Applied Catalysis B: Environmental 31 (2001) 145–157

Photocatalytic degradation pathway of methylene blue in water

Ammar Houas a , Hinda Lachheb a,b , Mohamed Ksibi a , Elimame Elaloui a ,
Chantal Guillard b , Jean-Marie Herrmann b,∗
a Laboratoire de Catalyse et Environnement, Ecole Nationale d’Ingénieurs de Gabès (ENIG), Gabès, Tunisia
b Laboratoire de Photocatalyse, Catalyse et Environnement LPCE, Ecole Centrale de Lyon, B.P. 163, 69131 Ecully Cedex, France
Received 10 August 2000; received in revised form 13 November 2000; accepted 13 November 2000

Abstract
The TiO2 /UV photocatalytic degradation of methylene blue (MB) has been investigated in aqueous heterogeneous suspen-
sions. In addition to a prompt removal of the color, TiO2 /UV-based photocatalysis was simultaneously able to oxidize the dye,
with an almost complete mineralization of carbon and of nitrogen and sulfur heteroatoms into CO2 , NH4 + , NO3 − and SO4 2− ,
respectively. A detailed degradation pathway has been determined by a careful identification of intermediate products, in
particular aromatics, whose successive hydroxylations lead to the aromatic ring opening. These results suggest that TiO2 /UV
photocatalysis may be envisaged as a method for treatment of diluted waste waters in textile industries. © 2001 Elsevier
Science B.V. All rights reserved.
Keywords: Titania; Photocatalysis; Photocatalytic degradation; Dyes; Methylene blue; Water decolorization; Water purification; Total
mineralization

1. Introduction [3,4] and chemical methods (chlorination, ozonation

[5]) are the most frequently used.
About 15% of the total world production of dyes Among the new oxidation methods or advanced oxi-
is lost during the dyeing process and is released in dation processes (AOP), heterogeneous photocatalysis
the textile effluents [1]. The release of those colored appears as an emerging destructive technology leading
waste waters in the ecosystem is a dramatic source of to the total mineralization of most of the organic pol-
non-aesthetic pollution, eutrophication and perturba- lutants [6–12], following the usually proposed mech-
tions in the aquatic life. As international environmen- anism.
tal standards are becoming more stringent (ISO 14001,
1. Absorption of efficient photons (hν≥E G =3.2 eV)
October 1996), technological systems for the removal
by titania
of organic pollutants, such as dyes have been recently
developed. Among them, physical methods, such as − +
(TiO2 ) + hν → eCB + hVB (1)
adsorption [2], biological methods (biodegradation)
2. Oxygen ionosorption (first step of oxygen re-
duction; oxygen’s oxidation degree passes from
∗ Corresponding author. Tel.: +33-4-72-18-64-93;
0 to −1/2)
fax: +33-4-78-33-03-37.
E-mail address: jean-marie.herrmann@ec-lyon.fr −
(J.-M. Herrmann). (O2 )ads + eCB → O2 •− (2)

0926-3373/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 6 - 3 3 7 3 ( 0 0 ) 0 0 2 7 6 - 9
146 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157

3. Neutralization of OH− groups by photoholes which 2. Experimental

produces OH◦ radicals
2.1. Materials
+
(H2 O ⇔ H+ + OH− )ads + hVB → H+ + OH• (3)
Methylene blue was supplied by a textile firm
4. Neutralization of O2 ◦− by protons
and used as received. Solutions were prepared using
O2 ◦− + H+ → HO2 ◦ (4) water from a Millipore Waters Milli Q purification
unit. The photocatalyst was titania Degussa P-25
5. Transient hydrogen peroxide formation and dismu- (anatase/rutile = 3.1; surface area = 50 m2 g−1 ,
tation of oxygen non-porous particles, mean size = 30 nm).

2HO2 ◦− → H2 O2 + O2 (5) 2.2. Photoreactor and light source

6. Decomposition of H2 O2 and second reduction of Two types of Pyrex reactors open to air were uti-
oxygen lized. Reactor 1 (90 ml) had a cylindrical shape with
H2 O2 + e− → OH◦ + OH− (6) a bottom optical window whose section area was ca.
11 cm2 , through which the suspension was irradiated.
7. Oxidation of the organic reactant via successive Constant agitation of the solution was insured by a
attacks by OH• radicals magnetic stirrer working perpendicularly to the re-
actor basis. The UV-irradiation was provided by a
R + OH◦ → R • + H2 O (7) high pressure mercury lamp (Philips HPK 125 W) and
passed through a 2.2 cm thick circulating-water cu-
8. Direct oxidation by reaction with holes vette to remove the heating IR beams. This cuvette
R + h+ → R+◦ → degradation products (8) was mounted with either Pyrex filters (transmittance >
290 nm) or with Corning 0–52 cut-off filters trans-
mitting wavelengths >340 nm. The intensity of the
As an example of the last process, holes can react UV radiation reaching the reactor was measured with
directly with carboxylic acids generating CO2 a radiometer and estimated to be 6.94 × 10−7 and
RCOO− + h+ → R• + CO2 (9) 2.39 × 10−7 Einstein s−1 (mol of photons per second).
Reactor 2 (1 l) was equipped with a plunging tube
A quasi-exhaustive list of various families of organic containing the same Philips HPK 125 W lamp. This
pollutants which can be treated by photocatalysis has tube had a Pyrex cylindrical jacket in which water
been given in [13]. In most cases, the degradation was circulated to avoid the heating of the solution.
is conducted for dissolved compounds in water with The intensity of the UV-radiation reaching the reactor
UV-illuminated titania. The possible extents of the (λ > 290 nm) delivered an efficient photonic flux (i.e.
technique concern the irradiation source and the phys- absorbable by titania) equal to 6 × 10−6 Einstein s−1
ical state of the pollutant. Recently, some works have (mol of photons per second).
reported the degradation of organic dyes induced by
visible light by photosensitization [14–17]. The inter- 2.3. Procedure
est is to use solar visible light which is free and inex-
haustible. A volume of 20 ml of the aqueous solution of MB
In the present article, it was attempted to establish (C0 = 72 ␮mol l−1 ) was introduced in reactor 1 with
the degradation pathway of a dye present in colored 50 mg of powder TiO2 (2.5 g l−1 ). In reactor 2, 750 ml
aqueous effluents from textile industries. The model of solution and 375 mg of TiO2 were used. The
dye chosen was methylene blue (MB). The kinetics of degradations were carried out at 293 K and at three
its disappearance was already studied in [18–38] but different pH values. The pH was adjusted either by
the detailed description of the main steps of the whole using NaOH or HNO3 . The suspension was first stirred
degradation process was not treated. in the dark for 60 min before irradiation. This was
 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157 147

sufficient to reach an equilibrated adsorption as de- carbon (TOC) was determined by using a Shimazu
duced from the steady-state concentrations. To deter- TOC analyzer (model 5050A).
mine the adsorption constants, different concentra- Intermediate compounds were identified by GC/MS
tions of MB were used. and LC/MS (photodiode arrays and API-ES nega-
tive). A CP-Sil 5CB (L = 25 m, ∅ = 0.32 mm,
2.4. Analyses e = 1.2 ␮m) and a Resteck Stabilwax (L = 30 m,
∅ = 0.25 mm, e = 0.5 ␮m) were used in GC/MS
analysis. The column temperature program was: 313 K
Before analysis, the aqueous samples were filtered
(2 min); 313–473 K (5 K min−1 , hold time: 10 min).
through 0.45 ␮m Millipore discs to remove TiO2 ag-
For LC/MS, a CAR-H (L = 300 mm, ∅ = 7.8 mm)
glomerates in suspension and to protect the various
and an Hypersil (L = 150 mm, V = 3.6 mm) were
chromatographic columns. A “Safas Monaco 2000”
employed.
UV/VIS spectrometer recording the spectra over the
190–750 nm range was used for the determination of
MB concentration to follow its kinetics of disappear- 3. Results and discussion
ance. A Beer–Lambert diagram was established to
correlate the absorbance at 670 nm to MB concentra- 3.1. Photocatalytic degradation of methylene blue in
tion. The chemical oxygen demand (COD) was deter- water by TiO2 /UV
mined by using a Bioblock COD analyzer based on
the method of acidic oxidation by dichromate. Anions 3.1.1. Adsorption of methylene blue on TiO2
and cations were analysed by HPLC using a Waters The kinetics of MB adsorption in the dark per-
501 isocratic pump, a Waters 431 conductivity detec- formed in reactor 1 for different initial concentrations
tor, and an IC-PAK HR anion column (l = 50 mm, is represented in Fig. 1. It can be seen that they are
i.d. = 4.6 mm, particules = 10 ␮m) or a Vydac of the Langmuirian type with an equilibrium adsorp-
Cation IC40 (l = 50 mm, i.d. = 4.6 mm). Eluents tion reached within 1 h, whatever the initial concen-
were, respectively, borate/gluconate at 0.9 ml min−1 trations. From the number of adsorbed molecules at
and HNO3 2.5 mM at 1.5 ml min−1 . Total organic a given final concentration, a linear transform can

Fig. 1. Kinetics of adsorption of MB in the dark.

148 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157

Fig. 2. Langmuirian linear transform of the curve in Fig. 1.

be determined by plotting (1/nads ) as a function of TiOH + OH− ⇔ TiO− (13)

Ceq .
According to the Langmuir model, the coverage θ Since MB has a cationic configuration, its adsorption
varies as is favored in alkaline solution.

nads KC 3.2. Real photocatalytic nature of the reaction

θ= = (10)
n0 1 + KC
Thence To obtain relevant information about the photo-

 catalytic degradation, it was necessary to carry out
1 1 1 1 experiments from which direct photolysis was ex-
= + (11)
nads n0 n0 K C cluded. Experiments were made in the absence of
TiO2 (neat photochemical regime) at two different
where n0 is the total number of adsorption sites and K
ranges of wavelengths (λ ≥ 290 and 340 nm). The
is the adsorption constant of MB. The linear transform
results are given in Fig. 4. It can be seen that both
is given in Fig. 2. It can be calculated that K is equal
photolyses could be neglected with respect to their
to 6.25 × 103 l mol−1 and n0 equal to 0.135 site nm−2 .
corresponding photocatalyses. As observed for other
The value of K is in agreement with that of other
aqueous pollutants, the photocatalytic nature of the
pollutants easily soluble in water. The value of n0 is
organic contaminant disappearance can presently be
quite equal to that of indigo carmine, which has an
confirmed. In addition, suspended titania can play the
analogous structure with two aromatic rings [39].
role of inner filter and the real photochemical part
Fig. 3 reports the increase in adsorption when the
in presence of a suspension should be much smaller
pH increases from 3 to 9. This is in agreement with
than that shown in homogeneous phase.
the influence of pH on the ionization state of titania.
At low pH (pH = 3), lower than the pzc of titania, the
surface becomes positively charged, whereas at pH > 3.3. Kinetics of methylene blue disappearance
pzc it becomes negative according to
The kinetics of disappearance of MB is represented
TiOH + H+ ⇔ TiOH2 + (12) in Fig. 4, with an adsorption period in the dark of
 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157 149

Fig. 3. Variations of MB adsorption at three different pH values.

90 min and with two λ ranges. It follows an appar- kKC

r = kθ = ≈ kKC = kapp C (14)
ent first order in agreement with a generally observed 1 + KC
Langmuir–Hinshelwood mechanism, with the rate r The total disappearance was achieved within 60 and
being proportional to the coverage θ which becomes 120 min for λ ≥ 290 and λ ≥ 340 nm, respectively.
proportional to C at low concentrations The apparent first order linear transforms are given in

Fig. 4. Disappearance of MB by photochemistry and by photocatalysis under UV-irradiation at λ = 290 nm and at λ = 340 nm.
150 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157

Fig. 5. First order linear transforms of disappearance of MB by photocatalysis under UV-irradiation at λ = 290 nm and at λ = 340 nm.

Fig. 5. The slopes yield apparent rate constants of 0.06 in our laboratory for other pollutants’ elimination. It
and 0.025 min−1 , respectively. has to be mentioned that, since the kinetics r = f (C)
follows a Langmuir–Hinshelwood mechanism, with a
3.4. Catalytic quantum yield linear increase of r at low concentrations [6], this cat-
alytic quantum yield could be consequently higher for
The catalytic quantum yield is defined as the ratio higher concentrations. It is high enough to ensure a
of the reaction rate r to the photonic flux ϕ, i.e. as reasonable degradation rate for diluted colored water
the ratio of the number of molecules transformed per solutions.
second to the number of incident efficient photons (i.e.
absorbable by titania) per second 3.5. Mineralization
r0
ρ0 = (15) The degradation leads to the conversion of organic
ϕ carbon into harmless gaseous CO2 and that of nitrogen
The initial rate of disappearance of MB was found to and sulfur heteroatoms into inorganic ions, such as
be equal to 1.1 and to 3.5 ␮mol l−1 min−1 for λ ≥ 340 nitrate and ammonium, and sulfate ions, respectively.
and 290 nm, respectively. The corresponding efficient The formation of inorganic ions is reported in Fig. 6
photonic fluxes ϕ were found equal to 2.39 × 10−7 for an irradiation period of 1000 min.
and to 6.94 × 10−7 mol photons s−1 . The resulting The shape of the curves of sulfate and ammonium
initial photocatalytic quantum yields were found equal ion evolution indicates that both products are initial
to ρ0 = r0 /ϕ = 0.12 and 0.14%, respectively. ones with a non-nil initial formation rate. Heteroatom
These similar values underline the intrinsic charac- S seems to undergo a direct oxidation from the oxi-
ter of the quantum yield. They are quite in good agree- dation degree −2 to the highest final stable +6 one in
ment with a previous value given by Matthews [18] SO4 2− . However, the quantity of sulfate ions released
and are of the same order of magnitude as those found is lower than that expected from stoichiometry. This
 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157 151

Fig. 6. Evolution of sulfate, nitrate and ammonium ions in the solution during the photocatalytic degradation of MB.

could be first explained by a loss of sulfur-containing ably via the formation of a sulfoxide, a sulfone and
volatile compounds such as H2 S and/or SO2 . However, a sulfonic group as observed in Fenamiphos degrada-
this is not probable since both gases are very soluble tion [40]. The adsorption mode of BM by a coulom-
in water and known as readily oxidizable into sulfate bic interaction between the C–S+ =C functional group
by photocatalysis [6]. The more probable explanation and basic TiOH or TiO− sites of the surface of ti-
for the quantity of SO4 2− obtained smaller than that tania confirms this hypothesis. The initial steps of
expected from stoichiometry is given by the partially the degradation do not concern the amino groups, al-
irreversible adsorption of some SO4 2− ions at the sur- lowing nitrogen to conserve its initial (−3) oxidation
face of titania as already observed [6,9]. However, state.
this partial adsorption of SO4 2− ions does not inhibit The mineralization of organic carbon was followed
the photocatalytic degradation of pollutants [6,9]. The by TOC disappearance (Fig. 7) at λ ≥ 340 nm for both
higher amount of ammonium and nitrate ions is due photochemistry and photocatalysis. Whereas photo-
to the stoichiometric ratio N/S = 3 in the initial BM chemistry gives a very limited initial decrease of TOC,
molecule. photocatalysis produces a TOC disappearance, which
Fig. 6 shows that the initial formation of nitrogen- follows an apparent first order kinetics with a conver-
containing ions is that of ammonium ions exclusively, sion in CO2 of ca. 80% in 6 h.
since the evolution curve of nitrate ions has a sigma Simultaneously, the oxidative mineralization pro-
shape with an initial rate close to zero. NH4 + ions are cess was followed by chemical oxygen demand
subsequently slowly oxidized into nitrate ions, corre- (COD) disappearance (Fig. 8). The kinetic curve
sponding to the stable maximum oxidation state of ni- has a sigma shape indicating that it is related to the
trogen (+5). This is in agreement with the molecular formation of final products. The inflection point cor-
structure of MB with all the three nitrogen atoms in responds to a reaction time of ca. 30 min, which is
the −3 oxidation state. equal to the major part of MB disappearance. It could
Therefore, the initial step of MB degradation can be correspond to the aromatic rings opening with the
ascribed to the cleavage of the bonds of the C–S+ =C transient formation of carboxylic acids followed by
functional group in MB, leading to sulfate ions, prob- the evolution of CO2 according to the “photo-Kolbe”
152 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157

Fig. 7. Kinetics of disappearance of total organic carbon (TOC) (a) by photochemistry (without TiO2 ) and (b) by photocatalysis (TiO2 +UV).

Fig. 8. Kinetics of disappearance of chemical oxygen demand (COD).

 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157 153

Fig. 9. Kinetics of the photocatalytic degradation of MB at three different pH values.

reaction length, oxygen pressure, temperature, mass and tex-

ture of the catalyst). If the protons have any kinetic
R–COO− + h+ → R• + CO2 (9) influence on the rate limiting step, therefore r should
be a function of [H+ ], with the following formal form:
3.6. Influence of the pH
r = f (C; [H+ ]) = kθθH+
The pH is a complex parameter since it is related

kKC
to the ionization state of the surface = θH+ ≈ kKCθH+ = kapp CθH+
1 + KC
TiOH + H+ ⇔ TiOH2 + (12) = kapp C[H+ ]n (16)

TiOH + OH+ ⇔ TiO− (13) where kapp is the pH-independent apparent rate con-
stant and n is the kinetic partial order with respect to
as well as to that of reactants and products such as
proton concentration. At constant C, the log–log plot
acids and amines. Presently, three pH values have been
of Eq. (16)
selected: one reasonably acidic (3), one reasonably
basic (9) and natural pH (6.7). The kinetics of MB dis- log r = log k + n log[H+ ] = log k − npH (17)
appearance for the three pH values are given in Fig. 9.
It can be seen at time t = 0, that the quantities of MB has given a linear transform log r = f (pH), whose
adsorbed increased with pH and that the disappear- slope (−n) was found equal to +0.08. This means that
ance of MB was favored at basic pH. As seen above, the kinetic partial order n equal to −0.08 is signifi-
the reaction rate r is given by cantly negative but its absolute value can be neglected.
kKC Similar results for n were obtained by plotting the log
r = kθ = ≈ kKC = kapp C (14) of the mean reaction rate of DCO disappearance as a
1 + KC
function of pH. This clearly demonstrates that no pro-
where k is the true rate constant and is a complex func- tons will intervene in the rate-determining step of the
tion of several physical parameters (light flux, wave- whole process.
154 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157

Scheme 1. Photocatalytic degradation pathway of MB.

4. Identification of intermediate products identified by comparison with commercial standards

and by interpretation of their fragment ions in the mass
The intermediates generated during the degradation spectra. It has been attempted to identify the main aro-
process were analysed by GC/MS and LC/MS and matic metabolites resulting from MB decomposition.
 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157 155

They are presented in Scheme 1, where they are logi- Actually, the sulfoxide was identified by GC/MS at
cally reported according to their decreasing molecular m/e = 303 (Scheme 1). The electrophilic attack of
weight. OH◦ concerned the free doublet of heteroatom S, mak-
The titania-based photocatalytic oxidation in wa- ing its oxidation degree pass from −2 to 0. However,
ter is not selective by contrast with the selective mild the passage from C–S+ =C to C–S(=O)–C requires the
oxidation in pure organic gaseous or liquid phase conservation of the double bond conjugation, which
of aliphatic or substituted aromatic hydrocarbons per- induces the opening of the central aromatic ring con-
formed with the same titania catalyst as discussed taining both heteroatoms, S and N. The origin of H
in [6]. Two oxidative agents can be considered: the atoms necessary to C–H and N–H bond formation can
photo-produced holes h+ (mainly involved in the de- be proposed from the proton reduction by photogen-
carboxylation reaction (“photo-Kolbe) (Eq. 9)) and/or erated electrons as already observed in alcohol dehy-
the OH◦ radicals, which are known as strongly ac- drogenation [41] and pesticide degradation [42].
tive and degrading but non-selective agents and can
be generated by the following reactions: H+ + e− → H◦ (19)

1. Oxidation of water by holes An alternative rearrangement of the phenothiazine

structure is presented in Scheme 2.
(H2 O ⇔ H+ + OH− ) + h+ → H+ + OH◦ (3) The sulfoxide group can undergo a second attack by
an OH◦ radical producing the sulfone (non-detected)
2. Transient formation of hydroperoxide radicals
and causing the definitive dissociation of the two ben-
O2 (g) + e− → O2 ◦− (2) zenic rings

O2 ◦− + H+ → HO2 ◦ (4) NH2 –C6 H3 (R)–S(=O)–C6 H4 –R + OH◦

2HO2 ◦ → H2 O2 + O2 (5) → NH2 –C6 H3 (R)–SO2 + C6 H5 –R (20)

H2 O2 + e− → OH− + OH◦ (6) and/or

NH2 –C6 H3 (R)–S(=O)–C6 H4 –R + OH◦

In the present case, photo-holes are certainly not con-
cerned by the initial step since the reactant is cationic → NH2 –C6 H4 –R + SO2 –C6 H4 –R (21)
and not electron donor. By contrast, the OH◦ radicals The oxidation degree of sulfur has now increased from
can attack the C–S+ =C functional group in MB, which 0 to +5. Subsequently, the sulfone can be attacked
is in direct coulombic interaction with titania’s surface itself by a third OH◦ radical for giving a sulfonic acid
as evidenced by the influence of the pH. Therefore, detected in metabolites at m/e = 230, 218 and 158.
the initial step of MB degradation can be ascribed to
the cleavage of the bonds of the C–S+ =C functional SO2 –C6 H4 –R + OH◦ → R–C6 H4 –SO3 H (22)
group in MB
Presently, sulfur has reached its final, stable and max-
R–S+ =R + OH◦ → R–S(=O)–R + H+ (18) imum oxidation degree (+6) and the final release

Scheme 2. Electronic reorganization during the passage of MB adsorbed to the sulfoxide form.
156 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157

of SO4 2− ions can be attributed to a fourth attack R–C6 H4 –N(CH3 )–CHO + OH◦
by OH◦ → R–C6 H4 –N(CH3 )–C◦ =O + H2 O (31)
◦ ◦ +
R–C6 H4 –SO3 H + OH →R–C6 H4 + SO4 2−
+2H
R–C6 H4 –N(CH3 )–C◦ =O + OH◦
(23)
→ R–C6 H4 –N(CH3 )–COOH (32)
Radical R–C6 H4 ◦ can subsequently react either
+
with OH◦ giving a phenolic compound (see Scheme 1) R–C6 H4 –N(CH3 )–COOH + h
or with a H◦ radical generated by reaction (19). → R–C6 H4 –N◦ –CH3 + H+ (33)
Concerning the mineralization of the three
nitrogen-containing groups in MB molecule, two cases Subsequently, the phenyl-methyl-amine radical is
have to be examined. First, the central imino-group degraded, probably by successive attacks by OH◦ rad-
undergoes a N=C double bond cleavage induced by icals. The aromatic ring will undergo hydroxylations
the cleavage of the double bond of the –S+ = group producing phenolic metabolites whose degradation
in para position in the central aromatic ring. The has already been studied, hydroxyhydroquinone being
saturation of the two amino bonds is obtained by the last aromatic compound detected before the ring
H◦ radicals mentioned above, yielding a substituted opening. The amino group can form ammonium ions
aniline as seen for products at m/e = 303, 230 and as developed above, which are slowly oxidized into
218 in Scheme 1. As previously observed, the amino nitrate or can be directly oxidized into hydroxylamine
group can be substituted by an OH◦ radical forming leading also to nitrate [43].
the corresponding phenol and releasing a NH2 ◦ rad- Many other hydroxylated intermediates have cer-
ical which generates ammonia and ammonium ions, tainly been formed but they were difficult to detect
estimated to be primary products as mentioned above. because of their poor extractibility owing to their hy-
drophilic character.
R–C6 H4 –NH2 + OH◦
→ R–C6 H4 –OH + NH2 ◦ (24)
◦ ◦ 5. Conclusions
NH2 + H → NH3 (25)
NH3 + H+ → NH4 + (26) Methylene blue could be successfully decolorized
and degraded by titania-based photocatalysis at room
The other two symmetrical dimethyl-phenyl-amino
temperature. A detailed reaction mechanism could be
groups undergo a progressive degrading oxidation of
presented from the initial step of adsorption involving
one methyl group by an attack by OH◦ radical pro-
the cationic functional group of MB molecule, which
ducing an alcohol (not detected), then an aldehyde
is probably adsorbed perpendicularly to the surface
(Scheme 1), which is spontaneously oxidized into acid,
down to the final products (CO2 , SO4 2− , NH4 + and
which decarboxylates into CO2 by photo-Kolbe reac-
NO3 − ). The degradation intermediates originated
tion.
from the initial opening of the central aromatic ring
R–C6 H4 –N(CH3 )2 + OH◦ and their subsequent metabolites were formed in
agreement with general rules already put in evidence
→ R–C6 H4 –N(CH3 )–CH2 ◦ + H2 O (27) in the degradation of other complex molecules in
◦ ◦
R–C6 H4 –N(CH3 )–CH2 + OH water. It can be concluded that photocatalysis can
→ R–C6 H4 –N(CH3 )–CH2 OH (28) decontaminate colored used waters. Photocatalysis
◦ appears as the only sub-discipline of heterogeneous
R–C6 H4 –N(CH3 )–CH2 OH + OH catalysis, which is able to convert organic pollutants
→ R–C6 H4 –N(CH3 )–CH◦ –OH + H2 O (29) to CO2 in water without heating nor using high pres-
◦ ◦ sure of oxygen nor requiring chemical reactants or
R–C6 H4 –N(CH3 )–CH –OH + OH
additives. In addition, since our experiments can be
→ R–C6 H4 –N(CH3 )–CH(OH)2 successfully extrapolated to solar pilot devices, this
→ R–C6 H4 –N(CH3 )–CHO + H2 O (30) advanced oxidation technology can be envisaged to
 A. Houas et al. / Applied Catalysis B: Environmental 31 (2001) 145–157 157

clean colored effluents in semi-arid countries such as [15] F. Zhang, J. Zhao, T. Shen, H. Hidaka, E. Pelizzetti, N.
Tunisia to re-use or recycle water, which is becoming Serpone, Appl. Catal. B: Environ. 15 (1998) 147.
a precious raw material all around the world. [16] G. Wu, T. Wu, J. Zhao, H. Hidaka, N. Serpone, Environ. Sci.
Technol. 33 (1999) 2081.
[17] G. Liu, X. Li, J. Zhao, S. Horikoshi, H. Hidaka, J. Mol.
Catal. A: Chem. 153 (2000) 221.
Acknowledgements [18] R.W. Matthews, J. Chem. Soc., Faraday Trans. 1 85 (1989)
1291.
This work was supported by a CMCU Project [19] R.W. Matthews, Water Res. 25 (1991) 1169.
[20] P. Reeves, R. Ohlhausen, D. Sloan, K. Pamplin, T. Scoggins,
(99F1201). J.M.H and C.G. thank Dr. M.F. Grenier-
C. Clark, B. Hutchinson, D. Green, Solar Energy 48 (1992)
Loustalot and Mr. O. Paisset for LC/MS assistance. 413.
[21] R.F.P. Nogueira, W.F. Jardim, J. Chem. Educ. 70 (1993) 861.
[22] J.E. Valladares, J.R. Bolton, in: Photocatalytic Purification
References and Treatment of Water and Air, H.A Al-Ekabi, D. Ollis
(Eds.), Elsevier, Amsterdam, 1993, p. 111.
[1] Color Chemistry. Synthesis, Properties and Applications of [23] Yu.M. Artem’ev, M.A. Art’emeva, M.G. Vinogradov, T.I.
Organic Dyes and Pigments, 2nd revised Edition, in: H. Ilika, Russ. J. Appl. Chem. 67 (1994) 1354.
Zollinger (Ed.), VCH, 1991. [24] S. Lakshami, R. Renganathan, S. Fujita, J. Photochem.
[2] P.B. Dejohn, R.A. Hutchins, Tex. Chem. Color. 8 (1976) 69. Photobiol. A: Chem. 88 (1995) 163.
[3] S.S. Patil, V.M. Shinde, Environ. Sci. Technol. 22 (1988) [25] N. Xu, Z. Shi, Y. Fan, J. Dong, J. Shi, M.Z.C. Hu, Ind. Eng.
1160. Chem. Res. 38 (1999) 373.
[4] A.T. More, A. Vira, S. Fogel, Environ. Sci. Technol. 23 (1989) [26] B.N. Lee, W.D. Liaw, J.C. Lou, Environ. Eng. Sci. 16 (1999)
403. 165.
[5] Y.M. Slokar, A.M. Le Marechal, Dyes Pigments 37 (1998) [27] K. Tennakone, S. Senadeera, A. Priyadharshana, Sol. Energy
335. Mater. Sol. Cells 29 (1993) 109.
[6] J.M. Herrmann, Water treatment by heterogeneous [28] S. Klementova, D.M. Wagnerova, J. Indian Chem. Soc. 69
photocatalysis, in: F. Jansen, R.A. van Santen (Eds.), (1992) 470.
Environmental Catalysis, Catalytic Science Series, Vol. [29] R.D. Lowe, R.D. Snook, Analyst 118 (1993) 613.
1, Imperial College Press, London, 1999, Chapter 9, [30] S. Caglio, P.G. Righetti, Electrophoresis 14 (1993) 997.
pp. 171–194. [31] J.E. Valladares, J.R. Bolton, Photocatalytic purification and
[7] Photocatalysis and Environment. Trends and Applications, in: treatment of water and air, Trace Met. Environ. 3 (1993) 111.
M. Schiavello (Ed.), Kluwer Academic Publishers, Dordrecht, [32] S.L. Mali, V. Vaidya, R.L. Pitliya, V.K. Vaidya, S.C. Ameta,
1988. Asian J. Chem. 6 (1994) 796.
[8] Photocatalysis, Fundamentals and Applications, in: N. [33] S. Lakshmi, R. Renganathan, S. Fujita, J. Photochem.
Serpone, E. Pelizzetti (Eds.), Wiley Interscience, New York, Photobiol. 88 (1995) 163.
1989. [34] H.M. Cao, R.-W. Zhang, Y.-H. Guo, Y.-Y. Wang, G.-Q. Yang,
[9] C. Guillard, J.M. Herrmann, P. Pichat, Catal. Today 17 (1993) S.-K. Wu, Mater. Sci. Forum 175–178 (2) (1995) 663.
7. [35] L. Zhang, C.-Y. Liu, X.-M. Ren, J. Photochem. Photobiol. A
[10] Photocatalytic Purification and Treatment of Water and Air, 88 (1995) 47.
in: H.A Al-Ekabi, D. Ollis (Eds.), Elsevier, Amsterdam, 1993. [36] R.F.P. Nogueira, W.F. Jardim, J. Chem. Educ. 70 (1993) 861.
[11] D.W. Bahnemann, J. Cunningham, M.A. Fox., E. Pelizzetti, [37] Y.M. Artem’ev, M.A. Artem’eva, M.G. Vinogradov, T.I. Ilika,
P. Pichat, N. Serpone, in: R.G. Zeep, G.R. Helz, D.G. Zh. Prikl. Khim. 67 (1994) 1542.
Crosby (Eds.), Aquatic Surface Photochemistry, F.L. Lewis [38] V.Yu. Gak, V.A. Nadtochenko, J. Kiwi, J. Photochem.
Publishers, Boca Raton, 1994, p. 261. Photobiol. A 116 (1998) 57.
[12] O. Legrini, E. Oliveros, A.M. Braun, Chem. Rev. 93 (1993) [39] M.Vautier, C. Guillard, J.M. Herrmann, in press.
671. [40] C. Guillard, M. Fischer, J.M. Herrmann, A. Agüera, A.
[13] D.M. Blake, Bibliography of work on the photocatalytic Tejedor, L. Piedra, A. Fernandez-Alba, in press.
removal of hazardous compounds from water and [41] P. Pichat, J.M. Herrmann, J. Disdier, H. Courbon, M.N.
air, NREL/TP-430-22197, National Renewable Energy Mozzanega, Nouv. J. Chim. 5 (1981) 627.
Laboratory, Golden Co., 1997. [42] M. Kerzhentsev, C. Guillard, J.M. Herrmann, Catal. Today
[14] C. Pulgarin, G.M. Pajonk, J. Bandara, J. Kiwi, Meeting ACS 27 (1996) 215.
Division of Environmental Chemistry, Anaheim CA Paper [43] K. Nohara, H. Hidaka, E. Pelizzetti, N. Serpone, J.
No. 232, 1995, p. 767. Photochem. Photobiol. A: Chem. 102 (1997) 265.