Quantum yield measurements for the photocatalytic oxidation of Acid Orange 7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness

Article abstract of DOI:10.1016/j.cattod.2014.04.019

This work comprises the photoactivity assessment of transparent sol-gel TiO₂ coatings of various thickness using two test systems. The initial rates of both photocatalytic reactions, namely the oxidative bleaching of Acid Orange 7 (AO7) and the reductive bleaching of 2,6-dichlorindophenol (DCIP) increase linearly with increasing titania film thickness as well as with increasing absorbed light flux. The latter work revealed quantum yields (QY) of 0.19% and 92% for the AO7 and DCIP test system, respectively. The low QY for the AO7 oxidation is due to the combination of a slow irreversible reduction of oxygen and also for the oxidation of AO7, thus favouring the high efficiency for electron-hole recombination that is typical for aqueous organic pollutants. In contrast, the very high QY for the photocatalysed reduction of DCIP is due to the presence of a vast excess of glycerol which traps the photogenerated holes efficiently and so allow time for the slower reduction of dye to take place. Furthermore, the oxidation of glycerol results in the generation of highly reducing R-hydroxyalkyl radicals that are able to also reduce DCIP. As a consequence of this 'current doubling' effect, the observed QY (92%) is much higher than the apparent theoretical value of 50%.

Full text of DOI:10.1016/j.cattod.2014.04.019

G Model
CATTOD-9047; No. of Pages6
ARTICLE IN PRESS
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Quantum yield measurements for the photocatalytic oxidation of Acid
Orange 7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on
transparent TiO2 films of various thickness
J. Kry´saa,∗, M. Baudysa, A. Millsb
a Department of Inorganic Technology, Institute of Chemical Technology Prague, Technická 5, 166 28 Prague 6, Czech Republic
b School of Chemistry and Chemical Engineering, Queen’s University Belfast, Stranmillis Road, Belfast BT9 5AG, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
This work comprises the photoactivity assessment of transparent sol–gel TiO2 coatings of various thick-
ness using two test systems. The initial rates of both photocatalytic reactions, namely the oxidative
bleaching of Acid Orange 7 (AO7) and the reductive bleaching of 2,6-dichlorindophenol (DCIP) increase
linearly with increasing titania film thickness as well as with increasing absorbed light flux. The latter
work revealed quantum yields (QY) of 0.19% and 92% for the AO7 and DCIP test system, respectively. The
low QY for the AO7 oxidation is due to the combination of a slow irreversible reduction of oxygen and
also for the oxidation of AO7, thus favouring the high efficiency for electron–hole recombination that is
typical for aqueous organic pollutants. In contrast, the very high QY for the photocatalysed reduction of
DCIP is due to the presence of a vast excess of glycerol which traps the photogenerated holes efficiently
and so allow time for the slower reduction of dye to take place. Furthermore, the oxidation of glycerol
results in the generation of highly reducing R-hydroxyalkyl radicals that are able to also reduce DCIP. As
a consequence of this ‘current doubling’ effect, the observed QY (92%) is much higher than the apparent
theoretical value of 50%.
Received 10 March 2014
Received in revised form 7 April 2014
Accepted 8 April 2014
Available online xxx
Keywords:
TiO2 film
Thickness
Acid Orange 7
2,6-Dichlorindophenol
Quantum yield
© 2014 Published by Elsevier B.V.
1. Introduction
superhydrohilic effect is quite controversial but its main feature
is that the surface of the photocatalyst becomes superhydrophilic,
A thin transparent titanium dioxide films prepared by sol–gel
method represent one type of photocatalytically active self-
cleaning surfaces. The self-cleaning action is derived from a
combination of photocatalytic activity and photoinduced superhy-
drophilicity. With both of these effects photons of light of energy
greater than or equal to the band gap energy of the semicon-
ductor photocatalyst generate electron hole pairs. In most cases
these pairs recombine to generate heat but the photocatalyst
phenomena is due to the photogenerated electrons and holes that
reach the surface of the photocatalytic particle where they can
participate in redox reactions [1]. Thus, photoinduced holes very
oxidising and are able, either directly or indirectly, to oxidise stable
organic pollutants and the photogenerated electrons are able to
reduce ambient oxygen. As a result, semiconductor photocatalysts
can be used to clean organic deposits from their surfaces [2].
The underlying mechanism associated with the photo-induced
i.e. exhibits a water droplet contact angle ꢀ10◦ upon irradiation
with ultra-bandgap light [3,4].
A number of organic compounds have been used for the deter-
mination of the photocatalytic activities of different films of TiO2
photocatalyst [5]. Most of them are oxidised by photogenerated
holes as the thionine dye Methylene Blue (MB) [6] or azo dye Acid
Orange 7 (AO7) [7–11]. The disadvantage of using the photoxidative
degradation of dyes (as MB or AO7) as a measure of photocat-
alytic activity is the long time which is needed to completely
decolourise the solution [12]. Another, usually faster, approach
to assess the photocatalytic activity of self cleaning glass is to
use model ink based on redox indicator. Such an ink consists
of dye which can be irreversible reduced by the photocatalyst
under test so as to produce a reduced, coloured or colourless form.
Dyes suitable for this purpose include: Resazurin (7-hydroxy-3H-
phenoxazin-3-one 10-oxide) (colour change from blue to pink)
[13] or 2,6-dichloroindophenol (DCIP) (colour change from blue to
colourless) [14–16]. In addition to the redox indicator, the pho-
tocatalyst activity indicator ink contains hydroxyethylcellulose,
which serves as a binder, and an excess (compared to the amount of
∗
Corresponding author. Tel.: +420 220444112; fax: +420 220444410.
E-mail address: josef.krysa@vscht.cz (J. Kry´sa).
http://dx.doi.org/10.1016/j.cattod.2014.04.019
0920-5861/© 2014 Published by Elsevier B.V.
Please cite this article in press as: J. Kry´sa, et al., Quantum yield measurements for the photocatalytic oxidation of Acid Orange
7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.04.019

G Model
CATTOD-9047; No. of Pages6
ARTICLE IN PRESS
2
J. Kry´sa et al. / Catalysis Today xxx (2014) xxx–xxx
redox dye used) of glycerol which acts as a hole trap. When the ink
is in contact with a photocatalyst indicator ink, and irradiated with
ultra-bandgap light, photogenerated holes, oxidise the glycerol to
glyceraldehydes or to glycerol acid (Eq. (1)), leaving the photogen-
erated electrons to reduce the redox indicator dye, such as DCIP,
which is colourless in its reduced form (Eq. (2)) [14,17].
TiO2 layers of various thicknesses (from 1 layer to 5 layers) were
created.
2.2. Film characterisation
For each different type of film produced, the mass of TiO2
deposited was obtained using a mass balance with a precision
of 0.01 mg. These largely non-porous films exhibited the typical
interference spectra associated with thin films; a feature which
allowed the thicknesses of the titania films to be calculated from
the reflectance spectra using the NanoCalc software (Ocean Optics)
[18]. The location and presence of characteristic diffraction lines
were observed by X-ray diffraction (Diffractometer: Seifert – XRD
3000; Panalytical HighScore Plus). The XRD data were also used
to estimate the crystalline size. Diffuse reflectance spectra of TiO2
films were measured by UV–vis spectrophotometry (Varian Cary
100 with Ulbricht integrating sphere).
(1)
2.3. Determination of photocatalytic activity
The photocatalytic activity of each TiO2 film was probed using
two model test systems, namely: (i) the oxidative bleaching of
Acid Orange 7 (AO7) in solution and (ii) the reductive bleach-
ing of 2,6-dichloroindophenol (DCIP) in an ink. AO7 belongs to
(2)
the class of azo dyes characterised by the presence of
N N
group and the initial concentration of AO7 aqueous solution was
1 × 10−5 mol dm−3. The actual AO7 concentrations in the course
of the photocatalytic degradation were determined by absorption
spectroscopy (CECIL INSTRUMENTS CE 201) with the absorbencies
being measured at the maximum of the corresponding absorption
band for AO7, i.e. at 485 nm. For the photocatalytic degrada-
tion of AO7 in solution a set of four batch reactors was used
[19], in which the reactors were employed simultaneously under
the same conditions in photocatalytic degradation tests. Each
reactor was made of borosilicate glass which transmits light of
wavelengths > 330 nm and the reaction solution therein (25 ml)
was stirred continuously using a magnetic stirrer. The temper-
ature of the reaction solution was kept at 20 ◦C using a glass
jacket containing thermostatted, circulating water and the sam-
ples with TiO2 films were fixed in to the lid of the reactor. Each
photoreactor was irradiated horizontally using two 11 W fluo-
rescent lamps Sylvania Lynx-S BLB, with a broad maximum at
365 nm (I0 = 2.7 mW/cm2). The irradiance of the UV light was mea-
sured by UV metre Luton 365 nm with sensitive in the range
320–390 nm.
The DCIP ink was prepared by mixing of 18 g of 0.5% (m/m)
solution of 2-hydroxyethyl cellulose (Fluka) with 1.8 g of glycerol
(Sigma–Aldrich) and 30 mg of 2,6-dichloroindophenol (DCIP). A
coating of this well mixed ink solution was applied on the sol–gel
titania film by dip-coating the film in the ink, and then allow-
ing the ink layer to dry in the oven (70 ◦C, 10 min). The amount
of photocatalysed bleaching of the DCIP in the ink was deter-
mined using visible spectroscopy, by monitoring the absorbance
of the film at 630 nm. The main disadvantage of a photocatalyst
activity indicator ink containing DCIP is its lack of long-term sta-
bility in aqueous solution. Thus, after about 1 month the colour
of the ink changes from blue to brown as documented by the
apparent shift in the DCIP UV/vis absorption maximum to shorter
wavelengths with time. This instability in light and even at dark
conditions has been reported already by Lojander [20]. Erdey et al.
[21] recommend that an 0.1 wt% aqueous solution of DCIP should
only be used for three weeks. Therefore, in this work we used
freshly prepared DCIP ink which was never more than 2 weeks
old.
Whatever the method used to assess the activity of a photo-
catalyst films, an important parameter affecting the quantity of
absorbed light and hence the self-cleaning performance of the
photocatalyst film is its thickness. The aim of this work was to
prepare transparent sol–gel TiO2 coatings of various thickness and
assess their photocatalytic activities using two test systems, namely
an aqueous solution of the azo dye, Acid Orange 7, and a solid
film of the photocatalyst activity indicating ink, which contained
2,6-dichlorindophenol as the redox dye. Furthermore, by precise
measurement of the amount of UV light absorbed by the TiO2 films,
the quantum yield of both photocatalytic reactions was evaluated.
2. Experimental
2.1. Preparation of transparent SiO2/TiO2 films
In all cases an initial SiO2 underlayer was deposited on the
microscopic glass, where the SiO2 film was prepared as follows:
15 ml of ethanol (ethanol absolute p.a., Penta) were mixed with
10 ml of tetraethylorthosilicate (purum 98% Fluka) and after 90 min
of stirring 2.5 ml of distilled water and concentrated hydrochloride
acid (p.a. 36%, Penta) were added to the mixture.
The subsequent TiO2 overlayer was prepared using Titanium(IV)
isopropoxide (97%, Sigma–Aldrich) as the TiO2 precursor using a
method based on that reported in [18] in which: 11.6 ml of abso-
lute ethanol (p.a. Penta) were added drop wise under stirring to
8 ml of Titanium isopropoxide. Then, 11.6 ml of pure ethanol were
mixed with 2.75 ml of ethyl acetylacetate (99% p.a. Fluka) and 0.2 ml
of nitric acid (p.a. 65% Penta) were added to the isopropoxide
mixture. Thus prepared, the titania sol was stirred vigorously for
24 h.
Microscopic soda lime glass samples (7.5 cm × 12 cm × 0.1 cm)
were then dip-coated in the SiO2 sol (withdrawal speed
60 mm/min) and a SiO2 protective layer created by calcining the
coated glass at 550 ◦C for 3 h. The subsequent TiO2 photoactive over
– layer(s) was coated onto the glass using the same procedure as
described for the SiO2 layer, but using the titania, instead of the
SiO2 sol. By repeating the above procedure using the titania sol
Please cite this article in press as: J. Kry´sa, et al., Quantum yield measurements for the photocatalytic oxidation of Acid Orange
7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.04.019

G Model
CATTOD-9047; No. of Pages6
ARTICLE IN PRESS
J. Kry´sa et al. / Catalysis Today xxx (2014) xxx–xxx
3
Fig. 2. Absorption spectra of AO7 solution (initial concentration 1 × 10−5 M) in con-
tact with a typical 1 layer TiO2 film for various irradiation times. Insert shows the
dependence of absorption in maximum (485 nm) as a function of irradiation time.
Incident UV light intensity 12 W/m2.
Fig. 1. Dependence of TiO2 film thickness on number of layers (dipcoating and
heating cycles).
3. Results and discussion
Thus, simple visible absorption spectrophotometry can be used
to determine reliably the concentration changes of AO7 during its
photocatalytic oxidation. A typical set of absorption spectra of the
AO7 reaction solution as a function of irradiation time is shown in
Fig. 2 for a titania film (1 layer), with the insert diagram showing
the dependence of A(485 nm) on irradiation time, from which a
value for initial rate can be calculated.
3.1. Film characterisation
The film thicknesses of the prepared samples comprising vari-
ous layers of titania, t, can be calculated from mass of the film using
Eq. (3).
mfilm
S
1
t =
(3)
ꢀanatase((100 − P)/100) + ꢀair(P/100)
For the evaluation of photocatalytic oxidation rate of AO7 the
initial rate, Ri is given by:
where t represents the film thickness, S represents film geometrical
surface, m mass of the film, ꢀanatase density of titanium diox-
ide anatase modification (3.9 g/cm3 [22]), ꢀair density of air voids
(0.0112 g/cm3) and P porosity in %. For the present films poros-
ity was previously determined to be 4% [23]. Using this approach
and an interference method, based on reflectance measurements
and the NanoCalc software, two different values for t were deter-
mined for each of the different titania films made and the results
are illustrated in Fig. 1. From these data it can be seen that there
is a good agreement in the range from 0 to 250 nm, but for lay-
ers with thickness more than 250 nm, the value calculated from
weight increment is higher than for the same samples measured by
NanoCalc. This can be explained by the formation of an increasingly
significant rough non transparent part of the film at the bottom of
the sample during the withdrawal step in dip coating technique.
The rough part appeared on the films consisting of 3 and more lay-
ers (thickness higher than 250 nm) and represent about 5–10% of
the total geometrical surface of films. This rougher surface is asso-
ciated with an increased level of deposition of titania and so is
responsible for the slight increase in the TiO2 mass and thus the
observed slight positive deviation from linearity in Fig. 1 for the
thickness (calculated by mass) vs number of layers plot. Consid-
ering both sets of data illustrated in Fig. 1 it appears that there is
an approximate linear dependence of layer thickness with number
of deposited TiO2 layers, i.e. the thickness of a single layer of tita-
nia was 70 5 nm and increased approximately of 70 nm for every
added layer. XRD data of TiO2 films revealed, as observed previ-
ously [18] that the titania films generated by calcining the sol are
in the anatase crystalline phase with a crystallite size ca. 13 nm.
ꢀ
ꢁ
ꢀ
ꢁ
V
S
dcAO7
dt
V
1
dAAO7
dt
Ri = −
= −
·
(4)
S lεAO7
i
i
where Ri as a number of moles of model compound degraded
per unit time per unit irradiated area [units: mol min−1 cm−2],
(dcAO7/dt)i is the initial slope of AO7 concentration dependence on
irradiation time [mol dm−3 min−1], S is the irradiated geometrical
area of photocatalyst layer [cm2], V is the volume of the reac-
tion mixture [dm3], εAO7 is molar absorptivity of AO7 in aqueous
solution [2.17 × 104 dm3 mol−1 cm−1], l is optical length [1 cm] and
(dAAO7/dt)i is the initial slope of the absorbance vs irradiation time
plot, as illustrated in the insert plot in Fig. 2.
3.3. Reductive photocatalytic degradation of redox ink DCIP
The same titania films, with a DCIP ink film coating were also
irradiated and Fig. 3 shows a typical set of recorded absorption
spectra of the dried DCIP ink film on a TiO2 film (1 layer) as a
function of irradiation time. The change in UV/vis absorption spec-
trum of the ink film illustrated in Fig. 3 reveals that the absorption
band corresponding to DCIP (wavelengths maximum at 645 nm)
decreases with irradiation time. The decrease is linear (see the
insert graph in Fig. 3a plot of the absorbance at the absorption max-
imum of the DCIP ink as a function of t) as reported previously for
similar titania films [14]. As a consequence, the initial rate of DCIP
decolorisation in an ink film, (dADCIP,f/dt)i can be evaluated as the
slope of the Absorbance vs t plot, as illustrated in the insert diagram
in Fig. 3, with unit A s−1
.
The same approach to evaluating the oxidation rate of AO7 (Eq.
(4)) was used for the evaluation of the reduction rate of DCIP.
Dip coating is a technique conveniently used for the deposition of
homogeneous, and sufficiently thin, ink films on the TiO2 sol gel
layer for irradiation tests. However, the disadvantage of this tech-
nique is that we do not evaluate directly the exact molar amount
of dye present in the ink film per unit surface area. Therefore, it
3.2. Oxidative photocatalytic degradation of Acid Orange 7
Previous work has demonstrated that UV/vis absorption
spectrophotometry is suitable for analysing the photocatalytic
degradation of AO7 and that no products of AO7 degradation
absorb around 485 nm are produced at a detectable level [20].
Please cite this article in press as: J. Kry´sa, et al., Quantum yield measurements for the photocatalytic oxidation of Acid Orange
7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.04.019

G Model
CATTOD-9047; No. of Pages6
ARTICLE IN PRESS
4
J. Kry´sa et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 4. Dependence of initial rate of DCIP photocatalytic reduction and AO7 photo-
catalytic oxidation on the TiO2 film thickness.
Fig. 3. Absorption spectra of DCIP dried ink deposited on 1 layer TiO2 film by dip
coating for various irradiation times. Insert shows the dependence of absorption
in maximum (630 nm) as a function of irradiation time. Incident UV light intensity
7.5 W/m2.
AO7 and reduction rate of DCIP. The data shows that both methods
can be used equally for photoactivity evaluation. The methods
using the DCIP reduction is about 2 orders higher than that of the
AO7 oxidation which means that the method is more suitable for
surfaces with small amount of photocatalyst (thin films, composite
films) and/or for fast photoactivity measurement.
was necessary to determine this parameter (and the corresponding
extinction coefficient) in the layer of ink deposited on the TiO2 film.
For this purpose we deposited a known amount of aqueous ink
of known concentration and used the similar approach as for the
absorbance of DCIP aqueous solution:
3.4. Determination of quantum yield
A = εDCIP,aq · c l
(5)
where c is DCIP concentration (mol dm−3), εDCIP,aq is extinction
coefficient (1.37 × 104 dm3 mol−1 cm−1) and l is optical length
(1 cm). Absorbance of solid ink film with DCIP dye can be expressed
similarly as in aqueous solution using concentration of DCIP in the
film, cDCIP,f (mol cm−3) and thickness of dried film, b (cm) as opti-
cal path. Thickness b is then expressed by means of the volume of
DCIP solid ink film, Vf (cm3) and geometrical surface area of DCIP
solid ink film, S (cm2). Then, the absorbance, A, of ink film can be
expressed as a function of the molar amount of DCIP related to unit
area 1 cm2, nDCIP,f (mol cm−2) and the molar absorptivity of the dye
in solid film, εDCIP,f (cm2 mol−1) using Eq. (6).
Dyes have relatively large molar absorptivities and therefore
measurements of their concentrations in solutions are easy even
if the concentrations involved are very low. On the other side there
are several problems (either general or specific for dyes) which
complicate precise measurement of the efficiency of the photocat-
alytic oxidation or reduction process, i.e. quantum yield.
Quantum yield is defined as a number ratio of reacted or
liberated molecules to absorbed photons. The first problem in
its determination for reactions involving a heterogeneous light-
absorbing species is the difficulty in determining the number
of absorbed photons. Unlike homogeneously dispersed light-
absorbing species, heterogeneous ones scatter and reflect as well
as absorb incident photons and so reduce the light intensity arriv-
ing at a detector in a spectrophotometer. In addition, even optically
transparent (i.e. non-scattering) TiO2 films often show interference
bands, which complicate the calculation of the intensity of absorbed
light by the photocatalytic film. In addition, quantum yield can
depend on the irradiation wavelength and often on the irradia-
tion intensity (since the latter can change the reaction mechanism),
and thus in calculating quantum yield data should be obtained for
different wavelengths and preferably different incident irradiances.
cDCIP,f · Vf
A = εDCIP,f · cDCIP,f b = εDCIP,f
= εDCIP,f nDCIP,f
(6)
S
It can be shown that the typical amount on DCIP in ink film
is 1 × 10−8 mol/cm2 which corresponds to absorbance 0.125. Thus,
from the absorption spectra of dried ink of known molar amount of
DCIP per cm2 the extinction coefficient of DCIP in dried ink can be
calculated applying Eq. (6) as εDCIP,f = 1.25 × 107 cm2 mol−1. Note:
the corresponding absorbance of DCIP aqueous solution in 1 cm
cuvette containing the same molar amount of DCIP as in solid film
is slightly higher 0.137. This difference can be explained by the
fact that in the film, the concentration of the dye is much higher
(since b is ≪1 cm) and some DCIP will form aggregate species,
dimers and trimmers which will have much lower molar absorp-
tivity values than the monomer (in solution). Using the extinction
coefficient for the dye in the dried ink (1.25 × 107 cm2 mol−1) it is
possible to calculate molar amount of dye DCIP related to unit area
and express the photocatalytic reduction rate of DCIP in dried ink
(units: mol cm−2 s−1):
ꢂ
ꢃ
ꢂ
ꢃ
Vf
dADCIP,f
dt
dADCIP,f
dt
1
1
Ri =
·
=
·
(7)
S b εDCIP,f
εDCIP,f
i
i
Since, for an ink film deposited onto the photocatalytic film:
Vf = S·b. The dependence of initial rate of DCIP photocatalytic
reduction as well as initial rate of AO7 photocatalytic oxidation on
layer thickness is shown in Fig. 4. Both rates increase linearly in
the whole range of thickness of the titania films from 0 up to ca.
490 nm. Fig. 5 then shows the relation between oxidation rate of
Fig. 5. Correlation between initial rate of DCIP photocatalytic reduction and AO7
photocatalytic oxidation on the TiO2 films of thickness 70–490 nm.
Please cite this article in press as: J. Kry´sa, et al., Quantum yield measurements for the photocatalytic oxidation of Acid Orange
7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.04.019

G Model
CATTOD-9047; No. of Pages6
ARTICLE IN PRESS
J. Kry´sa et al. / Catalysis Today xxx (2014) xxx–xxx
5
Fig. 8. Degradation rate of Acid Orange 7 (blue, left axis) and 2,6-dichlorindophenol
(red, right axis) as a function of absorbed UV light in the range from 340 to 400 nm.
Fig. 6. Emission spectra of UV light source used for irradiation tests and true UV-
absorption spectra of a 3 layer TiO2 film (thickness 210 nm).
Note that in Fig. 7 the photon flux absorbed by TiO2 films as a func-
tion of wavelength has rather sharp maximum at 365 nm which is
a result of the combination of the UV absorption spectra of the TiO2
films and the emission spectra of UVA lamp.
It is generally suggested that optical interference on thin films
prevents a sufficiently precise determination of the light intensity
absorbed by photocatalyst films. However, complementary mea-
surements of both transmission and reflection spectra, using an
integrating sphere, allows a good estimation of true absorption dif-
ferently thick films and so helps to largely overcome this problem.
Fig. 6 then shows the emission spectra of the BLB UVA light used
for irradiation tests and the true absorption spectra of 3 layer TiO2
film(thickness210 nm);Ii, is theincidentlightintensityin therange
340–400 nm to TiO2 surface.
In the case of AO7 photocatalytic oxidation the incident light
intensity is 1.2 mW/cm2. In the case of DCIP ink reduction the inci-
dent light intensity is 0.75 mW/cm2. Assuming the incident light
intensity is proportional to the area under the relative intensity
distribution of the UVA lamp shown in Fig. 6 it is possible to cal-
culate the incident flux of photons per cm2 per nm as a function
of wavelength (shown for incident light intensity 1.2 mW/cm2 as
black line in Fig. 7). The overall incident photon flux (obtained by
an integration of curve in Fig. 7) is 3.74 × 10−5 mol/m2/s. Using this
data set it is possible to calculate for any of the titania films of differ-
ent thicknesses the portion of light absorbed by each TiO2 film and
correspondingly plot the absorbed photons as a function of wave-
length for each of the different titania layer thicknesses. In the case
of incident light intensity 1.2 mW/cm2 these plots for the different
titania films are shown as coloured curves in Fig. 7. Table 1 then
gives the calculated absorbed photon flux which were obtained by
an integration of curves for individual TiO2 films shown in Fig. 7.
Another problem in the accurate calculation of photocatalytic
yield in heterogeneous photocatalytic systems that uses dye
bleaching as the method of assessing activity, is that the dye
molecules absorb photons, especially in the visible light range, and
thus photoexcited electrons from the dye may be injected into pho-
tocatalyst particles and lead to dye bleaching. This method of dye
bleaching can be shown to occur from a brief examination of the
action spectrum, since under the circumstances it will be similar to
the photoabsorption spectrum of the dye. As a consequence, many
have suggested that dye bleaching should be excluded from the list
of possible ways to evaluate the activity of a visible light absorbing
photocatalyst, although often such tests can be sued for assessing
the activities of UV absorbing photocatalyst, since often the dyes
used show little absorbance in the UVA region.
The data in Table 1 coupled with the initial rate data, for the
different titania films of different thickness allows plots of rate vs
absorbed light flux to be calculated for the AO7 and DCIP test system
and apparent quantum yields of 0.19% and 92% to be calculated (see
Fig. 8). It is perhaps not too surprising to note the very low quan-
tum yield value for the photobleaching of AO7, since the overall
mineralisation process is:
C16H11N2NaO4S + 21O2 → 16CO2 + 2HNO3 + NaHSO4 + 4H2O
(8)
and requires the transfer of 84 electrons from the AO7. However,
the initial photobleaching process probably involves a much lower
number of photogenerated holes; indeed, it may be as low as two
if it is simply the reduction of the azo group. This would imply a
quantum yield of ca. 50% is possible assuming no other processes
compete, such as electron–hole recombination. The reduction of
oxygen and oxidation of AO7 are electrochemically slow irre-
versible processes and it is for this reason the quantum yield is
ꢀ50%, i.e. ca. 0.19%.
In the case of the photocatalysed reduction of DCIP, the situ-
ation is very different because there is a vast excess of glycerol
present in the ink to trap the photogenerated holes efficiently and
so allow time for the slower reduction of dye to take place, see Eqs.
(1) and (2). However, the latter eqn. suggests that for the reduction
of 1 molecule of DCIP only two photons are required and, there-
fore, that the maximum quantum yield will be 50%, and not the
value of 92% calculated. However, this, almost double that expected,
value of quantum yield may be rationalised upon consideration
of the initial oxidation of the glycerol, which will generate an
additional, intermediate, highly reducing species, a R-hydroxyalkyl
Fig. 7. Incident photons (dotted line) and photons absorbed by TiO2 film consisting
of 1–7 layers (solid lines)
Please cite this article in press as: J. Kry´sa, et al., Quantum yield measurements for the photocatalytic oxidation of Acid Orange
7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.04.019

G Model
CATTOD-9047; No. of Pages6
ARTICLE IN PRESS
6
J. Kry´sa et al. / Catalysis Today xxx (2014) xxx–xxx
Table 1
Incident light and absorbed light obtained by integration of curves shown in Fig. 7.
No of layers
AO7 oxidation, Ii = 12 W/m2
DCIP reduction, Ii = 7.5 W/m2
Incident light (mol/m2/s)
Absorbed light (mol/m2/s)
Incident light (mol/m2/s)
Absorbed light (mol/m2/s)
1
2
3
5
7
3.74 × 10−5
3.74 × 10−5
3.74 × 10−5
3.74 × 10−5
3.74 × 10−5
1.62 × 10−6
3.19 × 10−6
4.36 × 10−6
6.75 × 10−6
8.22 × 10−6
2.32 × 10−5
2.32 × 10−5
2.32 × 10−5
2.32 × 10−5
2.32 × 10−5
1.00 × 10−6
1.98 × 10−6
2.70 × 10−6
4.18 × 10−6
5.09 × 10−6
radical, that is known, from work with other alcohols, to be able
to inject an electron into the conduction band of TiO2 and effect
the well-known “current-doubling” effect, or reduce susceptible
chemical species, such as DCIP, nearby [24–26]. As a result ink dye
molecule, DCIP, is reduced not only by the photogenerated elec-
trons but also by an intermediate hydroxyalkyl radical generated
in the oxidation of glycerol by a photogenerated hole. This cur-
rent doubling process thus results in a quantum yield of 92% for
the photocatalysed reduction of DCIP and concomitant oxidation
of glycerol.
Acknowledgements
Authors acknowledge the financial support by the EU 7th Frame-
work Programme for Research (FP7-NMP-2012-CSA-6, GA No.
319210, project INTEC), by the Ministry of Education, Youth and
Sport of the Czech Republic (specific university research, MSMT No.
20/2013) and by the Grant Agency of the Czech Republic (project
No. P108/12/2104).
References
[1] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995)
69–96.
4. Conclusions
[2] J.M. Herrmann, Top. Catal. 34 (2005) 49–65.
[3] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C 1 (2000) 1–21.
[4] A.L. Linsebigler, G.Q. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735–758.
[5] J. Ryu, W. Choi, Environ. Sci. Technol. 42 (2008) 294–300.
[6] A. Mills, M. McFarlane, Catal. Today 129 (2007) 22–28.
[7] F. Kiriakidou, D.I. Kondarides, X.E. Verykios, Catal. Today 54 (1999) 119–130.
[8] K.V. Subba Rao, B. Lavedrine, P. Boule, J. Environ. Photochem. Photobiol. A Chem.
154 (2003) 189–193.
[9] G.N. Kryukova, G.A. Zenkovets, A.A. Shutilov, M. Wilde, K. Günther, D. Fassler,
K. Richter, Appl. Catal. B – Environ. 71 (2007) 169–176.
[10] M. Zlámal, J. Kry´sa, J. Jirkovsky´, Catal. Lett. 133 (2009) 160–166.
[11] C. O’Rourke, A. Mills, J. Photochem. Photobiol. A 216 (2010) 261–267.
[12] M. Baudys, M. Zlamal, J. Krysa, J. Jirkovsky, P. Kluson, React. Kinet. Mech. Catal.
106 (2012) 297–311.
Transparent titania films prepared by dip-coating sol gel pro-
cess are in the anatase crystalline phase with a crystallite size ca.
13 nm. The thickness of TiO2 films depends linearly on the number
of deposited TiO2 layers (up to 7 dipcoating, drying and calcina-
tion cycles), with the thickness of a single layer of titania equal
to 70 5 nm. True absorption spectra combined with the incident
light intensity and the emission spectrum of the UV lamp, allow
the calculation of the number of photons absorbed by each titania
film.
The initial rate of oxidative bleaching of AO7 and reductive
bleaching of DCIP increases linearly with increasing thickness as
well as with increasing absorbed light flux. The quantum yields
(QY) for the AO7 and DCIP test systems were calculated as: 0.19%
and 92%, respectively. The very low QY for the AO7 oxidation is due
to the combination of a slow irreversible reduction of oxygen and
oxidation of AO7, which results in a highly efficient electron–hole
recombination process. The very high QY for the reduction of DCIP
is due to the vast excess of glycerol present in the ink which results
in the efficient trapping of the photogenerated holes and conse-
quently highly effective reduction of DCIP by the photogenerated
electrons. The observed QY value of 92% for photocatalysed reduc-
tion DCIP is significantly higher than theoretical maximum QY
(50%). This can be explained by the generation of highly reduc-
ing R-hydroxyalkyl radicals (due to oxidation of glycerol) that are
able to either inject an electron into the conduction band of TiO2 or
reduce directly DCIP.
[13] A. Mills, J.S. Wang, M. McGrady, J. Phys. Chem. B 110 (2006) 18324–18331.
[14] A. Mills, M. McGrady, J. Photochem. Photobiol. A 193 (2008) 228–236.
[15] A. Mills, M. McGrady, J. Wang, J. Hepburn, Int. J. Photoenergy (2008), article ID
504945, 6 pp.
[16] A. Mills, A. Cusick, J. Hepburn, J. Adv. Oxid. Technol. 12 (2009) 152–157.
[17] A. Mills, J. Hepburn, M. McFarlane, ACS Appl. Mater. Int. 1 (2009) 1163–1165.
[18] J. Zita, J. Krysa, A. Mills, J. Photochem. Photobiol. A – Chem. 203 (2009) 119–124.
[19] J. Krysa, S. Pausova, M. Zlamal, A. Mills, J. Photochem. Photobiol. A 250 (2012)
66–71.
[20] W. Lojander, Suomen Kemistilehti 14A (1941) 26.
[21] L. Erdey, G. Svehla, L. Koltai, Anal. Chim. Acta 27 (1962) 164–168.
[22] S.L. Phillips, D.L. Perry (Eds.), Handbook of Inorganic Compounds, CRC Press,
Boca Raton, FL, 1995.
[23] J. Zita, J. Krysa, U. Cernigoj, U. Lavrencic-Stangar, J. Jirkovsky, J. Rathousky, Catal.
Today 161 (2011) 29–34.
[24] N.W. Hykaway, N. Sears, H.S. Morisaki, R. Morrison, J. Phys. Chem. 90 (1986)
6663–6669.
[25] T.A. Madelbaum, A.E. Regazzoni, M.A. Blesa, S.A. Bilmes, J. Phys. Chem. B 103
(1999) 5505–5511.
[26] C.Y. Wang, J. Rabani, D.W. Bahnemann, J.K. Dohrmann, J. Photochem. Photobiol.
A: Chem. 148 (2002) 169–176.
Please cite this article in press as: J. Kry´sa, et al., Quantum yield measurements for the photocatalytic oxidation of Acid Orange
7 (AO7) and reduction of 2,6-dichlorindophenol (DCIP) on transparent TiO2 films of various thickness, Catal. Today (2014),
http://dx.doi.org/10.1016/j.cattod.2014.04.019