Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology C:
Photochemistry Reviews
journal homepage: www.elsevier.com/locate/jphotochemrev
Review
Heterogeneous photocatalytic treatment of organic dyes in air
and aqueous media
K. Rajeshwar a,∗ , M.E. Osugi b , W. Chanmanee c , C.R. Chenthamarakshan a ,
M.V.B. Zanoni b , P. Kajitvichyanukul d , R. Krishnan-Ayer a
a
Center for Renewable Energy Science & Technology, University of Texas at Arlington, Arlington, TX 76019, USA
Departamento de Quimica Analitica, Instituto de Quimica de Araraquara, CEP 14801-970, Araraquara, SP, Brazil
National Center of Excellence for Environmental and Hazardous Waste Management, Chulalongkorn University, Bangkok, Thailand
d
Biological Engineering Program, Faculty of Engineering, King Mongkut’s University of Technology Thonburi, Thailand
b
c
a r t i c l e
i n f o
Article history:
Received 24 April 2008
Received in revised form 14 July 2008
Accepted 21 September 2008
Available online 17 October 2008
Keywords:
Photoexcitation
Free radicals
Solar irradiation
Electron–hole pairs
Langmuir–Hinshelwood mechanism
a b s t r a c t
This review focuses on the heterogeneous photocatalytic treatment of organic dyes in air and water.
Representative studies spanning approximately three decades are included in this review. These studies have mostly used titanium dioxide (TiO2 ) as the inorganic semiconductor photocatalyst of choice
for decolorizing and decomposing the organic dye to mineralized products. Other semiconductors such
as ZnO, CdS, WO3 , and Fe2 O3 have also been used, albeit to a much smaller extent. The topics covered
include historical aspects, dark adsorption of the dye on the semiconductor surface and its role in the subsequent photoreaction, semiconductor preparation details, photoreactor configurations, photooxidation
kinetics/mechanisms and comparison with other Advanced Oxidation Processes (e.g., UV/H2 O2 , ozonation, UV/O3 , Fenton and photo-Fenton reactions), visible light-induced dye decomposition by sensitization
mechanism, reaction intermediates and toxicity issues, and real-world process scenarios.
© 2008 Elsevier B.V. All rights reserved.
Contents
1.
2.
3.
4.
5.
6.
7.
8.
Introduction and scope of review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
Categories of dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Evolution of the field of heterogeneous photocatalytic treatment of organic dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dark adsorption of the dye on the semiconductor surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dark adsorption of the dye and photocatalytic reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photocatalyst details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1.
Titanium dioxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2.
Other semiconductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photoreactor configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photodegradation of organic dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.1.
Process variants and mechanistic aspects involving light absorption by semiconductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.2.
Comparison with Advanced Oxidation Processes and combination with sonolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.3.
Photosensitized conversion of organic dyes in visible light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.4.
Photoreactions at the solid/air interface or in the gas phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.5.
Photoreaction kinetics, variables, and related aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.6.
Photoreaction intermediates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.7.
Toxicity studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Into the real world: tests with dye wastewaters, process scale-up and economic aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
∗ Corresponding author. Tel.: +1 817 272 3492; fax: +1 817 272 3511.
E-mail address: rajeshwar@uta.edu (K. Rajeshwar).
1389-5567/$20.00 © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochemrev.2008.09.001
172
173
173
173
175
177
177
177
180
180
181
181
183
183
184
184
186
186
189
172
9.
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Krishnan Rajeshwar was born in Trivandrum, Kerala State
in India. After completing his undergraduate and masters degrees in chemistry at the University College and
Indian Institute of Technology respectively, he specialized
in solid-state chemistry for his PhD degree at the Indian
Institute of Science in 1974. After a post-doctoral stint in
Canada (St. Francis Xavier University, Nova Scotia), he was
a senior research associate at Colorado State University
where he began working on energy and environmental research problems. He began his academic career at
the University of Texas at Arlington in 1983 where he is
currently a Distinguished Professor of Chemistry & Biochemistry and Associate Dean in the College of Science. His research interests span
a broad spectrum in energy, environmental, and materials chemistry problems.
Marly Eiko Osugi was born in São Paulo, Brazil in 1980.
She obtained her PhD in chemistry from Institute of Chemistry of University of São Paulo State, UNESP, Brazil in
2008 under the supervision of Prof. M. V. B. Zanoni.
She further developed her research at The University of
Texas at Arlington (2006–2007) under the supervision of
Prof. K. Rajeshwar. Her main research interest is on the
electrochemistry of dyes and degradation of dyes by photoelectrochemical and electrochemical processes.
Wilaiwan Chanmanee was born in Supanburi, Thailand in 1981. She graduated from the National Center
of Excellence for Environmental and Hazardous Waste
Management, Chulalongkorn University, Thailand with
PhD in 2007. Her work was under the joint supervision
of Prof. Krishnan Rajeshwar and Assoc. Prof. Puangrat
Kajitvichyanukul. Currently, she is doing a post doctoral
stint at the Department of Chemistry and Biochemistry,
University of Texas at Arlington. Her main research is on
the fabrication of titania nano-materials using electrochemical techniques.
C.R. Chenthamarakshan was born in Pala, Kerala, India
in 1969. He received his MS in chemistry in 1991 from
Mahatma Gandhi University, Kerala, India and his PhD
in chemistry in 1997 from Regional Research Laboratory,
Council of Scientific and Industrial Research (RRL-CSIR),
The University of Kerala, Trivandrum, India under the
supervision of Dr. A. Ajayaghosh (RRL-CSIR). After a one
year postdoctoral research associate appointment at RRLCSIR he was awarded a post doctoral position in 1998
in the group of Prof. Krishnan Rajeshwar, Department of
Chemistry and Biochemistry, The University of Texas at
Arlington (UT Arlington), Arlington, Texas (USA). Subsequently he was appointed as a research assistant professor at UT Arlington and from
2007-to date he is working as a senior research associate at Corsicana Technologies,
Corsicana, TX (USA). His research interests include semiconductor photoelectrochemistry, photocatalysis, conducting polymers, and fatty amine chemistry.
Maria Valnice Boldrin Zanoni was born in Tupi Paulistacity, São Paulo, Brazil in 1957. She graduated from Institute
of Chemistry of São Paulo University (USP) with PhD in
physical chemistry, 1989. Her academic position started
at University of São Paulo State (UNESP) as an associate professor in analytical chemistry since 1987. Her
main research interests are in organic electrochemistry
and environmental electrochemistry, investigating analytical methodologies and new treatment methods for dye
effluents. Recently, she expanded her research interests
to photoelectrochemistry using nanoporous, nanotubular and nanocomposite semiconductor electrodes for dye
degradation.
189
190
190
Puangrat Kajitvichyanukul was born in Chiang Mai, Thailand in 1970. She graduated from Department of Civil and
Environmental Engineering of The University of Texas at
Arlington with PhD in 2002. Her academic position started
at King Mongkut’s University of Technology Thonburi in
Bangkok (Thailand) as a lecturer (1996–2003), as an assistant professor (2003–2006), and as an associate professor
(2006–present). Her main research interests are on the
synthesis and characterization of TiO2 to be used as a
photocatalyst for environmental, biological, and health
protection applications. She was a recipient of Distinguished Alumni Award of Chiang Mai University, Thailand
in 2008.
Rebecca Krishnan-Ayer is a senior in high school at the
Hockaday School in Dallas, TX, USA. She was involved in
this project as a summer intern at the University of Texas
at Arlington. She is now looking at Universities starting
Fall, 2009 and hopes to complete an undergraduate degree
combining writing, art history, and science.
1. Introduction and scope of review
Millions of various colored chemical substances have been generated within the last century or so, 10,000 of which are industrially
produced [1]. On a global scale, over 0.7 million tons of organic
synthetic dyes are manufactured each year mainly for use in the
textile, leather goods, industrial painting, food, plastics, cosmetics, and consumer electronic sectors. A sizable fraction of this is
lost during the dying process and is released in the effluent water
streams from the above industries. Therefore, decolorization and
detoxification of organic dye effluents have taken an increasingly
important environmental significance in recent years [2,3].
Most conventional methods for the removal of dye pollutants
such as adsorption on activated carbon, ultrafiltration, reverse
osmosis, etc. are non-destructive and merely transfer pollutants
from one phase (for example, aqueous) to the other (for example, adsorbent). Biodegradation is slow and inefficient for many azo
dyes and does not work at all for others (for example, Acid Orange 7).
Chlorination and ozonation are also relatively inefficient and have
high operating costs. Thus, Advanced Oxidation Processes (AOPs),
which rely on the generation of hydroxyl and other radical species
for environmental remediation, have been successfully deployed
for the treatment of organic dye-laden waters [3–6].
This review focuses on one such methodology within the AOP
category, namely heterogeneous photocatalysis using an inorganic
semiconductor as photocatalyst [3]. Fig. 1 constructed from a
sampling of ca. 300 papers in the journal literature over a ∼35
year period, shows an almost exponential growth, attesting to the
popularity of this remediation approach. In many instances, as elaborated below, complete mineralization of the dye can be achieved
under mild process conditions. The process economics can be further improved if sunlight is used (instead of UV light) for excitation
of the semiconductor photocatalyst. Importantly, both oxidative
and reductive conversion of the organic dye (into benign and colorless products) is possible using this approach [7,8].
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Fig. 1. Evolution of the organic dye photocatalysis literature over approximately
three decades.
Interestingly, the publications on this topic originate from
countries all across the globe as Fig. 2 illustrates. Perhaps not
surprisingly, China and India dominate in this regard and both
economically advanced nations (for example, USA, Japan, Canada)
as well as emergent economies such as Brazil and Taiwan feature
prominently in Fig. 2.
To our knowledge, this is the first review article dealing with the
heterogeneous photocatalytic treatment of organic dyes. Previous
authors have discussed the kinetics and mechanistic aspects of this
process for the specific instance of azo dyes in aqueous solutions [9].
Many review articles also exist on the heterogeneous photocatalytic
treatment of organic pollutants in general [8,10–20] and limited discussions on dyestuffs are contained in some of these. We also note
the related use of electrochemical treatment of textile dyes and dyehouse effluents [3], a topic not further addressed in detail herein.
Because of the plethora of review articles, book chapters, and
books on heterogeneous photocatalysis [3–5,8,10–20], familiarity
with the underlying principles of this treatment approach will be
assumed in what follows. Finally, the literature sampling is representative rather than comprehensive and is almost exclusively
limited to journal articles rather than conference proceedings and
patents. This choice is deliberate in that we believe refereed journal articles generally contain credible data that have already been
vetted by the peer review process.
173
Fig. 3. Types of dyes featured in the studies on organic dye photocatalysis.
2. Background
2.1. Categories of dyes
Colorants, or additive substances causing variation in color or
visible light absorption, can be divided into two categories: dyes
and pigments. The distinct delineation between a dye and a pigment
should be noted. Dyes are soluble or partly soluble organic (carbonbased; plant and animal extracted) colored compounds suspended
in a medium, and represent one type of colorant [21]. The process
of dyeing can be loosely defined as imparting color to textile fiber
or leather. Pigments, on the other hand, typically are completely
insoluble substances that have no chemical affinity for the substrate
to be colored.
Dyes can be classified on the basis of structure, function, or both.
Table 1 contains a representative listing of dyes grouped according to their chemical structure. Also listed for each example in
Table 1 is the color index, which contains reference numbers on
the basis of the color and chemical classification [1]. Dyes can
also be classified as acid, basic, direct, disperse, reactive, anionic,
cationic, etc. and indeed this notation is often simultaneously used
with the dye chemical structure type, for example, Basic Blue 41
and Acid Yellow 23 are both monoazo dyes. Of the synthetic dyes
manufactured today, azo compounds are dominant (∼50–70%) with
anthraquinone dyes being a distant second. The dominance of the
azo dyes is reflected in the heterogeneous photocatalysis literature
as well, Fig. 3.
2.2. Evolution of the field of heterogeneous photocatalytic
treatment of organic dyes
Fig. 2. Countries of origin of the publications sampled in this review on organic dye
photocatalysis.
The early photochemical studies on aqueous solutions of organic
dyes in the presence of inorganic semiconductors such as silver
halides, ZnO, and the like were driven by applications in electroand photography (i.e., dyes were used as visible light sensitizers for
electro- or photographic materials, respectively) [22,23]. The first
instance for observing dye instability in the presence of an inorganic semiconductor (TiO2 ) and illumination appears to be in 1969
when the photocatalytic reduction of Methylene Blue (a thiazine
dye, Table 1) to the leuco form was reported [24,25]. This was subsequently followed by other studies which reported N-dealkylation of
dyes such as Rhodamine B and Methylene Blue in aqueous suspensions of CdS [26]. The N-dealkylation was preceded by dye radical
cation formation as a result of photoinduced electron transfer from
the dye to the semiconductor conduction band.
174
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 1
Representative organic dyes considered in this review.
Type of dye
Example
Color index
Azo
Reactive Orange 16
C.I. 17757
Xanthene
Basic Violet 10
C.I. 45.170
Thiazine
Methylene Blue
C.I. 52015
Anthraquinone
Rective Blue 4
C.I. 61205
Triphenylmethane
Basic Violet 4
C.I. 42.600
Phthalocyanine
Reactive Blue 15
C.I. 74459
Indigo
Indigo Carmine
C.I. 73015
Quinoline
D&C Yellow 10
C.I. Acid Yellow 3
Phenanthrene
D&C Green 8
C.I. Acid Green 9
Deliberate attempts to photochemically destroy the organic
dye appear to have been instigated only almost a decade later
when photoinduced electron transfer from TiO2 to Methyl Orange
(a monoazo dye, Fig. 4) was observed to result in bleaching of
the dye absorption (max = 470 nm) and its reductive conversion
to a hydrazine derivative [27,28]. Importantly, no dye bleaching
was observed in the absence of TiO2 or visible radiation alone
Structure
( > 400 nm) indicating that the photocatalysis process involved
initial light absorption (max = 310 nm) with the concomitant generation of e− –h+ pairs in the semiconductor.
These early studies were accompanied soon by findings in
other laboratories which showed that many organic compounds
could be decomposed in aqueous media with a combination of
TiO2 and near-UV light [3–5,8,10–20]. The photocatalytic oxida-
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
175
(Ceq ) according to the Langmuir adsorption isotherm:
Nads =
Ns KL Ceq
1 + KL Ceq
(1)
In Eq. (1) Ns is the total number of accessible adsorption sites and
KL is the adsorption constant (in M−1 ). From the linearized form of
Eq. (1):
Ceq
Ceq
1
=
+
Nads
KL Nmax
Nmax
(2)
Nmax (the maximum, saturation dye coverage) and KL can be
determined from the slope and intercept, respectively. Adsorption
isotherms (according to Eqs. (1), (2) or both) appear in Refs. [58–62]
for Methylene Blue and in Refs. [32,34,39,44,46,47,49,50,64] for
azo dyes. Unfortunately, the results of these analyses are not readily comparable because of the variant units employed for Nads
and Ceq and also because of the differing nature of the TiO2 surface in the various studies. However, the KL values are generally
high (in the 102 to 105 M−1 range) signaling strong interaction of
the dye with the oxide surface. The parameter KL has also been
reported to increase with a decrease in the adsorbent particle
size because of a higher driving force for adsorption on the finer
particles [68].
The shape classification of the adsorption isotherms [69] has
been used to deduce, from the L-shape usually observed, that there
is no strong competition between the solvent molecules and the dye
to occupy the adsorbent surface sites [49,50]. On the other hand, the
disparity between Nmax values obtained from the above isotherm
data analyses (∼10−5 mol g−1 ) and the number of adsorption sites
on the oxide surface (∼10−4 mol g−1 ) has often led to conclusion
that the adsorbed dye molecules are surrounded by many anchored
water molecules.
The value of the dimensionless separation factor (RL ) indicating
the shape of the Langmuir isotherm [70]:
Fig. 4. Four representative dyes in the photocatalysis literature. See also Table 1.
tion of dyes such as Methylene Blue, Rhodamine B, Fluorescein
and Methyl Orange was reported thereafter [29,30]; however the
field legitimately “took off” only in the late 1980s and early 1990s
(Fig. 1).
The preceding discussion should have indicated that a multitude of mechanisms exist for dye decomposition involving both
oxidative (electron transfer from the dye) or reductive (electron
transfer to the dye) routes. The light absorption can occur in the
dye or by the semiconductor or both. These mechanistic aspects
are addressed later in this review (Section 7.1).
RL =
1
1 + KC0
(3)
(where C0 is the highest initial dye concentration) has also been
used in some studies to indicate the strength of the adsorption
process [47]. Values for RL (0 < RL < 1) diagnose the adsorption to be
favorable, unfavorable (RL > 1), linear (RL = 1), or irreversible (RL = 0).
Generally, values for RL in the 0–1 range have been reported [47].
While the Langmuir isotherm [71] has been overwhelmingly
popular in the dye adsorption studies, the applicability of the Freundlich isotherm [72] model has also been considered in several
studies [44,62]. This empirical model has been used for non-ideal
sorption that includes heterogeneous surface energy systems and
is embodied in the equation:
1/n
(4)
Nads = KF Ceq
3. Dark adsorption of the dye on the semiconductor surface
Dye adsorption is studied by equilibrating various concentrations of the dye solution with the powder or colloidal suspensions
of the oxide semiconductor particles in the dark for time periods ranging from a few minutes (∼30 min) to several hours
depending on the kinetics of the adsorption process. The supernatant solutions are then spectrophotometrically sampled for
the amount of dye adsorbed on the semiconductor surface.
Tables 2 and 3 summarize the results of such measurements for
a variety of dyes. Titanium dioxide was the semiconductor used
in the overwhelming majority of these studies with exceptions as
noted.
Dark adsorption data are processed in the form of a plot of concentration adsorbed (Nads ) versus the equilibrium concentration
where KF is the adsorption constant and 1/n is the adsorption intensity. The magnitude of 1/n gives an indication of the favorability of
adsorption (similar to the RL parameter above).
A third isotherm (Sips adsorption model) combines some of the
features of the classical Langmuir and Freundlich models [73]:
1/ns
Nads =
Nmax Ks Ceq
1/ns
1 + Ks Ceq
(5)
where Ks is the adsorption constant and 1/ns is the Sips parameter
related to the adsorption intensity.
For Methylene Blue and Rhodamine B adsorption on TiO2 –SiO2
mesoporous composites, the Langmuir model provided the best
fit of the data trends [62]. The Langmuir model fitted the data
better for Reactive Yellow 145 while the Freundlich model was
176
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 2
Representative studies on azo dye adsorption on the semiconductor surfacea in the dark.
Entry no.
Azo dye
Comments
Reference(s)
1
Solvent Red 1
[31]
2
Acid Orange 7
3
4
Acid Orange 7
Acid Orange 7
5
6
Acid Orange 7
Acid Orange 20
Other aspects of solution spectra (than the peak maximum shift) altered such as fine
structure and wavelength range of tail.
Both TiO2 and Al2 O3 surfaces studied. Value of adsorption constant (K) from isotherm
data analyses reported as 2.48 × 102 M−1 for adsorption on TiO2 .
Influence of TiO2 dose on dye adsorption studied.
Detailed study using FT-IR, diffuse reflectance and adsorption isotherms revealing
adsorption mode of dye on TiO2 and ZnO surfaces. Adsorbed dye exists as hydrazone
tautomer.
Another FT-IR study of dye adsorption.
Adsorption of dye on TiO2 studied by FT-IR and found to be essential for the
photocatalytic and photosensitized degradations to be effective.
Insignificant (<4%) dye adsorption on TiO2 noted.
Similar study as in Ref. [37] by same group.
The effect of Au loading on TiO2 surface on dye adsorption studied.
Effect of pH on dye adsorption described.
K = 2 × 103 M−1 . Increased amount of dye adsorbed on Fe(III)-modified TiO2 surface.
Effect of pH noted.
Effect of pH noted.
Effect of pH noted.
Adsorption isotherms presented. Perhaps one of the most comprehensive studies on
dye adsorption (see text).
Adsorption isotherm studied.
Mesoporous TiO2 films used.
Adsorption isotherm presented and effect of pH noted.
Adsorption isotherm presented.
Adsorption followed by FT-IR spectroscopy.
7
8
9
10
11
12
13
14
15
Acid Orange 20
Acid Orange 52b
Acid Orange 52b
Methyl Red
Acid Red 1
Reactive Black 5
Reactive Black 5
Reactive Black 5
Reactive Brilliant Red (X-3B)
16
17
18
19
20
Remazol Black B
Basic Blue 41
Direct Black 38
Safira HE XL
Amaranth
a
b
[32]
[33]
[34,35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
TiO2 was used as the semiconductor in these experiments except in Entry 4 when ZnO was also used.
Also known as Methyl Orange.
better for Reactive Black 5 adsorption on TiO2 [44]. Values of the
enthalpy, entropy, and free energy of adsorption are available for
Methylene Blue [62], Rhodamine B [62], and Acid Orange 7 [34]
on TiO2 and TiO2 –SiO2 composite surfaces. The adsorption kinetics were also studied for the Methylene Blue and Rhodamine B
dyes [62].
Oxide surfaces in aqueous media are amphoteric and TiO2 is
no exception [74]. The surface groups on titania are involved in
acid–base equilibria as follows:
(6)
Table 3
Representative studies on dye adsorption (non-azo dyes) on the TiO2 surface in the dark.
Entry no.
Dye category/dye
Comments
Reference(s)
Xanthene
1
Rhodamine B
In acidic media, adsorption is inhibited by electrostatic repulsion. Addition of an
anionic surfactant induces binding of the cationic dye molecules.
In acidic media, adsorption is inhibited by electrostatic repulsion. Addition of an
anionic surfactant induces binding of the cationic dye molecules.
Adsorption tunable via the positively charged diethylamine group or the negatively
charged sulfonate group of the dye molecule.
Adsorption of the anionic dye found to be a pre-requisite for its photocatalytic
degradation.
[52–54]
2
Rhodamine B
3
Sulforhodamine B
4
Eosin
Thiazine
5
[55]
[56]
[57]
[58]
Methylene Blue
Methylene Blue
Methylene Blue
Methylene Blue
Methylene Blue
Methylene Blue
Probably one of the earliest studies of dye adsorption on TiO2 in a photocatalysis
context. The photocatalyst was attached to glass tubing wound in a spiral.
Adsorption studied in a batch-recirculated photoreactor with hollow fiber membrane.
Both monomeric and dimeric forms of the dye found to adsorb to the same extent.
pH effect on dye adsorption noted.
As in Entry 2.
Adsorption studied as a function of temperature.
See also Entry No. 10, Table 2.
Anthraquinone
12
13
14
Alizarin S
Alizarin Red
Acid Blue 25
As in Entry No. 10, Table 2.
Electrostatic aspects of dye adsorption discussed.
Adsorption found to be important in control of the dye photodegradation rate.
[41]
[63]
[64]
Arylmethane
15
Malachite Green
In the absence of surfactant, low dye adsorption on TiO2 surface. The role of both
anionic and cationic surfactants discussed.
[65]
Indigo
16
Indigo and Indigo Carmine
Adsorption on TiO2 monitored by FT-IR spectroscopy.
[66]
Cyanine
17
Astrazone Orange G
Effect of salt probed.
[67]
6
7
8
9
10
11
Methylene Blue
[59]
[60]
[61]
[55]
[62]
[41]
177
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
(7)
The point of zero surface charge (pzc) defines the solution pH
when the positively charged surface groups on the oxide are exactly
counterbalanced by the negatively charged groups. Values quoted
for the pHpzc for TiO2 vary from 5.1 to 6.8 (a value of 6.8 appears to be
most tenable) and for ZnO it is 9.0 ± 0.3. In view of the oxide surface
charge tunability as a function of solution pH, it is hardly surprising that dye adsorption on TiO2 (or ZnO) has been monitored as a
function of solution pH in numerous studies (see Tables 2 and 3).
In general, cationic dyes (e.g., Methylene Blue, Table 1) bind at pH
values greater than pHpzc while anionic dyes (that is, dyes bearing
sulfonate groups) show the opposite trend.
Other than the characteristics of the photocatalyst surface, solution additives can markedly influence the extent of dye adsorption
on the photocatalyst surfaces in the dark. Thus Fe(III) aquo ions
enhance azo dye adsorption [42] although the effect appears to be
related also to the charge borne by the dye molecule. The adsorption of anionic dyes (for example, Acid Orange 7) is promoted by
the presence of FeCl3 . On the other hand, inhibitory effects have
also been reported for anions and attributed to their competitive
adsorption on the photocatalyst surface [67].
A series of papers discuss the effect of added surfactant on
dye adsorption [52–54,56,65]. Thus an anionic surfactant such as
sodium dodecylbenzenesulfonate adsorbs strongly on the TiO2 particles and in turn, the bound sulfonate groups electrostatically
attract cationic dyes such as Rhodamine B or Malachite Green
[52–54,65]. Significantly, in the absence of surfactants, Malachite
Green is difficult to degrade in aqueous TiO2 suspensions because of
its poor solubility in water and low adsorption on the TiO2 surface.
Interestingly, even a cationic surfactant such as hexadecyltrimethylammonium bromide can be deployed at a basic pH (>6.8) when it
binds to the negatively charged TiO2 surface (Eq. (7)) [65].
A molecular level understanding of the mode of dye adsorption on the photocatalyst surface is afforded by FT-IR spectroscopy
and many examples of application of this spectroscopic tool in
dye adsorption studies are available [34,36,37,51]. A particularly
detailed probe of the adsorption mode is given for the azo dye, Acid
Orange 7, on TiO2 and ZnO surfaces [34]. Based on these FT-IR data,
a bi-dentate coordination mode is envisioned for the dye–TiO2 surface complex while only one oxygen atom from the dye sulfonate
group is postulated to be involved in binding with the ZnO surface
[34]. As discussed later in this review, FT-IR spectroscopy also furnishes useful insights into the course of the photocatalytic process
and the chemical nature of any intermediates generated.
4. Dark adsorption of the dye and photocatalytic reactions
The foregoing discussion begs the important question: what
role, if any, does adsorption of the dye on the photocatalyst surface
(in the dark) play in the efficacy of its subsequent photocatalytic
degradation? There is much literature on this issue for non-dye
substrates including many organic compounds and metal ions
[75–83]. At the outset, it may be noted that applicability of a
rate equation for the heterogeneous photocatalysis reaction such
as Langmuir–Hinshelwood mechanism (see below), can itself be
construed as strong evidence for substrate adsorption to be an
important element of the overall reaction mechanism. Further, the
use of high-surface area, inert adsorbents, such as the candidates
discussed earlier and activated carbon [84], silica gel [85] and zeolites [84,86] along with TiO2 for sequestering substrates prior to
their subsequent decomposition, contains an implicit recognition
of the important role that adsorption can play in the heterogeneous
photocatalysis process.
Table 4
Representative studies on the use of Degussa P-25 TiO2 for the photocatalytic degradation of various types of organic dyes.
Entry no.
Dye category
Reference(s)
1
2
3
4
5
6
7
8
9
Azo
Xanthene
Thiazine
Anthraquinone
Arylmethane
Phthalocyanine
Indigo
Cyanine
Phthalein
[31,32,34,39,42,43,45,46,47,49,51,88–104]
[53,54,56,57,92,98,105–108]
[59,60,61,109–112]
[63,64,89,108,113–115]
[65,108,116]
[47,89]
[66]
[117]
[29,118]
Given the above, it should not be surprising that many studies indeed confirm and reinforce the notion that adsorption of
dye molecules on the photocatalyst (for example, TiO2 ) is a prerequisite for photodegradation [31,41,46,49,58,64,65]. A corollary is
that the photoinduced reaction occurs on the catalyst surface rather
than in solution bulk. On the other hand, there are also isolated literature data which suggest that adsorption plays a negligible role
in the photodegradation of the dye [44,87].
5. Photocatalyst details
5.1. Titanium dioxide
On a laboratory scale, the most popular photocatalyst configuration has been in the form of powder suspensions. Commercially
available samples of TiO2 have been used in many of these studies and Tables 4 and 5 contain representative listings of these.
Table 6 lists the characteristics of these commercial TiO2 samples.
Head-to-head comparisons of their photcatalytic activity have also
been done in many cases (e.g., Refs. [45,118]). The trends are rather
complex and while Degussa P-25 works the best in many cases,
there are exceptions, for example, when either the Millenium or
the Hombicat samples outperform the benchmark photocatalyst.
Photocatalyst parameters such as the surface area and particle size
alone are not sufficient indicators of photocatalytic activity and
surface chemistry factors may also play a role. Of course, the foregoing comments are not exclusive to the case of organic dyes but
apply equally well to the photocatalytic conversion of other organic
compounds and metal ions.
The influence of crystallographic modification of the TiO2 sample appears to have been only sporadically examined. Generally, the
anatase modification is considered to be superior in its photocatalytic activity relative to the rutile counterpart. On the other hand,
the mixture of anatase (dominant form) and rutile modifications
in Degussa P-25 TiO2 has been touted [126] to play an important
Table 5
Representative studies on the use of commercial TiO2 samples (other than Degussa
P-25) for the photocatalytic degradation of various types of organic dyes.
TiO2 origin
Dye category
Reference(s)
Aldrich
Azo, cyanine, thiazine,
arylmethane, xanthene
Azo
Azo
Azo
Azo
Azo
Azo
Azo
?
Azo
Thiazine
Azo
[45,67,116,117,119,120]
Millenium (PC 50 and PC 500)
Merck
Hombikat (UV 100, Pt-UV 100)
DuPont
Wako Pure Chemical Industries
Riedel de Haen
Kerr-McGee
Fluka
Tranox A-K-1
STS-21 (Ishihara Sangyo)
Fujititan
[45]
[121–123]
[45]
[37]
[87]
[33]
[38]
[55]
[45]
[124]
[125]
178
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 6
Specifications and characteristics of commercial TiO2 samples.
Entry no.
TiO2 sample
BET surface area (m2 g−1 )
Average particle size (nm)
Composition
1
2
Degussa P-25
Millenium PC 500
PC 50
Aldrich
Kerr-McGee
Hombikat UV 100
Tranox A-K-1
Mikroanatas IF 9308/18
DuPont R-900
50
287
45
11
90
>250
90
271
NA
21
5–10
75% anatase, 25% rutile
Anatase
Anatase
Anatase
NA
Anatase
Anatase
Anatase doped with 0.1% (w/w) Pt
Rutile
3
4
5
6
7
8
a
NAa
20
5
20
NA
NA
NA = not available.
role in the excellent photocatalytic activity of this material although
this proposal must remain speculative pending further scrutiny. A
study on the characteristics of the Hombikat UV-100 photocatalyst
is also available [127].
Many studies on dye photodegradation by TiO2 involve the use of
TiO2 colloidal particles prepared by controlled hydrolysis of a precursor. Thus, titanium(IV) isopropoxide in 2-propanol is injected
into acetonitrile and stirred under N2 blanket [128]. This results
in a colloidal dispersion of TiO2 particles with an effective hydrodynamic diameter of 0.15 m [128]. This procedure was adopted
in many subsequent studies by the same group [129–132]. Other
variants of controlled hydrolysis start with titanium(IV) tetraisopropoxide, titanium(IV) butoxide or TiCl4 with details differing in
the solution mixing sequence, pH, etc.
Non-hydrolytic synthesis approaches have also been deployed
for TiO2 . This includes hydrothermal syntheses [133,134] and a
solution synthesis method based on reacting TiCl4 and titanium(IV)
tetraisopropoxide at 300 ◦ C in a mixture of heptadecane and tri-noctylphosphine oxide (TOPO) as a co-ordinating (capping) agent
[135,136]. Nanosized particles (∼7 nm) are “fished out” from the
reaction mixture and then re-dispersed in organic solvents. The
hydrothermal approach involves a combination of high temperature and pressure in a sealed container such that a constant
autogenerated pressure is generated within. Thermal hydrolysis is
yet another route to preparing TiO2 particles. Thus, sulfur-doped
TiO2 was prepared from titanyl sulfate solution by reacting with
ammonia at 85 ◦ C and controlled pH [137]. The product was repeatedly washed with de-ionized water till it was free of sulfate ions
and then dried prior to thermal anneal. The photocatalyst thus
prepared (along with samples also Pt-modified) were then used
for the UV-assisted photooxidation of Acid Orange 7 dye in aqueous media [137]. Another variant involves a “sol-solvothermal”
process wherein the polymer-stabilized TiO2 sol (prepared from
the tetraisopropoxide precursor as above) is subsequently heated
at temperatures in the range, 80–150 ◦ C, for several hours in a
stainless-steel autoclave [138].
The advantages accrued from immobilizing or “supporting” TiO2
particles onto a high-surface area support material were identified
in a preceding section and this is a lesson learnt from the heterogeneous (thermal) catalysis [139,140] community. Thus even early
studies examined the influence of support material on the photocatalytic activity of TiO2 [30,118,141] and a slew of subsequent
studies have utilized TiO2 nanoparticles supported on glass beads
[118], buoyant polystyrene beads [142], silica [44,143], clay [144],
polyoxometallate [145], alumina [141] and composite inert oxide
containing either Ru or Ln metal [146,147]. Invariably, solutionbased synthesis approaches (such as the ones outlined in the
preceding paragraph), were used to derive supported TiO2 composite photocatalysts.
Instead of using TiO2 alone, a “coupled semiconductor” configuration (Fig. 5) [148] improves charge separation in many
cases because of “vectorial” electron transfer [149] and reduced
carrier recombination as further elaborated below in the discussion on TiO2 photocatalyst films. However, coupled nanoparticles
of CdS and TiO2 (denoted as “CdS/TiO2 ”) have been synthesized via a sol–gel/precipitation technique [150]. Other studies
report “coupled” semiconductor powders (TiO2 + ZnO, TiO2 + SnO2 ,
ZnO + SnO2 ) [151,152] but it is not clear whether the two semiconductor particles are truly coupled in these cases. No details are given
on how the commercial powder samples were combined in Refs.
[151,152].
A logical extension to a supported photocatalyst configuration is
to completely immobilize the active material on an inert substrate
such as metal, polymer, cloth, etc. The photocatalyst layer is then
present as a film on the substrate surface. A major advantage here
is that the reaction product(s) and photocatalyst do not have to be
separated unlike in the cases with powder or colloidal suspensions
of the photocatalyst. The quantum efficiencies are also often low
in the latter because of carrier recombination at the photocatalyst
surface or in its bulk. A review is available on the strategy of fixing
or depositing TiO2 on a variety of polymer-based substrates and
the use of these materials in the photodegradation of azo dyes in
solution [153].
Perhaps the simplest approach is to make a slurry of TiO2
nanoparticles (for example, in water or alcohol) and then mix it
with an organic or polymeric binder (e.g., Nafion). The surface
targeted can then be dip-coated with this slurry, the number of dipcoating steps replicated to build up a requisite film thickness. Early
studies [29,58] have utilized a similar method to attach Degussa P25 TiO2 nanoparticles as a thin film to the inner wall of borosilicate
glass tubing wound in a spiral. Anatase TiO2 supported in fiberglass mesh or glass fiber cloth is available from commercial sources
[118]. Other examples of dip-coating Degussa P-25 or anatase TiO2
film on soda-lime glass supports, for applications relevant to dye
Fig. 5. Coupled semiconductor photocatalyst configuration.
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
photodegradation, are contained in Refs. [154,155]. Photoreactor
configurations will be discussed later in this review.
There are many examples in the dye photocatalysis literature for
colloidal TiO2 (sol–gel) coatings on various substrates [156–159]
including SiO2 -coated glass plates [156], glass slides [40,48,157],
quartz slides [158–161], titanium foil [158–164], glass reactor tube
or ring [165,166], borosilicate Petri dish [167], and indium tin oxide
(ITO) or TiO2 glass [168]. Commercial products comprising of colorless TiO2 layers coated on clear glass (for example, Pilkington
ActiveTM , Pittsburgh Plate and Glass or PPG) are also available
and have been used for the photooxidation of organic compounds
including organic dyes such as Acid Blue 9 and Reactive Black 5
[169,170].
The electronically conductive substrates amongst the candidates listed above (i.e., Ti, ITO) offer the crucial advantage of
facilitating the imposition of a bias potential across the oxide
semiconductor–solution interfere to drive the photocatalytic process. For n-type semiconductor, positive potentials result in the
interface being reverse biased, driving the photogenerated holes
to the interface and the electrons to the rear contact [171,172].
The result is improved e− –h+ separation and consequently, a
lower carrier recombination—a key handicap with the use of TiO2
powder suspensions or TiO2 nanoparticulate films on insulating
substrates (see above). This process is termed “photoelectrocatalytic” to distinguish it from its non-biased “photocatalytic” process
counterpart. Many examples of photoelectrocatalysis are available
in the literature on organic dyes [158–162,168], including studies on coupled semiconductor films involving TiO2 and another
semiconductor such as SnO2 [173,174], or even combinations of
three semiconductors, TiO2 /CdO–ZnO [175]. Generally, the coupled
semiconductor films have been deposited on optically transparent
electrodes such as ITO.
Electrochemical or electrophoretic deposition [148,176,177]
offers a versatile and low-temperature approach to coating TiO2 (or
composites) on conductive substrates. Both anodic [178,179] and
cathodic [180,181] approaches can be utilized for depositing TiO2
although only the anodic electrosynthesis variant appears to have
been deployed for applications related to the photoelectrocatalytic
(PEC) degradation of organic dyes. Thus anodization of Ti mesh in
aqueous media results in a microporous TiO2 film [182,183] but
under carefully tuned conditions can also facilitate the growth and
self-assembly of nanotubular arrays on the Ti surface [178,179]. Nanotube arrays facilitate vectorial charge transfer (Fig. 6) relative to
the tortuous path undergone by photogenerated electrons in nanoor mesoporous electrode film geometries (c.f., Fig. 6a and b). Thus
it is not surprising that such nanotubular TiO2 arrays have afforded
179
faster photoelectrocatalytic degradation rates for three model disperse azo dyes (Disperse Orange I, Disperse Red 1, Disperse Red
13) relative to the nanoporous electrode assemblies prepared via
colloidal deposition [184,185].
This brings up the important issues: how do powder suspensions stack up vis-á-vis supported TiO2 photocatalysts in terms of
process efficiency? How does the dark electrocatalytic dye oxidation process counterpart perform relative to its photo-variants?
Two sets of studies on non-dye substrates have concluded that
TiO2 powder suspensions and immobilized configurations perform
equally well [186,187]. On the other hand, the photocatalytic activity of TiO2 coated on glass (by the sol–gel method) is found to be
intermediate between that of Millenium PC-500 and Degussa P-25
TiO2 powders [41]. Five TiO2 candidates and three different types
of organic dyes were included in this particular study [41].
The PEC process was compared with the photocatalysis counterpart for the degradation of Rose Bengal dye [182]. The bias
potential served to accelerate the dye photooxidation relative to
all the photocatalysis cases where the TiO2 solution dose was varied (from 0.10% to 0.30%) except when the solution dose was 1.0%
[182]. Another study on TiO2 –Ti mesh (prepared by anodization, see
above) [183] found the PEC variant to be superior to UV photolysis, electrooxidation, and photocatalysis for the conversion of an azo
dye. Essentially similar trends have been observed for Remazol Brilliant Orange 3R (azo dye) [158], Reactive Orange 16 (azo dye) [159],
Remazol Turquoise Blue 15 (copper phthalocyanine dye) [160,162],
and Reactive Blue 4 (anthraquinone dye) [161].
Direct UV photolysis and the PEC and photocatalysis process
variants were also compared for SnO2 –TiO2 coupled semiconductor films for Naphthol Blue Black and Acid Orange 7 azo dyes
[173,174]. The PEC variant performed the best relative to the other
two approaches. The effects of increasing bias potential and of O2
or N2 on the photoelectrode (or photocatalyst) compartment were
also investigated in these studies [173,174]. As expected, the more
positive the bias potential is, the better the PEC process performs.
Unlike in the photocatalyis process where air (or O2 ) is needed to
scavenge the photogenerated electrons [8,188,189], the PEC process can be carried out in anaerobic conditions since the (dark)
reduction occurs in the counterelectrode compartment of the cell.
In electrophoretic deposition, the surface charge on the suspended particles in the solution drives them to the appropriately
polarized working electrode surface; this tactic is utilized, for
example, for solvent-less painting in the automobile industry.
A variant of this is occlusion electrodeposition [176,177] where
electrophoretic particle occlusion is combined with continual
deposition of a metal “glue” (for example, nickel) which serves to
Fig. 6. Schematic diagrams contrasting: (a) nanostructured and (b) nanotubular photocatalyst film configurations and vectorial electron transfer in the latter.
180
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 7
Use of inorganic semiconductors other than TiO2 for the photocatalytic or photoelectrocatalytic degradation of organic dyes.
Entry no.
Inorganic semiconductor
Optical bandgap (eV)
Organic dye(s) studied
Reference(s)
1
2
3
4
5
6
7
8
9
10
11
12
CdS
CdS
ZnO
ZnO
ZnO
ZnO
ZnO
ZnO
SnO2
WO3
WO3
WO3
2.4
2.4
3.2–3.4
3.2–3.4
3.2–3.4
3.2–3.4
3.2–3.4
3.2–3.4
3.5
2.5–2.8
2.5–2.8
2.5–2.8
Methyl Orange
Reactive Black 5
Rose Bengal
Acid Orange 7
Acid Orange 20
Reactive Black 5
Methyl Red
Procion Red MX-5B, Amaranth
As in Entry No. 8
As in Entry No. 6
Naphthol Blue Black
Methylene Blue
[27,28]
[87]
[120]
[34]
[37]
[43]
[136]
[150]
[150]
[43]
[197,198]
[199]
immobilize the occluded particles. More modest voltages are used
in this process variant relative to the pure electrophoretic approach.
Thus Degussa P-25 TiO2 nanoparticles were electrophoretically
deposited on stainless-steel substrates and used for the PEC decoloration of Methylene Blue [190]. A similar approach was used in
another study along with carbon black to prepare composite films
on conductive glass (ITO) [191]. These films were then used for the
decoloration and degradation of Acid Orange 7 azo dye [191]. Silver
was used as the glue to occlude TiO2 nanoparticles and these composite films were used on Brilliant Green dye [192]. A -PbO2 film
was electrodeposited along with the occlusion of TiO2 nanoparticles to yield a -PbO2 –TiO2 composite film [193,194]. The efficacy
of these materials was established with Acid Orange 7 azo dye
[193,194]. Rounding out the electrosynthetically prepared photocatalysts is a study on anodic alumina (AAO) membranes which
were used as a template to house colloidal TiO2 nanoparticles [195].
These immobilized photocatalysts were used for the conversion of
Direct Black 168 dye [196].
We end this section on TiO2 photocatalyst with a brief mention of a potpourri of other preparative methods which include
magnetron sputtering, chemical vapor deposition or CVD, electron
beam evaporation and metal organic CVD or MOCVD. Titania films
synthesized by these methods were used for the photodegradation
of dyes such as Rhodamine B [196].
corrosion (see Section 7.7). Thus, of the photocatalysts that appear
in Table 7, TiO2 , WO3 , and SnO2 are immune from photocorrosion
while ZnO and CdS undergo photoanodic corrosion under bandgap
illumination [13].
In most of the studies considered in Table 7, commercial samples
of the photocatalysts were deployed. WO3 films were prepared by
cathodic electrodeposition [197,198] while they were deposited on
glass substrates by RF magnetron sputtering in another study [196].
A high PEC activity was observed for the electrodeposited WO3 film
electrodes than for TiO2 nanoparticulate films for the degradation
of Naphthol Blue Black dye [198]. Finally, anodic nanoporous WO3
films have shown greater photocatalytic activity for Methylene Blue
than their cathodically electrosynthesized counterparts [199]. A
crucial advantage with WO3 (relative to TiO2 ) is that its lower optical bandgap (Table 7) results in a much greater utilization of the
solar spectrum for solar photocatalysis or solar photoelectrocatalysis applications.
In summation of the results on the other semiconductors, comparisons of photocatalytic activity between different semiconductor
materials may be confounded by variables related to their morphology. We have seen how important this is even for a single
semiconductor (TiO2 ) which shows wide variations in photocatalytic activity for various dyes.
6. Photoreactor configurations
5.2. Other semiconductors
After TiO2 , ZnO perhaps is the most studied inorganic semiconductor in the dye photocatalysis community. We have already seen
(in Section 2.2) that ZnO actually preceded TiO2 as a semiconductor
of choice in the dye sensitization field. Therefore, it is not surprising
that early studies on photooxidative degradation of organic compounds via sensitized charge transfer by a dye (a process variant
elaborated later) used ZnO as the inorganic semiconductor (e.g.,
Ref. [120]).
Table 7 provides a listing of studies on ZnO and other semiconductors. Head-to-head comparisons of ZnO and TiO2 yield variable
results. Thus ZnO was seen to outperform Degussa P-25 TiO2 in the
photodegradation of Reactive Black 5 [43]. On the other hand, in
terms of CO2 formation as a function of time, ZnO was slower than
Degussa P-25 TiO2 [43]. Interestingly, ZnO outperformed Degussa
P-25 TiO2 in the UV-assisted photocatalytic conversion of Acid
Orange 22 but the trend was reversed in the photosensitization
process [37]. For azo dyes, ZnO is reported to be better than TiO2 in
some studies (e.g., Ref. [150]).
A comparison between TiO2 and CdS photocatalysts also appears
for a reactive azo dye [87]. The photocatalytic conversion of Reactive Black 5 obeyed Langmuir–Hinshelwood equation (see below)
for CdS but not for TiO2 [87]. The dye wastewater was decolorized
in both cases but the toxicity increased for CdS because of photo-
The simplest photoreactor design, at a laboratory scale level,
is the immersion well type schematically shown in Fig. 7a
[45,200–202]. Provisions are made in this set-up for flushing the
solution with air or inert gas (for example, N2 ) as needed and also
for withdrawing solution aliquots periodically for analyses. In situ
measurements within a diode array spectrophotometer can also
be performed with a photocell design consisting of an optically
transparent electrode substrate for the (transparent) photocatalyst
film [203]. The electrical connections for applying bias potentials
to the working electrode are incorporated into this design and the
illumination light source is placed orthogonal to the light beam
path of the spectrophotometer. Diffuse reflectance spectroscopy
measurements have been performed [31,141,173,174] with similar
cell designs. Fig. 7b illustrates a twin-compartment cell for photoelectrocatalysis experiments [162,184,185]. A single compartment
cell design has also been used [182] although this is less desirable
because of possible interference from the products generated at the
counterelectrode.
Batch recirculation is employed in many dye photocatalysis
studies and this can be done with either powder photocatalyst
suspensions or immobilized photocatalyst configurations. In the
former case, a membrane-based particle separation step can also
be incorporated for photocatalyst reuse [59,111]. A popular configuration involves the photocatalyst immobilized on a fabric or
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Fig. 7. Schematic diagrams of: (a) an immersion well photoreactor design and (b)
twin-compartment electrochemical cell for photoelectrocatalysis. See Ref. [162] for
identification of the set-up components in (b).
fiberglass mesh wrapped around the inner wall of borosilicate glass
tubing. The UV lamp (usually medium-pressure Hg arc) is housed
in a quartz inner tube such that the water to be treated flows
in the annular space between the lamp housing and the (illuminated) photocatalyst surface. This design can be deployed both
for batch recirculation and for continuous flow modes—the latter being amenable for pilot-scale commercial set-ups. Examples
(with minor variations in design details) may be found in Refs.
[29,58,118,166] in applications related to a variety of organic dyes.
The shallow pond or dish photoreactor is another arrangement
compatible with both slurry and immobilized photocatalysts; fluid
recirculation can be easily built into this photoreactor design. An
example for organic dye degradation applications may be found
in Ref. [30]. The use of naturally buoyant TiO2 -coated beads is an
intriguing approach that was first touted for the solar treatment
and remediation of oil spills [205]. A similar design incorporating immobilized TiO2 on polystyrene beads has been combined
with a pulsed baffled tube photochemical reactor and used for the
photocatalytic mineralization of Methylene Blue [142].
A study incorporating a 1 L plunging tube laboratory photoreactor has been compared with results from field experiments
performed at the Solar Platform in Almeria, Spain [41]. Two types
of photoreactor designs available at this facility were used [41,204].
One is a cascade falling film photoreactor with the photocatalyst (for example, TiO2 ) supported on non-woven inorganic fibers.
181
Fig. 8. (a) Photooxidation and (b) photosensitized mechanisms for organic dye photocatalysis. Note that light absorption occurs in the semiconductor in (a) and by the
dye in (b).
The second is a compound parabolic collector photoreactor for
experiments and demonstrations with powdered photocatalyst
suspensions. A variety of dyes (thiazine, azo, anthraquinone) were
included in the study described in Ref. [41].
7. Photodegradation of organic dyes
7.1. Process variants and mechanistic aspects involving light
absorption by semiconductor
Consider first, light absorption by the inorganic semiconductor with TiO2 as a prototype (Fig. 8a). When photons with energy
(h) equal to or greater than the semiconductor bandgap (Eg ) are
incident, electrons (ecb − ) and holes (hvb + ) are generated in the conduction and valence band, respectively [3–5] (Eq. (8)):
TiO2 + h → hvb + + ecb −
(8)
The photogenerated holes that escape direct recombination (Eq.
(9)):
hvb + + ecb − → heat
(9)
diffuse to the TiO2 surface (or migrate, if there is an imposed electric
field) and react with surface adsorbed hydroxyl groups or water
molecules to form trapped holes, Eq. (10):
hvb + + ≡ TiIV OH → [≡ TiIV OH• ]+ → ≡ TiIV O• + H+
(10a)
182
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
hvb + + TiO2 (H2 O) → TiO2 (• OH) + H+
(10b)
The trapped holes may be regarded as surface-bound hydroxyl
radicals. The bound radicals can also diffuse away from the surface
toward the solution bulk and exist transiently as free • OH. Either
the surface bound or free hydroxyl radicals can react with adsorbed
or free dye molecules to oxidatively decompose the latter [3–5].
A preponderance of literature evidence (albeit not much of it on
organic dyes, see below) exists for this mechanistic picture based
on detection of hydroxylated products and intermediates and also
spin-trapping with subsequent EPR detection [206–208].
Complicating the sequence of oxidative steps identified above is
direct oxidation of the dye by the photogenerated holes. This pathway is thermodynamically feasible because the redox potential of
the dye frequently lies above the semiconductor (e.g., TiO2 ) valence
bandedge. The quantum yields for • OH and hole generation have
been estimated (7 × 10−5 and 5.7 × 10−2 , respectively) [209] and
this suggests that the direct hole transfer pathway should dominate. However, the • OH mediated oxidation route will be important
if the dye is adsorbed on the semiconductor surface. Indeed, which
of the two competing mechanisms prevail may well depend on the
particular organic compound and the experimental conditions. A
study on the use of • OH radical or hole scavengers on the TiO2 assisted UV photodegradation of Acid Orange 7 [210] indicated
that the direct hole transfer pathway played a major role. We have
already seen that in cases where • OH mediated oxidation is dominant, whether the reaction occurs on the semiconductor surface or
in the solution bulk, depends crucially on the particular dye.
What happens to the photogenerated electrons in Reaction (8)
that escape subsequent recombination? They can be trapped at
the titanium lattice sites to generate Ti3+ defects (color centers).
Recall that photocatalysis usually occurs in an oxygenated aqueous
medium so that the following chain of reactions can occur:
∼40–50 times smaller than that for the photoreduction of Methyl
Orange by ecb − in CdS suspensions [27,28]. This reduction was promoted by a hole scavenger such as ethylenediamine tetraacetic acid
(EDTA) from amongst several electron donors tested [27,28].
The typical UV flux near the terrestrial surface of the earth
is ∼20–30 W/m2 in the 300–400 nm wavelength range. Unfortunately a semiconductor such as TiO2 (with an optical bandgap of
3.0–3.23 eV, absorption cut-off: ∼380 nm) harnesses only a small
fraction (∼5%) of the entire solar spectrum. For practical considerations, the electrical costs associated with driving a UV lamp
(see below) will have to be factored into the overall process economics [212] so that making the process amenable to using sunlight
(by using a lower bandgap oxide semiconductor or doping TiO2 )
becomes eminently attractive. Thus it is not surprising that many
studies as exemplified by Refs. [37,97,101,109,116,118] (see also the
review on azo dyes, Ref. [9]) on the photocatalytic conversion of
organic dyes have been performed in natural or simulated sunlight.
It is pertinent to mention some cautionary remarks here on
the use of light sources and UV cut-off filters (as suggested by a
reviewer). Not all the UV component is cut-off with the use of filters and many “visible light” sources (e.g., tungsten halogen lamp)
contain a UV component in their output such that the resultant
data may lead to erroneous conclusions. It is therefore important
to establish the wavelength response of a particular photocatalysis
procedure (via photoaction spectroscopy) before conslusions are
drawn.
As also reviewed in Ref. [9] for azo dyes, solution additives can
significantly influence the photooxidation of the organic dye. Thus
additives such as H2 O2 and persulfate (S2 O8 2− ) can exert a dual
function, first as strong oxidants themselves (see Reactions (13)
and (14)) and Reactions (17) and (19):
S2 O8 2− + ecb − → SO4 2− + SO4 •−
(17)
(18)
ecb + O2(ads) → O2 •−
(11)
SO4 •− + H2 O → SO4 2− + • OH + H+
−
+ ecb + 2H → H2 O2
(12)
SO4
O2 •− + H2 O2 → • OH + OH− + O2
(13)
H2 O2 + ecb →
H2 O2 + h → 2• OH
(14)
Second, as Reactions (17) and (20) show, these additives can
function as electron scavengers minimizing electron–hole recombination at the semiconductor surface and thereby improving the
quantum yield for • OH and hvb + production.
Consistent with the above mechanistic picture, negligible chemical oxygen demand (COD) abatement was found for the textile dye,
Orange 7, using TiO2 in the absence of S2 O8 2− while over 70% COD
abatement was obtained with a combination of TiO2 , potassium
persulfate, and sunlight [101]. Other studies attest to the beneficial effect of H2 O2 and (NH)4 S2 O8 addition for reactive azo dye
decolorization with TiO2 and sunlight [97].
Too high a peroxide level can have an adverse effect on dye
photooxidation. This is because H2 O2 is also a scavenger of semiconductor valence band holes and • OH:
−
O2
•−
+
We are ignoring back-reactions, for example:
ecb − + ≡ TiIV O• + H+ → ≡ TiIV OH
(15)
in the above scheme. Incidentally, Reaction (15) may be regarded
as a trap-mediated recombination of the electron with a surfaceimmobilized hole. Species such as peroxide and the superoxide
radical (Reactions (11) and (12)) are important in AOPs and in
photosensitized dye degradation as elaborated later. Evidence for
H2 O2 generation under UV irradiation of oxide semicondcutors in
oxygenated aqueous media was presented even in the early photocatalysis literature, albeit not in the context of organic dyes [211].
The organic dye can also be involved in an irreversible reductive photoreaction pathway involving the photogenerated electrons
in the semiconductor conduction band. Thus laser flash photolysis data show evidence for formation of the radical anion of
Phenosafranin dye in colloidal TiO2 and CdS suspensions [129] consistent with a one-electron reduction process:
ecb − + D → D•−
(16)
where D is the dye molecule. Bandgap excitation of WO3 and TiO2
suspensions was shown to result in irreversible reduction of two
azo dyes, Naphthol Blue Black and Disperse Blue 79 via a trapped
electron pathway [132]. The reaction between the dyes and trapped
electrons was reported to be diffusion-limited with rate constants
of 1.1 × 108 and 4.0 × 107 M−1 s−1 for the above two dyes, respectively [132]. The rate constant for Reaction (11) was found to be
•−
+ D → SO4
−
2−
• OH
+ oxidized products of dye
+ OH
−
(20)
H2 O2 + 2hvb + → O2 + 2H+
H2 O2
+ • OH
HO2 +
•
• OH
→ H2 O + HO2
→ H2 O + O2
(19)
(21)
•
(22)
(23)
Thus as also pointed out in Ref. [9], the reported influence
of oxidant additives such as S2 O8 2− and H2 O2 can be conflicting
depending on the dose and the specific experimental conditions.
Other chemicals such as Na2 CO3 and NaCl are inherent in the
dyeing process itself. For example, Na2 CO3 is added to adjust the
pH of the dyeing bath—a crucial variable for fixing the dye onto the
fabric and for color fastness. Sodium chloride is mainly used for
transferring the dyestuff to the fabric. Therefore, the dye industry
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
wastewater has the propensity to contain a considerable amount
of CO3 2− and Cl− making an examination of the influence of these
species on the photooxidative dye degradation quite relevant. Both
species can scavenge • OH and hvb + :
CO3 2− + • OH → OH− + CO3 •−
HCO3
−
−
+ • OH
+
Cl + hvb →
→ H2 O + CO3
•−
Cl•
Cl• + Cl− → Cl2 •−
(24a)
(24b)
(25a)
(25b)
Other anions such as SO4 2− and phosphate also have an
inhibitory effect [213]. Interestingly, however, no adverse effects
were observed in chloride- and sulfate-containing media on the
extent of color removal and mineralization of a reactive textile azo
dye as long as the pH was carefully controlled in the acidic range
[159]. Cations such as Fe3+ can have a beneficial effect because of
the possibility of Fenton reactions consuming H2 O2 in the aqueous
phase [214]:
183
UV/H2 O2 relies on Reaction (14) with H2 O2 absorbing mainly in
the deep UV region (wavelengths shorter than ∼200 nm) [3,218].
Thus while the reaction has been attempted in sunlight because of
process energy efficiency considerations (see above), the yield of
• OH will be very low.
The use of ozone for water treatment is not a new concept, dating back to ∼1907 in the drinking water industry [3]. However,
photolytic decomposition of ozone [3]:
h
O3 −→O(1 D) + O2
(27a)
O(1 D) + H2 O → H2 O2
−
(27b)
+
H2 O2 + H2 O ↔ HO2 + H3 O
(27c)
accompanied by Reactions (14) and (28)–(30):
HO2 − + O3 → O3 − + HO2 •
(28)
O3 − + H2 O → HO3 • + OH−
(29)
Fe3+ + H2 O2 → Fe2+ + HO2 • + H+
(26a)
HO3 +
Fe2+ + H2 O2 + H+ → Fe3+ + H2 O + • OH
(26b)
results in the generation of • OH. Ozone has a strong absorption
band centered at 260 nm with a molar extinction coefficient of
∼300 M−1 cm−1 [218]. The combination of UV radiation with H2 O2
and O3 (the so-called Peroxone process) increases the fraction of
available ozone and possibly improves the mass transfer rate of
ozone in the fluid phase compared to the use of ozone alone [3].
The Fenton reaction (Eqs. (26a) and (26b)) was discussed earlier.
The photo-Fenton variant [3] provides for an added source of catalytic Fe2+ species (which are consumed via excess H2 O2 in Reaction
(26b)) and • OH as follows:
Thus the effect of iron species on the photodegradation of Acid
Red 1, an azo dye, was studied in TiO2 aqueous suspensions [42]. The
accumulation of H2 O2 was found to be completely suppressed during the photocatalytic process because of Reactions (26a) and (26b)
above [42]. However, a confounding decrease in the rate of H2 O2
formation (via Reaction (12)) because of the competition between
adsorbed Fe(III) species and O2 also cannot be ruled out here as
pointed out by the authors themselves of this study [82]. This
example may be taken as illustrative of the complexity of reaction
pathways in a heterogeneous photocatalysis system.
We close this section with a brief mention of miscellaneous
aspects related to the photooxidation of organic dyes. First,
attempts have been made to extend the wavelength response of
TiO2 toward the visible so as to increase its absorption overlap
with the solar spectrum. Thus, precipitation of anatase TiO2 layers
on perovskite-type (Sr0.95 La0.05 )TiO3+ı particles yielded a composite that was able to photobleach Methylene Blue at wavelengths
longer than ∼420 nm [147]. Another strategy is to dope TiO2 with
non-metallic species such as boron, carbon, or nitrogen [215].
Dye photooxidation has been combined with Cr(VI) reduction
in a TiO2 slurry reactor [216]. In this scheme, photogenerated electrons are utilized to reduce the toxic Cr(VI) to the environmentally
benign and immobile Cr(III) species. This strategy is particularly
relevant to practical remediation scenarios involving wastewater
effluent streams containing both reducible species (such as metal
ions) and oxidizable organic dyes. Yet another value-added innovation to dye photodegradation is combination with photoinduced
hydrogen generation as exemplified by a study on the simultaneous
degradation of azo dyes and H2 generation using irradiated Pt/TiO2
suspensions [100].
7.2. Comparison with Advanced Oxidation Processes and
combination with sonolysis
It was mentioned at the outset of this article (Section 1) that
heterogeneous photocatalysis fell under the umbrella of technologies called AOPs that rely on the generation and use of free radicals
(predominantly • OH). We shall consider here how heterogeneous
photocatalysis fares in comparison with their homogeneous process counterparts, namely UV/H2 O2 , UV/O3 , UV/H2 O2 /O3 , and
electro-Fenton and photo-Fenton processes for treating organic
dyes. A brief outline of each of these process candidates is given
first. Further details may be found in Refs. [3–5,217].
•
• OH
+ O2
(30)
Fe3+ + h → Fe2+
(31)
Fe(OH)2+ + h → Fe2+ + • OH
(32)
The electro-Fenton scheme involves the electrolytic generation
of H2 O2 at the cathode (usually made of carbon) via the 2e− electroreduction of O2 :
O2 + 2H+ + 2e− → H2 O2
(33)
together with the generation of Fe2+ species at an iron anode (see,
for example, Ref. [183], and citations therein).
Sonolysis, like photolysis, is a proven method for generating
free radicals in aqueous media [219,220]. Propagation of ultrasonic waves, especially at high frequencies (e.g., ∼600 kHz), leads
to the formation of cavitation bubbles in the presence of dissolved gas in the medium [219,220]. The collapse of these bubbles
spawns extremely high local temperatures (10,000 K) and pressures
(10,000 atm). Under these extreme conditions, water molecules
dissociate to form hydrogen and hydroxyl radicals:
H2 O → H• + • OH
(34)
While sonolysis can be used in isolation for the oxidative treatment of textile dyes, the beneficial effect of combining sonolysis
and photolysis [221] has been recognized in other studies on
organic dyes [45,95,222,223]. Examples of organic dyes treated by
the combination of UV/TiO2 and sonolysis include Reactive Black 5
[45], Naphthol Blue Black [95], and Remazol Black B [222].
7.3. Photosensitized conversion of organic dyes in visible light
Inspired by early work on the sensitization of wide bandgap
oxide semiconductors (Section 2.2), visible light has been used to
decolorize and decompose the dye, hopefully to completely mineralized products. Light absorption, unlike in the cases considered
184
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 8
Examples of visible light-induced photosensitized degradation of organic dyesa , b , c .
Entry no.
Organic dye/dye category
Reference(s)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Rose Bengal/xanthene
Methylene Blue/thiazine
Basic Blue 4/azo
Brilliant Red X-3B/azo
15 dyes (various types)
Acid Orange 7/azo
Acid Orange 7/azo
Nile Blue A/oxazine
Solvent Red 1/azo
Rose Bengal/xanthene
Erythrosine/xanthene
Rhodamine B/xanthene
Malachite Green/arylmethane
Eosin/xanthene
7 dyes/various types
Alizarin Red/anthraquinone
Sulforhodamine B/xanthene
3 dyes/various types
[120]
[119]
[48]
[46]
[98]
[99]
[143]
[130]
[31]
[141]
[106]
[52–54,106]
[65]
[57]
[92]
[63,114]
[56,107]
[108]
a
See also Ref. [9] for examples of azo dyes.
TiO2 used as the oxide semiconductor in all cases except in Entry 1 where ZnO
was also used.
c
See also Ref. [60] for a review of work on Methylene Blue.
b
in the preceding two sections, now occurs by the dye molecule, D,
adsorbed on the semiconductor surface (Fig. 8b):
D + h(vis) → 1 D ∗ or3 D∗
(35)
1
(36)
D ∗ or3 D ∗ + TiO2 → D•+ + TiO2 (ecb )
D+• → decomposition products
+•
D
−
+ OH → D +
• OH
D + • OH → decomposition products
(37a)
(38)
(37b)
Reaction (36) is followed by Reaction (11) and the following
sequence:
O2 •− + H+ → HO2 •
+•
+ O2
+•
+ HO2 → decomposition products
D
D
•−
•
→ DO2 → decomposition products
(39)
(37c)
(37d)
Excited states of the dye are highly reactive (relative to the
ground state) and in addition to the oxidative (electron injection)
scheme outlined above, the photoexcited dye can also undergo
reduction by accepting an electron from the semiconductor conduction band. This is exemplified by the humic acid sensitized
photoreduction of oxazine type dyes such as Oxazine 725 and Nile
Blue [130] and dyes such as Methylene Blue and Thionine, which are
converted to their (colorless) leuco form followed by irreversible
decomposition of the dye [24,25,60].
Table 8 lists representative studies on the visible light-induced
(photosensitized) degradation of organic dyes using inorganic
semiconductor supports such as TiO2 . Reference [9] also provides
examples of azo dyes that are converted to cation radicals and
decomposition products in this manner. It is worth underlining that
close proximity of the dye to the semiconductor surface (whether
via adsorption or surfactant-induced binding, see above) is a
pre-requisite for this photosensitization degradation mechanism
(Fig. 8b). Contrastingly and illustratively, the dye can be stabilized
against photodegradation by capping the semiconductor nanoparticles with an inert coating (such as polystyrene sulfonate, Ref. [91])
such that photoinduced electron injection from the excited dye
molecule is inhibited.
Interestingly, dye sensitization using visible light and an inorganic semiconductor (e.g., TiO2 ) can be used to decompose other
organic pollutants (e.g., pesticides and herbicides); this topic has
been reviewed [224]. Examples of this environmental remediation
approach includes the use of visible light and TiO2 modified with
Rose Bengal for the treatment of terbutylazine [225] and with azo
and thiazine dyes for treating bromacil [94]. In many of these cases,
it is not entirely clear to what extent the oxidized dye is regenerated to sustain the remediation process. This process could well be
dye-limited recalling there is only a monolayer of the dye (or at
best, a few monolayers) to begin with. In any case, this process is
not central to the theme of this review article and is not considered
further.
Finally, simultaneous and synergistic conversion of organic dyes
and metal ions in aqueous TiO2 suspensions under visible light illumination has been described [108]. The presence of Cu2+ or Fe3+ was
found to have a deleterious effect on the visible light degradation
of Sulforhodamine B, Alizarin Red and Malachite Green while Zn2+ ,
Cd2+ or Al3+ only had a “slight” influence [108]. These differences
were attributed [108] to the decreased reduction of O2 by the semiconductor conduction band electrons in the presence of reducible
metal ions and consequently diminished formation of reactive oxygen species (O2 − , • OH) (Reactions (12) and (13)). On the other hand,
Zn2+ , Cd2+ or Al3+ are not reduced by the conduction band electrons
in TiO2 since their redox levels are not located below the conduction
bandedge.
7.4. Photoreactions at the solid/air interface or in the gas phase
The oxide semiconductor particles can be stirred with a nonaqueous solution of the dye so as to adsorb the dye onto the surface
(Section 3) (for example, Ref. [141]). The stirred particles can be
vacuum-treated for evaporating off the solvent and then air dried.
Photochemical reactions can then be performed on the dye loaded
solid samples in air [31,32,90,141]. For the dye photodegradation
to be effective (via either of the mechanisms discussed above), it
is imperative that O2 be adsorbed on the semiconductor surface to
function as an electron scavenger (see Reaction (11)). Importantly,
control experiments performed with Al2 O3 show no photodegradation of the dye signaling that the semiconductor characteristics
of the oxide support play a crucial role. Thus organic dyes such
as Solvent Red 1 [31], Acid Orange [32,102], Naphthol Blue Black
[90], Rose Bengal [141], Acid Blue 9 [102,169], Reactive Blue 19
[102], Reactive Black 5 [102,169], and Indigo [66] have been photodegraded in this manner. Instead of nanosized TiO2 particles,
films of the oxide semiconductor on glass (for example, Pilkington ActivTM photocatalytic glass) can also be used (for example,
Ref. [169]).
Remote bleaching of Methylene Blue by UV-irradiated TiO2
has been reported in the gas phase [124]. A rear illuminated
TiO2 -coated glass plate was arranged to face another glass plate
coated with the dye with the two surfaces separated by a small
gap (12.5–500 m). The authors posit that the dye is not simply
bleached to its leuco form but is oxygenated or decomposed by
active oxygen species generated on the irradiated TiO2 surface and
transported through the gaseous inter-annular space to the Methylene Blue dye layer [124].
7.5. Photoreaction kinetics, variables, and related aspects
In this section, we review data on the kinetics of organic dye
photodegradation in UV- or visible light-irradiated semiconductor
suspensions and films. Like in all the case discussed above, these
studies have mostly featured TiO2 as the semiconductor photocatalyst. The kinetics experiments involve monitoring of either changes
in the dye absorbance (at a specific wavelength, usually, max ) or the
total organic carbon (TOC) (measured via chromatographic proce-
185
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
dures) as a function of time. The CO2 evolved, from organic carbon
decomposition, can also be monitored as a function of time (see
below). It must be borne in mind the temporal changes in these
three parameters are not identical (see below).
The kinetics data are then fit to specific models of which the
Langmuir–Hinshelwood (L–H) mechanism [3,5,226] appears to be
the most popular:
R=−
dC
k1 KC
=
1 + KC
dt
(40a)
where k1 and K are constants. At low dye concentrations and when
the product KC ≪ 1, Eq. (40a) collapses to an expression similar to
that applicable for first-order kinetics:
−
dC
= k1 KC = k′ C
dt
(40b)
In Eqs. (40a) and (40b), C is the (time-dependent) dye concentration
and k′ in Eq. (40b) is a pseudo-first-order rate constant. At the other
extreme of high dye concentrations when KC ≫ 1, Eq. (40a) reduces
to:
−
dC
= k1
dt
(40c)
This is the zero-order reaction kinetics limit with a constant
reaction rate. This behavior is diagnosed by linear plots of C versus t
unlike the exponential decay seen in first-order kinetics processes.
First-order or zero-order kinetics behavior has been seen in a
number of instances in the organic dye photocatalysis literature.
Tables 9 and 10 list studies where the temporal photoconversion
of the dye has been monitored. As with their counterparts in
Tables 2 and 3, azo dyes dominate the literature (with an extremely
diverse group) and they are considered in Table 9 while non-azo
dyes are considered in Table 10.
Rearrangement and integration of Eq. (40b) yields:
ln
C0
= k′ t
C
(40d)
In Eq. (40d), C0 is the initial concentration of the dye. Semi-log plots
of C0 versus t yield straight-line fits of the kinetics data from which
values for k1 (in units of min−1 ) can be extracted from the slopes.
Alternately, a linear transform of Equation (40a) yields:
1
1
1 1
=
+
R
k1
k1 K C
(40e)
where R is the photodegradation (or decolorization) rate embodied
by the left-hand term in Eq. (40a). Thus an inverse plot of the rate
versus the equilibrium dye concentration (C) would yield values
for k′ and K from the intercept and slope, respectively from which
corresponding values for k′ can be extracted (k′ = k1 K). Eqs. (40a)
and (40e) predict (a Langmuir-type) asymptotic behavior where
R increases non-linearly with C ultimately reaching a (surfacelimited) plateau value (given by k′ ) (Eq. (40c)). This can also be seen
from Eq. (40e); as C → ∞, R → k1 .
Literature values for the pseudo first-order rate constant (k′ )
show a very wide variation for the azo and other dyes, respectively.
Thus even if some of the anomalously high values for k′ are discounted, there is a ∼3 orders of magnitude variance in the reported
k′ values for the azo dyes and a corresponding variance of ∼500 for
the non-azo dyes. Some of these variations can be rationalized on
the dependence of k′ on process variables such as the type of photocatalyst used, the illumination details, the initial dye concentration,
photon flux, solution pH, the presence of oxidants in the solution
such as H2 O2 or Na2 S2 O8 , etc.
The effect of inorganic ions on dye adsorption and free radical
photochemistry was mentioned in the preceding sections. Their
effect on the kinetics of dye photoreactions has been addressed
in several studies. Thus Ag+ ions aided in the photodecolorization
of Methyl Orange to a greater extent than Cu2+ , Co2+ , Fe3+ , and
Ce4+ ions [122]. Similarly, NO3 − was better as a dye counterion
than SO4 2− [122]. Transition metal ions such as Cu2+ and Fe3+ were
shown to markedly depress the visible light-induced photodegradation of Sulforhodamine B, Alizarian Red, and Malachite Green
while other metal ions such as Zn2+ , Cd2+ , and Al3+ had only a slight
influence [108]. Sulfate and chloride ions were reported to not significantly affect the reaction rate for Methyl Orange [97]. SO4 2− ,
Cl− and NO3 − ions decreased the photocatalytic degradation rate
Table 9
Representative kinetics studies on the photodegradation of azo dyes in TiO2 suspensions or filmsa .
Organic dye
Comments
Reference(s)
Methyl Orange
The first azo dye to be studied from a kinetics perspective. Both photooxidation and
photoreduction of dye observed. Both the L–H model and other (steady-state approximation
derived) models used for analyzing photodegradation kinetics.
Pseudo first-order kinetics (Eqs. (40b) and (40d)) observed.
Both TOC and dye concentration monitored.
Pseudo first-order kinetics observed for both TiO2 suspensions and immobilized photocatalyst.
Exponential decay of dye concentration with time seen.
Pseudo first-order rate law seen to be obeyed.
Pseudo first-order rate law seen to be obeyed.
Exponential decay in dye absorbance seen in the photoelectrocatalytic mode.
Pseudo first-order kinetics seen.
Pseudo first-order kinetics seen.
A popular azo dye for kinetics studies.
–
Degradation said to “approximately” follow first-order kinetics.
–
–
First-order kinetics seen for TiO2 films.
–
–
–
–
L–H model adherence and pseudo first-order kinetics.
Photoelectrocatalytic data shown in terms of dye absorbance vs. time.
Both TOC and COD disappearance show initially zero-order kinetics.
[27,28,39,40,97,122,123,146,199]
Acid Orange 7
Acid Orange 20
Orange G
Orange IIb
Methyl Red
Reactive Red 2
Reactive Orange 16
Reactive Red 120
Reactive Red 141
Reactive Black 5
Congo Red
Solvent Red 1
Direct Blue 160
Basic Yellow 15
Basic Blue 41
Disperse Orange 1
Disperse Red 1
Disperse Red 13
Direct Black 38
Safira HEXL
Remazol Brilliant Orange 3 R
Amaranth
a
b
TiO2 was used as the photocatalyst unless otherwise stated.
Also known as Acid Orange 7.
[88,93,100,143,166,193,210]
[37]
[88,96]
[92,97]
[96]
[151]
[159]
[89]
[133]
[43–45,87,169]
[96,97]
[88]
[89]
[89]
[166]
[184,185]
[184,185]
[184,185]
[49]
[50]
[158]
[51]
186
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 10
Representative kinetics studies on the photodegradation of non-azo dyes in TiO2 suspensions or films.
Organic dye
Comments
Reference(s)
Thiazine
Methylene Blue
One of the earliest dyes to be studied from a photodegradation kinetics perspective.
[29,30,58,59,61,96,98,112,119]
A kinetics model incorporating O2 and sacrificial electron donor concentration developed in Ref.
[105]. First-order kinetics used in Ref. [30].
Pseudo first-order kinetics seen.
Degradation kinetics follows L–H model under visible light irradiation.
[30,105,155]
Anthraquinone
Anthraquinone 2-sulfuric acid
Acid Blue 25
Acid Blue 80
Acid Blue 40
Alizarin S
Pseudo first-order kinetics seen.
Kinetics plots shown using UV–vis spectrophotometry, TOC, and HPLC-diode array detection data.
Pseudo first-order kinetics seen.
As above but at pH 3 only.
–
[113]
[64]
[167]
[89]
[96]
Arylmethane
Malachite Green
L–H model and its linear transform (Eq. (40e)) used.
[65]
Indigo
Indigo and Indigo Carmine
L–H model and its linear transform (Eq. (40e)) used.
[66]
Phthalocyanine
Direct Blue 87
L–H model adherence found at certain pHs but not at others.
[89]
Cyanine
Fluorescein
L–H model and pseudo first-order kinetics used in continuous recirculation reactor.
[29]
Xanthene
Rhodamine B
Rose Bengal
Eosin
for Acid Orange 7 [210]. The above discussion makes it clear that
these effects (as with the other variables) are very much specific to
each dye system and possibly very sensitive to the process conditions employed in a particular study such that generalizations are
difficult.
Chemometrics [227,228] offers a versatile route to unraveling the effects of these variables on the photocatalysis rate and
also for process optimization. Thus a study on Safira HEXL dye
[50] shows a response surface (see Fig. 7 in Ref. [50]) of the initial rate as a function of initial dye concentration and solution
pH that is curved and complex diagnosing interaction between
the two variables. In general, multivariate statistical analyses and
central composite experimental designs are powerful tools for
screening and optimization of photodegradation rates. Especially,
interactions between process variables translates to the fact that
traditional experimental designs relying on changing one variable level at a time (while holding other variables constant)
will simply not be efficient and may even lead to misleading
conclusions.
7.6. Photoreaction intermediates
Decolorization of the dye-laden wastewater is but one central
objective of heterogeneous photocatalytic treatment. Total mineralization of the dye is an equally crucial end goal. Thus all the
original carbon and hydrogen in the organic dye must be converted to CO2 and water. Any nitrogen in the dye (as in the azo
compounds) must be similarly converted to N2 , NH4 + and nitrate
ions, sulfur to sulfate, phosphorus to phosphate, chlorine to chloride etc. Thus aside from monitoring absorbance changes during
the course of the photocatalytic remediation process, other complementary measurements must be carried out to get a total picture of
the process efficacy. These include the monitoring of the release of
inorganic ions and evolution of CO2 from the dye. Determination of
the total organic carbon and/or measurement of the chemical oxygen demand (COD) and biological oxygen demand (BOD) will afford
reliable indicators of the extent of mineralization of the dye. In particular, total decolorization of the dye may still be accompanied
by significant residual TOC in the solution, signaling that color-
[98,120,182]
[57,98]
less (organic) intermediates are formed during the process. Often
these intermediates may be as toxic as the original dye. There are
even cases where the original dye is not toxic but the photocatalytic
reaction intermediates are (see below).
The fate of the original dye has not been tracked in all the studies included in this review. However, many studies do discuss this
aspect and Tables 11 and 12 contain examples of these, again for
azo compounds (Table 11) and for non-azo dyes (Table 12). Reference [9] contains a good discussion of the nature and evolution of
organic intermediates for mono-, di- and triazo dyes and for triazine ring-containing azo dyes. The major degradation pathways
are identified for two azo dyes, Acid Orange 7 and Acid Orange 52,
respectively [9]. Rather than duplicate the discussion, the reader is
instead referred to this prior review article [9] for details.
The triazine ring is a recurrent feature in the chemical structure of many azo dyes of the reactive dye family. Triazines are
six-membered rings containing three nitrogen atoms in the aromatic ring structure [229]. Instead of terminating in CO2 and other
mineralized products (see above), the triazine ring typically yields
cyanuric acid via ring hydroxylation [9,229] attesting to its recalcitrant nature. Fortunately, this compound is not toxic as further
elaborated in the next section.
7.7. Toxicity studies
An early study on the extraction (with methylene chloride)
of organic compounds from a dye manufacturing plant wastewater, revealed the presence of several toxic chemicals in it [230].
More recently, a textile azo dye processing plant effluent was
identified as a major source of mutagenic activity [231]. Benzidine, a known carcinogen, was additionally found in this effluent
[231]. These results underscore the need for proper treatment of
the dye wastewater prior to discharging it into the environment.
Pertinent to this issue is the question of whether heterogeneous
photocatalysis is capable of mitigating the toxicity of the dye
effluent.
It is therefore surprising that, notwithstanding the existence of
hundreds of papers on the treatment of dye-laden water samples
using this approach (e.g., Figs. 1 and 2), only a handful of studies
187
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 11
Representative studies probing the formation of reaction intermediates during the heterogeneous photocatalytic conversiona of azo dyes.
Dye(s)
Technique(s) used
Comments
Reference(s)
Methyl Orange
HPLC–MS
[97]
Acid Orange 7b
COD monitored H2 O2
formation via a
spectrophotometric assay
GC–MS
FT-IR spectroscopy
Progressively demethylated products found.
Hydroxylated products found in laboratory
experiments but not observed in field tests.
–
TOC content only ∼80% after 3 h.
C2 aliphatic acids, carbonated and oxygenated
sulfur compounds detected.
Both qualitative and quantitative detection of
intermediates reported.
Visible light irradiation leads to decolorization
but not complete mineralization.
TOC removal found to be pH dependent.
UV–vis spectra during reaction indicate
quinonic intermediate formation.
–
Decomposition of the azo group more effective
than that of the naphthalene ring structure.
Mineralization of the nitrogen more effective
than sulfur (from the sulfonate group).
Original dye transformed to CO2 and aliphatic
(formic, acetic, oxalic) acids which are
converted more slowly by trapped holes and
•
OH.
A multiplicity of pathways including
demethylation and • OH attack of the aromatic
rings or to the azo-linkage bearing carbon.
Oxalic acid identified as main degradation
product from photoelectrocatalysis, 56–62%
TOC reduction observed.
Ion evolution found not to track one another
for sulfate and nitrate.
–
Compounds containing aromatic rings as well
as some short chain aliphatic acids (butanoic
and propanoic) detected.
Detailed degradation pathway proposed.
[88]
[35]
∼70% TOC reduction obtained after ∼3 h of
photoelectrocatalysis.
Aromatic intermediates detected.
[184,185]
Complete mineralization reported although
color removal slower than that of Amaranth
itself diagnosing the intermediate formation of
hydroxylated or desulfonated Amaranth.
Sulfate evolution rate approaches overall dye
degradation rate diagnosing the first step to be
C–S bond cleavage.
–
–
[51]
Acid Orange 7b
Acid Orange 7b
Acid Orange 7b
Acid Orange 7b
Acid Orange 7b
COD measurement, GC/MS,
FT-IR and ion chromatography
H2 O2 determination via
colorimetry
HPLC, TOC measurement,
UV–vis and FT-IR spectroscopy
Acid Orange 7b
Orange G
COD measurement
Ion chromatography and TOC
measurement
Acid Orange 20
FT-IR and TOC
Acid Orange 52b , c
GC–MS, HPLC and 1 H NMR
Reactive Orange 16
HPLC and UV–vis
spectrophotometry
Reactive Yellow 145 and Reactive Black 5
TOC measurement and ion
chromatography
Ion chromatography
GC–MS
Reactive Black 5
Reactive Black 5
Congo Red
Disperse Orange 1, Disperse Red 1 and Disperse Red 13
Direct Black 38
Amaranth
Ion chromatography and
electrospray-MS
TOC measurement
FT-IR spectroscopy and UV–vis
spectroscopy
TOC and COD measurements,
HPLC and ion chromatography
Amaranth
Ion chromatography
8 azo dyes
Commercial leather dye (not identified)
COD measurement
NH4 + assay by colorimetry
FT-Raman and FT-IR
spectroscopy
UV–vis spectrophotometry
Diffuse reflectance FT-IR
spectroscopy
Blue textile dye (not identified)
Naphthol Blue Black
Naphthol Blue Black
Procion Red MX-5B
a
b
c
HPLC
Ion chromatography
–
A quinonic product speculated as a result of
the photosensitized degradation of the dye on
TiO2 .
As above.
Quantity of sulfate and chloride ions found to
be sub-stoichiometric.
[92]
[36]
[99]
[194]
[93]
[183]
[37]
[39]
[159]
[44]
[43]
[45]
[97]
[49]
[150]
[33]
[168]
[175]
[90]
[197]
[150]
TiO2 photocatalyst was used in all the cases except in Refs. [150,175,197] where ZnO, CdO–ZnO, and WO3 were used, respectively.
See also Ref. [9] for further details.
Also known as Methyl Orange.
have specifically addressed the toxicity issue. A variety of toxicological assays exist based on monitoring the inhibition of growth of
bacteria, algae, mammalian cell lines, or aquatic organisms. Two
of the more popular candidates are the Microtox® method and
inhibition of Escherichia coli growth. In the former assay, luminescence from the bacterium, Vibrio fisheri, is used as an indicator of
medium toxicity. Thus Microtox® assay was used in photocataly-
sis studies on Reactive Black 5 [87], Remazol Turquoise Blue [162],
and in an electrochemical treatment study of textile dyes and dyehouse effluents [232]. Inhibition of E. coli respiration was used
as a toxicity indicator in a study on combined photocatalysis and
ozonation of a textile effluent [233]. Other types of assays have
also been used: inhibition of metabolic degradation of phenol [44],
death of Artemia salina cysts [49], and Salmonella/microsome assay
188
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
Table 12
Representative studies probing the formation of reaction intermediates during the heterogeneous photcatalytic conversiona of non-azo dyes.
Dye/dye category
Technique(s) used
Comments
Reference(s)
Methylene Blue/thiazine
CO2 analyses by conductivity
measurement.
COD and TOC measurements,
HPLC, ion chromatography,
GC–MS, LC–MS
CO2 formation seen to lag considerably behind
the rate of dye disappearance.
A detailed degradation pathway developed
incorporating ring opening and formation of
aromatic intermediates. Almost complete
mineralization of carbon, nitrogen, sulfur into
CO2 , NH4 + , NO3 − and SO4 2− observed.
Photobleaching found at low dissolved oxygen
concentrations although colored intermediates
also found.
A comparative study also involving
anthraquinone and azo dyes.
As above.
[58]
Methylene Blue/thiazine
Methylene Blue/thiazine
TOC measurement and UV–vis
spectrophotometry.
Methylene Blue/thiazine
HPLC, ion chromatography,
TOC and COD measurements.
HPLC, ion chromatography,
TOC and COD measurements.
COD, 1 H NMR, and EPR.
Methylene Blue/thiazine
Rhodamine B/xanthene
Rhodamine B/xanthene
Sulforhodamine B/xanthene
Eosin/xanthene
Alizarin Red/anthraquinone
Alizarin Red/anthraquinone
Alizarin Red/anthraquinone
Acid Blue 80/anthraquinone
Acid Blue 25/anthraquinone
Reactive Blue 4/anthraquinone
Indigo Carmine/indigo
Remazol Turquoise Blue 15/metallo-phthalocyanine
Squarylium cyanine dye
a
H2 O2 evolution by colorimetry
and COD measurements.
TOC measurements and above
techniques combined with 1 H
NMR, EPR, and GC–MS.
GC-TCD and ion
chromatography
COD measurements combined
with 1 H NMR, UV–vis
spectrophotometry, IR
spectroscopy, GC–MS and ESR
ESR and GC–MS along with MO
calculations
As in Ref. [61] above
CO2 by headspace GC, ion
chromatography, and HPLC–MS
HPLC-diode array detection.
TOC measurement, HPLC–MS
TOC measurement, 1 H NMR
and IR spectroscopy
TOC measurement, CO2 by GC
headspace analysis, GC–MS,
ion chromatography
HPLC-diode array detection, Cu
assay by anodic stripping
voltammetry
1
H NHR, UV–vis
spectrophotometry, ESR and
GC–MS
De-ethylation and oxidative degradation found
to occur in stepwise fashion.
Degradation rate correlated with H2 O2
accumulation.
Extent of N-de-ethylation found to depend on
mode of dye binding to TiO2 surface.
[61]
[112]
[96]
[41]
[54]
[92,108]
[56,107]
Bromide ion release along with CO2 evolution
found concomitantly with the
photodecomposition of dye.
Peroxides and carbonyl species along with
phthalic acid found as first intermediates. No
organo-peroxides found.
[57]
Pathway predicted by frontier electron
densities and point charges on all the dye
atoms.
–
Substrate hydroxylation and C–N bond
cleavage lead to unstable intermediates.
Two different kinetics found for decolorization
and dye degradation diagnosing the
occurrence of colored reaction intermediates.
After 1 h of photoelectrolysis, total color
removal observed but only 37% mineralization.
Dye decolorization not accompanied by TOC
reduction or release of inorganic anions.
[63]
100% color removal accompanied by 83% TOC
reduction. 69% of copper in the macrocycle is
released as free metal ions.
H2 O2 generated during visible light
photo-degradation but no organo-peroxides
found. Cleavage of the cyanine C C bond
dominated yielding 1-sulfopropyl-3,3dimethyl-5-bromoindolenium-2-one. This
intermediate not excited by visible light and
photo-degradation stops.
[160,162]
[114]
[41,96]
[115]
[64]
[161]
[66]
[117]
TiO2 was used as the photocatalyst.
[231]. The above studies were all performed on azo dye-laden water
samples.
How effective is heterogeneous photocatalysis in reducing the
toxicity of dye-laden water? Toxicity reduction was seen for Reactive Black 5 but after a treatment period of several hours, residual
toxicity was still found [45,87]. Toxicity data are often quantified in terms of the parameter, EC50 , which is the effective
concentration of a toxin causing 50% reduction in luminescent
light emission in the assay. Thus the 15 min EC50 values for Procion Red MX-5B and Reactive Brilliant Red K-2G were found
to be 23.57% and 21.68%, respectively at an initial dye concentration of 50 mg/L [234]. Complete detoxification (but not
mineralization) was obtained in these cases diagnosing that the
final product, cyanuric acid, is not toxic. Effective detoxification was also noted for an unspecified leather dye after UV/TiO2
treatment [49].
Very recently, we have studied a novel class of mutagenic
compounds, derived from dinitrophenyl azo dyes assigned as: C.I.
Disperse 373, C.I. Disperse violet 93 and C. I. Disperse Orange
37, respectively [235]. These dyes were identified as contaminants in rivers in Brazil, accounting, in some cases, for at least
50% of the mutagenic activity detected in the water bodies. The
degradation of these mutagenic dyes was instigated by photoelectrocatalytic oxidation. All photoelectrochemical experiments
were conducted using Ti/TiO2 thin film photoelectrodes prepared
by the sol–gel method and their mutagenic responses were compared and analyzed by using the strains YG1041 and YG1042
of Salmonella typhimurium. Using optimized conditions for photoelectrocatalytic oxidation carried out on Ti/TiO2 anode and
UV illumination in 0.25 M NaCl, 100% of decoloration and 60%
of total organic carbon reduction were obtained after 120 min
of photoelectrocatalysis [235]. In addition, the method pro-
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
moted an amelioration of the mutagenic properties of these dyes
(83% revertants/L),
A variety of confounding factors not related to the organic dye
itself can influence toxicity results. The use of ZnO as photocatalyst leads to a very toxic medium for the bacterium, V. fisheri, as a
result of ZnO photodissolution, which releases Zn2+ ions into the
aqueous medium [236]. Similarly, CdS photocatalysis increased the
medium toxicity because of CdS photocorrosion and release of Cd2+
ions [87]. Release of copper ions from the phthalocyanine macrocyclic assembly because of photodegradation caused an increase in
the acute toxicity for V. fisheri [162]. Finally, the use of EmulsogenTM
surfactant to solubilize Disperse Orange 1, Disperse Red 1 and Disperse Red 13 azo dyes resulted in acute toxicity for 293F cell lines
[184,185].
8. Into the real world: tests with dye wastewaters, process
scale-up and economic aspects
Real world dye wastewaters often contain high levels of suspended solids as well as other additives including salt (see above).
Relatively few studies exist where the efficacy of heterogeneous
photocatalysis has been assessed to treat wastewater effluents from
dyeing industries. This void is exacerbated by the fact that many of
the titles of papers in the relevant literature are misleading in that
they contain the word “wastewater” in them, although the test samples used in these studies were actually pristine water to which dye
had been added (e.g., Refs. [196,237]). Further, even in some cases
where real-world dye wastewater samples were used, their source
was not fully described nor was their characteristics (TDS, pH, TOC,
COD, BOD, etc.) identified.
A study on dye waste containing a mixture of reactive azo dyes
compared the efficacy of UV/TiO2 with the photo-Fenton process;
UV/TiO2 was found to be less effective for bleaching and degradation of the dye waste [237]. On the other hand, another study on
municipal wastewater contaminated with textile dyes found that
TiO2 -based photocatalysis was a viable method for decolorizing and
oxidizing the organics in the dye wastewater [104]. Wastes from
dye rinse baths were compared and blue-colored rinse was found
to degrade faster than pink, orange, or yellow colored wastewater [238]. Yet another study reports a lower degree of effectiveness
and this study also compares raw, coagulated and biologically pretreated textile effluent [239]. UV/TiO2 (or even solar/TiO2 ) could
prove to be a versatile polishing step for water streams to be posttreated by other (e.g., biodegradation) methods.
Interestingly, TiO2 -based photocatalysis can also improve the
biodegradability of dyeing wastewater. The increase in the BOD
value observed on photocatalysis can be taken to signal this possibility. This was confirmed in a study on anthraquinone 2-sulfonic
acid (sodium salt) [113]. Another study on Reactive Yellow 145 [44]
subjected the products of photocatalysis to bacterial degradation
after 6 h of UV/TiO2 pre-treatment. A 47% TOC reduction was noted
after 4 h of biological post-treatment while the original (unirradiated) sample showed only a 4% reduction [44].
Switching focus to process scale-up efforts, the section on photoreactor designs (Section 6) already alluded to studies going
beyond the laboratory scale [41,204]. Water throughput volume
is always an important design criterion in a practical water treatment system and in this regard, photoreactor designs incorporating
immobilized photocatalyst films are likely to be more practical.
However, films are also prone to fouling and residue accumulation
on their surface.
The final aspect considered in this penultimate section on the
more pragmatic aspects of organic dye photocatalysis, concerns
process economics. It may be argued that costing exercises may
189
be a little premature at this juncture considering the lack of many
pilot-scale studies (see above). However, examining how UV/TiO2
would fare in an energy-efficient and economic sense against other
AOPs is still relevant given that a common thread here is the cost of
electrically driving the UV lamps. A useful parameter here is EE/O,
which is defined as the electrical energy in kilowatt hours (kWh)
required to instigate the degradation of a pollutant by one order
of magnitude in 1000 L of contaminated water or air [212]. For the
purpose of this review, we can adapt EE/O to be the electrical energy
required to remove the initial color of the organic dye by one order
of magnitude.
In one study, EE/O was compared for two inorganic semiconductor candidates (TiO2 and ZnO) [240] while in two other studies,
this parameter was estimated and compared for UV/TiO2 against
other AOP variants [241,242]. Thus EE/O was three times higher
for TiO2 than for ZnO in the case of Reactive Blue 19 dye [240].
The use of UV alone resulted in very high EE/O values for Acid
Orange 7 and the values were ordered thus for other process
variants: UV/TiO2 > UV/TiO2 /H2 O2 > UV/TiO2 /IO4 − with the specific
values being dependent on the oxidant (H2 O2 and IO4 − ) concentrations [241]. In another study, the following ordering was
reported for the EE/O values for Reactive Orange 4 and Reactive Yellow 14 azo dyes: UV/H2 O2 > UV/TiO2 > Fe2+ /H2 O2 /UV [242].
The EE/O calculations, of course, are necessary but not sufficient, and the costs of chemicals and capital outlays must be
factored in to gain an overall comparative picture of the process
economics.
9. Concluding remarks
This review has shown that heterogeneous photocatalytic treatment of organic dyes is a mature science that has percolated to
all parts of the globe. On the scientific front, while much has
been learnt about the process (in all its interesting variations) and
mechanistic aspects, research on new generations of photocatalyst
materials (moving beyond TiO2 ), reaction intermediates/pathways,
and toxicity issues must continue. Especially much can be learnt
from corresponding advances in the discovery of new oxide photocatalyst materials for solar-driven hydrogen generation from water
[215].
While effluent streams from dyeing processes are fairly nontoxic in many cases (relative to trends with other organic materials
such as pesticides), the potential exists for generating toxic
by-products and intermediates. Photocatalysis appears to be unsurpassed in its capability to remove color quickly (unlike biological
treatment) and it appears to be energy-efficient relative to AOP
competitors such as UV/H2 O2 . Other AOP candidates such as ozonation or UV/O3 (and variants thereof), which are also very effective
for dye decolorization and degradation, have accompanying handicaps. For example, the chemical instability of ozone necessitates its
generation on site and its hazardous nature requires proper venting
to be built into the water treatment system. On the other hand, the
visible light-induced photosensitization approach based on TiO2
(and other oxide semiconductors) is particularly amenable to use
in developing country scenarios and in tropical or desert locales
where solar insolation is plentiful.
On the technological front, the scientific advances summarized
in this review article, must be translated to large-scale systems
capable of continuously and reliably treating large volumes of
dyeing wastewaters. Adaptation of the photocatalytic process to
treat poorly soluble dyes and dye waste sludge/cake must also be
developed. In this regard, very few pilot and/or field studies exist
at present to assess the large-scale feasibility of the photocatalysis approach. Indeed the field is now poised to move from the
190
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
chemistry research laboratories to the chemical engineering/civil
engineering/environmental engineering realm for further breakthroughs and advances.
Finally, process economics comparisons pitting UV/TiO2 and
other process variations (discussed above) against other more
conventional methods such as biological treatment, adsorption,
chemical coagulation, etc. for treatment and remediation of dye
wastewaters, are lacking as are many studies aimed at evaluating
the use of heterogeneous photocatalysis to supplement these existing methods as a pre- or post-treatment priming or polishing step,
respectively in a hybrid scheme.
Acknowledgments
One of us (K.R.) thanks the University of Texas System for a Science and Technology Acquisition and Retention (STAR) grant and
the U.S. Department of Energy (Basic Energy Sciences) for partial
funding support. We also thank Rohini Krishnan, Reena Krishnan and Asha S. Nair for assistance in manuscript preparation.
Profs. Akira Fujishima, Prashant Kamat, Leonardo Palmisano, and
Nick Serpone are thanked for comments on an initial draft of the
manuscript. Finally, the two anonymous reviewers are thanked for
constructive criticisms of an earlier manuscript version.
References
[1] H. Zollinger, Color Chemistry, VCH, Weinheim, 1991.
[2] D. Brown, A. Mark, S.C. De Vito, Crit. Rev. Environ. Sci. Technol. 23 (1993)
249.
[3] K. Rajeshwar, J. Ibanez, Environmental Electrochemistry, Academic Press, San
Diego, 1997.
[4] O. Legrini, E. Oliveros, A.M. Braun, Chem. Rev 93 (1993) 671.
[5] G.R. Helz, R.G. Zepp, D.G. Crosby (Eds.), Aquatic and Surface Photochemistry,
Lewis Publishers, Boca Raton, FL, 1994.
[6] J.T. Spadaro, L. Isabelle, V. Renganathan, Environ. Sci. Technol. 28 (1994) 1389.
[7] K. Rajeshwar, C.R. Chenthamarakshan, S. Goeringer, M. Djukic, Pure Appl.
Chem. 73 (2001 1849).
[8] K. Rajeshwar, J.G. Ibanez, J. Chem. Educ. 72 (1995 1044).
[9] I.K. Konstantinou, T.A. Albanis, Appl. Catal. B: Environ. 49 (2004) 1.
[10] A. Mills, R.H. Davies, D. Worsley, Chem. Soc. Rev 22 (1993) 417.
[11] P.V. Kamat, Chem. Rev 93 (1993) 267.
[12] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995)
69.
[13] K. Rajeshwar, J. Appl. Electrochem. 25 (1995 1067).
[14] A.L. Linsebigler, G. Lu, J.T. Yates Jr., Chem. Rev. 95 (1995) 735.
[15] A. Mills, S. LeHunte, J. Photochem. Photobiol. A: Chem. 108 (1997) 1.
[16] M.I. Litter, Appl. Catal. B: Environ. 23 (1999) 89.
[17] D.A. Tryk, A. Fujishima, K. Honda, Electrochim. Acta 45 (2000) 2363.
[18] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C: Photochem. Rev.
1 (2000) 1.
[19] D.M. Blake, Bibliography of Work on the Heterogeneous Photocatalytic
Removal of Hazardous Compounds from Water and Air, NREL/TP-570-26797,
update No. 4 (to October 2001), National Renewable Energy Laboratory,
Golden, CO, 2001.
[20] K. Rajeshwar, Guest Ed., J. Appl. Electrochem., vol. 35 (2005) (Special Issue on
Heterogeneous Photocatalysis).
[21] F. Delamare, B. Guineau, Colors-The Story of Dyes and Pigments, Harry N.
Abrams Inc., New York, 2000 (Engl. Translation by S. Hawkes).
[22] H. Gerischer, F. Willig, Top. Curr. Chem. 61 (1976) 31.
[23] R. Memming, Prog. Surf. Sci. 17 (1984) 7.
[24] A.V. Pamfilov, Ya.S. Mazurkevich, E.P. Pakhomova, Kinet. Catal. (USSR) 10
(1969) 915.
[25] H. Yoneyama, Y. Toyoguchi, H. Tamura, J. Phys. Chem. 76 (1972) 3460.
[26] T. Takizawa, T. Watanabe, K. Honda, J. Phys. Chem. 84 (1980) 51 (see also
references therein).
[27] G.T. Brown, J.R. Darwent, J. Chem. Soc., Faraday Trans. I 80 (1984) 1631.
[28] G.T. Brown, J.R. Darwent, J. Phys. Chem. 88 (1984) 4955.
[29] R.W. Matthews, J. Phys. Chem. 91 (1987) 3328.
[30] R.W. Matthews, Water Res. 25 (1991) 1169.
[31] M.S. Dieckmann, K.A. Gray, R.G. Zepp, Chemosphere 28 (1994) 1021.
[32] K. Vinodgopal, D.E. Wynkoop, P.V. Kamat, Environ. Sci. Technol. 30 (1996)
1680.
[33] M.S.T. Gonçalves, A.M.F. Oliveira-Campos, E.M.M.S. Pinto, P.M.S. Plasência,
M.J.R.P. Queiroz, Chemosphere 39 (1999) 781.
[34] C. Bauer, P. Jacques, A. Kalt, Chem. Phys. Lett. 307 (1999) 397.
[35] C. Bauer, P. Jacques, A. Kalt, J. Photochem. Photobiol. A: Chem. 140 (2001) 87.
[36] M. Stylidi, D.I. Kondarides, X.E. Verykios, Appl. Catal. B: Environ. 40 (2003)
271.
[37] C. Galindo, P. Jaques, A. Kalt, Chemosphere 45 (2001) 997.
[38] T. Papadam, N.P. Xekoukoulotakis, I. Poulios, D. Mantzavinos, J. Photochem.
Photobiol. A: Chem. 186 (2007) 308.
[39] C. Galindo, P. Jaques, A. Kalt, J. Photochem. Photobiol. A: Chem. 130 (2000) 35.
[40] I.M. Arabatzis, T. Stergiopoulos, D. Andreeva, S. Kitova, S.G. Neophytides, P.
Falaras, J. Catal. 220 (2003) 127.
[41] C. Guillard, H. Lachheb, A. Houas, M. Ksibi, E. Elaloui, J.M. Herrmann, J. Photochem. Photobiol. A: Chem. 158 (2003) 27.
[42] M. Mrowetz, E. Selli, J. Photochem. Photobiol. A: Chem. 162 (2004) 89.
[43] I. Poulios, I. Tsachpinis, J. Chem. Technol. Biotechnol. 74 (1999) 349.
[44] A. Aguedach, S. Brosillon, J. Morvan, E.K. Lhadi, Appl. Catal. B: Environ. 57
(2005) 55.
[45] D.E. Kritikos, X. Xekoukoulotakis, E. Psillakis, D. Mantzavinos, Water Res. 41
(2007) 2236.
[46] Y. Xu, C.H. Langford, Langmuir 17 (2001) 897.
[47] I. Arslan, I.A. Balcioğlu, Dyes Pigments 43 (1999) 95.
[48] E. Stathatos, T. Petrova, P. Lianos, Langmuir 17 (2001) 5025.
[49] R.F.P.M. Moreira, T.P. Sauer, L. Casaril, E. Humeres, J. Appl. Electrochem. 35
(2005) 821.
[50] T. Sauer, G.C. Neto, H.J. José, R.F.P.M. Moreira, J. Photochem. Photobiol. A: Chem.
149 (2002) 147.
[51] M. Karkmaz, E. Puzenat, C. Guillard, J.M. Hermann, Appl. Catal. B: Environ. 51
(2004) 183.
[52] P. Qu, J. Zhao, L. Zang, T. Shen, H. Hidaka, Colloid Surf. A: Physicochem. Eng.
Aspects 138 (1998) 39.
[53] J. Zhao, T. Wu, K. Wu, K. Oikawa, H. Hidaka, N. Serpone, Environ. Sci. Technol.
32 (1998) 2394.
[54] T. Wu, G. Liu, J. Zhao, H. Hidaka, N. Serpone, J. Phys. Chem. B 102 (1998) 5845.
[55] D. Chatterjee, A. Mahata, J. Photochem. Photobiol. A: Chem. 153 (2002) 199.
[56] G. Liu, X. Li, J. Zhao, H. Hidaka, N. Serpone, Environ. Sci. Technol. 34 (2000)
3982.
[57] F. Zhang, J. Zhao, T. Shen, H. Hidaka, E. Pelizzetti, N. Serpone, Appl. Catal. B:
Environ. 15 (1998) 147.
[58] R.W. Matthews, J. Chem. Soc., Faraday Trans. I 85 (1989) 1291.
[59] K. Sopajaree, S.A. Qasim, S. Basak, K. Rajeshwar, J. Appl. Electrochem. 29 (1999)
533.
[60] A. Mills, J. Wang, J. Photochem. Photobiol. A: Chem. 127 (1999) 123.
[61] A. Houas, H. Lachheb, M. Ksibi, E. Elaloui, C. Guillard, J.M. Herrmann, Appl.
Catal. B: Environ. 31 (2001) 145.
[62] P.V. Messina, P.C. Schulz, J. Colloid Interface Sci. 299 (2006) 305.
[63] G. Liu, X. Li, J. Zhao, S. Horikoshi, H. Hidaka, J. Mol. Catal. A: Chem. 153 (2000)
221.
[64] I. Bouzaida, C. Ferronato, J.M. Chovelon, M.E. Rammah, J.M. Herrmann, J. Photochem. Photobiol. A: Chem. 168 (2004) 23.
[65] J. Zhao, K. Wu, T. Wu, H. Hidaka, N. Serpone, J. Chem. Soc., Faraday Trans. 94
(1998) 673.
[66] M. Vautier, C. Guillard, J.M. Herrmann, J. Catal. 201 (2001) 46.
[67] M. Sökmen, A. Özkan, J. Photochem. Photobiol. A: Chem. 147 (2002) 77.
[68] H.Z. Zhang, R. LeePenn, R.J. Hamers, J. Phys. Chem. B 103 (1999) 4656.
[69] C.H. Giles, A.P. D’Silva, I.A. Easton, J. Colloid Interface Sci. 47 (1974) 766.
[70] R.H. Perry, D. Green (Eds.), Perry’s Chemical Engineers’ Handbook, McGrawHill, New York, 1984.
[71] I. Langmuir, J. Am. Chem. Soc. 38 (1916) 2221.
[72] H. Freundlich, Z. Phys. Chem. A 57 (1906) 228.
[73] M.F. Sawalha, J.R. Peralta-Videa, J. Romero-González, J.L. Gardea-Torresdey, J.
Colloid Interface Sci. 300 (2006) 100.
[74] G.D. Parfitt, Prog. Surf. Sci. 11 (1976) 181.
[75] J. Cunningham, G. Al-Sayyed, J. Chem. Soc., Faraday Trans. 86 (1990) 3935.
[76] J. Cunningham, S. Srijaranai, J. Photochem. Photobiol. A: Chem. 58 (1991)
361.
[77] J. Cunningham, G. Al-Sayyed, S. Srijaranai in Ref. [5], Ch. 22, pp. 317–348.
[78] C.R. Chenthamarakshan, K. Rajeshwar, Electrochem. Commun. 2 (2000) 527.
[79] C.R. Chenthamarakshan, H. Yang, Y. Ming, K. Rajeshwar, J. Electroanal. Chem.
494 (2000) 79.
[80] P. Kajitvichyanukul, C.R. Chenthamarakshan, K. Rajeshwar, S.R. Qasim, J. Electroanal. Chem. 519 (2002) 25.
[81] Y. Ming, C.R. Chenthamarakshan, K. Rajeshwar, J. Photochem. Photobiol. A:
Chem. 147 (2002) 199.
[82] P. Kajitvichyanukul, C.R. Chenthamarakshan, S.R. Qasim, K. Rajeshwar, Langmuir 17 (2001) 3792.
[83] P. Kajitvichyanukul, C.R. Chenthamarakshan, K. Rajeshwar, S.R. Qasim, Adsorp.
Sci. Technol. 21 (2003) 217.
[84] T. Torimoto, S. Ito, S. Kuwabata, H. Yoneyama, Environ. Sci. Technol. 30 (1996)
1275.
[85] H. Tada, M. Akazawa, Y. Kubo, S. Ito, J. Phys. Chem. B 102 (1998) 6360.
[86] Y. Xu, C.H. Langford, J. Phys. Chem. B 101 (1997) 3115.
[87] L.B. Reutergárdh, M. Iangphasuk, Chemosphere 35 (1997) 585.
[88] K. Hustert, R.G. Zepp, Chemosphere 24 (1992) 335.
[89] W.Z. Tang, H. An, Chemosphere 31 (1995) 4157.
[90] C. Nasr, K. Vinodgopal, L. Fisher, S. Hotchandani, A.K. Chattopadhyay, P.V.
Kamat, J. Phys. Chem. 100 (1996) 8436.
[91] L. Ziolkowski, K. Vinodgopal, P.V. Kamat, Langmuir 13 (1997) 3124.
[92] T. Wu, G. Liu, J. Zhao, H. Hidaka, N. Serpone, J. Phys. Chem. B 103 (1994) 4862.
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
[93] F. Kiriakidou, D.I. Kondravides, X.E. Verykios, Catal. Today 54 (1999) 119.
[94] L. Feigelson, L. Muszkat, L. Bir, K.A. Muszkat, Water Sci. Technol. 42 (2000)
275.
[95] N.L. Stock, J. Peller, K. Vinodgopal, P.V. Kamat, Environ. Sci. Technol. 34 (2000
1747).
[96] H. Lachheb, E. Puzenat, A. Houas, M. Ksibi, E. Elaloui, C. Guillard, J.M. Herrmann, Appl. Catal. B: Environ. 39 (2002) 75.
[97] V. Augugliaro, C. Baiocchi, A.B. Prevot, E. García-López, V. Loddo, S. Malato, G.
Marci, L. Palmisano, M. Pazzi, E. Pramauro, Chemosphere 49 (2002) 1223.
[98] G.A. Epling, C. Lin, Chemosphere 46 (2002) 561.
[99] M. Stylidi, D.I. Kondarides, X.E. Verykios, Appl. Catal. B: Environ. 47 (2004)
189.
[100] A. Patsoura, D.I. Kondarides, X.E. Verykios, Appl. Catal. B: Environ. 64 (2006)
171.
[101] S.F. Villanueva, S.S. Martinez, Solar Energy Mater. Solar Cells 91 (2007) 1492.
[102] A.J. Julson, D.F. Ollis, Appl. Catal. B: Environ. 65 (2006) 315.
[103] Y. Zhiyang, M. Bensimon, V. Sarria, I. Stolitchnov, W. Jardim, D. Laub, E. Mielczarski, J. Mielczarski, L. Kiwi-Minsker, J. Kiwi, Appl. Catal. B: Environ. 76
(2007) 185.
[104] R.J. Davis, J.L. Gainer, G. O’Neal, I.W. Wu, Water Environ. Res. 66 (1994)
50.
[105] A. Mills, A. Belghazi, R.H. Davis, D. Worsley, S. Morris, J. Photochem. Photobiol.
A: Chem. 79 (1994) 131.
[106] F. Zhang, J. Zhao, L. Zang, T. Shen, H. Hidaka, E. Pelizzetti, N. Serpone, J. Mol.
Catal. A: Chem. 120 (1997) 173.
[107] G. Liu, J. Zhao, New J. Chem. 24 (2000) 411.
[108] C. Chen, X. Li, W. Ma, J. Zhao, H. Hidaka, N. Serpone, J. Phys. Chem. B 106 (2002)
318.
[109] R.F.P. Nogeira, W.F. Jardim, J. Chem. Educ. 70 (1993) 861.
[110] S. Lekshmi, R. Renganathan, S. Fujita, J. Photochem. Photobiol. A: Chem. 88
(1995) 163.
[111] K. Sopajaree, S.A. Qasim, S. Basak, K. Rajeshwar, J. Appl. Electrochem. 29 (1999)
1111.
[112] T. Zhang, T. Oyama, A. Aoshima, H. Hidaka, J. Zhao, N. Serpone, J. Photochem.
Photobiol. A: Chem. 140 (2001) 163.
[113] J. Kiwi, C. Pulgarin, M. Grätzel, New J. Chem. 17 (1993) 487.
[114] G. Liu, T. Wu, J. Zhao, H. Hidaka, N. Serpone, Environ. Sci. Technol. 33 (1999
2081).
[115] A.B. Prevot, C. Baiocchi, M.C. Brussino, E. Pramauro, P. Savarino, V. Augugliaro,
G. Marci, L. Palmisano, Environ. Sci. Technol. 35 (2001) 971.
[116] K.D. Giglio, D.B. Green, B. Hutchinson, J. Chem. Educ. 72 (1995) 352.
[117] T. Wu, T. Lin, J. Zhao, H. Hidaka, N. Serpone, Environ. Sci. Technol. 33 (1999)
1379.
[118] P. Reeves, R. Ohlhausen, D. Sloan, K. Pamplin, T. Scoggins, C. Clark, B. Hutchinson, D. Green, Solar Energy 48 (1992) 413.
[119] K. Tennakone, S. Senadeera, A. Priyadharshana, Solar Energy Mater. Solar Cells
29 (1993) 109.
[120] K. Tennakone, W.C.B. Kiridena, S. Punchihewa, J. Photochem. Photobiol. A:
Chem. 68 (1992) 389.
[121] L.C. Chen, T.C. Chou, J. Mol. Catal. 85 (1993) 201.
[122] L.C. Chen, T.C. Chou, Ind. Eng. Chem. Res. 33 (1994) 1436.
[123] L.C. Chen, T.C. Chou, Ind. Eng. Chem. Res. 32 (1993) 1520.
[124] T. Tatsuma, S.I. Tachibana, T. Miwa, D.A. Tryk, A. Fujishima, J. Phys. Chem. B
103 (1999) 8033.
[125] K. Tanaka, K. Padermpole, T. Hisanaga, Water Res. 34 (2000) 327.
[126] R.I. Bickley, T.G. Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid State Chem.
92 (1991) 178.
[127] M. Lindner, D.W. Bahnemann, B. Hirthe, W.D. Griebler, J. Solar Energy Eng. 119
(1997) 120.
[128] P.V. Kamat, M.A. Fox, Chem. Phys. Lett. 102 (1983) 379.
[129] K.R. Gopidas, P.V. Kamat, Langmuir 5 (1989) 22.
[130] K. Vinodgopal, Res. Chem. Intermed. 20 (1994) 825.
[131] K. Vinodgopal, I. Bedja, S. Hotchandani, P.V. Kamat, Langmuir 10 (1994 1767).
[132] C. Nasr, K. Vinodgopal, S. Hotchandani, A.K. Chattopadhyay, P.V. Kamat, Res.
Chem. Intermed. 23 (1997) 219.
[133] F. Sayilkan, S. Erdemoğlu, M. Asiltůrk, M. Akarsu, Ş. Şener, H. Sayilkan, M.
Erdemoğlu, E. Arpac, Mater. Res. Bull. 41 (2006) 2276.
[134] M. Nag, P. Basak, S.V. Manorama, Mater. Res. Bull. 42 (2007) 1691.
[135] M.L. Curri, R. Comparelli, P.D. Cozzoli, G. Mascolo, A. Agostiano, Mater. Sci.
Eng. C 23 (2003) 285.
[136] J.E.B. Katari, V.L. Colvin, A.P. Alivisatos, J. Phys. Chem. 98 (1994) 4109.
[137] G.N. Kryukova, G.A. Zenkovets, A.A. Shutilov, M. Wilde, K. Günther, D. Fassler,
K. Richter, Appl. Catal. B: Environ. 71 (2007) 169.
[138] L. Wu, J.C. Yu, X. Wang, L. Zhang, J. Yu, J. Solid State Chem. 178 (2005) 321.
[139] G.A. Somorjai, Introduction to Surface Chemistry and Catalysis, J. Wiley &
Sons, New York, 1994.
[140] A.W. Adamson, A.P. Gast, Physical Chemistry of Surfaces, J. Wiley & Sons, New
York, 1997.
[141] K.R. Gopidas, P.V. Kamat, J. Phys. Chem. 93 (1989) 6428.
[142] M.E. Fabiyi, R.L. Skelton, J. Photochem. Photobiol. A: Chem. 132 (2000)
121.
[143] Y. Chen, K. Wang, L. Lou, J. Photochem. Photobiol. A: Chem. 163 (2004) 281.
[144] J. Feng, R.S.K. Wong, X. Hu, P.L. Yue, Catal. Today 98 (2004) 441.
[145] D. Li, Y. Guo, C. Hu, C. Jiang, E. Wang, J. Mol. Catal. A: Chem. 207 (2004) 181.
[146] A.O. Ibhadon, G.M. Greenway, Y. Yue, Catal. Commun. 9 (2008) 153.
191
[147] S. Otsuka-Yao-Matsuo, M. Ueda, J. Photochem. Photobiol. A: Chem. 168 (2004)
1.
[148] N.R. de Tacconi, R.O. Lezna, K. Rajeshwar, Chem. Mater. 15 (2003) 3046.
[149] E.S. Smotkin, A.J. Bard, A. Campion, M.A. Fox, T. Mallouk, S.E. Webber, J.M.
White, J. Phys. Chem. 90 (1986) 4604.
[150] Y. Bessekhouad, N. Chaoui, M. Trzpit, N. Ghazzal, D. Robert, J.V. Weber, J.
Photochem. Photobiol. A: Chem. 183 (2006) 218.
[151] C.H. Wu, Chemosphere 57 (2004) 601.
[152] C.H. Wu, C.L. Chang, J. Hazard. Mater. B 128 (2006) 265.
[153] Y. Zhiyong, E. Mielczarski, J. Mielczarski, D. Laub, Ph. Buffat, U. Klehm, P. Albers,
K. Lee, A. Kulik, L. Kiwi-Minsker, A. Renken, J. Kiwi, Water Res. 41 (2007) 862
(see also references therein).
[154] H. Tada, M. Tanaka, Langmuir 13 (1997) 360.
[155] Y. Ma, J. Yao, Chemosphere 38 (1999) 2407.
[156] M. Miyauchi, A. Nkajima, A. Fujishima, K. Hashimoto, T. Watanabe, Chem.
Mater. 12 (2000) 3.
[157] S. Sitkiewitz, A. Heller, New J. Chem. 20 (1996) 233.
[158] M.V.B. Zanoni, J.J. Sene, M.A. Anderson, J. Photochem. Photobiol. A: Chem. 157
(2003) 55.
[159] P.A. Carneiro, M.E. Osugi, J.J. Sene, M.A. Anderson, M.V.B. Zanoni, Electrochim.
Acta 49 (2004) 3807.
[160] M.E. Osugi, G.A. Umbuzeiro, M.A. Anderson, M.V.B. Zanoni, Electrochim. Acta
50 (2005) 5261.
[161] P.A. Carneiro, M.E. Osugi, C.S. Fugivara, N. Boralle, M. Furlan, M.V.B. Zanoni,
Chemosphere 59 (2005) 431.
[162] M.E. Osugi, G.A. Umbuzeiro, F.J.V. De Castro, M.V.B. Zanoni, J. Hazard. Mater.
B 137 (2006) 871.
[163] M.A. Anderson, M.J. Gieselmann, Q. Xu, J. Membr. Sci. 39 (1988) 243.
[164] D.H. Kim, M.A. Anderson, Environ. Sci. Technol. 28 (1994) 479.
[165] T. Watanabe, A. Nakajima, R. Wang, M. Minabe, S. Koizumi, A. Fujishima, K.
Hashimoto, Thin Solid Films 351 (1999) 260.
[166] P. Bouras, P. Lianos in Ref. [20], p. 831.
[167] C.H. Ao, M.K.H. Leung, R.C.W. Lam, D.Y.C. Leung, L.L.P. Vrijmoed, W.C. Yam, S.P.
Ng, Chem. Eng. J. 129 (2007) 153.
[168] L.C. Macedo, D.A.M. Zaia, G.J. Moore, H. de Santana, J. Photochem. Photobiol.
A: Chem. 185 (2007) 86.
[169] P. Chin, D.F. Ollis, Catal. Today 123 (2007) 177.
[170] A. Mills, A. Lepre, N. Elliott, S. Bhopal, I.P. Parkin, S.A. O’Neill, J. Photochem.
Photobiol. A: Chem. 160 (2003) 213.
[171] K. Rajeshwar, in: V. Balzani (Ed.), Electron Transfer in Chemistry, Wiley–VCH,
Weinheim, 2001.
[172] K. Rajeshwar, in: S. Licht (Ed.), Encyclopedia of Electrochemistry, Wiley–VCH,
Weinheim, 2001, pp. 3–53, Ch. 1.
[173] K. Vinodgopal, I. Bedja, P.V. Kamat, Chem. Mater. 8 (1996) 2180.
[174] K. Vinodgopal, P.V. Kamat, Environ. Sci. Technol. 29 (1995) 841.
[175] R. Suárez-Parra, I. Hernández-Pérez, M.E. Rincón, S. López-Ayala, M.C. RoldánAhumada, Solar Energy Mater. Solar Cells 76 (2003) 189.
[176] K. Rajeshwar, N.R. de Tacconi, in: A. Wieckowski (Ed.), Interfacial Electrochemistry, Marcel Dekker, New York, 1999, pp. 721–736, Ch. 40.
[177] K. Rajeshwar, N.R. de Tacconi, C.R. Chenthamarakshan, Curr. Opin. Solid State
Mater. Sci. 8 (2005) 173.
[178] N.R. de Tacconi, C.R. Chenthmarakshan, G. Yogeeswaran, A. Watcherenwong,
R.S. de Zoysa, N.A. Basit, K. Rajeshwar, J. Phys. Chem. B 110 (2006) 25347.
[179] C.A. Grimes, O.K. Varghese, S. Ranjan, Light, Water, Hydrogen, Springer, New
York, 2008 (ch. 5), pp. 257–369.
[180] G.H.A. Therese, P.V. Kamath, Chem. Mater. 12 (2000) 1195.
[181] C. Natarajan, G. Nogami, J. Electrochem. Soc. 143 (1996) 1547.
[182] X.Z. Li, H.L. Liu, P.T. Yue, Y.P. Sun, Environ. Sci. Technol. 34 (2000) 4401.
[183] X.B. Xie, X.Z. Li, Mater. Chem. Phys. 95 (2006) 39.
[184] M.E. Osugi, C.R. Chenthamarakshan, N.R. de Tacconi, G.A. Woldermariam, S.S.
Mandal, M.V.B. Zanoni, K. Rajeshwar, Proceedings of the International Society
of Electrochemistry (ISE) Meeting Foz da Iguazu, Brazil, 2008.
[185] M.E. Osugi, C.R. Chenthamarakshan, N.R. de Tacconi, G.A. Woldermariam, S.S.
Mandal, M.V.B. Zanoni, K. Rajeshwar, J. Adv. Oxid. Technol 11 (2008) 425.
[186] M.F.J. Dijktra, A. Michorius, H. Buwalda, H.J. Panneman, J.G.M. Winkelman,
A.A.C.M. Beenacker, Catal. Today 66 (2001) 487.
[187] S. Parra, S.E. Stanca, I. Guasaquillo, K.R. Thampi, Appl. Catal. B: Environ. 51
(2004) 73.
[188] H. Gerischer, A. Heller, J. Phys. Chem. 95 (1991) 5261.
[189] H. Gerischer, A. Heller, J. Electrochem. Soc. 139 (1992) 113.
[190] E. Vlatka, Ž. Kulešius, J. Appl. Electrochem. 37 (2007) 415.
[191] J.M. Peralta-Hernández, J. Manríquez, Y. Meas-Vong, F.J. Rodriguez, T.W. Chapman, M.I. Maldonado, L.A. Godínez, J. Hazard. Mater. 147 (2007) 588.
[192] K. Ui, T. Fujita, N. Koura, F. Yamaguchi, J. Electrochem. Soc. 153 (2006) C449.
[193] G. Li, J. Qu, X. Zhang, J. Ge, Water Res. 40 (2006) 213.
[194] G. Li, J. Qu, X. Zhang, H. Liu, H. Liu, J. Mol. Catal. A: Chem. 259 (2006) 238.
[195] H. Zhang, X. Quan, S. Chen, H. Zhao, Y. Zhao, Sep. Purif. Technol. 50 (2006) 147.
[196] T. Wang, H. Wang, P. Xu, X. Zhao, Y. Liu, S. Chao, Thin Solid Films 334 (1998)
103.
[197] M. Hepel, J. Luo, Electrochim. Acta 47 (2001) 729.
[198] J. Luo, M. Hepel, Electrochim. Acta 46 (2001) 2913.
[199] A. Watcherenwong, W. Chanmanee, N.R. de Tacconi, C.R. Chenthamarakshan,
P. Kajitvichyanukul, K. Rajeshwar, J. Electroanal. Chem. 612 (2008) 112.
[200] W.Y. Lin, C. Wei, K. Rajeshwar, J. Electrochem. Soc. 140 (1993) 2477.
[201] W.Y. Lin, K. Rajeshwar, J. Electrochem. Soc. 144 (1997 2751).
192
K. Rajeshwar et al. / Journal of Photochemistry and Photobiology C: Photochemistry Reviews 9 (2008) 171–192
[202] C.R. Chenthamarakshan, K. Rajeshwar, E.J. Wolfrum, Langmuir 16 (2000 2715).
[203] N.R. de Tacconi, J. Carmona, K. Rajeshwar, J. Electrochem. Soc 144 (1997)
2486.
[204] J.M. Herrmann, Catal. Today 53 (1999) 115.
[205] N.B. Jackson, C.M. Wang, Z. Luo, J. Schwitzgebel, J.G. Eckerdt, J.R. Brock, A.
Heller, J. Electrochem. Soc. 138 (1991) 3660.
[206] M.A. Grela, M.E.J. Coronel, A.J. Colussi, J. Phys. Chem. 100 (1996) 16940.
[207] L.Z. Sun, J.R. Bolton, J. Phys. Chem. 100 (1996) 4127.
[208] Y. Nosaka, S. Komori, K. Yawat, T. Hirakawa, A.Y. Nosaka, Phys. Chem. Chem.
Phys. 5 (2003) 4731.
[209] K.I. Ishibashi, A. Fujishima, T. Watanabe, K. Hashimoto, J. Photochem. Photobiol. A: Chem. 134 (2000) 139.
[210] Y. Chen, S. Yang, K. Wang, L. Lou, J. Photochem. Photobiol. A: Chem. 172 (2005)
47.
[211] V. Rao, K. Rajeshwar, V. Pai Verneker, J. Dubow, J. Phys. Chem. 84 (1980 1987).
[212] J.R. Bolton, K.G. Bircher, W. Tuma, C.A. Tolman, J. Adv. Oxid. Technol. 1 (1996)
13.
[213] M. Abdullah, G.K.C. Low, R.W. Matthews, J. Phys. Chem. 94 (1990) 6820.
[214] C. Walling, Acc. Chem. Res. 8 (1975) 125.
[215] K. Rajeshwar, J. Appl. Electrochem. 37 (2007) 765.
[216] S.G. Schrank, H.J. José, R.F.P.M. Moreira, J. Photochem. Photobiol. A: Chem. 147
(2002) 71.
[217] W.W. Eckenfelder, A.R. Bowers, J.A. Roth (Eds.), Chemical Oxidation: Technologies for the Nineties, vol. 2, Technomic Publishing, Lancaster and Basel,
1992;
W.W. Eckenfelder, A.R. Bowers, J.A. Roth (Eds.), Chemical Oxidation: Technologies for the Nineties, vol.3, Technomic Publishing, Lancaster and Basel,
1993.
[218] J. R. Bolton and S. R. Cater in Ref. [5], Ch. 33, pp. 467–490.
[219] K.S. Suslick, Ulrasound: Its Chemical, Physical and Biological Effects, VCH, New
York, 1988.
[220] K.S. Suslick, Science 247 (1990) 1439.
[221] M. Stevenson, K. Bullock, W.Y. Lin, K. Rajeshwar, Res. Chem. Intermed. 23
(1997) 311.
[222] K. Vinodgopal, J. Peller, O. Makogon, P.V. Kamat, Water Res. 32 (1998) 3646.
[223] T. Velegraki, I. Poulios, M. Charalabaki, N. Kalogerakis, P. Samaras, D. Mantzavinos, Appl. Catal. B: Environ. 62 (2006) 159.
[224] D. Chatterjee, S. Dasgupta, J. Photochem. Photobiol. C: Photochem. Rev. 6
(2005) 186.
[225] H. Ross, J. Bendig, S. Hecht, Solar Energy Mater. Solar Cells 33 (1994) 475.
[226] D.F. Ollis, J. Phys. Chem. B 109 (2005) 2439 (see also references therein).
[227] S.N. Demming, S.L. Morgan, Experimental Design: A Chemometric Approach,
Elsevier, Amsterdam, 1987.
[228] G. Box, W.G. Hunter, Statistics for Experimenters: An Introduction to Design,
Data Analysis, and Model Building, Wiley, New York, 1987.
[229] N. Watanabe, S. Horikoshi, H. Hidaka, N. Serpone, J. Photochem. Photobiol. A:
Chem. 174 (2005) 229.
[230] L.M. Games, R.A. Hites, Anal. Chem. 49 (1977) 1433.
[231] R.O. Alves de Lima, A.P. Bazo, D.M.F. Salvadori, C.M. Rech, D. de Palma Oliveira,
G. de Aragão Umbuzeiro, Mutat. Res. 626 (2007) 53.
[232] E. Chatzisymeon, N.P. Xekoukoulotakis, A. Coz, N. Kalogerakis, D. Mantzavinos,
J. Hazard. Mater. B 137 (2006) 998.
[233] S. Gomes de Moraes, R.S. Freire, N. Duran, Chemosphere 40 (2000) 369.
[234] C. Hu, J.C. Yu, Z. Hao, P.K. Wong, Appl. Catal. B: Environ. 42 (2003) 47.
[235] M.V.B. Zanoni, P.A. Carneiro, G.A. Umbuzeiro, M.A. Anderson, in press.
[236] M. Kositzi, I. Poulios, K. Samara, E. Tsatsaroni, E. Darakas, J. Hazard. Mater. 146
(2007) 680.
[237] G. Ruppert, R. Bauer, G. Heisler, Chemosphere 28 (1994) 1447.
[238] A.N. Gopalakrishnan, V. Mohan, Ind. J. Environ. Protect. 17 (1997) 268.
[239] I.A. Balcioglu, I. Arslan, J. Adv. Oxid. Technol. 4 (1999) 189.
[240] C. Lizama, J. Freer, J. Baeza, H.D. Mansilla, Catal. Today 76 (2002) 235.
[241] W.A. Sadik, J. Photochem. Photobiol. A: Chem., in press.
[242] M. Muruganandham, K. Selvam, M. Swaminathan, J. Hazard. Mater. 144 (2007)
316.