On the photocatalytic degradation of phenol and dichloroacetate by BiVO4: The need of a sacrificial electron acceptor

Article abstract of DOI:10.1016/j.jphotochem.2010.08.021

The photodegradation of phenol and dichloroacetic acid (DCAA) by BiVO ₄ was studied in the absence as well as presence of selected electron scavengers. The experiments were performed under the visible (vis) irradiation of aqueous solutions over a wide pH range (1-13). Phenol was photocatalytically degraded by BiVO₄ into p-benzoquinone below pH 3 and into an open-ring structure at pH 13. Methylene blue (MB) accelerated the reaction below the isoelectric point of BiVO₄ and did not undergo significant degradation. In presence of H₂O₂, phenol was rapidly degraded up to pH 9. The degradation rates are two orders of magnitude higher than in absence of electron scavenger. The degradation of dichloroacetic acid was only possible in presence of H₂O₂. High initial concentrations of H₂O₂ inhibit the reaction and its consumption is very fast. Sequential additions of this sacrificial electron acceptor (SEA) enables the total degradation of a 1 mM DCAA solution.

Full text of DOI:10.1016/j.jphotochem.2010.08.021

Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
On the photocatalytic degradation of phenol and dichloroacetate by BiVO4: The
need of a sacrificial electron acceptor
Nikola C. Castilloa,b,∗, Laura Dinga, Andre Heelb,1, Thomas Grauleb, Cesar Pulgarina,∗∗
a École polytechnique fédérale de Lausanne (EPFL), Faculté des Sciences de Base, Institut des Sciences et Ingénierie Chimiques, Groupe de Génie Électrochimique,
CH-1015 Lausanne, Switzerland
b Laboratory for High Performance Ceramics, Empa - Swiss Federal Laboratories for Materials Science and Technology, Überlandstrasse 129, 8600 Dübendorf, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 16 September 2010
The photodegradation of phenol and dichloroacetic acid (DCAA) by BiVO4 was studied in the absence as
well as presence of selected electron scavengers. The experiments were performed under the visible (vis)
irradiation of aqueous solutions over a wide pH range (1–13). Phenol was photocatalytically degraded by
BiVO4 into p-benzoquinone below pH 3 and into an open-ring structure at pH 13. Methylene blue (MB)
accelerated the reaction below the isoelectric point of BiVO4 and did not undergo significant degradation.
In presence of H2O2, phenol was rapidly degraded up to pH 9. The degradation rates are two orders of
magnitude higher than in absence of electron scavenger. The degradation of dichloroacetic acid was only
possible in presence of H2O2. High initial concentrations of H2O2 inhibit the reaction and its consumption
is very fast. Sequential additions of this sacrificial electron acceptor (SEA) enables the total degradation
of a 1 mM DCAA solution.
Keywords:
Bismuth vanadate
Visible light photocatalysis
Hydrogen peroxide, Electron scavenger
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
be detected. This was ascribed to the weaker oxidative power of
BiVO4 photoholes than of TiO2 ones [8,9].
Since the discovery of the Honda–Fujishima effect in 1972 [1],
semiconductor photochemistry has evolved – among others – into
twin disciplines solar energy conversion water photoelectrolysis
[2,3] and photocatalytic oxidation of pollutants [4,5]. TiO2 has
proved the best material for application eversince but researchers
are still looking for better performing ones. Thus BiVO4 was pre-
sented in 1998 as a promising candidate for oxygen evolution in
the half reaction of water decomposition due to its electronic band
structure allowing the absorption of visible light up to 540 nm and
an unusually high valence-band edge for a metal oxide [6,7]. As
a side effect, Kohtani et al. reported the photocatalytic degrada-
tion of endocrine disruptor 4-n-nonylphenol at pH 13 to favor the
phenolate form of the compound. The alkylphenolates yielded an
open-ring structure with one acid and one aldehyde group (direct
hole transfer), or a catechol (•OH attack) after 80% degradation of
the initial compound. However no smaller products nor CO2 could
Other pollutants such as rhodamine B [10], methyl orange [11]
and methylene blue [12] were also reportedly degraded by pure
monoclinic BiVO4 although the experimental conditions were not
always optimal (e.g. possible photosensitization by the dye not
mentioned, reductive bleaching not excluded, minor oxidation of
easily degraded functional groups; see Table 1). Recombination is
a matter of concern for BiVO4 [13], so a number of metals have
been loaded as electron traps on the surface of the semiconductor
such as Ag [14] and Cu [15]. Alternatively, Long et al. suggested
to use composite semiconducting particles by loading Co3O4 onto
BiVO4 for phenol degradation. Those should also suppress recom-
bination because charge separation is favored by an electric field
created within each nanoparticle [16]. Finally, the use of AgNO3 as
sacrificial electron acceptor to test O2 evolution activity may have
inspired Xie et al. to reduce Cr(VI) as they oxidized phenol [17].
Apart from the work by Kohtani et al. at pH 13 [9,14,18], BiVO4
has not yet been reported as a reliable photocatalyst for the degra-
dation of recalcitrant organic pollutants more than 10 years after
the discovery of its photoactivity [6]. Several reasons could explain
this: (a) photocatalytic activity has mainly been tested by the
simple-to-measure decolorization of dyestuff beside oxygen evo-
lution and has not risen much interest among the community
specialized in the degradation of recalcitrant organic pollutants;
(b) BiVO4 is not that good a photocatalyst, and results showing
its inability to degrade typical model pollutants well known to be
degraded by UV-driven TiO2 photocatalysis remained unpublished.
∗
Corresponding author at: École polytechnique fédérale de Lausanne (EPFL), Fac-
ulté des Sciences de Base, Institut des Sciences et Ingénierie Chimiques, Groupe de
Génie Électrochimique, CH-1015 Lausanne, Switzerland.
∗∗
Corresponding author.
E-mail addresses: nikola.castillo@epfl.ch (N.C. Castillo), cesar.pulgarin@epfl.ch
(C. Pulgarin).
1
Present address: Marketing, Knowledge and Technology Transfer, Empa - Swiss
Federal Laboratories for Materials Science and Technology, Überlandstrasse 129,
8600 Dübendorf, Switzerland.
1010-6030/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2010.08.021

222
N.C. Castillo et al. / Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
Table 2
Physico-chemical properties of BiVO4 nanoparticles.
Nomenclature
Crystal phase
Monoclinic
Eg (eV)
SSA (m2g−1
)
IEP
BQ
p-benzoquinone, C6H4O2
2.50
24.4
2.3
DCA
DCAA
e−cb
dichloroacetate anion, CHCl2COO−
dichloroacetic acid, CHCl2COOH
conduction-band electron
(pure, Riedel-de Haën, Germany) and K2S2O8 (≥99%, for analysis
ACS, Acros Organics, Belgium), were all used as received. MilliQ
water was used throughout the experiments.
h+vb
valence-band hole
hydroperoxyl radical
HOO•
HQ
hydroquinone, C6H4(OH)2
isoelectric point
leuco methylene blue, C16H20N3SCl
methylene blue, C16H18N3SCl
semi-reduced methylene blue
superoxide anion
free hydroxyl radical
surface-bound hydroxyl radical
phenol, C6H5OH
IEP
2.1. BiVO4 nanopowders
LMB
MB
BiVO4 nanoparticles were made by flame spray synthesis
with in situ crystallization as reported in
[21]. Crystal phase, bandgap energy (Eg), BET specific surface
area (SSA) and surface isoelectric point (IEP) are summarized
in Table 2.
•MB−
a previous work
•O2
−
•OH
•OHs
PhOH
2.2. Photocatalytic degradations
Some of us have been studying the degradation of model as
well as real organic pollutants by a number of advanced oxidation
processes (AOP) [22,23]. In particular, phenol and dichloroacetic
acid (DCAA) are degraded over TiO2 under UV irradiation by the
two major possible pathways, i.e. electrophilic attack of the aro-
matic ring by surface hydroxyl radical, and direct hole transfer,
respectively [24]. In order to reach total mineralization of an
organic pollutant, small aliphatic acid intermediates will eventu-
ally undergo a photo-Kolbe oxidation. Such reaction is therefore of
major importance and can be well assessed by the degradation of
DCAA because the presence of the two Cl atoms make decarboxyla-
tion the most dominant initial degradation pathway [25,26]. In this
contribution, we discuss the conditions required to degrade such
pollutants by visible-light driven photocatalysis over flame-made
BiVO4 nanoparticles and the use of a number of electron scavengers
to improve the reaction.
Photocatalytic activity was evaluated in cylindrical 60 mL Pyrex
glass reactors with a top septum lid for sampling. Aqueous solu-
tions were aerated under magnetic stirring (500 rpm) and exposed
to the irradiation of five 18 W Philips TLD blue SLV fluorescent
tubes (emission maximum at ꢀ = 445 nm; no overlap with any of
the chemicals’ absorption but BiVO4; global radiation flux at reactor
distance = 58 W m−2). Two ventilation fans ensured a stable tem-
perature of 25 ◦C inside the reactors throughout the experiments
[21].
BiVO4 powder was dispersed ultrasonically in the reactors filled
with MilliQ water, and then a phenol or DCAA solution was added
to the reactor. The pH was adjusted with HClO4 and NaOH. The
suspensions were then stirred in the dark to reach adsorption equi-
librium. Concentration of the photocatalyst was set to 0.5 g L−1
in all phenol degradations and to 1.0 g L−1 in all DCAA degra-
dations. Initial phenol and DCAA concentrations were 53.1 ␮M
(5 ppm) and 1.06 mM (137 ppm), respectively. The pH of DCAA solu-
tions was naturally 4.5 so the acid (pKa = 1.35 [27]) is dissociated
into dichloroacetate (DCA). 0.7 mL aliquots were collected with
a syringe at pre-established reaction times and subsequently fil-
tered (0.22 ␮m millipore membranes) prior to HPLC analysis. High
Performance Liquid Chromatography with a Diode Array Detector
(Agilent 1100 series) and a reverse phase Spherisorb silica col-
2. Experimental
Phenol (Reag. ACS, Riedel-de Haën, Germany), dichloroacetic
acid (99%, Fluka, Germany), hydrogen peroxide solution 35% by
weight (puriss., stabilized Sigma–Aldrich, Germany), methylene
blue (pure USP, AppliChem GmbH, Germany), Cu(ClO4)2·6H2O
Table 1
Chronological list of pollutants degraded by BiVO4-based photocatalysts.
Year
Photocatalyst
BiVO4
Pollutant
Comment
pH 13, solar simulator. Ring opening into aldehyde-acid/catechols but no
Ref.
2003
4-n-Alkylphenols
[9]
smaller aliphatic acids
2005
Ag–BiVO4
4-n-Alkylphenols
pH 13, solar simulator. Degradation rates increased by Ag loading on
long-chain alkylphenols but mineralization still negligible; no effect on PhOH
In acetonitrile, solar simulator
0.8% Co3O4 loaded as co-catalyst. Efficient mineralization and recycling of
photocatalyst
[14]
2005
2006
Ag–BiVO4
Co3O4–BiVO4
PAHsa
Phenol
[18]
[16]
2006
2006
2006
BiVO4
BiVO4
BiVO4
Phenol
PhOH degradation at pH 1.5. Enhancement of degradation and mineralization
by simultaneous reduction of Cr(VI) below IEP
Natural sunlight, dye sensitization not excluded. Dealkylation of the R–N(Et)2
dye
High-pressure Hg lamp, dye sensitization not excluded. Decolorization of the
azo dye
[17]
[10]
[11]
Rhodamine B
Methyl orange
2007
2008
2009
BiVO4
Cu–BiVO4
BiVO4
Methylene blue
Methylene blue
Methylene blue
W halogen lamp, dye sensitization not excluded
Hg lamp. Dealkylation of the R–N(Me)2 dye
Solar simulator, dye sensitization not excluded. Dye degradation by
amorphous BiVO4
[12]
[15]
[19]
2010
2010
BiVO4
BiVO4
Methylene blue
Methylene blue
UV, blue, and white light sources. Surface poisoning upon recycling
Blue light, dye sensitization and reductive photobleaching excluded.
Dealkylation of the R–N(Me)2 dye by amorphous BiVO4
[20]
[21]
a
Polycyclic aromatic hydrocarbons.

N.C. Castillo et al. / Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
223
Fig. 3. Degradation of a 53.1 ␮M phenol solution in a 0.5 g L−1 BiVO4 suspension
with different initial [H2O2] at pH 7. Concentration of PhOH (–), H2O2 (· · ·) and BQ
in a.u. (- - -) are plotted.
Fig. 1. Degradation of a 53.1 ␮M phenol solution (–) in a 0.5 g L−1 BiVO4 suspension
at pH 1, 2 and 13. The build-up of intermediates in arbitrary units (- - -) are ascribed
to BQ for pH 1 and 2 and to an unidentified compound at pH 13.
amount of phenol was degraded in a more linear fashion suggesting
zero order kinetics (Fig. 1).
umn C-18 (Macherey-Nagel) was used to detect PhOH, H2O2 and
p-benzoquinone (BQ) by UV absorption at 220, 220 and 244 nm,
respectively. The mobile phase consisted of an acetonitrile:water
gradient. [H2O2] was confirmed by Merckoquant Peroxide Test
The fact that two different reaction rate laws apply at oppo-
site pH values suggests two different reaction pathways. On the
one hand, phenol is degraded following pseudo-first order kinetics
below the IEP of the photocatalyst. The build-up of p-benzoquinone,
BQ, as main intermediate, symmetric to phenol disappearance
(Fig. 1), is in good agreement with the widely accepted electrophilic
hydroxylation of phenol (reaction (1)) [28]. On the other hand, the
zero order found at pH 13 is reminiscent of the rate laws found for
the degradation of long-chained alkylphenols with BiVO4 [9]. The
deprotonated state of phenol at such high pH (pKa = 9.99) may favor
a different pathway. Indeed, Kohtani et al. suggested a 6-carbon
open-ring structure with a carboxylic group on C1 and an aldehyde
one on C62 as their main degradation product from direct hole
transfer [9]. The linear and very symmetric build-up of the HPLC
peak of a fairly high polar compound in the current work may well
correspond to such type of muconic acid-like compound (Fig. 1).
As a matter of fact, the absence of absorption band at ꢀ > 230 nm
is characteristic of a molecule that has lost its aromaticity by ring
cleavage. The oxidation through direct hole transfer to adsorbed
phenolate is consistent with the suggestion made above: the oxi-
dation of adsorbed water is unfavored above the IEP of BiVO4.
Moreover, Kohtani et al. observed a negligible production of free
•OH at pH 12 [29], which is consistent with the lack of phenol
hydroxylation products (BQ) at pH 13.
strips at levels between 0.5 and 25 mg L−1
.
DCAA was isolated with a Supelcogel H column and its concentra-
tion was measured with a refraction index detector. The mobile
phase was a 5 mM H2SO4 solution.
3. Results and discussion
3.1. Phenol degradation
Phenol was degraded down to 10% of its initial concentration
after a 300-min long irradiation of a BiVO4 suspension at initial
pH 1 (Fig. 1). At pH 2, phenol was degraded to a lower extent, the
activity was even more reduced at pH 3 and became null for the
range 4 < pH < 13 (Fig. 2). These results are in agreement with the
experiments by Xie et al. at pH 1.5 [17]. It is worth to note that
the dramatic drop in activity occurs across the isoelectric point of
BiVO4 (IEP = 2.3). Phenol is uncharged in acidic and neutral media
(pKa = 9.99 [27]) and weakly adsorbs onto hydrophilic oxides sur-
face so it cannot react with valence-band holes, h+vb. It rather reacts
with surface-bound hydroxyl radicals, •OHs [24]:
•OHs + PhOH → •Ph(OH)2
(1)
3.2. Phenol degradation in the presence of electron scavengers
The sole activity of BiVO4 below its IEP must thus be related to
the formation of reactive oxygen species (ROS) on the surface of
positively charged BiVO4. It is not clear whether such species can
form on negatively charged BiVO4 and if so, why they cannot oxi-
dize phenol. One possibility is that the oxidation of adsorbed water
is not favored on negatively charged BiVO4. At pH 13, a significant
3.2.1. H2O2
Addition of 3 mM H2O2 into the photocatalytic reactor prior to
irradiation lead to the total degradation of phenol at pH 7 within
20 min (Fig. 3). However, for lower H2O2 initial concentrations, the
reaction is immediately quenched after the entire consumption of
H2O2. The evolution of [H2O2] from several initial concentrations
shows that the decomposition is linear, with a consumption rate
of 3 × 10−6 M s−1. The degradation of PhOH occurs in the presence
of H2O2: up to t = 7 min for [H2O2]0 = 1 mM; up to t = 20 min for
[H2O2]0 = 2 mM (Fig. 3). BQ is concomitantly formed upon PhOH
removal. BQ reaches its maximum concentration, [BQ]max, at 7 min
for [H2O2]0 = 1 mM, i.e. when [H2O2] reaches zero. [BQ]max is shifted
to 12 min when [H2O2]0 is doubled; H2O2 obviously lasts longer in
the reactor. It is worth to note that [BQ] starts to decrease after
the total consumption of H2O2 in these two runs. This slow decre-
ment may not reflect photocatalytic degradation but rather redox
2
The aldehydic acid was wrongly named cis,cis-4-alkyl-6-oxo-2,6-hexadienoic
acid. The correct nomenclature should be cis,cis-4-alkyl-6-oxo-2,4-hexadienoic acid
since the second double bond lies between C4 and C5.
Fig. 2. Phenol removal as a function of pH from a 53.1 ␮M solution after 300 min of
irradiation with 0.5 g L−1 BiVO4.

224
N.C. Castillo et al. / Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
for •OH in a chain reaction both on the surface and in the solution
bulk [35]:
H2O2 + •OH → H2O + HOO•
HOO• + •OH → H2O + O2
(10)
(11)
Note that one H2O2 molecule scavenges two •OH. Not only is H2O2
consumed in this scheme but •OH are lost to yield the same reac-
tion products as in reaction (7). One last option is the reaction of
H2O2 with HOO•, which can produce both surface-adsorbed and
free hydroxyl radicals [5]:
H2O2 + HOO•
→
•OH + H2O + O2
(12)
From the above considerations, the fate of •OHs from the one-
electron reduction of H2O2 (reaction (3)) is either: (a) further
reduction to H2O (reaction (4)), (b) decomposition of H2O2 (reac-
tions (10) and (11)), or (c) the electrophilic attack of a pollutant
(e.g. reaction (1)). A competition between these four reactions is
thus expected. Similarly, the fate of HOO• from the one-electron
oxidation of H2O2 (reaction (5)) is either further oxidation to O2
(reaction (6)) or other deactivation pathways (e.g.: reaction (11)).
The competition between those reactions will depend on the
availability of the pollutants, their nature, pH, and on the surface
properties of the photocatalyst such as degree of hydroxylation and
the nature of the active sites. The high values of the H2O2/PhOH
ratio and its different values under different pH conditions are
ascribed to the complexity of the system. Assuming pseudo-first
order kinetics for phenol degradation, the reaction rates with H2O2
are around two orders of magnitude higher than in absence of elec-
tron acceptor at pH < IEP. The decomposition of H2O2 is faster for
pH values closer to the IEP with a maximum decomposition rate
at pH 2 which could not be measured because H2O2 was entirely
decomposed within 3 min (Fig. 4).
Fig. 4. Degradation of a 53.1 ␮M phenol solution in a 0.5 g L−1 BiVO4 suspension at
several initial pH with 1 mM H2O2.
equilibrium reactions with hydroquinone, HQ:
BQ + 2e− + 2H+ ꢀ HQ E◦ = +0.6992 V
(2)
The shift towards the right is consistent with the pH drop
observed at the end of the run: 4.5 < pHfinal < 5.0. At higher [H2O2]0,
[BQ] rapidly starts to decrease before [H2O2] reaches zero. The
H2O2/PhOH molar ratio for the complete degradation of phenol at
pH 7 is about 55. If complete degradation of BQ is sought too, the
ratio becomes 75.
The H2O2/PhOH ratio and the reaction kinetics are pH depen-
dent. Higher rates are observed at low pH and thus lower
H2O2/PhOH ratios are required (Fig. 4). The ratio for full degradation
is ≤20 at pH 2, while it rises to about 55 at pH 7. The accumulation of
BQ in all runs with significant PhOH degradation shows that 1 mM
H2O2 is not sufficient to completely degrade both PhOH and BQ
even at pH 2 (data not shown). The use of H2O2 as electron scav-
enger for TiO2 photocatalysis, along with its decomposition kinetics
was studied in the early 1990s [30,31] and later reviewed by Ilisz et
al. [32]. The photocatalytic decomposition of H2O2 on TiO2 surface
can proceed either by reduction
H2O2 can considerably speed up the degradation of phenol,
which was achieved by only BiVO4 at pH < IEP. Above the IEP, the
function of H2O2 is to activate BiVO4 and extend its activity to more
neutral pH values by working as sacrificial electron acceptor.
H2O2 + e−cb + H+
→
•OHs + H2O
(3)
(4)
3.2.2. Methylene blue
•OHs + e−cb + H+ → H2O
The decolorization of methylene blue (MB) is widely used to
test the activity of novel photocatalysts, including our flame-made
BiVO4 [21] (Table 1). MB is stable in acidic media, where both its
oxidation andreduction are reversible. Thepresence ofMB(5.7 ␮M)
accelerates the degradation of a 53.1 ␮M phenol solution at pH 2.
The rate is not as fast as with H2O2 but MB is not consumed since
it is regenerated within ␮s [36]. The proposed reaction mechanism
between the conduction-band electron and MB is as follows:
or by oxidation
H2O2 + h+ → HOO• + H+
(5)
(6)
vb
HOO• + h+vb → O2 + H+
in two-electron transfer processes. The combination of Eqs. (3)–(6)
yields:
2H2O22h−ꢁ,→TiO22H2O + O2
(7)
2e−cb + 2MB → 2•MB−
2•MB− + 2H+ → MB + LMB
LMB + O2 → MB + H2O
(13)
(14)
(15)
which is nothing else than a photocatalyzed disproportionation of
H2O2 [33]. The amount of H2O2 can thus affect both the forma-
tion of HOO•/•O2− and •OH [34], which are usually formed through
reactions summarized by:
The combination of Eqs. (13)–(15) yields:
e−cb + O2 + H+ → HOO•
h+vb + H2O → •OHs + H+
(8)
(9)
2hꢁ,MB,BiVO
2e− + 2H+ + O2
−→ 4H2O
(16)
MB is neither consumed nor produced in this reaction, as opposed
to H2O2 in reaction (7). Therefore it should not be regarded as a
sacrificial electron acceptor (SEA) and is better described as a redox
mediator or as a co-catalyst. Unlike H2O2, MB does not enable the
degradation of phenol at pH > IEP of BiVO4. This issue is addressed
elsewhere in detail [36].
Note that the radicals produced in reactions (8) and (9) are identical
to those from (5) and (3), respectively. However, reactions (3) and
(8) are reductions while (5) and (9) are oxidations, i.e. opposite
types of electron transfer are expected to yield the same species.
If we assume that reactions (8) and (9) occur upon irradiation of
BiVO4 under certain pH conditions (Figs. 1 and 2), it is difficult to
establish whether •OHs comes from hole (9) or electron (3) transfer
when H2O2 has been added.
3.3. Dichloroacetic acid degradation
As reviewed by Dionysiou et al. high concentrations of H2O2 can
also be detrimental because it may compete with trace pollutants
Dichloroacetate anion (DCA) could not be degraded at pH 4.5 in
absence of electron acceptor (Fig. 5). The kinetics of the complete

N.C. Castillo et al. / Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
225
With [H2O2]0 = 23.3 and 37.2 mM, an induction period was
observed before the degradation started (Fig. 5). Besides, final
[DCA] was around 60% of [DCA]0, i.e. very close to the run with
[H2O2]0 = 10.6 mM. This indicates that high concentrations of H2O2
inhibit the degradation of DCA. It can be argued that surface sites
are not available to DCA adsorption until a large fraction of H2O2
is decomposed [35] by reaction (7), but also mainly by reactions
(10) and (11) – which are favored by excess H2O2 – combined to
•OHs generation through reactions (3), (9) and/or (12). Indeed, the
induction time reaches 60 min for [H2O2]0 = 37.2 mM. This tran-
sient poisoning of the photocatalyst is in agreement with the claims
of Bahnemann et al., who found that DCA adsorption on colloidal
TiO2 (Eq. (17)) is a prerequisite for its oxidation [38]. Hirakawa
& Nosaka also found that high amounts of adsorbed H2O2 would
inhibit reaction (8) on adsorbed O2 [34].
Fig. 5. Degradation of a 1.06 mM DCA solution in a 1.0 g L−1 BiVO4 suspension with
varying initial [H2O2] at pH 4.5. The run with stepwise additions included 10 loads
of H2O2 at 30 min intervals.
The continuous availability of a reactant in low amounts can
be provided by a fed-batch or by dropwise addition. Fig. 5 shows
the effect of several additions of concentrated H2O2 to renew
[H2O2] = 2.33 mM after each addition and try to supply the reactor
with fresh H2O2 as soon as it is totally consumed. Total degradation
of a 1 mM DCAA solution is thus achieved after 6 h of irradiation. It
is worth to note that the total amount of added H2O2 is equivalent
to the run with [H2O2]0 = 23.3 mM, which was largely inefficient,
with <40% of the initial concentration of DCA being degraded. This
is an important observation. It shows that most H2O2 is wasted
if not added at a proper time and in an adequate amount. [H2O2]
should hence be carefully controlled and kept at a constant level
either by continuous addition or in situ generation. The formation
of H2O2 as a byproduct in reaction (21) may even lower the demand
of added H2O2. Dionysiou et al. found that the optimal H2O2 to con-
taminant molar ratio for the degradation of 4-chlorobenzoic acid
(C7H5ClO2) in a continuous-mode configuration was equal to 5.33,
above which inhibitory effects were observed [35]. In the present
work, total degradation was achieved with a 21.7 molar ratio by
repeated H2O2 additions. This value is much higher and DCA is min-
eralized after one single hole transfer [37] while 4-chlorobenzoic
acid is a larger molecule. This observation emphasizes the ease of
decomposition of H2O2 and the potential improvements to be made
by adjusting some reaction engineering parameters.
mechanism of DCA degradation by TiO2 photocatalysis was studied
by Zalazar et al. [37]. Only one hole is sufficient to mineralize one
DCA molecule by direct hole transfer, although adsorption on the
photocatalyst surface is a prerequisite [38]:
TiO2 + CHCl2COO− ꢀ CHCl2COO−
(17)
ads
CHCl2COO− + hv+b → CHCl2COO•
CHCl2COO• → HCl2C• + CO2
(18)
(19)
(20)
(21)
(22)
ads
O2ads + HCl2C• → CHCl2O2
•
•
2CHCl2O2 → 2COCl2 + H2O2
COCl2 + H2O → CO2 + 2HCl
Note that reaction (21) is a bimolecular step that requires the vicin-
•
ity of two CHCl2O2 radicals to yield two short-lived phosgene
molecules and one H2O2. In the present work, it seems that pho-
togenerated holes from BiVO4, h+vb, were not powerful enough to
oxidize DCA (reaction (18)) or that DCA did not adsorb on BiVO4 (Eq.
(17)). The latter is a realistic scenario from an electrostatic point of
view because both DCA and BiVO4 surface are negatively charged
at pH 4.5.
Based on the results found for phenol, DCA degradation was
attempted with the best electron acceptor, i.e. sacrificial H2O2 (see
Subsection 3.2). With an initial H2O2 concentration of 2.33 mM,
DCA was degraded to 82% of its initial concentration within 30 min
(Fig. 5). Accordingly, with [H2O2]0 = 10.6 mM, DCA was degraded
to 67% of its initial concentration within 60 min, until H2O2 was
totally consumed, as it was the case for PhOH. The role of H2O2
can be manifold. Reaction (3) can prevent charge recombination
and extend the lifetime of holes, and favor reaction (18). In the
mean time, H2O2 can contribute to the formation of •OHs through
reaction (3) and/or reaction (9). DCA is known to be oxidized by free
•OH in a homogeneous H2O2-UV system [39], but surface-bound
hydroxyl radicals are reportedly unreactive towards DCA [38]. It
is concluded that DCA is either oxidized by valence-band holes,
whose lifetime is extended by the addition of H2O2, or by free •OH
resulting from the acid-catalyzed hydrolysis of surface ␮-peroxo
species, as suggested by Nakamura & Nakato [40]:
3.4. Choice of the electron acceptor on the basis of band edges
In an ideal photocatalytic system, only the photocatalyst needs
to be supplied. All other agents are expected to be “naturally”
present: sunlight, water/organic pollutants as hole scavenger, and
dissolved atmospheric oxygen as electron scavenger. The alter-
native electron acceptors studied in this work present different
features: hydrogen peroxide is consumed in reaction (3) but it
is efficient and “green” (reaction products are only H2O and O2,
reaction (7)). One problem related to larger scale application is
the liquid state and facile thermal decomposition of H2O2, which
require special handling. Methylene blue is a cheap organic dye,
readily available as an easy-to-handle powder. Its role as electron
acceptor is not sacrificial because it is regenerated to its initial
state (reactions (13)–(15)), so it can be used in catalytic amounts:
MB/PhOH molar ratio is 0.1 in the present study. Once the target
pollutant is totally degraded, the degradation of the dye itself could
become an issue, as in dye-sensitized solar cells (DSSC) [42]. Aque-
ous Cu2+ had no effect on the degradation reaction under any of our
experimental conditions (1 ␮M ≤c0(Cu2+) ≤1 mM, pH 7; data not
shown). Its redox potential is not pH dependent, although oxides
and/or hydroxides may precipitate in basic conditions. Its reduction
is expected to yield metallic copper that might deposit on the pho-
tocatalyst unless some parallel Fenton-like redox process allows
its regeneration [30,43]. Peroxydisulfate anion can be used for pol-
lutant oxidation but its actual role is not clear. This compound is
[Ti − O − O − Ti]s + H2O → [Ti − O − OH HO − Ti]s
(23)
When the O–O bond in TiO–OH breaks, a free hydroxyl radical is
released [5]:
[Ti − O − OH HO − Ti]s → [Ti − O• HO − Ti]s + •OH
(24)
This issue is still under debate but it is becoming clear that the
generation of free •OH from metal oxide surfaces requires the pres-
ence of surface peroxo groups. The geometry of these groups seems
critical for the process [40,41].

226
N.C. Castillo et al. / Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
addition. A steady state formation of H2O2 inside the reactor would
best solve this problem and allow the extended use of BiVO4 as
a powerful, non-selective, visible-light photocatalyst for the com-
plete mineralization of recalcitrant organic pollutants. BiVO4 shows
best activity close by the isoelectric point of 2.3 and below. The
use of MB as a regenerating electron acceptor is promising and
deserves further investigation. Simultaneous reduction of Cu2+ was
not successful at natural pH.
4. Conclusion
Aqueous phenol was degraded by BiVO4 photocatalysis under
visible light below the isoelectric point of BiVO4 (pH 2.3). The degra-
dation rate was enhanced by the addition of MB as a regenerating
redox mediator. A more effective and even highly efficient degra-
dation was achieved in the presence of sacrificial H2O2 up to pH
9. But degradation is then limited by the kinetics of H2O2 con-
sumption. Dichloroacetate was also degraded in presence of low
H2O2 concentrations. Competitive adsorption seems to inhibit DCA
degradation when high H2O2 concentrations are used, so the sys-
tem is more efficient with controlled and consecutive addition of
low H2O2 amounts. The need of hydrogen peroxide as a sacrificial
electron acceptor has been identified as a critical issue for the pho-
tocatalytic degradation of phenol and dichloroacetic acid by pure
BiVO4. H2O2 is required if this material is to be used in realistic
advanced oxidation processes and not limited to applications under
extreme pH conditions.
Fig. 6. Redox potential – pH diagram of selected couples used as electron scavenger.
Band edge energies of TiO2 and BiVO4 are also displayed. The one-electron reduction
2−
potential of S2O8 – is equal to that of H2O2. Data from [5,27,47,48]. Values for the
conduction band of BiVO4 are calculated from the onset potential of dark reduction
performed on a thin film electrode at pH 4 and extrapolated by the Nernst equation
(E(pH) = EpH 4 − 0.059(pH − 4); unpublished data). The potential of the valence-band
edge is 2.50 V more positive than the potential of the conduction-band edge as
measured by DRS [21].
Acknowledgments
known to oxidize phenol into hydroquinone (HQ) in aqueous solu-
tion through the so-called Elbs persulfate oxidation [44], it was
formerly used as a precursor in the synthesis of H2O2 under very
acidic conditions [45], and its slow thermal homolysis leads to •OH
radicals [46]. Its redox potential is pH independent and the reac-
tion product is sulfate. In this study, DCA was partially degraded
in a peroxydisulfate solution with and without photocatalyst and
hence no further investigations were carried out (data not shown).
The redox potentials of all these electron scavengers are plot-
ted vs. the solution pH in Fig. 6 along with the band edges of
BiVO4. The valence-band edge is practically at the same level as
free hydroxyl radical formation up to pH 12, where •OH genera-
tion becomes thermodynamically unfavored. On the other hand O2
reduction seems viable down to pH 8. The work of Kohtani et al. on
The authors thank the Secrétariat d’État Suisse à l’éducation et
à la recherche (SER) for its support to the project SER C06.0074
in the framework of the Cost Action 540 “PHONASUM” from the
European community; N. Leresche for the degradation experiments
with MB. Dr J.A. Rengifo-Herrera contributed to fruitful discussions
and experimental setups.
References
[1] A. Fujishima, K. Honda, Electrochemical photolysis of water at a semiconductor
electrode, Nature 238 (5358) (1972) 37–38.
[2] A. Kudo, Y. Miseki, Heterogeneous photocatalyst materials for water splitting,
Chem. Soc. Rev. 38 (1) (2009) 253–278.
[3] J. Tang, J.R. Durrant, D.R. Klug, Mechanism of photocatalytic water splitting in
TiO2. Reaction of water with photoholes, importance of charge carrier dynam-
ics, and evidence for four-hole chemistry, J. Am. Chem. Soc. 130 (42) (2008)
13885–13891.
[4] A.G. Agrios, P. Pichat, State of the art and perspectives on materials and applica-
tions of photocatalysis over TiO2, J. Appl. Electrochem. 35 (7–8) (2005) 655–663.
[5] A. Fujishima, X. Zhang, D.A. Tryk, TiO2 photocatalysis and related surface phe-
nomena, Surf. Sci. Rep. 63 (12) (2008) 515–582.
[6] A. Kudo, K. Ueda, H. Kato, I. Mikami, Photocatalytic O2 evolution under visi-
ble light irradiation on BiVO4 in aqueous AgNO3 solution, Catal. Lett. 53 (3–4)
(1998) 229–230.
[7] A. Kudo, K. Omori, H. Kato, A novel aqueous process for preparation of crystal
form-controlled and highly crystalline BiVO4 powder from layered vanadates
at room temperature and its photocatalytic and photophysical properties, J.
Am. Chem. Soc. 121 (49) (1999) 11459–11467.
[8] S. Kohtani, S. Makino, A. Kudo, K. Tokumura, Y. Ishigaki, T. Mat-
sunaga, O. Nikaido, K. Hayakawa, R. Nakagaki, Photocatalytic degradation
of 4-n-nonylphenol under irradiation from solar simulator: compari-
son between BiVO4 and TiO2 photocatalysts, Chem. Lett. 31 (7) (2002)
660–661.
[9] S. Kohtani, M. Koshiko, A. Kudo, K. Tokumura, Y. Ishigaki, A. Toriba, K. Hayakawa,
R. Nakagaki, Photodegradation of 4-alkylphenols using BiVO4 photocatalyst
under irradiation with visible light from a solar simulator, Appl. Catal. B 46 (3)
(2003) 573–586.
the detection of free •OH and •O2 carried out at pH 12 with the
−
same method as Hirakawa et al. [34] fits well this picture: evidence
for •O2− generation but not for •OH [29]. The reduction potentials of
the acceptors used in this work are all more positive than EO
−
•
/ O
2
2
and than the potential of conduction-band electrons. This means
that the electron transfer reaction should be thermodynamically
favored. All energies and potentials, with the exception of Cu2+/Cu,
vary linearly between pH 0 and 4.5. Therefore, the energy picture
cannot explain the differences in activities found over this range
across the IEP of the photocatalyst. Indeed, other parameters such
as adsorption energies, steric hindrance and electrostatic interac-
tions do also play a role. The authors believe that reactivity more
strongly depends on surface interactions because phenol degrada-
tion in absence of an electron scavenger and in presence of MB
is only observed below the IEP. On the other hand the photocat-
alytic activity in presence of H2O2 is determined by its consumption
rate, which is a kinetic parameter rather than a thermodynamic
one.
BiVO4 turns out to be an efficient photocatalyst under visible
irradiation (ꢀmax = 450 nm) for the degradation of model pollutants
phenol and DCAA at natural pH provided that H2O2 is available on
the photocatalyst surface. Due to the high decomposition rate and
the price of H2O2, the use of this SEA is only viable by continuous
[10] L. Zhang, D. Chen, X. Jiao, Monoclinic structured BiVO4 nanosheets: hydrother-
mal preparation, formation mechanism, and coloristic and photocatalytic
properties, J. Phys. Chem. B 110 (6) (2006) 2668–2673.
[11] L. Zhou, W. Wang, S. Liu, L. Zhang, H. Xu, W. Zhu, A sonochemical route to
visible-light-driven high-activity BiVO4 photocatalyst, J. Mol. Catal. A: Chem.
252 (1–2) (2006) 120–124.

N.C. Castillo et al. / Journal of Photochemistry and Photobiology A: Chemistry 216 (2010) 221–227
227
[12] X. Zhang, Z. Ai, F. Jia, L. Zhang, X. Fan, Z. Zou, Selective synthesis and visible-
light photocatalytic activities of BiVO4 with different crystalline phases, Mater.
Chem. Phys. 103 (1) (2007) 162–167.
[13] M. Long, W. Cai, H. Kisch, Visible light induced photoelectrochemical properties
of n-BiVO4 and n-BiVO4/p-Co3O4, J. Phys. Chem. C 112 (2) (2008) 548–554.
[14] S. Kohtani, J. Hiro, N. Yamamoto, A. Kudo, K. Tokumura, R. Nakagaki, Adsorp-
tive and photocatalytic properties of Ag-loaded BiVO4 on the degradation of
4-n-alkylphenols under visible light irradiation, Catal. Commun. 6 (3) (2005)
185–189.
[15] H. Xu, H. Li, C. Wu, J. Chu, Y. Yan, H. Shu, Z. Gu, Preparation, characterization
and photocatalytic properties of Cu-loaded BiVO4, J. Hazard. Mater. 153 (1–2)
(2008) 877–884.
[16] M. Long, W. Cai, J. Cai, B. Zhou, X. Chai, Y. Wu, Efficient photocatalytic degra-
dation of phenol over Co3O4/BiVO4 composite under visible light irradiation, J.
Phys. Chem. B 110 (41) (2006) 20211–20216.
species in semiconductor photocatalysis, Phys. Chem. Chem. Phys. 10 (2008)
2986–2992.
[30] T.Y. Wei, Y.Y. Wang, C.C. Wan, Photocatalytic oxidation of phenol in the
presence of hydrogen peroxide and titanium dioxide powders, J. Photochem.
Photobiol. A 55 (1990) 115–126.
[31] V. Auguliaro, E. Davi, L. Palmisano, M. Schiavello, A. Sclafani, Influence of hydro-
gen peroxide on the kinetics of phenol photodegradation in aqueous titanium
dioxide dispersion, Appl. Catal. 65 (1) (1990) 101–116.
[32] I. Ilisz, K. Föglein, A. Dombi, The photochemical behavior of hydrogen peroxide
in near UV-irradiated aqueous TiO2 suspensions, J. Mol. Catal. A: Chem. 135 (1)
(1998) 55–61.
[33] B. Jenny, P. Pichat, Determination of the actual photocatalytic rate of
hydrogen peroxide decomposition over suspended titania. fitting to the
Langmuir–Hinshelwood form, Langmuir 7 (5) (1991) 947–954.
−
[34] T. Hirakawa, Y. Nosaka, Properties of O2 and OH formed in TiO2 aqueous sus-
[17] B. Xie, H. Zhang, P. Cai, R. Qiu, Y. Xiong, Simultaneous photocatalytic reduction
of Cr(VI) and oxidation of phenol over monoclinic BiVO4 under visible light
irradiation, Chemosphere 63 (6) (2006) 956–963.
[18] S. Kohtani, M. Tomohiro, K. Tokumura, R. Nakagaki, Photooxidation reactions
of polycyclic aromatic hydrocarbons over pure and Ag-loaded BiVO4 photocat-
alysts, Appl. Catal. B 58 (3–4) (2005) 265–272.
[19] J. Yu, Y. Zhang, A. Kudo, Synthesis and photocatalytic performances of BiVO4
by ammonia co-precipitation process, J. Solid State Chem. 182 (2) (2009) 223–
228.
[20] Y. Zhou, K. Vuille, A. Heel, B. Probst, R. Kontic, G.R. Patzke, An inorganic
hydrothermal route to photocatalytically active bismuth vanadate, Appl. Catal.
A 375 (2010) 140–148.
[21] N.C. Castillo, A. Heel, T. Graule, C. Pulgarin, Flame-assisted synthesis of
nanoscale, amorphous and crystalline, spherical BiVO4 with visible-light pho-
tocatalytic activity, Appl. Catal. B 96 (3–4) (2010) 335–347.
[22] D. Gumy, S.A. Giraldo, J. Rengifo, C. Pulgarin, Effect of suspended TiO2 physic-
ochemical characteristics on benzene derivatives photocatalytic degradation,
Appl. Catal. B 78 (1–2) (2008) 19–29.
[23] M. Lapertot, C. Pulgarín, P. Fernández-Ibán˜ez, M.I. Maldonado, L. Pérez-Estrada,
I. Oller, W. Gernjak, S. Malato, Enhancing biodegradability of priority substances
(pesticides) by solar photo-Fenton, Water Res. 40 (5) (2006) 1086–1094.
[24] R. Enríquez, A.G. Agrios, P. Pichat, Probing multiple effects of TiO2 sintering
temperature on photocatalytic activity in water by use of a series of organic
pollutant molecules, Catal. Today 120 (2) (2007) 196–202.
[25] D.W. Bahnemann, S.N. Kholuiskaya, R. Dillert, A.I. Kulak, A.I. Kokorin, Photode-
struction of dichloroacetic acid catalyzed by nano-sized TiO2 particles, Appl.
Catal. B 36 (2) (2002) 161–169.
pensions by photocatalytic reaction and the influence of H2O2 and some ions,
Langmuir 18 (8) (2002) 3247–3254.
[35] D.D. Dionysiou, M.T. Suidan, I. Baudin, J.-M. Laîné, Effect of hydrogen perox-
ide on the destruction of organic contaminants-synergism and inhibition in a
continuous-mode photocatalytic reactor, Appl. Catal. B 50 (4) (2004) 259–269.
[36] N.C. Castillo, N. Leresche, J. Teuscher, J.-E. Moser, A. Heel, T. Graule, C. Pulgarín,
Adsorbed methylene blue on BiVO4 nanoparticles catalyzes the photocatalytic
degradation of aqueous phenol, Langmuir, in press.
[37] C.S. Zalazar, C.A. Martin, A.E. Cassano, Photocatalytic intrinsic reaction kinet-
ics. II: Effects of oxygen concentration on the kinetics of the photocatalytic
degradation of dichloroacetic acid, Chem. Eng. Sci. 60 (15) (2005) 4311–4322.
[38] D.W. Bahnemann, M. Hilgendorff, R. Memming, Charge carrier dynamics at TiO2
particles: reactivity of free and trapped holes, J. Phys. Chem. B 101 (21) (1997)
4265–4275.
[39] C.S. Zalazar, M.D. Labas, R.J. Brandi, A.E. Cassano, Dichloroacetic acid degra-
dation employing hydrogen peroxide and UV radiation, Chemosphere 66 (5)
(2007) 808–815.
[40] R. Nakamura, Y. Nakato, Primary intermediates of oxygen photoevolution
reaction on TiO2 (rutile) particles, revealed by in situ FTIR absorption and pho-
toluminescence measurements, J. Am. Chem. Soc. 126 (4) (2004) 1290–1298.
[41] T. Hirakawa, K. Yawata, Y. Nosaka, Photocatalytic reactivity for O2•−and OH•
radical formation in anatase and rutile TiO2 suspension as the effect of H2O2
addition, Appl. Catal. A 325 (2007) 105–111.
[42] A. Nozik, R. Memming, Physical chemistry of semiconductor–liquid interfaces,
J. Phys. Chem. 100 (31) (1996) 13061–13078.
[43] R. Cai, Y. Kubota, A. Fujishima, Effect of copper ions on the formation of hydro-
gen peroxide from photocatalytic titanium dioxide particles, J. Catal. 219 (1)
(2003) 214–218.
[44] S.M. Sethna, The Elbs persulfate oxidation, Chem. Rev. 49 (1) (1951) 91–101.
[45] G. Goor, J. Glenneberg, S. Jacobi, Ullmann’s Encyclopedia of Industrial Chem-
istry, Wiley-VCH, Weinheim, 2007 (Ch. Hydrogen peroxide, pp. 18–20).
[46] R.L. Johnson, P.G. Tratnyek, R.O. Johnson, Persulfate persistence under thermal
activation conditions, Environ. Sci. Technol. 42 (24) (2008) 9350–9356.
[47] R. Memming, Mechanism of the electrochemical reduction of persulfates and
hydrogen peroxide, J. Electrochem. Soc. 116 (6) (1969) 785–790.
[48] A. Mills, J. Wang, Photobleaching of methylene blue sensitised by TiO2: an
ambiguous system? J. Photochem. Photobiol. A 127 (1–3) (1999) 123–134.
[26] R. Enriquez, B. Beaugiraud, P. Pichat, Mechanistic implications of the effect of
TiO2 accessibility in TiO2–SiO2 coatings upon chlorinated organics photocat-
alytic removal in water, Water Sci. Technol. 49 (4) (2004) 147–152.
[27] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 90th ed., CRC Press,
Boca Raton, 2009.
[28] K.-i. Okamoto, Y. Yamamoto, H. Tanaka, M. Tanaka, A. Itaya, Heterogeneous
photocatalytic decomposition of phenol over TiO2 powder, Bull. Chem. Soc.
Jpn. 58 (7) (1985) 2015–2022.
[29] S. Kohtani, K. Yoshida, T. Maekawa, A. Iwase, A. Kudo, H. Miyabe, R. Nak-
agaki, Loading effects of silver oxides upon generation of reactive oxygen