
no changes are observed in the pres-
ence of other anions, such as Fꢀ, Clꢀ,
to fluoride, acetate and dihydrogen
phosphate. Each of these anions indu-
ces a different colour change. In the
second part of the paper, we report the
attachment of one of these dyes onto
nanostructured TiO2 and Al2O3 films.
The stability of these sensitised films to
pH was studied and we concluded that
the sensitised Al2O3 films are more
robust, and hence, better than the TiO2
for anion sensing. The dye-sensitised
Al2O3 films were immersed in solutions
of different anions and their response
studied. The films can detect cyanide
down to 3 ppm in aqueous solution
with relatively good selectivity over
other anions.
Keywords: anions · cyanide detec-
tion · sensitizers · sensors · thin
films
ꢀ
ꢀ
Brꢀ, CH3COOꢀ, H2PO4 , HSO4 ). In-
terestingly, in DMSO these dyes are re-
sponsive not only to cyanide, but also
Introduction
sorption or emission) of a receptor are especially attractive.
Recent examples that use this approach have been reported
by the groups of Geddes,[19–21] Martinez-MaÇez,[22–25] Palo-
mares[26,27] and Raymo[9,28] amongst others.[29–33] In spite of
these developments there are still relatively few examples of
selective probes for cyanide owing to high interference from
other anions, in particular, fluoride.[33,34] Another drawback
with most of the current methods is that the detection needs
to be carried out in organic solvents (or mixtures of organic
solvents and water), which limits their application to the
analysis of “real” samples that are often aqueous solutions.
In addition, in most of the current methods the chemical
sensors have to be in solution to perform sensing. Only in a
few cases has sensing been reported to take place over a sur-
face or film,[27,35,36] which is very convenient for quick and
easy detection of the anion.
The recognition and sensing of anions has received consider-
able attention over recent years.[1–7] Part of this interest
stems from the important roles anions play in biological pro-
cesses (e.g., phosphates).[8] In addition, there is interest in
finding better and more efficient ways of detecting anions
that can be potentially harmful to the environment or
human health. One such anion is cyanide, which is lethal to
humans at concentrations of 0.5–3.5 mg per kg of body
weight.[9] Consequently, the presence of this anion in drink-
ing water (e.g., as a result of spillage from industrial waste
or mining activities) can pose a very serious risk to human
health. It is therefore evident that reliable and efficient
ways of detecting the presence of cyanide in water are
needed.
Several methods have previously been developed to
detect cyanide by using a wide diversity of experimental
protocols and detection techniques.[10–18] Methods based on
anion-induced changes of the optical properties (either ab-
Herein, we report the synthesis of two new azo phenyl
thiourea compounds and their ability to selectively detect
cyanide. The first studies we present have been carried out
in solution (in methanol and DMSO), which show that these
dyes are indeed sensitive to the presence of anionic species.
The second part of our investigations focused on incorporat-
ing one of these dye compounds onto nanostructured semi-
conducting surfaces to generate cyanide-sensing films.
[a] N. Gimeno, X. Li, J. R. Durrant, R. Vilar
Department of Chemistry
Imperial College London
London SW7 2AZ (UK)
Fax : (+44)20-7594-1139
E-mail: r.vilar@imperial.ac.uk
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
3006
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3006 – 3012

FULL PAPER
Results and Discussion
Synthesis of thiourea dyes 1 and 2: 4-Nitrophenylazo deriva-
tives have previously been used as chromogenic subunits in
the development of receptors for anion sensing.[37] These
systems contain an acceptor nitro-phenyl moiety connected
to a donor atom (N or O) through a conjugated azo system.
Modification of the nature of the donor through hydrogen
bonding can result in a redshift in the charge-transfer (CT)
band in the UV-visible region, and consequently, in a
change in colour. With this in mind, compounds 1 and 2
were prepared (see Scheme 1). The synthesis was carried
out by treating the commercial dye Disperse Orange 3 (4-
(4-nitrophenylazo)aniline) with thiophosgene, followed by
treatment with the corresponding amine, either 4-aminoben-
zoic acid or aniline.
Figure 1. Normalised UV/Vis spectra of a 0.5 mm solution of 1 in MeOH
upon addition of different anions (30 equiv).
sorption spectrum of the dye (the lmax of the dye shifted
from 390 to 414 nm). This was also evident by naked-eye in-
spection: a colour change from pale orange to red/orange
was only observed upon addition of cyanide (Figure 2).
Scheme 1.
These compounds were characterised by spectroscopic
1
and analytical techniques. The H and 13C NMR spectra of
both samples were consistent with the proposed formula-
tions. The integrations and multiplicity of the aromatic reso-
nances in the H NMR spectra were indicative of the pres-
Figure 2. Photograph of solutions of 1 with different anions in methanol
(top) and DMSO (bottom).
1
ence of both aromatic rings in 1 and 2. A resonance in the
13C NMR spectra of both samples at d=180 ppm was as-
signed to the thiourea carbon. The formulations of these
two molecules were confirmed by FAB+ mass spectrometry
(which gave a molecular peak at 422 amu for 1 and at
378amu for 2) and elemental analyses.
Once the selectivity of the dye for cyanide sensing was es-
tablished, it was of interest to determine its sensitivity to-
wards this anion. Titration of a 0.5 mm solution of 1 with cy-
anide showed a progressive shift from 390 to 414 nm
(Figure 3).
Studies of the anion–dye interactions in solution: The anion
sensing properties of 1 and 2 were studied in solution by
using methanol and DMSO as solvents. An intense band for
the p!p* transition of the 4-nitroazophenyl chromophore
appears at 390 (in methanol) and 412 nm (in DMSO) for 1
and at 414 nm for 2 (in DMSO). The anions selected for
Since the optical properties of azo dyes can change as a
function of the solvent in which they are dissolved, analo-
gous studies to those described above were carried out in
DMSO. In this case, more striking colour changes were ob-
served, not only for cyanide, but also for other anions
(Figure 2). Addition of CNꢀ or Fꢀ to a solution of 1 in
DMSO, changed its colour from pale orange to dark violet
ꢀ
ꢀ
these studies were Fꢀ, Clꢀ, Brꢀ, CH3COOꢀ, H2PO4 , HSO4
ꢀ
and CNꢀ. A solution of 1 in methanol (0.5 mm) was treated
with a fixed amount of the different anions (30 equiv, added
as tetrabutyl ammonium salts). As shown in Figure 1, only
the addition of CNꢀ caused a change in the UV-visible ab-
and blue, respectively, whereas CH3COOꢀ and H2PO4 in-
duced a change to dark red/purple. In contrast, no changes
in colour were detected upon addition of other anions, such
ꢀ
as Clꢀ, Brꢀ or HSO4 .
Chem. Eur. J. 2008, 14, 3006 – 3012
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3007

R. Vilar et al.
origin of the colour changes and differences in the behav-
iour of 1 in methanol and DMSO are not immediately evi-
dent. Previous studies with urea- and thiourea-containing
chromophores have shown that, in principle, anions can
induce changes in colour through two distinct processes: hy-
ꢀ
drogen bonding or deprotonation of the N H groups. These
two processes are highly dependant on the solvent in which
the studies are carried out and the basicity of the anions
under study. To investigate this further, several studies were
carried out.
First, analogous dye 2 (see Scheme 1), which lacks a ter-
minal COOH group, was prepared and its optical response
to anions was studied. In DMSO, by naked-eye inspection,
this dye behaves largely as compound 1. However, closer in-
spection of the absorption spectra upon the addition of in-
creasing amounts of the corresponding anions revealed an
isosbestic point, which indicated that the conversion of the
initial product to the final blue-purple compound does not
go via a detectable intermediate (see Figure 5 for an exam-
Figure 3. Changes in the UV/Vis spectra upon the addition of increasing
amounts of cyanide to a 0.5 mm solution of 1 in methanol. With this data
a logK value of 2.02ꢁ0.13 was obtained.
To investigate quantitatively the changes observed in
ꢀ
DMSO upon addition of CNꢀ, Fꢀ, CH3COOꢀ and H2PO4 , a
series of titrations using UV-visible spectroscopy were car-
ried out. An example of the titration curves obtained upon
addition of increasing amounts of cyanide to 1 is shown in
Figure 4. The changes observed in the absorption spectrum
Figure 4. Changes to the spectrum of 1 upon the addition of increasing
amounts of cyanide, showing the lmax shift from 412 to 542 nm. The spec-
tra were obtained by using a 0.5 mm solution of 1 in DMSO.
of 1 suggest that three different species are present, the pro-
portions of which depend on the amount of cyanide added.
In the absence of anion, compound 1 in DMSO has a lmax
value of 412 nm. Upon addition of one equivalent of cya-
nide, a second species with maximum at around 460 nm (the
peak is too broad to determine the lmax value accurately) ap-
pears. Increasing amounts of cyanide yield a third species
with a lmax of 542 nm, which is the compound associated
with the dark blue/violet colour of the final solution. The
Figure 5. a) Changes in the spectrum of 2 upon the addition of increasing
amounts of cyanide, showing the lmax shift from 414 to 520 nm. The spec-
tra were obtained by using a 0.1 mm solution of 2 in DMSO. b) Plot
showing the change in absorbance of 2 at 414 nm (as a DA) with the ad-
dition of increasing amounts of cyanide.
3008
www.chemeurj.org
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3006 – 3012

Cyanide Sensing Organic Dyes
FULL PAPER
ple in which cyanide was used as the titrant; plots for the ti-
trations with other anions can be found in the Supporting
Information).
From the plot of DA versus equivalents of cyanide added
(Figure 5b) the overall equilibrium constant of the system
was determined to be logK=4.2ꢁ0.15, which, as discussed
below, corresponds to the deprotonation of the thiourea
group by the corresponding anion.
With anions such as fluoride or cyanide, it is likely that the
second NH group is also deprotonated to yield a fully conju-
gated species with an intense blue-violet colour. Our spec-
troscopic observations (both by H NMR and UV/Vis spec-
troscopies) are consistent with this type of deprotonation
process.
1
Sensitivity and detection limit in solution: Once the anion
selectivity of 1 and 2 in MeOH and DMSO was established,
it was of interest to determine the sensitivity and the detec-
tion limit of the system. The former can be defined as the
rate at which the absorbance changes upon addition of the
From these results, it is clear that the major colour
changes of 1 upon addition of anions cannot be the result of
simple deprotonation of the carboxylic acid of the dye be-
cause almost the same colour changes were observed for
both 1 and 2 when cyanide was added. However, it is likely
that the intermediate species observed by UV-visible spec-
troscopy in the titration of 1 (but not in 2) with cyanide
(lmax ꢂ460 nm.) is indeed associated to the deprotonation of
the carboxylic acid group in 1.
analyte, namely, d(A)/d[anion]. On the other hand, the de-
R
tection limit depends on the instrumentation (i.e., the mini-
mum change in absorbance that can be reliably measured
spectroscopically, which is 0.01 in this case) and on the sen-
sitivity of the dye (as defined above). By using these two pa-
rameters, the detection limit can then be calculated as 0.01/
sensitivity. With the titration data shown in Figures 3, 4 and
5, it was possible to determine the sensitivity and detection
limit of 1 (in MeOH) to cyanide, and of 1 and 2 (in DMSO)
to cyanide, fluoride, acetate and phosphate. These results
are summarised in Table 1.
1H NMR spectroscopic studies were carried out to gain
some insight into the structural changes of the dyes upon
addition of different anions (namely, Clꢀ, CH3COOꢀ,
ꢀ
H2PO4 and Fꢀ; cyanide was avoided owing to its toxicity
and the fact that in DMSO solution its behaviour was analo-
gous to that of fluoride). The corresponding anions
(30 equiv) were added to solutions of 1 in [D6]DMSO and
1
the H NMR spectra were measured. Upon the addition of
Table 1. Detection limit (defined as 0.01/sensitivity in which sensitivity=
the anions, the orange dye solution retained its original
colour for chloride, turned red for acetate and phosphate,
d(A)/d[anion]) in ppm for 1 and 2 for different anions. In the case of 1 in
G
MeOH the detection limit is only reported for cyanide because no optical
changes were detected with other anions.
1
and blue for fluoride. The H NMR spectra of the solutions
ꢀ
Dye Solvent
CNꢀ
[ppm]
Fꢀ
[ppm]
CH3COOꢀ
[ppm]
H2PO4
[ppm]
that contained fluoride and acetate showed very broad
peaks, whereas for chloride and phosphate sharp and clearly
identifiable peaks were observed.
1
1
2
MeOH[a] 8.00
DMSO[a] 0.95
DMSO[b] 0.06
–
0.33
0.23
–
0.92
0.17
–
0.57
1.04
The spectrum of the initial anion-free solution of 1
showed six doublets between d=7.69 and 8.44 ppm, which
corresponded to the protons of the three aromatic rings
(two signals for each) and two singlets at d=10.44 and
10.49 ppm, which corresponded to the protons of the thiour-
ea. Upon addition of chloride to this solution, the aromatic
protons of 1 appeared to be slightly shifted, but the original
pattern was retained; the resonances associated with the thi-
ourea NH protons shifted to d=11.70 and 11.83 ppm, re-
spectively, which suggested the formation of hydrogen
bonds between the chloride and the dye. In contrast, addi-
tion of dihydrogen phosphate to 1 (which produces a colour
[a] Concentration of the dye was 0.5 mm. [b] Concentration of the dye
was 0.1 mm.
Dye sensitised TiO2 and Al2O3 films: Once the anion-sens-
ing properties of 1 were established in solution, it was of in-
terest to incorporate this dye onto nanostructured semicon-
ducting films. The advantages of doing this is that dip-in
sensing is then possible (which is much more practical), and
as we and others have previously demonstrated,[26,27] recep-
tors that are not soluble in water (as is the case with 1) once
supported onto films, can be exposed to aqueous solutions
of the analytes to be measured.
Compound 1 was first supported on a TiO2 film by im-
mersing it in a solution of the dye, which yielded an orange
film with a lmax value of 366 nm. The film was then exposed
to aqueous solutions of various anions (the same as those
used in solution) and the optical changes of the film investi-
gated. An initial qualitative naked-eye study indicated that
the film only changed colour (to a darker orange-red) in the
presence of cyanide. This behaviour is analogous to that ob-
served for 1 in methanol (see above). However, a closer in-
spection of the system showed that in the presence of cya-
nide, not only does the adsorbed dye change colour, but it
also desorbs from the film. This was confirmed by measuring
1
change to red) generated greater changes in the H NMR
spectrum; two of the doublets associated with the aromatic
protons of 1 and the two NH thiourea resonances disap-
peared. This clearly indicates more important changes to the
structure of 1 than simple hydrogen-bonding interactions.
Based on the previous work reported by the groups of
Fabrizzi[39–41] and Gale[42–44] for similar systems to that pre-
sented herein, the striking change in colour observed for sol-
utions of 1 and 2 in DMSO can be associated with deproto-
nation of the NH groups of the thiourea. In the first step
and with the less basic anions, such as CH3COOꢀ and
ꢀ
H2PO4 , only one of the NH groups of the thiourea is de-
protonated (in the case of 1 the carboxylic acid should also
be deprotonated) to yield a red-coloured conjugated species.
Chem. Eur. J. 2008, 14, 3006 – 3012
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3009

R. Vilar et al.
the UV-visible absorption spectrum of the sensitised TiO2
films in the presence of increasing amounts of cyanide. The
intensity of the lmax band decreased with increasing amounts
of the anion; in addition the aqueous solution in which the
film was immersed, turned orange, which indicated that the
dye was desorbing.
It is likely that this problem is caused by an increase in
the pH (more basic) of water upon the addition of anions.
To investigate this hypothesis, the 1-sensitised film was im-
mersed in aqueous solutions of different acidity (namely,
pH 5 or 9) and left in such solutions for some time. As
shown in Figure 6, a constant decrease in the intensity of the
lmax of the film was observed at pH 9 with time. Considering
that this pH value can be easily reached in unbuffered aque-
ous solutions of anions, it was concluded that TiO2-sensitised
films were not suitable for sensing anions with this dye.
Therefore, we investigated whether Al2O3 could be suita-
ble as nanostructured support for sensing anions with 1.
Al2O3 is of particular interest
Figure 6. Plot showing the reduction in intensity of the lmax absorption
for the 1/TiO2 and 1/Al2O3 films as a function of pH.
because it has a basic point of
zero charge relative to TiO2 (at
pHꢂ4 for TiO2 and pHꢂ9 for
Al2O3).[38,45] Hence, at pH
values below ꢂ9 the Al2O3 sur-
face will be positively charged,
which allows strong interaction
with the carboxylate group of
the dye. The result of this is
that 1 does not easily desorb
from Al2O3 at pH<9 (in con-
trast with TiO2 in which the sur-
face is only positive below pH
ꢂ4, and therefore, the dye de-
sorbs at neutral or basic pH
values).
The methodology to prepare
the mesoporous Al2O3 films is
given in the Experimental Sec-
tion. Similar dye loading was
observed for these films com-
pared to TiO2.
As can be seen in Figure 6,
dye-sensitised Al2O3 films
Figure 7. Normalised absorption spectra of 1/Al2O3 films immersed in 5 mm aqueous solutions of different
anions.
(using 1) were indeed much
more stable towards desorption
at basic pH values, even after leaving the film immersed in
the basic solution for 20 min.
was also visible by naked eye inspection of the film (see
inset in Figure 7); the colour changed from yellow/orange to
red/orange.
Anion response of Al2O3-sensitised films: Having estab-
lished the stability of the 1/Al2O3 film towards desorption
under basic conditions, its response to different anions was
studied by UV-visible spectroscopy. The film was immersed
in 1 and 5 mm aqueous solutions of different anions (namely,
To establish the sensitivity of 1/Al2O3 to cyanide and the
detection limit of the system, the absorbance of the film at
different cyanide concentrations was recorded (Figure 8).
Plotting DA versus [CN] allowed us to calculate the sensitiv-
ity (namely, d(DA)/d(CN)) to be 0.3 mmꢀ1 and the detection
limit to be 2.6 ppm (using 0.01/sensitivity as explained
above).
ꢀ
ꢀ
Fꢀ, Clꢀ, Brꢀ, CH3COOꢀ, H2PO4 , HSO4 and CNꢀ) and the
absorption spectra recorded. As shown in Figure 7, immer-
sion of the film in a 5 mm solution of cyanide caused a lmax
shift of the sensitised film from 376 to 406 nm. This shift
3010
www.chemeurj.org
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3006 – 3012

Cyanide Sensing Organic Dyes
FULL PAPER
UV/Vis absorption spectra were monitored at room temperature by
using a Thermo Genesys 10 UV/Vis spectrophotometer. Elemental analy-
ses were carried out at the London Metropolitan University. The associa-
tion constants were calculated from non-linear curve fitting of the spec-
troscopic data by using a 1:1 binding model with the SPECFIT pro-
gram.[46]
Synthesis of 1: Step 1: Disperse Orange 3 (2.12 g, 8.74 mmol) was dis-
solved in THF (150 mL), and thiophosgene (1.35 mL, 17.58mmol) was
added. The reaction mixture was heated to reflux and stirred for 1 h
while argon was bubbled through the solution. All volatiles were evapo-
rated under reduced pressure to yield 4-nitroazobenzene-4’-isothiocya-
nate as a red solid. This was used in the next step without further purifi-
cation.
Step 2: 4-Carboxyaniline (1.0 g 7.28mmol) was dissolved in dried THF
(50 mL). 4-Nitroazobenzene-4’-isothiocyanate (2.480 g 8.74 mmol; ob-
tained in the first step) was dissolved in THF (100 mL) and added to the
solution of 4-carboxyaniline. The reaction mixture was heated at reflux
for 6 h under N2 and stirred at room temperature for a further 12 h. The
solid was removed by filtration and the solvent evaporated to yield an
orange-red solid. The crude solid was washed with methanol to yield the
pure product as an orange solid (1.84 g, 60%). 1H NMR (400 MHz,
[D6]DMSO): d=8.44 (d, J=9.2 Hz, 2H), 8.07 (d, J=9.2 Hz, 2H), 7.99 (d,
J=9.2 Hz, 2H), 7.93 (d, J=8.4 Hz, 2H), 7.84 (d, J=9.2 Hz, 2H),
7.69 ppm (d, J=8.4 Hz, 2H); 13C NMR (400 MHz, CDCl3): d=179.5,
167.4, 155.8, 148.6, 148.5, 144.3, 143.9, 130.5, 126.6, 125.6, 124.5, 123.8,
123.1, 122.5 ppm; UV/Vis: lmax (e)=376 nm (8,300mꢀ1 cmꢀ1); MS-
Figure 8. Normalised absorption spectra of 1/Al2O3 films immersed in
aqueous solutions of cyanide at different concentrations.
Conclusions
A
C20H15N5O4S: C 57.00, H 3.59, N 16.62; found: C 57.25, H 3.72, N 16.42.
Two new dyes have been prepared and their optical re-
sponse to anions has been studied both in solution and sup-
ported onto Al2O3 nanostructured films. In solution, the
colour changes displayed by the dye strongly depend on the
solvents employed. In methanol, the dye only changes
colour in the presence of cyanide (with a detection limit of
8ppm); even when a large excess of the other anions was
used, no colour changes were observed. In DMSO, not only
CNꢀ induces colour changes (in this case to dark purple),
Synthesis of 2: Aniline (0.082 g, 80 mL 0.88 mmol) was dissolved in dried
THF (10 mL). 4-Nitroazobenzene-4’-isothiocyanate (0.300 g, 1.06 mmol;
obtained as described in the synthesis of 1) dissolved in dried THF
(10 mL) was added and the mixture was heated at reflux for 2 h under
N2. The mixture was then allowed to cool, the solid was removed by fil-
tration and the solvent was evaporated to yield an orange-red solid. The
crude solid was washed with methanol to yield the pure product as an
orange solid (0.17 g, 50%). 1H NMR (400 MHz, [D6]DMSO): d=10.28
(s, 1H), 10.16 (s, 1H), 8.43 (d, J=8.8 Hz, 2H), 8.06 (d, J=8.8 Hz, 2H),
7.97 (d, J=8.8 Hz, 2H), 7.84 (d, J=8.8 Hz, 2H), 7.51 (d, J=8.0 Hz, 2H),
7.36 (t, J=8.0 Hz, 2H), 7.16 ppm (t, J=8.0 Hz, 1H); 13C NMR
(400 MHz, CDCl3): d=179.7, 155.8, 148.6, 148.3, 144.7, 139.6, 129.0,
125.6, 125.3, 124.4, 124.1, 123.8, 122.9 ppm; UV/Vis: lmax (e)=412 nm
(e=47,300mꢀ1 cmꢀ1) MS (FAB+): m/z (%): 378(100) [ M+1]+; elemen-
tal analysis calcd (%) for C19H15N5O2S: C 60.46, H 4.01, N 18.56; found:
C 60.34, H 4.30, N 18.90.
ꢀ
but also Fꢀ (to blue), CH3COOꢀand H2PO4 (both to violet/
red). Our spectroscopic data, together with recent reports in
the literature from the groups of Fabrizzi[39–41] and Gale,[44]
strongly suggests that the origin of the colour changes is not
a simple hydrogen-bonding interaction, but a process that
involves deprotonation of the thiourea NH groups.
Compound 1 was then supported on nanostructured
Al2O3 films and its optical properties studied. These sensi-
tised films were stable to desorption over a wide pH range
(in contrast with the behaviour observed with TiO2 films).
Immersion of the 1-sensitised Al2O3 film in an aqueous solu-
tion of cyanide induced a colour change that can be used for
the detection of cyanide down to 2.6 ppm. These easy-to-use
films also showed a good degree of selectivity (although
these will need further improvements in future work).
Preparation of mesoporous nanocrystalline TiO2 films: Mesoporous
nanocrystalline films consisting of anatase TiO2 particles (15–25 nm in
size) were prepared as follows: titanium isopropoxide (40 mL) was mixed
with glacial acetic acid (9.1 g) under a nitrogen atmosphere and stirred
for 10 min. The mixture was then poured into a conical flask that con-
tained 0.1m aqueous nitric acid (240 mL) at room temperature and it was
subsequently stirred at 808C for 8h. The TiO 2 colloidal was filtered by
using a 0.45 mm syringe filter and then autoclaved at 2208C for 12 h. The
colloids were re-dispersed with a 2 min cycle burst from an LDU Sonip-
robe horn, as reported previously.[26] The solution was then concentrated
to 12.5% (in TiO2 weight) on a rotary evaporator by using a membrane
vacuum pump at 45–508C. Carbowax 20000 (50% TiO2 by weight) was
added and the resulting paste was stirred slowly overnight before distilled
water (50:50 w/w to the TiO2 paste) was added to dilute the TiO2 paste
for ensuring a thin film. The mixture was stirred slowly overnight prior to
film deposition to ensure homogeneity and prevent trapped air bubbles.
The paste was spread on the FTO substrates using a glass rod and 3M ad-
hesive tape was used as spacer. After drying in air, the film was sintered
at 4508C for 30 minutes in a furnace. The resulting film thickness, using
one layer of 3M adhesive tape, was of 1.4ꢁ0.1 mm (measured by Alpha
200 Profilometer).
Experimental Section
Materials and instrumentation: Solvents were dried from the appropriate
drying agent, degassed and stored under nitrogen. All commercially
available starting materials were purchased from Aldrich and not further
purified unless otherwise stated. 1H and 13C NMR spectra were recorded
by using a Bruker Avance 400 Ultrashield NMR spectrometer with TMS
as the internal reference. IR spectra were recorded by using a Research
Preparation of mesoporous nanocrystalline Al2O3 films: Colloidal alumi-
nium oxide (50 nm, Alfa Aesar) was diluted with distilled H2O (50:50)
and hydroxypropyl cellulose (2 wt%, 370,000 gmolꢀ1; Sigma Aldrich)
Series FTIR instrument as KBr disks in the range from 4000 to 500 cmꢀ1
.
Chem. Eur. J. 2008, 14, 3006 – 3012
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3011

R. Vilar et al.
was added. The mixture was stirred for a week to allow the polymer to
dissolve completely prior to film deposition, which ensured homogeneity
and that no air bubbles were trapped within the film. The Al2O3 paste
was doctor bladed onto the microscope slide with one layer of Scotch
Tape. The film was dried at room temperature for about 20 min, and
then calcined at 4008C for 30 in. Typical Al2O3 films were (3.8ꢁ0.2) mm
thick. Films were recalcined for 20 min at 4008C before dye sensitization.
[17] M. J. Schçning, J. P. Kloock, D. T. Knobbe, J. Raede, M. Keusgen,
Chem. Sens. 2004, 20, 98–99.
[18] D. Shan, C. Mousty, S. Cosnier, Anal. Chem. 2004, 76, 178–183.
[19] R. Badugu, J. R. Lakowicz, C. D. Geddes, Anal. Biochem. 2004, 327,
82–90.
[20] R. Badugu, J. R. Lakowicz, C. D. Geddes, Curr. Anal. Chem. 2005,
1, 157–170.
[21] R. Badugu, J. R. Lakowicz, C. D. Geddes, J. Am. Chem. Soc. 2005,
127, 3635–3641.
[22] R. Martinez-Manez, F. Sancenon, Coord. Chem. Rev. 2006, 250,
3081–3093.
[23] J. V. Ros-Lis, R. Martinez-Manez, J. Soto, Chem. Commun. 2005,
5260–5262.
[24] J. V. Ros-Lis, R. Martinez-Manez, J. Soto, Chem. Commun. 2002,
2248–2249.
[25] F. Garcia, J. M. Garcia, B. Garcia-Acosta, R. Martinez-Manez, F.
Sancenon, J. Soto, Chem. Commun. 2005, 2790–2792.
[26] E. Palomares, R. Vilar, A. Green, J. R. Durrant, Adv. Funct. Mater.
2004, 14, 111–115.
[27] E. Palomares, M. V. Martinez-Diaz, T. Torres, E. Coronado, Adv.
Funct. Mater. 2006, 16, 1166–1170.
[28] M. Tomasulo, F. M. Raymo, Org. Lett. 2005, 7, 4633–4636.
[29] A. Afkhami, N. Sarlak, Sens. Actuators B 2007, 122, 437–441.
[30] R. Badugu, J. R. Lakowicz, C. D. Geddes, Anal. Chim. Acta 2004,
522, 9–17.
[31] R. Badugu, J. R. Lakowicz, C. D. Geddes, Dyes Pigm. 2005, 64, 49–
55.
The nanocrystalline Al2O3 material was further analysed by X-ray
powder diffraction, which showed that the structure of the alumina was
g-Al2O3. This was carried out by using a Bruker Advance D8instrument,
which measured at a 2q range of 20 to 808 with CuKa radiation.
Sensitising the TiO2 films: Compound 1 was adsorbed onto a TiO2 film
by soaking the film in a solution of the dye (1 mm) in a 1:1 mixture of
acetonitrile/tert-butanol at room temperature overnight. A yellow 1/TiO2
film was obtained. The dye loading on the film was determined to be
159 nmolcmꢀ2
.
Sensitising the Al2O3 films: Compound 1 was adsorbed onto a 4 mm-thick
Al2O3 film by soaking the film in a solution of the dye (0.1 mm) in a 1:1
mixture of acetonitrile/tert-butanol at room temperature overnight. A
yellow 1/Al2O3 film was obtained. The dye loading on the film was deter-
mined to be 152 nmolcmꢀ2
.
Acknowledgements
This work has been supported by the EU project STRP 516982 (HET-
EROMOLMAT). We are very grateful to Dr. Brian OꢁRegan for his
help in the preparation of the Al2O3 films and to Dr. Eugenia Martinez
for her help with the XRD data.
[32] C.-F. Chow, M. H. W. Lam, W.-Y. Wong, Inorg. Chem. 2004, 43,
8387–8393.
[33] P. Anzenbacher, Jr., D. S. Tyson, K. Jursikova, F. N. Castellano, J.
Am. Chem. Soc. 2002, 124, 6232–6233.
[34] H. Miyaji, J. L. Sessler, Angew. Chem. 2001, 113, 158–161; Angew.
Chem. Int. Ed. 2001, 40, 154–157.
[35] K. Poland, E. Topoglidis, J. R. Durrant, E. Palomares, Inorg. Chem.
Commun. 2006, 9, 1239–1242.
[1] T. Gunnlaugsson, M. Glynn, G. M. Tocci, P. E. Kruger, F. M. Pfeffer,
Coord. Chem. Rev. 2006, 250, 3094–3117.
[2] J. W. Steed, Chem. Commun. 2006, 2637–2649.
[3] C. Suksai, T. Tuntulani, Top. Curr. Chem. 2005, 255, 163–198.
[4] P. D. Beer, S. R. Bayly, Top. Curr. Chem. 2005, 255, 125–162.
[5] R. Martinez-Manez, F. Sancenon, Chem. Rev. 2003, 103, 4419–4476.
[6] P. A. Gale, Coord. Chem. Rev. 2003, 240, 191–221.
[7] P. D. Beer, P. A. Gale, Angew. Chem. 2001, 113, 502–532; Angew.
Chem. Int. Ed. 2001, 40, 486–516.
[8] E. A. Katayev, Y. A. Ustynyuk, J. L. Sessler, Coord. Chem. Rev.
2006, 250, 3004–3037.
[9] M. Tomasulo, S. Sortino, A. J. P. White, F. M. Raymo, J. Org. Chem.
2006, 71, 744–753.
[10] C.-L. Chen, Y.-H. Chen, C.-Y. Chen, S.-S. Sun, Org. Lett. 2006, 8,
5053–5056.
[11] V. Ganesh, M. P. Calatayud Sanz, J. C. Mareque-Rivas, Chem.
Commun. 2007, 5010–5012.
[12] A. E. Lindsay, D. OꢁHare, Anal. Chim. Acta 2006, 558, 158–163.
[13] D. Shan, S. Cosnier, C. Mousty, Biosens. Bioelectron. 2004, 20, 390–
396.
[14] C. G. Siontorou, D. P. Nikolelis, Anal. Chim. Acta 1997, 355, 227–
234.
[36] J. A. Legako, B. J. White, H. J. Harmon, Sens. Actuators B 2003, 91,
128–132.
[37] F.-Y. Wu, X.-F. Tan, Y.-M. Wu, Y.-Q. Zhao, Spectrochim. Acta Part
A 2006, 65, 925–929.
[38] C. E. McNamee, Y. Tsujii, M. Matsumoto, Langmuir 2005, 21,
11283–11288.
[39] V. Amendola, D. Esteban-Gomez, L. Fabbrizzi, M. Licchelli, Acc.
Chem. Res. 2006, 39, 343–353.
[40] M. Boiocchi, L. Del Boca, D. Esteban-Gomez, L. Fabbrizzi, M. Lic-
chelli, E. Monzani, Chem. Eur. J. 2005, 11, 3097–3104.
[41] D. Esteban-Gomez, L. Fabbrizzi, M. Licchelli, J. Org. Chem. 2005,
70, 5717–5720.
[42] L. S. Evans, P. A. Gale, M. E. Light, R. Quesada, NewJ. Chem.
2006, 30, 1019–1025.
[43] P. A. Gale, Acc. Chem. Res. 2006, 39, 465–475.
[44] L. S. Evans, P. A. Gale, M. E. Light, R. Quesada, Chem. Commun.
2006, 965–967.
[45] N. Omura, Y. Hotta, K. Sato, Y. Kinemuchi, S. Kume, K. Watari, J.
Am. Ceram. Soc. 2006, 89, 2738–2743.
[46] SPECFIT, version 3.0.36, Spectrum Software Associates, USA, 2006.
[15] A. R. Surleva, V. D. Nikolova, M. T. Neshkova, Anal. Chim. Acta
2007, 583, 174–181.
[16] M. Keusgen, J. P. Kloock, D. T. Knobbe, M. Junger, I. Krest, M.
Goldbach, W. Klein, M. J. Schçning, Sens. Actuators, B 2004, 103,
380–385.
Received: May 9, 2007
Revised: December 17, 2007
Published online: February 6, 2008
3012
www.chemeurj.org
ꢀ 2008Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 3006 – 3012