Hydroxyl radical and free and shallowly trapped electron generation and electron/hole recombination rates in TiO2 photocatalysis using different combinations of anatase and rutile

Article abstract of DOI:10.1016/j.apcata.2018.07.045

Hydroxyl radical (OH[rad]) and electron (e^?) generation in aqueous suspensions during TiO₂ photocatalysis was studied, varying the properties of the photocatalyst and using the molecular probes terephthalic acid (TA) for OH[rad] and resazurin dye (Rz) for electrons. Overall photocatalytic activity was calculated from rhodamine B dye degradation. The percentage of e^? recombination was found as the difference between the results for aqueous Rz in the presence and absence of a hole scavenger (glycerol). The findings show that the properties of the TiO₂ particles affect e^? and OH[rad] generation rates differently. The generation rates of these active species and their overall photocatalytic activity were observed to be heavily dependent upon the physical properties of the TiO₂ used and the recombination percentage.

Full text of DOI:10.1016/j.apcata.2018.07.045

Applied Catalysis A, General 565 (2018) 20–25
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Hydroxyl radical and free and shallowly trapped electron generation and
electron/hole recombination rates in TiO photocatalysis using different
2
combinations of anatase and rutile
⁎
⁎
Eva Jimenez-Relinque , Marta Castellote
Eduardo Torroja Institute for Construction Science (IETcc), Serrano Galvache, 4, Madrid, Spain
A R T I C L E I N F O
A B S T R A C T
−
Keywords:
Hydroxyl radical (OH%) and electron (e ) generation in aqueous suspensions during TiO
studied, varying the properties of the photocatalyst and using the molecular probes terephthalic acid (TA) for OH
and resazurin dye (Rz) for electrons. Overall photocatalytic activity was calculated from rhodamine B dye
2
photocatalysis was
TiO
2
Properties
Hydroxyl radical
Electron
Recombination
Photocatalytic activity
%
−
degradation. The percentage of e recombination was found as the difference between the results for aqueous Rz
in the presence and absence of a hole scavenger (glycerol). The findings show that the properties of the TiO
particles affect e and OH% generation rates differently. The generation rates of these active species and their
overall photocatalytic activity were observed to be heavily dependent upon the physical properties of the TiO
used and the recombination percentage.
2
−
2
−
+
1. Introduction
e /h recombination on reactions photocatalyzed by TiO
2
with dif-
ferent properties. Overall photocatalytic activity is measured as the
rhodamine B (RhB) photodecomposition rate constant, for although
part of the dye may be sensitized under visible light, RhB is often used
to trace photocatalytic activity in construction materials (UNI
11259:2008). The relationships between the generation rates of the
active species and their effect on overall photocatalytic performance are
discussed.
2
The mechanism governing TiO -mediated photocatalysis has been
−
+
widely studied [1,2]. The free carriers (e and h ) generated have
been shown to participate in the photocatalytic degradation of ad-
sorbed pollutants, directly or via the formation of reactive oxygen
species (ROS) [3,4]. The identification of such species (charge carriers
and ROS) and their generation and recombination rates are conse-
quently of scientific and technological significance, for they may con-
tribute to the design of new materials with practical applications.
This study analyzes the oxidative reaction induced by photo-
2. Materials and methods
−
generated OH% and the reductive potential of photogenerated e . The
2
2.1. TiO powder
−
2
rates of e and OH% photocatalytic generation from TiO powder ob-
tained by calcining commercial anatase were respectively quantified
with terephthalic acid (TA) [3–5] and resazurin dye (Rz) [6–9]. Despite
the many reports on the use of Rz dye as an indicator of photocatalytic
activity, its utility in that respect has been confined to research on the
rate of change in properties such as the absorbance/color associated
with photocatalytic reactions Inasmuch as Rz can be specifically re-
duced by electrons, it may be deemed to be a suitable method for
Commercial titanium dioxide powder was used as supplied
(Hombikat UV100 TiO ; TiO -C). In addition, a number of blends of
anatase/rutile TiO were obtained by heating a crucible containing 1 g
of powder in a muffle furnace for 4 h at temperatures ranging from
400 °C to 1000 °C.
2
2
2
2
2.2. TiO characterization
−
quantifying the rate of e formation in photocatalysis. To the authors’
−
knowledge, this is the first study to quantify the e generation rate by
TiO
conducted on a Bruker Advance-D8 diffractometer. The specific surface
of the powder was found with the Brunauer-Emmett-Teller (SBET
method using Micromeritics NOVA2000 nitrogen absorption
2
crystallography was determined by X-ray diffraction (XRD)
measuring the decline in Rz concentration with UV–vis spectro-
photometry. The use of data gathered with this dye to calculate the
electronic recombination rate also provides insight into the effect of
)
a
⁎
Corresponding authors.
E-mail addresses: eva.jimenez@csic.es (E. Jimenez-Relinque), martaca@ietcc.csic.es (M. Castellote).
https://doi.org/10.1016/j.apcata.2018.07.045
Received 4 April 2018; Received in revised form 30 July 2018; Accepted 31 July 2018
Available online 01 August 2018
0926-860X/ © 2018 Elsevier B.V. All rights reserved.

E. Jimenez-Relinque, M. Castellote
Applied Catalysis A, General 565 (2018) 20–25
analyzer. The optical properties of the compound were found with
diffuse reflectance spectroscopy (DRS) conducted on a Shimadzu UV-
2
2
2
600 UV–vis spectrophotometer.
.3. Photocatalytic experiments
.3.1. Hydroxyl radical (OH%) quantification: terephthalic acid (TA)
fluorescence (FL)
The amount of OH% forming at the illuminated photocatalyst/water
interface was quantified with terephthalic acid fluorescence (TA-FL)
[
3,5]. Fluorescence was applied to detect the OH% forming on the ir-
radiated sample with highly reactive TA as the chemical probe to
speedily generate fluorescent 2-hydroxyterephthalic acid (TAOH) [5].
Earlier studies showed that this method reliably quantifies free OH%
+
Fig. 1. Light spectrum for Actinic BL 15 W/10 SLV lamps.
generation based on the h photogenerated in photocatalytic materials
[
3,4]. OH% formation was determined in a suspension of 0.0055 g of
TiO powder in 100 mL of a 2 mM TA aqueous solution containing
mM NaOH in an aerated Pyrex glass cell, experimental conditions
under which TA hydroxylation has been shown to be attributable ex-
clusively to OH% [3].
The suspension was stirred in a dark box until photocatalyst-TA
adsorption–desorption equilibrium was reached, after which stirring
continued and the powder sample was irradiated uninterruptedly for
2
2
fluorescence and Rz absorbance, dissolving 0.05 g of TiO powder in an
2
aqueous solution containing 8 ppm of RhB. The solution was stored in
the dark and stirred to saturate RhB adsorption prior to UV irradiation.
The change in absorbance recorded by a spectrophotometer during ir-
radiation at a wavelength of 550 nm was used to determine RhB con-
centration in the solution. As the relationship between the logarithm of
0
the relative RhB concentration in the solution (ln(C/C )) and irradia-
2
0 min. The solution was subsequently centrifuged to remove the
tion time was approximately linear, the slope of the expression, i.e., the
rate constant (kRhB), was determined for each sample and used as the
representative parameter for photocatalytic behavior.
photocatalyst and FL intensity in the supernatant solution was mea-
sured at 425 nm after excitation at 315 nm. The amount of TAOH
generated, determined from a calibration curve, was used to estimate
the amount of free OH% formed, assuming a trapping factor of 35%
2.3.4. Irradiation set up
[
3,10].
The light source consisted of two Philips Actinic BL 15 W/10 SLV
fluorescent lamps emitting at 350 nm–400 nm (optimally at 360 nm),
with four additional narrow peaks at 406 nm, 435 nm, 545 nm and
−
2
.3.2. Electron (e ) reduction: resazurin (Rz) absorbance
−
The e formation rate at the illuminated photocatalyst/water in-
terface was detected by UV–vis absorbance using Rz as the probe mo-
lecule. Resazurin (Rz, blue) readily reacts with e to yield resorufin (Rf,
576 nm. The light spectrum for the lamps is reproduced in Fig. 1. The
mean light intensity at the surface of the reaction solution was about
−
2
1
0 W/m .
pink) (Eq. (1)).
−
Resazurin(Rz) + 2e → Resorufin(Rf )
(1)
3
. Results and discussion
Rz absorbance at 600 nm was measured for several standard con-
centrations and the absorbance – concentration values plotted to obtain
a calibration curve. The concentration of the e generated in the
3.1. Characterization of TiO
2
samples
−
photocatalytic reaction studied was found by extrapolating absorbance
The XRD- and BET-determined physical properties of the TiO
2
at 600 nm of the irradiated TiO
2
powder solution on the x-axis (Rz
powder used are summarized in Table 1. Anatase was stable up to a
temperature of 800 °C, with small amounts (1%) of rutile first appearing
in the 900 °C samples and rising to 51% in the 950 °C and to 71% in the
960 °C samples. Anatase disappeared altogether at 1000 °C. Entering the
anatase (101) and rutile (110) reflection broadening values into the
Scherrer equation [14] showed that the anatase crystallites grew with
concentration). In addition to the aforementioned reaction mechanism
−
(
4
Rz+2e →Rf), electron calculations envisaged a trapping factor of
0% to account for the direct or indirect competition between O and
2
−
Rz for the photogenerated e [11,12].
Two ink formulations were used: (i) 3 g of glycerol and 0.04 g of
resazurin dye in 1 L of aqueous solution as described by Mills et al. [13];
and (ii) the same formulation in the absence of glycerol. The glycerol-
containing system is known as “intelligent ink” [7] for electron-hole
Table 1
2
Physical-chemical and optical properties of TiO samples (A: anatase; R: rutile).
−
+
(
e -h ) recombination is controlled by the hole scavenger (glycerol),
Crystallite size FWHM (○)a
(nm)
Sample
Phase
BET
surface
Band
gap
(eV)
a
affording the photogenerated electrons more time reduce the dye. The
degree of electron recombination was calculated by comparing the
values for the two formulations.
percentage
a
(%)
area
2
(
m /g)
A
R
A
R
A
R
The experimental procedure involved magnetically stirring 0.0055 g
of TiO
2
powder in 50 mL of Rz solution (with or without glycerol) in a
TiO
TiO
TiO
TiO
TiO
2
2
2
2
2
-C
100
100
100
100
100
100
99.1
44.9
28.5
–
–
–
–
–
–
–
1
51.1
71.5
100
8
–
–
–
–
–
–
163
138
136
137
0.99
0.65
0.51
0.37
0.26
0.18
0.17
0.18
0.21
–
–
–
–
–
–
–
310
109
79
49
27
3.32
3.30
3.29
3.26
3.25
3.22
3.22
2.94
2.91
2.99
-400
-500
-600
-700
11
16
32
49
69
75
90
91
–
Pyrex vial stored in the dark until photocatalyst-Rz ad-
sorption–desorption equilibrium was reached. Stirring then continued
as the suspensions were uninterruptedly irradiated. The TiO powder
2
was removed by centrifugation at different irradiation times and the
degree of reduction measured in the supernatant solution. UV–vis ab-
sorption spectra were collected on a Shimadzu UV-2600 spectro-
photometer.
TiO -800
2
13
TiO
TiO
TiO
TiO
2
2
2
2
-900
-950
-960
-1000
0.074
0.075
0.076
0.075
9.5
3.4
3.2
2.8
a
2.3.3. Photocatalytic degradation of rhodamine B
Estimated from the broadening of anatase (101) and rutile (110) XRD re-
RhB discoloration was conducted under the same conditions as TA
flections.
21

E. Jimenez-Relinque, M. Castellote
Applied Catalysis A, General 565 (2018) 20–25
1
/2
Fig. 2. Tauc plot: (FR·hν)
[(absorption coefficient × Planck’s constant x
1
/2
frequency) ] vs. incident photon energy (eV).
rising temperature from 8 nm to 91 nm at 960 °C, the last at which they
were detected. In contrast, rutile crystallinity and crystal size exhibited
no particular temperature-dependent pattern. Sample surface area de-
clined with rising calcination temperatures.
The diffuse reflectance spectroscopic (DRS) values for TiO nano-
2
particles are represented in Fig. 2, where the Tauc equation was used to
plot the Kubelka-Munk optical absorption coefficients [F(R)]. The
findings show a progressive red shift in the absorption edges of the TiO
2
samples with rising temperature. The slight changes observed in the
absorption edges at temperatures of under 900 °C became more per-
ceptible at 950 °C to 1000 °C due to the gradual transition from anatase
to rutile. The progressive decline observed in UV light absorption re-
gion (> 4 eV), in turn, could be attributed to the greater light scattering
characterizing larger crystallites [15]. These DRS findings were con-
sistent with the XRD results discussed above.
Fig. 3. (a) FL intensity vs. time (mean of two measurements); (b) OH% forma-
tion rate (ROH·) by photocatalyst.
The TiO
2
powder band gap (Eg) was found by using the Tauc
equation [16] to plot the Kubelka − Munk optical absorption coeffi-
1
/b
Abs600 values on the calibration curve, assuming the reaction me-
chanism described in Eq. (1) and a trapping factor of 40% (Section
cients [F(R)] (Fig. 2). (FRhν) (b = 2 for indirect band gaps such as in
TiO ) vs. hν curves are linear in the valence to conduction band tran-
2
2
.3.2.)
The linearity of Rz Abs600 – irradiation time plots for all the TiO
suspensions is clearly visible in Fig. 4(b) (with glycerol) and (c)
without glycerol). At irradiation times of over 10 min decline was non-
sition region (VB-CB) (Fig. 2). The Eg value was found by linear ex-
trapolation to the x-axis. Table 1 also gives the band gap values for the
samples.
2
(
linear in the most active TiO−
2
suspensions, a finding attributed to Rz
3
2
.2. Rate of OH% formation in irradiated TiO photocatalysts
depletion. In this study, the e formation rates calculated from the data
in Fig. 4(b) and (c) for the two Rz formulations are shown in Fig. 4(d),
which also includes the recombination percentages found with Eq. (2):
FL intensity was observed to be linearly related to the duration of
UV irradiation in the TiO suspensions (Fig. 3(a)). TAOH concentration
2
−
was quantified from the calibration curve obtained by plotting FL in-
tensity at 425 nm against previously reported standard TAOH con-
centrations (FL(a.u.) = 143.66·[TAOH](μmol)) [3] assuming a trapping
factor of 35% [3,10]. The OH% formation rate (Fig. 3(b)), ROH% (μmol/
min), was calculated from the slope of the FL intensity – irradiation
time curve.
e
recombination (%) = [(Re−g−Re−)/Re−g] × 100
(2)
−
−
−
where erecombination is the e recombination percentage; Re-g is the e
formation rate using Rz with, and Re- the rate using Rz without glycerol.
−
The differences in the e formation rate with and without the
scavenger (glycerol) were observed to vary widely depending on the
temperature, as discussed in detail in Section 3.5.
ROH% rose with temperature up to values of 700 °C–800 °C, began to
decline with the appearance of the rutile phase at 800 °C and sharply
2
thereafter. The effect of the intrinsic properties of TiO samples on OH%
formation is discussed in detail in Section 3.5.
3.4. Rhodamine B (RhB) test
The fluctuations in relative RhB concentration in water versus ir-
−
3
.3. Rate of e formation in irradiated TiO
2
photocatalysts
radiation time are plotted for all the samples in Fig. 5(a). The RhB
−
1
degradation rate (KRhB, min ) determined from the slopes of the lines
in Fig. 5(a) is graphed in Fig. 5(b). The higher the calcination tem-
perature, the higher was rate constant KRhB up to around 700 °C to
800 °C, when it began to decline with the appearance of rutile, i.e.,
exhibiting the same pattern as observed for OH% formation (ROH%),
discussed in Section 3.2. The effect of the intrinsic properties of cal-
As the calibration graph in Fig. 4(a) shows, Rz concentration is
linearly related to its absorbance at 600 nm (Abs600). Given that Rz
Abs600 in the TiO suspensions declined at longer irradiation times,
likewise linearly (Fig. 4(b) and (c)), e formation at the TiO
could be inferred to be proportional to irradiation time. The formation
rate (Re−, μmol/min) was consequently determined from the slope of
the Abs600- irradiation time curve by extrapolating the experimental
2
−
2
surface
cined TiO
2
samples on overall photocatalysis is discussed in detail in
Section 3.5.
22

E. Jimenez-Relinque, M. Castellote
Applied Catalysis A, General 565 (2018) 20–25
Fig. 4. (a) Regression line for Rz absorbance; (b) absorbance (Abs600) in Rz samples containing glycerol; (c) absorbance in Rz samples without glycerol; and (d)
−
logarithm of the e formation rate (Re-) The percentages of e- recombination are also included.
3.5. Factors affecting photocatalytic generation of active species and overall
effect on overall photocatalytic activity. Nonetheless, the highest re-
photocatalytic activity
combination was observed for the pure anatase samples, while kRhB and
the e -h recombination rate were inversely and linearly related, with
the former declining when the latter rose.
−
+
Variations in the calcination temperature had a significant effect on
the intrinsic properties of the photocatalyst, such as crystal size, type
and content of phases present and surface area, all of which impacted
its activity substantially, although the respective relationships were not
straightforward. An earlier study of photocatalytic activity in 35 titania
samples with different physical properties found the dependence of
activity on each property to be ambiguous [17].
Fig. 6 shows that in the anatase samples, larger crystal size in
conjunction with a smaller surface area induced by rising calcination
temperatures resulted in higher ROH% and KRhB. As smaller particle size
and larger surface area generally translate into higher catalyst effi-
ciency due to the larger number of active sites per square meter, re-
activity would have been expected to be lower under the conditions
prevailing here [20]. Furthermore, ROH% and KRhB photoactivity must
also be related to the percentage of surface recombination taking place.
The graph in Fig. 6 depicts the RhB photocatalytic degradation rate
(KRhB), the OH% formation rate (ROH%), the e (Re−) formation rate
−
−
+
before and after recombination and the surface recombination percen-
In nanocrystalline semiconductor particles, e -h recombination is a
more significant factor at smaller particle sizes, due to the rise in re-
combination site density [21–23]. The findings described here therefore
suggest that recombination prevailed over crystallographic and micro-
structural parameters in the anatase samples. In keeping with that
−
tage (erecombination)values found in this study versus calcination tem-
perature, BET area, phase content and anatase crystal size.
Further to the figure, KRhB and ROH% followed the same pattern and
each behaved fairly similarly in all the samples irrespective of their
differing physical properties. Those data were consistent with the de-
gradation pathway described in earlier studies, according to which RhB
photocatalytic oxidation under UV light may proceed either by OH% or
OOH% attack (or both) [18,19]. The high correlation between these two
parameters also suggested that RhB bleaching was related to the pho-
tocatalytic generation of this active species by the photocatalyst tested,
for while dye sensitization cannot be ruled out completely, it may be
−
+
reasoning, e- formation with controlled e -h recombination (Fig. 6,
e- with glycerol) was higher at smaller particle sizes due to higher
R
photonic efficiency.
In the rutile-containing (alone or in combination with anatase)
samples, ROH% and KRhB were observed to decline with rising calcina-
tion temperature. That behavior appeared to be related to the increase
in rutile phase content. Kim et al. suggested that as rutile adsorbs OH%
−
deemed negligible in the cases studied here. In contrast, the e for-
more effectively than anatase, the amount of OH% generated on TiO
2
mation rate (Re-) detected by the molecular probes with and without
glycerol did not follow the same pattern as KRhB/ROH%. The surface
recombination percentage was much lower in the rutile- than in the
anatase-containing samples and taken separately appeared to have no
and diffusing to the bulk solution would be lower in rutile crystallites
[24]. That difference in OH% adsorption might explain the present
findings, whereby ROH%, as found with the TA-FL method, was higher in
anatase than rutile. Other authors reported that although more OH%
23

E. Jimenez-Relinque, M. Castellote
Applied Catalysis A, General 565 (2018) 20–25
That argument is consistent with the present findings.
These results conflict, however, with studies in which synergies
were observed in the photoactivity induced by thoroughly blended
anatase–rutile powders. The theoretical model to explain enhanced
photocatalysis with such a combination is based on charge separation
[
29–32]. Assuming the presence of well-defined rutile and anatase
particles in the TiO calcined samples analyzed, electronic contact be-
2
−
+
tween the two phases would efficiently separate the e h pair, low-
ering carrier recombination. According to the findings of this study,
however, separation efficiency expressed as surface recombination (%)
did not enhance the photocatalytic efficiency parameters expressed as
K
RhB photodegradation or ROH%. Sun et al. [33] showed that the bene-
ficial synergies of anatase-rutile blends are not observed consistently,
but vary with the relative anatase and rutile Fermi levels and particle
shape. The findings discussed hereunder are consistent with the model
set out in [23,28,34]. As valence band edge energy is higher in rutile
than in anatase [23,28,34,35], when charge carriers are generated by
UV excitation in the latter, h+ migrates toward the rutile phase. The
−
outcome is longer conduction band e lifetime in anatase as well as a
decline in the effective h+ concentration on its surface [23]. As dis-
cussed earlier, OH% formation via water oxidation is far less efficient in
rutile than in anatase particles. Therefore, although the presence of
rutile may render charge carrier separation more efficient, neither the
ROH% in aqueous photocatalytic reactions nor KRhB photodegradation
would be favored.
Moreover, while the surface recombination values were low in ru-
tile-containing samples, a steep decline in Re- (with and without gly-
cerol) was also observed with rising calcination temperature due to the
increase in rutile phase content. Such differences in behavior between
the two phases may be the result of the proportion of free or shallowly
trapped e-, which have a longer lifetime and are highly reactive. As the
2
proportion is higher in anatase TiO , photocatalytic reduction capacity
is greater in it than in rutile,where the deeply trapped free electrons are
less reactive and reduction activity lower [36]. That reasoning is con-
sistent with the present experimental results, in which the proportion of
photogenerated electrons declined with rising rutile content. The be-
havior of the three parameters analyzed (ROH%, KRhB and Re-) in rutile-
containing samples can be said to be induced by crystallographic factors
such as rutile phase content, and exhibit lower sensitivity to re-
combination.
Fig. 5. (a) RhB photodegradation and (b) pseudo first order kinetic rate con-
stant (KRhB) for RhB degradation in the presence of TiO
two measurements).
2
samples (average of
In follow-up research these results are being compared to electro-
chemical findings to validate the above hypothesis.
4. Conclusions
−
The hydroxyl radicals (OH%) and electrons (e ) forming on photo-
irradiated TiO powders were respectively quantified with terephthalic
2
acid fluorescence (TA-FL) and resazurin (Rz). The generation of these
two active species and their effect on the overall activity expressed as
the rate of oxidative degradation of rhodamine B were studied in con-
nection with the physical-chemical properties of TiO
showed that crystallite size and phase content play an important role in
nanocrystalline TiO -based photocatalysis primarily by affecting pho-
2
. The findings
2
−
+
tonic efficiency and e -h recombination dynamics. In anatase sam-
ples, recombination seems to prevail over the photonic efficiency gov-
erned by crystallographic and microstructural factors. In contrast, in
mixed anatase-rutile samples, crystallography and microstructure ap-
pear to predominate.
−
Fig. 6. OH% formation rate (ROH%, μmol/min), e formation rate with and
without glycerol (Re-, μmol/min), e- recombination (%) and RhB photo-
−
1
degradation (KRhB, min ) vs. physical-chemical parameters.
was found near the surface than in the bulk solution for the two types of
crystals, both surface and bulk solution OH% yields were extremely
small in rutile. Those data would support scant photocatalytic-mediated
OH% formation in rutile [23–26]. The possible explanation would be
that on the anatase surface, water oxidation would convert trapped h+
to adsorbed OH%, generating significant amounts of OH%. In contrast,
−
The decline in OH% and e generation rates with the appearance of
rutile is the outcome of inefficient OH% generation during water oxi-
dation on the rutile particles, and the shorter lifetime and lower re-
activity of the free e- in rutile samples. These findings may suggest an
answer to the long-unresolved question of why anatase is more active
than rutile in many photocatalytic reactions. This study provides fur-
ther support for the premise that the photocatalytic synergies attributed
to anatase-rutile blends are not invariably present, but depend on the
on the rutile surface water would be oxidized to H
to O , generating a certain amount of OH% in the latter process [25–28].
2 2
O and subsequently
2
24

E. Jimenez-Relinque, M. Castellote
Applied Catalysis A, General 565 (2018) 20–25
relationship between anatase and rutile particle Fermi levels as well as
particle characteristics. In addition to contributing to an understanding
of TiO photocatalytic mechanisms, these findings may prove useful for
2
[14] P. Scherrer, Nach Ges Wiss Gottingen (1918) 96–100.
[
15] J.-G. Yu, H.-G. Yu, B. Cheng, X.-J. Zhao, J.C. Yu, W.-K. Ho, J. Phys. Chem. B 107
2003) 13871–13879.
16] E. Jimenez-Relinque, I. Llorente, M. Castellote, Catal. Today (2016).
(
[
the development of highly active TiO
2
photocatalysts.
[17] O.-O. Prieto-Mahaney, N. Murakami, R. Abe, B. Ohtani, Chem. Lett. 38 (2009)
38–239.
18] A. Folli, U.H. Jakobsen, G.L. Guerrini, D.E. Macphee, J. Adv. Oxid. Technol. 12
2009) 1.
2
[
[
Acknowledgments
(
19] T. Wu, G. Liu, J. Zhao, H. Hidaka, N. Serpone, J. Phys. Chem. B 102 (1998)
This study was funded by the LIFE Photoscaling European project
LIFE-PHOTOSCALING 13/ENV/ES/001221) and benefitted from a pre-
doctoral fellowship awarded by the Spanish National Research
Council’s Post-graduate Studies Board. The experimental assistance
provided by Miriam Bravo Pérez at the Eduardo Torroja Institute for
Construction Science is gratefully acknowledged.
5845–5851.
[
[
[
20] A. Ibhadon, P. Fitzpatrick, Catalysts 3 (2013) 189.
21] H. Lin, C. Huang, W. Li, C. Ni, S.I. Shah, Y.-H. Tseng, Appl. Catal. B 68 (2006) 1–11.
22] Z. Zhang, C.-C. Wang, R. Zakaria, J.Y. Ying, J. Phys. Chem. B 102 (1998)
10871–10878.
(
[
23] J.T. Carneiro, T.J. Savenije, J.A. Moulijn, G. Mul, J. Phys. Chem. C 115 (2011)
2
211–2217.
24] T. Kim, M. Lee, S. Lee, Y. Park, C. Jung, J.-H. Boo, Thin Solid Films 475 (2005)
71–177.
[
1
[25] T. Hirakawa, K. Yawata, Y. Nosaka, Appl. Catal. A Gen. 325 (2007) 105–111.
[26] J. Zhang, Y. Nosaka, J. Phys. Chem. C 118 (2014) 10824–10832.
[27] R. Nakamura, Y. Nakato, J. Am. Chem. Soc. 126 (2004) 1290–1298.
References
[
1] J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang, Y. Horiuchi, M. Anpo,
[28] M. Buchalska, M. Kobielusz, A. Matuszek, M. Pacia, S. Wojtyła, W. Macyk, ACS
D.W. Bahnemann, Chem. Rev. 114 (2014) 9919–9986.
Catal. 5 (2015) 7424–7431.
[
[
[
[
2] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735–758.
3] E. Jimenez-Relinque, M. Castellote, Cem. Concr. Res. 74 (2015) 108–115.
4] T. Hirakawa, Y. Nosaka, Langmuir 18 (2002) 3247–3254.
[29] H. Hou, M. Shang, L. Wang, W. Li, B. Tang, W. Yang, Sci. Rep. 5 (2015) 15228.
[30] T. Kawahara, Y. Konishi, H. Tada, N. Tohge, J. Nishii, S. Ito, Angew. Chemie 114
(2002) 2935–2937.
[31] X. Zhang, Y. Lin, D. He, J. Zhang, Z. Fan, T. Xie, Chem. Phys. Lett. 504 (2011)
71–75.
5] K.-I. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, Electrochem. Commun. 2
(2000) 207–210.
[
[
[
[
6] R. Mohan, K. Krishnamoorthy, S.-J. Kim, Solid State Commun. 152 (2012) 375–380.
7] A. Mills, J. Wang, S.-K. Lee, M. Simonsen, Chem. Commun. (2005) 2721–2723.
8] A. Mills, J. Wang, M. McGrady, J. Phys. Chem. B 110 (2006) 18324–18331.
9] A. Mills, J. Hepburn, D. Hazafy, C. O’Rourke, N. Wells, J. Krysa, M. Baudys,
M. Zlamal, H. Bartkova, C.E. Hill, J. Photochem. Photobiol. A Chem. 290 (2014)
[32] J. Yan, G. Wu, N. Guan, L. Li, Z. Li, X. Cao, J. Chem. Soc. Faraday Trans. 15 (2013)
10978–10988.
[33] B. Sun, P.G. Smirniotis, Catal. Today 88 (2003) 49–59.
[34] D.O. Scanlon, C.W. Dunnill, J. Buckeridge, S.A. Shevlin, A.J. Logsdail,
S.M. Woodley, C.R.A. Catlow, M.J. Powell, R.G. Palgrave, I.P. Parkin, Nat. Mater.
12 (2013) 798–801.
6
3–71.
[
[
[
10] M. Sahni, B.R. Locke, Ind. Eng. Chem. Res. (2006) 45.
11] A. Mills, N. Wells, J. MacKenzie, G. MacDonald, Catal. Today 281 (2017) 14–20.
12] A. Mills, N. Wells, C. O’Rourke, J. Photochem. Photobiol. A Chem. 338 (2017)
[35] X. Sun, W. Dai, G. Wu, L. Li, N. Guan, M. Hunger, Chem. Commun. 51 (2015)
13779–13782.
[36] A. Yamakata, J.J.M. Vequizo, H. Matsunaga, J. Phys. Chem. C 119 (2015)
24538–24545.
1
23–133.
[13] A. Mills, N. Wells, C. O’Rourke, Catal. Today 230 (2014) 245–249.
25